1. IntroductionRickettsia rickettsii is a tick-borne pathogen that causes Rocky Mountain spotted fever (RMSF) and Brazilian spotted fever in North, Central, and South America. In the United States and Canada, the American dog tick (Dermacentor variabilis) and the Rocky Mountain wood tick (Dermacentor andersoni) are reputed as major vectors of this pathogen (Burgdorfer, 1975) while from Mexico to Argentina various ticks of the genus Amblyomma have been implicated in natural maintenance of R. rickettsii and transmission to humans. These include A. aureolatum, A. cajennense sensu lato, A. dubitatum, A. tenellum (published as A. imitator), A. parvum and A. scultptum (de Rodaniche, 1953; Guedes et al., 2005; Pinter and Labruna, 2006; Labruna et al., 2008; Labruna, 2009; Oliveira et al., 2010; Guedes et al., 2011; Labruna et al., 2011; Soares et al., 2011; Brites-Neto et al., 2013; Dzul-Rosado et al., 2013; Krawczak et al., 2014). In addition, brown dog ticks (Rhipicephalus sanguineus sensu lato) are involved in transmission of R. rickettsii in Mexico, the south-western US, and Brazil (Bustamante and Varela, 1947; Demma et al., 2005; de Almeida et al., 2013).
In the United States, the lone star tick (Amblyomma americanum) is one of the most aggressive human-biting ticks. Normally, larvae and nymphs of this species feed on a variety of medium-sized to large mammals and ground-feeding birds, while adult ticks prefer feeding on ungulates. However, all three life stages readily attach to and feed on humans, often in large numbers. A. americanum is a known vector of a number of bacterial agents including Francisella tularensis, Ehrlichia chaffeensis, E. ewingii, and R. amblyommatis (reviewed by (Childs and Paddock, 2003)), E. ruminantium-like organism referred to as the Panola Mountain Ehrlichia (PME) (Loftis et al., 2006, 2008) and Heartland virus (Savage et al., 2013). The geographical range of A. americanum has dramatically increased in the past 30 years, ostensibly due to expansion of white-tailed deer populations (Childs and Paddock, 2003). It is now distributed from west-central Texas, north to Iowa, and eastward spanning all the eastern United States. Along the Atlantic Coast, the range of this species extends from northern Mexico as far north as New York, New Hampshire and Maine (Springer et al., 2014) (http://www.cdc.gov/ticks/maps/lone_star_tick.pdf).
A large part of the A. americanum geographical range overlaps with the known distribution of both D. variabilis and reported cases of RMSF. The two tick species utilize the same species of vertebrate hosts as both A. americanum immatures and D. variabilis adults feed on medium-sized mammals. Therefore, there is a potential for A. americanum larvae and nymphs to become naturally exposed to R. rickettsii whether by acquiring blood meal from infected animals or while feeding side-by-side with infected D. variabilis adults.
Indeed in the pre-PCR era, Parker and coauthors described identification of the RMSF agent in questing A. americanum nymphs in Oklahoma (Parker et al., 1943). They also summarized a dozen spotted fever cases recognized in Louisiana, Maryland, Missouri, Oklahoma and Texas between 1926 and 1942, for which available evidence suggested transmission by lone star ticks (Parker et al., 1943). More recently, R. rickettsii DNA was detected in several A. americanum collected in Kansas (Berrada et al., 2011). Moreover, at least one PCR- and sequence-confirmed human case of RMSF in North Carolina has been associated with the bite of an A. americanum tick (Breitschwerdt et al., 2011) suggesting that the lone star tick may indeed serve as a vector of R. rickettsii. Yet, vector competence of the lone star tick for R. rickettsii has not been properly evaluated in a controlled environment unlike that of several other Amblyomma spp. distributed from Mexico to South America. Therefore, we assessed the ability of A. americanum to acquire, maintain and transmit R. rickettsii under laboratory conditions.
2. Methods2.1. Rickettsial isolatesTwo isolates of R. rickettsii originating from different geographical regions of the US were used in this study. Isolate BSF-Di-6 was originally isolated in 1961 from a spleen and liver homogenate taken from a Virginia opossum (Didelphis virginiana) trapped in Hanover County, VA (Bozeman et al., 1967). Since then, the isolate was propagated and stored in yolk-sack culture for a total of five passages. The AZ-3 isolate was cultured in 2004 from blood of an Arizona patient who died of RMSF (Eremeeva et al., 2006). Prior to the current study, this isolate was propagated and stored in Vero cells for a total of three passages (Karpathy et al., 2007). According to a molecular analysis, BSF-Di-6 and AZ-3 belong to different genotypes within the Clade I of R. rickettsii isolates (Karpathy et al., 2007). In preliminary trials, both isolates have been shown to cause nonlethal clinical infection in guinea pigs when inoculated with dosages containing approximately 103 to 105 copies of rickettsial DNA. Survival of laboratory animals through the infection was necessary to ensure successful feeding of ticks. Our preliminary experiments demonstrated that guinea pigs infected with low dosages of either isolate exhibited self-limiting fever after incubation period of 5–9 days, scrotal edema with or without necrosis, erythematous dermatitis of ears, and swollen, hemorrhagic footpads.
