Bovine herpesvirus-1: Genetic diversity of field strains from cattle with respiratory disease, genital, fetal disease and systemic neonatal disease and their relationship to vaccine strains.

Bovine herpesvirus-1 (BoHV-1) causes disease in cattle with varied clinical forms. In the U.S. there are two BoHV1 subtypes, BoHV-1.1 and BoHV-1.2b. Control programs in North America incorporate modified live (MLV) or killed (KV) viral vaccines. However, BoHV-1 strains continue to be isolated from diseased animals or fetuses after vaccination. It is possible to differentiate BoHV-1 wild-type from MLV vaccine strains by determining their single nucleotide polymorphism (SNP) patterns through either whole-genome sequencing or PCR sequencing of genomic regions containing vaccine-defining SNPs. To determine the BoHV-1 subtype in clinical isolates and their relationship to MLV strains, 8 isolates from varied clinical disease at three different laboratories in the U.S. were sequenced and phylogenetically analyzed. Five samples were isolated within the past 5 years from New York and 3 were archived samples recovered 35 years prior from Oklahoma and Louisiana. Based on phylogenetic analysis, four of the cases appeared to be due to an MLV vaccine: 3 cases of aborted fetuses and one neonate with systemic BoHV-1 disease. One aborted fetus was from a herd with no reported history of MLV vaccination in two years. The remaining four isolates did not group with any MLV vaccines: two were associated with bovine respiratory disease, one with vulvovaginitis, and a fourth was determined to be a BoHV-1.2b respiratory isolate. Recovery of BoHV-1.1 that is very closely related to an MLV vaccine virus from a herd not receiving vaccines in an extended period prior to its isolation suggests that MLV viruses may remain latent or circulate within herds for long periods.

Disease risk surface for Coxiella burnetii seroprevalence in white-tailed deer.

Coxiella burnetii is considered a re-emerging zoonosis in many countries. The bacterium is enzootic in livestock and wildlife in the United States, and environmental contamination is widespread. Despite the potential for exposure, the estimated prevalence of Q fever in humans and animals is not well elucidated, and reported human infections in the United States are relatively rare. Zoonotic transmission of the bacterium is usually associated with abortions in domestic ruminants, but other modes of transmission, such as contact with infected blood and/or milk during field dressing of infected wildlife, have not been thoroughly investigated. Studies of zoonotic pathogen transmission between animal reservoir hosts and humans are usually established in response to documented emergence or re-emergence of a zoonosis in a particular locale, and, as such, the prevalence of infection in wildlife is largely unknown for many zoonotic pathogens, including C. burnetii. The objective of this study was to create a disease risk surface for C. burnetii seroprevalence in wild white-tailed deer (Odocoileus virginianus) in New York State. Blood samples were collected from hunter-harvested deer from across New York State in 2009 and 2010. The samples were processed and tested for the presence of anti-C. burnetii antibodies via indirect microimmunofluorescence assays using phase II C. burnetii strain RSA439. Overall, 14.50% of the tested white-tailed deer were C. burnetii phase II seropositive. The dual Kernel density estimation method was used to create a smoothed disease risk surface, which revealed variation in seroprevalence ranging from 0% to 32.0%. Areas of higher seroprevalence were detected in four discrete areas of Central New York and in one additional area in the southwest corner of the northern part of the state. This suggests certain locales where humans may be at increased risk for exposure to the bacterium secondary to contact with potentially infected deer.

Genome characterization of Salem virus reveals its evolutionary intermediate status in the subfamily Paramyxovirinae.

Salem virus (SalPV) was originally isolated from a horse during a disease outbreak in 1992 in the USA. In this study, we complete the genome characterization of SalPV and confirm the classification of this virus as a member of the subfamily Paramyxovirinae. The SalPV genome is 16,698 nucleotides in length, with six transcriptional units in the order 3'-N/P/V/C-M-F-G-L-5'. SalPV has a fixed 3-nt intergenic region, which is the same as in all Paramyxovirinae members except rubulaviruses and avulaviruses. The genome size of SalPV is greater than those of the "old" members of the subfamily, such as morbilliviruses and respiroviruses, but smaller than those of the "new" members, including henipaviruses and several unclassified novel viruses. Interestingly, this evolutionary "middle point" was also reflected in the phylogeny, suggesting that SalPV represents an important intermediate between the previously well-characterized paramyxoviruses and the recently emerged members of this subfamily.

Accuracy of a point-of-care ELISA test kit for predicting the presence of protective canine parvovirus and canine distemper virus antibody concentrations in dogs.

Canine parvovirus (CPV) and canine distemper virus (CDV) are highly infectious and often fatal diseases with worldwide distributions, and are important population management considerations in animal shelters. A point-of-care ELISA test kit is available to detect serum antibodies to CPV and CDV, and presumptively to predict protective status. The aim of this study was to determine the diagnostic accuracy of the test compared to CPV hemagglutination inhibition titers and CDV serum neutralization titers determined by a reference laboratory, using sera collected from dogs housed at animal shelters. The ELISA test was used under both field and laboratory conditions and duplicate specimens were processed using an extra wash step. The test kit yielded accurate results (CPV: sensitivity 92.3%, specificity 93.5%; CDV: sensitivity 75.7%, specificity 91.8%) under field conditions. CDV sensitivity was improved by performing the test under laboratory conditions and using an optical density (OD) meter (laboratory performed 94.0%; OD 88.1%). Point-of-care ELISA testing for serum CPV and CDV antibody titers was demonstrated to be a useful tool for determining antibody status when making decisions regarding the need for CPV and/or CDV vaccination and also in animal shelters for population management.

Long-term clinicopathological characteristics of alpacas naturally infected with bovine viral diarrhea virus type Ib.

BACKGROUND: Substantial bovine viral diarrhea virus (BVDV)-related production losses in North American alpaca herds have been associated with BVDV type Ib infection. OBJECTIVES: To classify and differentiate the long-term clinicopathological characteristics of BVDV type Ib infection of alpaca crias, after natural virus exposure. We hypothesized that persistently infected (PI) alpacas specifically demonstrate growth retardation, clinicopathological evidence of opportunistic infections, and early mortality. ANIMALS: Thirty-five crias naturally exposed to BVDV (18 acute, 3 chronic, 14 PIs), and 19 healthy cohort controls of 5 northeastern alpaca farms were prospectively evaluated over 2 years (September 2005-September 2008). METHODS: Observational cohort-control study. RESULTS: Chronically (viremia >3 weeks) and PI crias demonstrated significantly lower birth weights, decreased growth rates, anemia, and monocytosis compared with control animals. