Serological survey of Ehrlichia species in dogs, horses and humans: zoonotic scenery in a rural settlement from southern Brazil.

The aims of this study were to determine the seroprevalence of Ehrlichia spp. and risk factors for exposure in a restricted population of dogs, horses, and humans highly exposed to tick bites in a Brazilian rural settlement using a commercial ELISA rapid test and two indirect immunofluorescent assays (IFA) with E. canis and E. chaffeensis crude antigens. Serum samples from 132 dogs, 16 horses and 100 humans were used. Fifty-six out of 132 (42.4%) dogs were seropositive for E. canis. Dogs > one year were more likely to be seropositive for E. canis than dogs ≤ one year (p = 0.0051). Ten/16 (62.5%) and 8/16 (50%) horses were seropositive by the commercial ELISA and IFA, respectively. Five out of 100 (5%) humans were seropositive for E. canis and E. chaffeensis. Rhipicephalus sanguineus (n = 291, 97.98%) on dogs and Amblyomma cajennense (n = 25, 96.15%) on horses were the most common ticks found. In conclusion, anti-Ehrlichia spp. antibodies were found in horses; however, the lack of a molecular characterization precludes any conclusion regarding the agent involved. Additionally, the higher seroprevalence of E. canis in dogs and the evidence of anti-Ehrlichia spp. antibodies in humans suggest that human cases of ehrlichiosis in Brazil might be caused by E. canis, or other closely related species.

INTRODUCTION

Ehrlichiosis is a tick-borne disease caused by Ehrlichia spp. that affects animals and humans worldwide9 27 39 54. The disease is historically endemic in tropical and subtropical regions and has increasingly been recognized, not only in traditionally endemic areas, but also in temperate regions25. This may be attributed to several factors, including the improved diagnostic tools, and both environmental and climate changes which directly influences the distribution of ticks24.

In some regions of Brazil, dogs and horses are frequently exposed to ticks29 30. Dogs and humans are exposed and susceptible to infection by many of the very same tick-borne bacterial pathogens in the order Rickettsiales, including Ehrlichia spp.24. Ehrlichia canis is the causative agent of canine monocytic ehrlichiosis and is the main Ehrlichia species present in dogs in Brazil54. Additionally, E. canis DNA was also amplified from the blood of six human patients with clinical signs of human monocytic ehrlichiosis in Venezuela, suggesting that E. canis can also be associated with clinical manifestation in humans44.

In humans there are two recognized diseases to date caused by Ehrlichia species; human monocytic ehrlichiosis (HME), caused by Ehrlichia chaffeensis; and human granulocytic ehrlichiosis (HGE) due to Ehrlichia ewingii 41. Human ehrlichiosis cases have been serologically identified in Brazil since 19806 11 12; however, the Ehrlichia species associated with these cases were not identified. Additional cases of human ehrlichiosis have been serologically diagnosed in other South American countries, including Argentina45, Chile31, Peru38 and Venezuela43 44.

Equine monocytic ehrlichiosis (EME), caused by Neorickettsia risticii (formerly E. risticii), and equine granulocytic anaplasmosis (EGA) caused by Anaplasma phagocytophilum (formerly E. equi) are the two recognized diseases caused by ehrlichial species16. Ticks have never been implicated in the transmission of N. risticii 3 16, whereas ticks belonging to the Ixodes genus are the vectors of A. phagocytophilum 56. While clinical cases of EME have been reported in southern and southeastern regions of Brazil10 17 18, horses serologically reactive to A. phagocytophilum were reported with clinical alterations in the central-west region of the country48.

The increasing number of people living in rural settlements in Brazil, with poor-resources and precarious living conditions, inadequate sanitary care, and sanitary education, associated to the presence of pets, production animals, wild animals, and ticks sharing the same environment, may represent an important source of several zoonotic pathogens. Thus, the aims of the present study were to i) determine the seroprevalence of Ehrlichia species in a restricted population of dogs, horses, and humans highly exposed to tick bites, ii) identify the tick species parasitizing dogs and horses, and iii) determine risk factors for exposure in a rural settlement from Paraná State, southern Brazil.

MATERIALS AND METHODS

Ethical principles: The study was approved by the Ethics Committee in Animal Experimentation and Animal Welfare at Universidade Estadual de Londrina (UEL) (protocol number 34/2011), and was conducted according to the ethical principles of animal experimentation, adopted by the Brazilian College of Animal Experimentation. Collection of human blood samples was approved by the Ethics Committee on Human Research at UEL (protocol number 53/2011).