2.2. Preparation of tissue homogenatesIn advance of the study, groups of male guinea pigs were inoculated intraperitoneally with either of the two isolates. Each inoculum consisted of infected Vero cells containing approximately 105 copies of rickettsial DNA resuspended in 3 ml of in Snyder-1 media (sodium phosphate glutamate buffer). These guinea pigs were euthanized at the height of infection – 7–9 days post infection (DPI). Spleen and liver homogenates were prepared from individual guinea pigs by triturating tissues in SRM buffer (sucrose glutamate buffer (0.22 M sucrose, 0.1 M potassium phosphate, 0.005 M sodium l-glutamate, pH 7.0)) supplemented with 5 mM MgCl2 and 1% Hypaque-76 (Nycomed, Inc., Princeton, NJ). The amount of pathogen in homogenates derived from individual animals was evaluated by quantitative PCR. A single homogenate with the highest amount of rickettsial DNA was chosen from among guinea pig groups infected with each of the two isolates and aliquoted to contain 105 copies of R. rickettsii DNA per aliquot. These aliquots were stored in liquid nitrogen until used to infect the study animals.
2.3. Amblyomma americanum colonyA. americanum ticks used in this study were derived from Rickettsia-free colonies maintained at the CDC Medical Entomology Laboratory by feeding on pathogen-free naïve New Zealand white rabbits as previously described (Troughton and Levin, 2007). The colony originated from adult ticks collected from vegetation at a State Park 15 miles SW of Atlanta, Georgia and has been maintained in the laboratory for six generations. Absence of either pathogenic or endosymbiotic Rickettsia spp., as well as known tick-borne Anaplasma, Borrelia, Ehrlichia species and Heartland virus, in this colony has been assured in every generation by PCR-testing samples of larvae from progeny of every female and serological assessment of all rabbits used for tick feeding.
2.4. Model animalsPathogen free and tick-naïve male guinea pigs 6–9 wks.-old were used as model animals. Throughout the study, guinea pigs were monitored daily for clinical signs of infection including fever (defined as body temperature ≥ 39.5 °C), scrotal edema, dermatitis and edema of ears and foot pads (Walker et al., 1977). Samples of skin from the ear and EDTA-treated whole blood from the ear vein were collected from experimental guinea pigs every 2–3 days using sterile 2mm-diameter ear punches (Kent Scientific Corporation, Torrington, CT) and 27-gauge needles respectively. Animals were euthanized at 10–14 days after infection using barbiturate overdose (Beuthanasia®-D,Intervet International B.V.). Samples of serum, liver, spleen, testis with epididymis, lung and heart were collected at the time of euthanasia. Blood and tissue samples were stored at −20° C until tested by PCR for the presence of R. rickettsii DNA and sera samples were refrigerated at 4° C for serologic testing.
In addition, two mix-breed (hound-type) 1 year-old male dogs were used to assess the ability of R. rickettsii-infected adult ticks to transmit infection to canine hosts. Throughout the study, the dogs were housed indoors – in a climate-controlled animal facility as described before (Levin et al., 2014a). Dogs were provided with water and a standard dry food diet (Laboratory Canine Diet 5006; Purina, Fairburn, GA, USA) ad libitum. The appetite, behavior, disposition, and level of activity of each dog were monitored twice daily throughout the study. Dogs exhibiting decreased appetite were fed soft canned dog food (Pedigree, McLean, VA, USA).
2.5. Assessment of vector competenceFirst, we evaluated the ability of A. americanum to become infected with R. rickettsii as a result of feeding together on the same host with infected D. variabilis ticks. This experiment was conducted with the Di-6 isolate of R. rickettsii, which originated from an area within the overlapping ranges of D. variabilis and A. americanum. Groups of approximately 200 uninfected A. americanum larvae were placed on three naïve guinea pigs together with 25 Rickettsia-infected (38.0 ± 6.9%) D. variabilis nymphs from the laboratory colony described before (Schumacher et al., 2016). Engorged larvae were collected, allowed to molt and tested for the presence of R. rickettsii DNA.
Next, we assessed the ability of A. americanum larvae to acquire R. rickettsii from systemically infected model animals. Two groups of three guinea pigs were intraperitoneally inoculated with spleen/liver homogenates of either DI-6 or AZ-3 isolates of R. rickettsii. The inoculated animals were each infested with approximately 500 uninfected larvae of A. americanum on either the seventh or fifth day post-inoculation depending on the rickettsial isolate used so that tick feeding period would coincide with the peak of infection in the hosts. Engorged larvae were collected daily and allowed to molt. Twenty five freshly molted nymphs fed as larvae upon each needle-inoculated guinea pig were tested individually by PCR.
To evaluate the capability of infected nymphs for transmission of the pathogen to susceptible hosts, remaining nymphs from each of the inoculated guinea pigs were fed on three naïve guinea pigs in pools of 50. Simultaneously, uninfected A. americanum larvae were placed on the same guinea pigs for assessment of rickettsial transmission between infected and uninfected ticks as a result of side-by-side feeding and to propagate R. rickettsii-infected tick colonies. Engorged ticks were allowed to molt and 25 molted adults and 25 nymphs from each guinea pig were tested for the presence of rickettsial DNA.
For assessment of transovarial transmission (TOT), adult ticks derived from guinea pig-fed nymphs were allowed to feed on New Zealand white rabbits as previously described (Troughton and Levin, 2007; Levin and Schumacher, 2016). Rabbits were infested with 20 pairs (male/female) of A. americanum ticks from either Di-6- or AZ-3-infected cohorts. Engorged female ticks were kept in an incubator until they laid eggs and tested for the presence of R. rickettsii DNA soon after oviposition. Approximately 100 larvae from the progeny of each individual female were tested in a single pool by PCR to assess frequency of TOT.
We also tested the ability of R. rickettsii-infected adult ticks to transmit infection to canine hosts. Each of the two dogs was infested with 20 pairs of A. americanum ticks as described by Zemtsova et al. (2010). One dog was infested with ticks from the cohort infected with the DI-6 isolate of R. rickettsii and the other – with the AZ-3 isolate. The absence of antibodies to spotted fever group (SFG) rickettsiae in each dog was confirmed prior to enrollment into the study by the indirect immunofluorescence assay (IFA) as described below. Throughout the 2-wk study period, the body temperature of each dog was measured every morning; and temperatures ≥ 39.5 °C were defined as febrile (Aiello and Moses, 2016). Ear skin biopsies 2 mm diameter were collected aseptically from the ear twice per week under general anesthesia for PCR as skin biopsies appear to be significantly more useful in detecting rickettsial infection than blood-PCR (Levin et al., 2016). At 14 days post infestation, dogs were treated with oral doxycycline as previously described (Levin et al., 2014a) and transferred to an unrelated study.
2.6. PCR and serologyDNA extraction and PCR procedures were carried out in separate facilities. DNA was extracted from tick and tissue samples using the Qiagen DNEasy Blood & Tissue kit, whereas blood samples were processed with the Flexi Gene kit (Qiagen Inc., Valencia, CA) according to manufacturer’s protocols. The presence of Rickettsial DNA was detected by PCR using primers RR190-547F and RR190-701R to amplify a 154-bp fragment of the rompA gene of Rickettsia spp. as described by Eremeeva et al. (2003). All samples were tested in duplicate. Two negative (distilled water) and two positive (R. massiliae plasmid) samples were included in each run. Samples demonstrating amplification prior to 40 cycles with appropriate amplicon melting temperature in both replicates were considered positive.
IFA was performed on guinea pig and dog sera, as previously described (Lennette et al., 1995), using FITC labeled goat anti-guinea pig and anti-dog IgG (γ) conjugates diluted per manufacturer’s recommendations (KPL, Inc. Gaithersburg, Maryland, USA). Slides were spotted with whole cell R. rickettsii antigen, air-dried and fixed in acetone. Serum samples were initially screened at 1/16 and 1/256 dilutions, and positive samples were titrated to the endpoint in a two-fold dilution series. Serologic data are reported as the reciprocal of the last dilution showing positive fluorescence. Titers ≥ 32 were considered positive.
Prevalence of rickettsial infection in cohorts of ticks and frequencies of TOT were compared by Chi-squared and Fisher exact tests (2-tailed) at the 95% confidence level.