Area:The rural settlement is located in Alvorada do Sul county (22° 54′ 34.4″ S 51° 13′ 49.1″ W), Paraná State, southern Brazil. The area is located within the rural perimeter of Alvorada do Sul, 16 km from downtown, 380 m above sea level. The region presents a subtropical climate with rainfall throughout the year, with a higher concentration in summer months (average temperature of 25 °C)26.

The region is subdivided in 60 homesteads, each with an area of approximately 12 hectares, amounting to 786 hectares. The main activity of subsistence is the cultivation of grains and vegetables. The area also comprises 20% of a native forest having a diverse fauna, with populations of capuchin monkeys, capybaras, opossums, coatis, and wild canids, as well as a wide variety of birds and fishes. There is a barrage close to the area, where the habitants often fish and bath. This region provides the maintenance of various ticks, to which dogs, horses, and humans are continually exposed.

Study design: According to the seasonal dynamics of adult ticks50 52, samples were collected in March 2011, which represents the end of the summer in the South hemisphere. Sampling was performed house-to-house, comprising all 60 homesteads of the area.

A questionnaire focused on epidemiological aspects was given to each owner. Breed, age, gender of their dogs and horses, and presence or previous contact with ticks were evaluated. The age of the dogs was classified into groups of ≤ one year and > one year. Age, gender and history of previous contact with ticks were also addressed for humans.

Collection of ticks: A total of 297 and 26 tick specimens were collected from dogs and horses, respectively. Ticks were removed and kept in 70% ethanol-labeled tubes in order to identify each host. Ticks were classified according to taxonomic keys2 20 34 42.

Sampling: Dog (n = 132) and horse (n = 16) blood samples (up to 10 mL per individual) were collected by venipuncture of the jugular vein. Blood specimens from human (n = 100) were collected by nurses by venipuncture of the brachial vein. All samples were collected in tubes without anti-coagulant and kept at room temperature (25 °C) until visible clot retraction, centrifuged at 1500 g for five min, and the serum was separated and kept at -20 °C until processing.

Detection of anti-Ehrlichia spp antibodies: Serum samples of 138 dogs and 16 horses were tested for E. canis using a commercial ELISA rapid test (SNAP® 4Dx®, IDEXX Laboratories Inc., Westbrook, ME, USA), according to the manufacturer's instructions. The kit also detects antibodies anti-A. phagocytophilum, anti-B. burgdorferi (s.l.), and anti-Dirofilaria immitis antigen.

Anti-Ehrlichia spp. antibodies in horses and human serum samples were evaluated by indirect immunofluorescent assay (IFA) using E. canis (São Paulo strain) and E. chaffeensis (Arkansas strain) as antigens. Crude antigens were produced by culturing ehrlichiae in DH82 cells, as previously described1 51. IFA was performed with 10 µL of serum samples incubated at 37 °C for 30 min in slides previously seeded with E. canis or E. chaffeensis, washed three times for five min in phosphate buffered saline (PBS, pH 7.2), additionally washed by distilled water, then dried at room temperature. Twenty microliters of fluorescein isothiocyanate-conjugated rabbit anti-human IgG (Sigma-Aldrich, St. Louis, MO) at 1:800 dilution in PBS with 1% of bovine serum albumin and 1% Evans blue were applied onto the slide. Horse samples were titred with a dilution of 1:1200 of fluorescein isothiocyanate-conjugated rabbit anti-horse IgG (Sigma-Aldrich, St. Louis, MO). Slides were then incubated at 37 °C for 30 min, three times washed for five min, additionally washed by distilled water, allowed to air dry and subsequently examined using a microscope with a fluorescent light source. Samples were considered positive when reacting with dilution ≥ 1:645 11 19 39. Titers were determined to the largest dilution in which fluorescence was visualized around the bacteria (endpoint titers).

Statistical analysis: The Chi-square or Fisher's exact test were used to determine the difference between whether individual factors were associated with seropositivity to Ehrlichia spp. Odds ratio (OR), 95% confidence interval and p values were calculated separately for each variable. Results were considered significantly different when p < 0.05. Data were compiled and analyzed in Epi Info™ Software (version 3.5.3).