2.7. Ethics statementThe study was undertaken at a facility fully-accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care International. Care and husbandry of animals were performed in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals 8th edition. All procedures of this study were pre-approved by the Centers for Disease Control Institutional Animal Care and Use Committee and monitored by a veterinarian stationed on-site. The US Animal Welfare Act and all associated regulations were strictly followed.
3. ResultsA. americanum larvae fed on guinea pigs simultaneously with Di-6–infected D. variabilis nymphs were tested (25 per guinea pig) by PCR after they molted into the nymphal stage. In total, five (6.7 ± 2.9%) out of 75 tested ticks contained R. rickettsii DNA. As the rickettsial isolate used in this experiment originated from an area occupied by both tick species, results demonstrate that R. rickettsii infection in A. americanum can result from interspecies spill-over when the American dog tick and the lone star tick utilize the same vertebrate hosts.
Uninfected A. americanum larvae acquired R. rickettsii from guinea pigs needle-inoculated with either of the two isolates and transmitted the pathogen transstadially to the nymphal stage. Approximately 49% and 33% of freshly molted nymphs were PCR-positive after feeding as larvae upon animals infected with DI-6 and AZ-3 isolates, respectively (Table 1). The prevalence of infection was significantly different between the two cohorts (p = 0.04662).
According to our previous observations, clinical signs of rickettsial infection in guinea pigs infected with the Di-6 and AZ-3 isolates include fever (evident after 5–9 DPI), scrotal inflammation and dermatitis of ears and foot-pads. In addition, rickettsial DNA can be detected in guinea pig skin or venous blood starting 6–8 DPI (Levin et al., 2016). When A. americanum nymphs from infected cohorts were placed on naïve guinea pigs in the current study, the majority of animals displayed similar clinical signs of rickettsial infection. Of the nine guinea pigs infested with ticks infected with the isolate DI-6, seven developed fever above 39.5 °C starting on 7–9 DPI and lasting 2–4 days. The same seven animals had scrotal and foot-pad edema as well as ear and foot-pad dermatitis of varying severity. Additionally, rickettsial DNA was detected in skin, blood and internal organs of these seven guinea pigs (Table 2). Of the nine guinea pigs fed upon by AZ-3-infected nymphal ticks, five developed fever starting 6–8 DPI, but eight exhibited scrotal edema, as well as dermatitis and edema of ears and foot pads. Rickettsial DNA was also detected in skin, blood and internal organs of eight of the nine guinea pigs (Table 2). One of the animals exposed to A. americanum nymphs infected with AZ-3 isolate of R. rickettsii became morbidly ill by day 9 post-infestation and had to be euthanized. All 18 guinea pigs fed upon by A. americanum nymphs infected with DI-6 and AZ-3 isolates of R. rickettsii developed IgG antibodies reactive with R. rickettsii antigen within 14 days after placement of ticks, including the one euthanized on day 9 post infestation (Table 2). Reciprocal IFA titers ranged from 64 to 2048 indicating that all guinea pigs were exposed to the pathogen delivered via tick bite. Geometric mean titers in groups exposed to DI-6 and AZ-3 isolates were 362 and 420, respectively.
Uninfected A. americanum larvae fed side by side with infected nymphs acquired both DI-6 and AZ-3 isolates and transmitted the pathogen transstadially. The prevalence of infection in resulting nymphs varied significantly between individual guinea pigs – from 0% to 40% (as detected by testing 25 ticks per guinea pig). Overall, 11% and 9% of freshly molted nymphs were PCR-positive after feeding as larvae side by side with ticks infected with DI-6 and AZ-3 isolates, respectively (Table 1). The prevalence of infection in resulting nymphs was not significantly different between DI-6 and AZ-3 cohorts.
Infected A. americanum nymphs used in the co-feeding experiment were allowed to molt and 25 freshly molted adult ticks (12 males and 13 females per guinea pig) from the Di-6 and AZ-3 cohorts were tested by PCR. The prevalence of infection in unfed adult ticks was 68% and 52% in the 2 cohorts, respectively (Table 3) and did not differ between males and females. For both cohorts, prevalence of infection in unfed adult ticks was higher than in unfed nymphs (Tables 1 and 3), though the differences were not statistically significant (p = 0.096 and p = 0.105, respectively). Even higher percentages of adult ticks from the same cohorts tested positive after their feeding upon naïve New Zealand rabbits. In either cohort, 100% of males tested PCR-positive after feeding. Female ticks were tested at the end of oviposition. Among the 99 ovipositing females from the Di-6 cohort, 97 (98%) contained rickettsial DNA; and 65 out of 92 (70.7%) females were positive in the AZ-3 cohort (Table 3). The described step-by-step increase in the prevalence of rickettsial infection in cohorts of A. americanum from unfed nymphs to engorged adults also confirms efficient transmission of both R. rickettsii isolates between simultaneously feeding ticks at all stages of their life cycle.
Rickettsial DNA was also detected in larval batches produced by some of the infected females indicating successful TOT. A total of 27 (27.8%) out of 97 individual larval batches produced by females infected with the Di-6 isolate tested positive (Table 3). Of the 65 females infected with the AZ-3 isolate, 9 (13.8%) produced infected progeny. The frequency of TOT of R. rickettsii was significantly higher (p = 0.036) among ticks infected with the Di-6 isolate than those infected with AZ-3 isolate.
Dogs infested with adult ticks from either Di-6- or AZ-3-infected cohorts became infected and clinically ill. The body temperature rose above 39.5 °C on day 4 and 3 post-infestation, respectively, and peaked at 41.2 °C and 41.5 °C on day 7. Both dogs developed abundant petechiae on the gums, buccal mucosa, and in the conjunctiva of both eyes. Widespread macular skin rash appeared on all exposed areas of the skin including ears, limbs, testes and the abdomen. In addition, dogs became anorectic and lethargic within a week of tick placement. Taken together, clinical signs exhibited by dogs after infestation with adult A. americanum infected with either Di-6 or AZ-3 isolates of R. rickettsii were identical to those of canine RMSF (Levin et al., 2014a). In addition, R. rickettsii DNA was identified in skin biopsies from the dog exposed to Di-6 on days 7 and 9 post-infestation and the dog exposed to AZ-3 on days 5, 7, and 12.
4. DiscussionThe geographical range of R. rickettsii includes most the North and South Americas, from Canada to Argentina. Some strains of R. rickettsii can cause up to 80% mortality in humans. Different tick species serve as vectors of R. rickettsii including A. aureolatum, A. parvum, A. sculptum, D. andersoni, D. variabilis, and Rh. sanguineus. In addition, varying degrees of vector competence have been demonstrated under laboratory conditions for A. tenellum (published as A. imitator), D. occidentalis, and H. leporispalustris (Parker et al., 1933, 1951; Freitas et al., 2009; Oliveira et al., 2010). In the United States, the Rocky Mountain wood tick – D. andersoni and the American dog tick – D. variabilis are the best known vectors of R. rickettsii while Rh. sanguineus s.l. acts as the primary vector in parts of the Southwest. Conversely, in Central and South America several species of the genus Amblyomma act as vectors and reservoirs of the pathogen (Bustamante and Varela, 1946; de Rodaniche, 1953; Lima et al., 1995; de Lemos et al., 1997; Pinter and Labruna, 2006; Labruna et al., 2008; Paddock et al., 2008; Labruna, 2009; Dzul-Rosado et al., 2013).
In the US, A. americanum is the predominant human-biting tick throughout the Southeast (Merten and Durden, 2000; Stromdahl et al., 2015). Its geographical range overlaps with that of RMSF in the southern and southeastern states, where it is often much more numerous than D. variabilis (Apperson et al., 2008). A. americanum represented 53% of all ticks recovered from humans in Mississippi from 1990 through 1999 and approximately 56% of ticks collected from humans between 2008 and 2014 in Texas (Goddard, 2002; Mitchell et al., 2016), while a study in Georgia and South Carolina found that 83% of ticks attached to 460 patients were lone star ticks (Felz et al., 1996). A recent survey in North Carolina showed that A. americanum accounted for 93–95% of ticks removed from the skin of outdoor workers (Lee et al., 2014). The potential role of A. americanum in transmission of R. rickettsii to humans has been suspected in Oklahoma and North Carolina (Cooley and Kohls, 1944; Breitschwerdt et al., 2011) and the pathogen has been occasionally identified in questing lone star ticks (Parker et al., 1943; Heise et al., 2010; Berrada et al., 2011). However only scant experimental evidence has been published to support transmission of R. rickettsii by A. americanum ticks. For example, transmission of R. rickettsii to guinea pigs by A. americanum nymphs and adults was demonstrated under laboratory conditions (Maver, 1911), but the efficiency of acquisition or transmission could not be assessed. In 1933, Parker with coauthors reported that the lone star tick can occasionally transmit the agent of RMSF transovarially (Parker et al., 1933). Yet, follow-up experiments by Burgdorfer and Briton provided equivocal results as A. americanum infected with a highly virulent Sawtooth-♀2 strain of R. rickettsii time after time failed to maintain infection transovarially because of high mortality in both females and eggs (Burgdorfer and Brinton, 1975). With the discovery of other pathogenic and endosymbiotic bacteria transmitted by the lone star tick (Childs and Paddock, 2003) and associated obstacles to establishment and maintenance of conclusively uninfected tick colonies, the vector competence of this species for R. rickettsii has not been sufficiently studied.