RESULTS

A total of 132 dogs were sampled, 83 (62.8%) of which were males and 49 (37.2%) females. All were mixed breed, with ages varying from six months to 12 years. By the commercial ELISA rapid test, 56/132 (42.4%; 95%CI: 33.9 - 51.3%) dogs were seropositive for E. canis. Dogs > one year were more likely to be seropositive for E. canis than dogs ≤ one year (OR = 3.13, 95% CI, 1.45 - 6.75%). No significant association was found between gender or presence of ticks, and seropositivity to E. canis. Data on E. canis seroprevalence are shown in Table 1. Additionally, anti-A. phagocytophilum antibodies were found in 3/132 (2.3%; 95% CI: 0.5 - 6.5%) dogs. Anti-B. burgdorferi antibodies and D. immitis antigens were not found in dogs.

Table 1

Seroprevalence of E. canis in dogs within each variable studied from a rural settlement, Paraná State, southern Brazil

Variable	+/N (%)	OR	95% CI	p-value
Presence of ticks
Yes	35/73 (47.9%)	1.66	0.82-3.36	0.1534
No	21/59 (35.6%)
Age (years)
> 1	43/82 (52.4%)	3.13	1.45-6.75	0.0051
≤1	13/50 (26%)
Gender
Male	33/83 (39.8%)	0.74	0.36-1.52	0.4200
Female	23/49 (46.9%)

+, Number of positive animals; N, number of samples per variable; OR, odds ratio; 95% CI, 95% confidence interval.

Horses included nine (56.2%) males and seven (43.8%) females, all mixed breed, with ages ranging from three to 15 years. When serum samples were evaluated by each test, 10/16 (62.5%; 95% CI: 35.4 - 84.8%) and 8/16 (50%; 95% CI: 24.7 - 75.3%) were positive for E. canis antigen by the commercial ELISA rapid test and IFA, respectively. Ten out of 16 (62.5%; 95% CI: 35.4 - 84.8%) horses were positive for E. chaffeensis antigen. Antibodies titers ranged from 256 to 2048 for E. canis and from 256 to 1024 for E. chaffeensis by IFA. Additionally, antibodies anti-B. burgdorferi were found in 2/16 (12.5%; 95% CI: 1.6 - 38.3%) horses. Antibodies anti-A. phagocytophilum were not found.

Humans included 48 (48%) males and 52 (52%) females, with ages varying from two to 79 years. Five out of 100 (5%; 95% CI: 1.6 - 11.3%) humans were seropositive for E. canis and E. chaffeensis antibodies, respectively. From the total of seropositive samples, two reacted for both agents. Antibodies titers ranged from 64 to 512 for E. canis and from 64 to 256 for E. chaffeensis by IFA. Seventy-five out of 100 (75%) humans recollected having been bitten by ticks. No significant association was found between age, gender or previous exposure to tick bites, and seropositivity to Ehrlichia spp.

Seropositive human samples were: female, 12 years-old, and kept five dogs at home, all seropositive for E. canis; male, 13 years-old, owned a dog seronegative for E. canis; female, 29 years-old, kept two dogs at home, seronegative for E. canis; male, 30 years-old, kept three dogs at home and all were seronegative for E. canis; female, 72 years-old, owned a dog seropositive for E. canis. All humans recalled tick and insect bites, and recollected acute febrile syndromes in the past.

A total of 297 ticks (154 males, 104 females, 34 nymphs and five larvae) were collected from 73/132 (55.3%; 95% CI, 46.4 - 64%) dogs, ranging from one to 26 ticks per animal. Three tick species were identified: Rhipicephalus sanguineus (n = 291, 97.98%), Amblyomma ovale (n = 5, 1.68%) and Amblyomma cajennense (n = 1, 0.34%). From the total of 73 dogs found infested by ticks, 68/73 (93.15%; 95% CI, 84.95 - 97%) were infested by R. sanguineus, 3/73 (4.11%; 95% CI, 1.41 - 11.4%) by A. ovale, and 1/73 (1.37%; 95% CI, 0.24 - 7.36%) by A. ovale and A. cajennense.

Twenty-six ticks (18 males and eight females) were collected from 7/16 (43.7%; 95% CI, 23.1 - 66.8%) horses, ranging from one to nine ticks per animal. Two tick species were identified: A. cajennense (n = 25, 96.15%) and Rhipicephalus (Boophilus) microplus (n = 1, 3.85%). A. cajennense ticks were found infesting 6/7 (85.7%; 95% CI, 48.6 - 97.4%) horses, and R. (B.) microplus in 1/7 (14.3%; 95% CI, 2.5 - 5.1%) horse.