To evaluate the ability of A. americanum for acquisition, maintenance and transmission of R. rickettsii, we used a specific pathogen-free (SPF) tick colony maintained in our laboratory for several generations. This colony is routinely tested in every generation to insure absence of known tick-borne bacteria including Anaplasma, Borrelia, Ehrlichia, and Rickettsia species. Use of SPF ticks and laboratory animals assured that any clinical and pathological signs in vertebrate hosts as well as all positive PCR results in either tick or animal samples can be unequivocally attributed to a single pathogen – R. rickettsii. Different isolates of a rickettsial pathogen can cause widely disparate effects on the survival of its vector (Levin et al., 2009). This suggested that the ability of a tick species to acquire, maintain and transmit a pathogen may vary between different strains circulating within or outside of the geographical range of that tick. Therefore, one of the rickettsial isolates used in this study (Di-6) originated from the area within the geographical range of A. americanum while the other isolate (AZ-3) was derived from Arizona – outside of the lone star tick range. Guinea pigs served as model animals for the bulk of the experiments as a less costly alternative to natural hosts of A. americanum. Yet, dogs were used for the final confirmation that infected A. americanum are capable of causing clinically recognizable infections in a natural host.
Our experiments demonstrate that A. americanum can be a competent vector for both of these two isolates of R. rickettsii, although the degree of competence differs somewhat between the isolates. Uninfected larval ticks are able to acquire the pathogen from either systemically infected (rickettsiemic) hosts or simultaneously feeding infected nymphs. Prior to this study, we determined that in naïve guinea pigs infected with Di-6 and AZ-3 isolates of R. rickettsii the first rise of body temperature occurs at 5–9 days post-infection (Levin et al., 2016) – at 5–7 DPI in needle-inoculated animals and 7–9 DPI in those infected by ticks. In addition, rickettsial DNA is not detectable in either the venous blood or skin biopsies of guinea pigs until 6–8 after placement of R. rickettsii-infected ticks (Levin et al., 2016). Meanwhile, A. americanum larvae and nymphs complete their engorgement on guinea pigs within 3–4 and 5–6 days, respectively. Thus, both larval and nymphal ticks in the current study detached from the hosts prior to either clinical or molecular manifestations of a systemic infection and rickettsemia. Based on this, we concluded that the actual sharing of the pathogen between simultaneously feeding ticks in the present study took place prior to the development of a systemic infection and rickettsemia in the hosts – i.e. via the co-feeding route. The efficiency of rickettsial proliferation by this route is relatively low – fewer larvae acquire either isolate via co-feeding than from rickettsiemic hosts (11% vs 49% and 9% vs 33% for the Di-6 and AZ-3 isolates, respectively). Yet, additional sharing of Rickettsia takes place within both nymphal and adult cohorts of ticks during their feeding, which is demonstrated by the rising prevalence of infection from the unfed nymphs (33–49%) to blood-fed male ticks derived from those nymphs (100%).
Placement of nymphs infected with either rickettsial isolate upon naïve guinea pigs induced typical clinical infection and/or seroconversion confirming the competence of A. americanum nymphs in transmission of R. rickettsii to susceptible hosts. This corroborates results reported by Maria Maver a century ago (Maver, 1911). Likewise, two dogs infested with Rickettsia-positive adult A. americanum both developed canine RMSF. Thus lone star ticks acquiring R. rickettsii with larval or nymphal blood-meal successfully retain the pathogen transstadially and transmit it to susceptible hosts during subsequent feedings. Our results confirm vector competence of A. americanum for R. rickettsii under laboratory conditions, although the efficiency of pathogen acquisition is lower than what had been reported for some other vector species including D. variabilis, A. aureolatum, and H. leporispalustris (Labruna et al., 2008; Freitas et al., 2009; Schumacher et al., 2016).
Ticks are often referred to as both vectors and reservoirs of pathogenic rickettsiae due to their ability to maintain bacteria from generation to generation via TOT. In some vector species, TOT of R. rickettsii can be highly effective with up to 100% of D. andersoni and A. aureolatum progenies being infected under laboratory conditions (Burgdorfer, 1963; Burgdorfer and Brinton, 1975; Burgdorfer, 1988; Labruna et al., 2011). Conversely, rickettsial DNA was detected in larval samples from only nine (47.4%) progenies of the 19 D. variabilis females infected with the Di-6 isolate (Schumacher et al., 2016), and the frequency of TOT of R. rickettsii (Taiaçu strain) in A. sculptum (reported at the time of publication as Amblyomma cajennense) varied between 6.2% and 42.8% in cohorts of ticks exposed to the pathogen at different life stages (Soares et al., 2011). Unfortunately, these published TOT rates are not directly comparable to each-other because individual studies utilized different methods to infect ticks, different host species, and isolates of Rickettsia that were sympatric with the vector in some cases and allopatric in others. In our experiments, R. rickettsii DNA was detected in progenies of approximately 28% and 14% of A. americanum females infected with the Di-6 and AZ-3 isolates, respectively. Thus, an isolate originating from within the geographic range of A. americanum seems to be transmitted transovarially twice as often as an isolate circulating outside of that range. It must be pointed out, however, that an A. americanum female produces several thousand eggs and only 100 larvae were tested from each progeny. Therefore, the true frequency of TOT in our study might have been underestimated if the filial prevalence of infection was below 1%.
A number of previous studies have reported higher rates of TOT when ticks acquired R. rickettsii as larvae or nymphs, whereas low to no transmission was observed in those becoming infected only during the adult feeding (Parker et al., 1933; Burgdorfer and Brinton, 1975,; Piranda et al., 2011; Soares et al., 2011). It has been proposed that ticks infected only at the adult stage may fail to transmit R. rickettsii transovarially because of the inability of the pathogen to sufficiently colonize the ovaries in females (Burgdorfer and Brinton, 1975). In our experiments, the prevalence of infection with either Di-6 or AZ-3 isolates in all cohorts of unfed freshly molted nymphal or adult A. americanum was always below 100%, and the prevalence of infection increased with every successive feeding due to sharing of the pathogen between simultaneously feeding ticks. Therefore, it was impossible to separate individual females, which acquired the Rickettsia during larval feeding and successfully maintained infection throughout the life cycle, from those acquiring bacteria during nymphal or adult feedings. Still, it is noteworthy that the proportion of infected A. americanum females producing infected progenies was approximately 2–3 times lower than the prevalence of infection in the preceding nymphs and approximately 4–5 times lower than in engorged adults. We infer that even among ticks acquiring R. rickettsii as larvae, only a fraction was capable of TOT. Overall, it appears that the ability of A. americanum to carry R. rickettsii from generation to generation is lower than those reported for D. andersoni, D. variabilis, and A. aureolatum, but comparable to that in A. sculptum.
In our experiments, we noted significant differences between the Di-6 and AZ-3 isolates in their infectivity for larval A. americanum as well as in the frequency of TOT. In both cases, ticks were more successful in acquiring and transmitting the eastern Di-6 isolate, which is derived from the area within the geographical range of A. americanum. Considering the continent-wide distribution of R. rickettsii correlating with molecular differentiation of its isolates (Karpathy et al., 2007), it is likely that diverse strains of the pathogen are adapted to different vectors. Conversely, different populations of a widely distributed tick species may vary in the vector competence for the different strains of a pathogen (Levin et al., 2009; Moraes-Filho et al., 2015).
Considering the relatively low frequency of TOT in this study, A. americanum may not be able to perpetually maintain the infection in the wild from generation to generation. However, new introductions of R. rickettsii into A. americanum populations are possible as a result of interspecies spill-over events when lone star ticks feed on already rickettsiemic hosts or side by side with infected D. variabilis, as demonstrated in this study. Moreover, the prevalence of R. rickettsii infection in all experimental cohorts of A. americanum steadily increased from larval to adult stage due to sharing of the pathogen between ticks feeding together. Although the current study was primarily performed in guinea pigs, existence and efficiency of the co-feeding route of rickettsial transmission had been demonstrated in natural hosts – dogs (Zemtsova et al., 2010). This co-feeding route can be realized even when ticks feed on insusceptible or previously immunized animals (Zemtsova et al., 2010; Levin et al., 2014b). The fact that two geographically and genetically distinct isolates of R. rickettsii were both transmitted between simultaneously-feeding ticks, which completed engorgement prior to development of rickettsemia in the hosts, indicates pervasiveness of the co-feeding route of rickettsial transmission in A. americanum.