DISCUSSION

Anti-Ehrlichia spp. antibodies were found in 62.5% of horses from southern Brazil by using the commercial ELISA rapid test. Despite this test has been developed for screening canine samples8, the assay uses antigen-specific conjugate and was previously validated for screening B. burgdorferi and A. phagocytophilum in horses7 28. Using this same commercial ELISA in horses from Denmark, France, French Guyana and Africa, anti-Ehrlichia spp. antibodies were not found21 35 36. Previous studies using this point-of-care ELISA assay in dog serum samples reported cross-reactivity between E. canis and E. chaffeensis antibodies40. In the present study, 50% of the horses were positive by IFA using E. canis and E. chaffeensis as antigens. Thus, differences in the number of seropositive horses found by each test were somewhat expected, since the commercial ELISA rapid test utilizes synthetic peptides from p30 and p30-1 outer membrane proteins of E. canis as antigen40, while IFA used E. canis and E. chaffeensis crude antigens.

A weak cross-reactivity between N. risticii and E. canis by IFA, ELISA and Western blot methods has been reported4 23 46 49. Recently developed Western blot tests coupled with serum absorption techniques have been used in order to identify the organisms involved and solve the serological cross-reactivity between these agents22. In the present study, horse antibody endpoint titers ranged from 256 to 2048 for E. canis and from 256 to 1024 for E. chaffeensis by IFA. Moreover, 43.7% of the horses were infested by ticks, in the majority A. cajennense (96.1%). Ehrlichia canis is transmitted through the bite of the brown dog tick R. sanguineus 15, while E. chaffeensis is mainly transmitted by A. americanum ticks55. Neither horse infestations by R. sanguineus nor E. canis infection in Amblyomma ticks are natural host-parasite interactions. In addition, no data are currently available on the vector competence of A. cajennense for E. chaffeensis 55. On the other hand, ticks have never been implicated in the transmission of N. risticii 16, which is transmitted upon ingestion of this bacterium in the metacercarial stage of trematodes encysting in aquatic insects by horses47. Thus, authors did not exclude the possibility on the involvement of a not-yet-described Ehrlichia species in the population of horses herein studied, which should be further molecularly identified and characterized.

IFA is considered the gold standard test for diagnosis of HME, although cross-reactivity between E. canis and E. chaffeensis by serological methods commonly occur54. In the present study, 5% of the humans were seropositive by IFA for either E. canis or E. chaffeensis antigens. Although clinical cases of human ehrlichiosis have been reported in Brazil by using E. chaffeensis antigens6 11 12, the existing cross-reactions between Ehrlichia species in serological assays precluded the determination of the etiological agent of human patients. Although molecular detection of E. chaffeensis DNA has already been related in wildlife in Brazil32 33, E. canis is indeed the most prevalent Ehrlichia species in the country54. Since 75% of humans of the present study reported past tick bites, and R. sanguineus ticks, the main vector of E. canis- were the most common tick species found (90%)- it is possible that anti-Ehrlichia spp. antibodies detected in this study were due to exposure to E. canis-infected R. sanguineus ticks, although this statement needs to be confirmed by direct detection of the agent in vertebrate blood.

Seropositivity to E. canis was found in 42.4% of the dogs. Serological surveys of E. canis in dogs from rural areas have found prevalence data ranging from 24.7% to 65.6% by different methods37 54. We found that age (> one year-old) is associated with seropositivity to E. canis (p = 0.0051), corroborating with other studies performed in the veterinary teaching hospital in Londrina City, southern Brazil53. Previous studies have reported that male dogs previously exposed to tick bites were at high risk of being seropositive for E. canis 13. In the present study, besides 56.2% of the dogs were males infested by ticks; association between gender or presence of ticks and seropositivity to E. canis was not observed. In addition, previous studies have stated that the commercial ELISA rapid test is able to identify dogs with titers > 32040. Thus, seroprevalence for E. canis in dogs from the studied area might be higher, since dogs with low titers may have not been recognized when a point-of-care ELISA assay is used14.

CONCLUSION

Antibodies anti-Ehrlichia species were found in horses by two different serological methods. However, the lack of a molecular characterization precludes any conclusion regarding the agent involved. The higher E. canis seroprevalence in dogs and the detection of anti-Ehrlichia spp. antibodies in humans suggest that human cases of ehrlichiosis in Brazil might be caused by E. canis, or other related species.