Lone star ticks are notorious for attaching to their hosts in extremely high numbers. There are reports of over 3000 A. americanum including larvae, nymphs and adults being collected from a single jack rabbit in Texas, and white-tailed deer sampled in Arkansas had as many as 2550 A. americanum ticks per ear (Goddard, 1989; Goddard and McHugh, 1990). Such dense infestations provide suitable conditions for transmission of pathogens – including R. rickettsii – between simultaneously feeding ticks. In addition, different life stages of this tick utilize the same host species as D. variabilis including eastern cottontail rabbits, dogs, feral swine, deer, horses, raccoons, and even squirrels (Koch, 1982; Demarais et al., 1987; Mock et al., 1991; Pung et al., 1994; Kollars et al., 2000; Allan et al., 2001; Goldberg et al., 2002; Cooney et al., 2005; Duell et al., 2013; Barrett et al., 2014). It is unknown how often these hosts become systemically infected with R. rickettsii but utilization of the same hosts and large numbers of simultaneously feeding ticks provide opportunity for A. americanum to acquire the pathogen from infected individual ticks of either the same or a different species. Once R. rickettsii is picked up by A. americanum with larval or nymphal blood-meal, it will be sustained transstadially and further transmitted to other ticks or susceptible hosts during the next feeding.
On the other hand, the prevalence of a pathogenic Rickettsia within a tick population may be diminished or even suppressed when a significant proportion of ticks is infected with an endosymbiotic Rickettsia species due to transovarial interference (Burgdorfer et al., 1981a, 1988; Macaluso et al., 2002). In 1973, W. Burgdorfer and colleagues identified a novel spotted fever group Rickettsia in questing lone star ticks, which was provisionally designated as “Rickettsia amblyommii” (Burgdorfer et al., 1981b; Stothard and Fuerst, 1995). This endosymbiotic bacterium, only recently formally named Rickettsia amblyommatis (Karpathy et al., 2016), has since been detected in most of the examined populations of A. americanum with prevalence of infection often exceeding 40% (reviewed by Karpathy et al., 2016). As transovarial maintenance of multiple infections of Rickettsia spp. in ticks is believed to be all but impossible, A. americanum is not currently considered as a vector of RMSF (Goddard and Norment, 1986; Weller et al., 1998; Childs and Paddock, 2003). However, recent studies identified dual and even triple rickettsial coinfections in questing ticks. A single D. variabilis tick collected in Ohio was found coinfected with Rickettsia bellii, R. montanensis, and R. rickettsii (Carmichael and Fuerst, 2010); two A. americanum collected in Kansas were concurrently infected by R. rickettsii and R. amblyommatis (Berrada et al., 2011); and two Peruvian A. maculatum were co-infected with both Candidatus R. andeanae and R. parkeri (Flores-Mendoza et al., 2013). In addition, laboratory-reared A. dubitatum were able to acquire and transmit R. rickettsii transstadially despite being 100% infected with R. bellii (Sakai et al., 2014). These findings indicate that presence of one Rickettsia in a tick does not automatically preclude acquisition or transstadial transmission of another bacterial species in nature. The ability of R. amblyommatis to interfere in acquisition, maintenance and transmission of R. rickettsii by the lone star tick needs to be verified and evaluated.
In conclusion, our results correspond with earlier deductions that the lone star tick may be an epidemiologically significant vector of Rocky Mountain spotted fever (Parker et al., 1943; Berrada et al., 2011; Breitschwerdt et al., 2011). We have demonstrated that A. americanum is a competent vector of R. rickettsii. It can acquire the pathogen from systemically infected hosts as well as from (simultaneously feeding) infected ticks of either the same or different species. Interspecies transmission of Rickettsia via co-feeding can allow new introductions of the pathogen into A. americanum populations as spill-over from the primary vectors. Ticks acquiring R. rickettsii during feeding successfully retain the pathogen during the molting process from larval to nymphal and from nymphal to adult stages, and transmit it to susceptible hosts during the subsequent feedings. Despite a relatively low efficiency of TOT, the aggressive feeding behavior of A. americanum, all stages of which can feed on humans, and its tendency to cluster on a host in high numbers may allow the lone star tick to sustain the transmission cycle of R. rickettsii in nature. Its role in the ecology and epidemiology of RMSF in North America may be underestimated. A number of recent studies demonstrated paucity of R. rickettsii in D. variabilis – the reputed primary vector of the causative agent of RMSF in the eastern US (Ammerman et al., 2004; Moncayo et al., 2010; Stromdahl et al., 2011; Nadolny et al., 2014). These findings triggered a revision of vector competence of tick species other than D. variabilis, including the work present here. However, a comparative analysis of A. americanum and D. variabilis vector competence was not an objective of the current study, thus authors refrain from making any conclusions or speculations in that regard. A separate study is warranted to further elucidate relative roles of different tick species as vectors of R. rickettsii in nature and its transmission to humans.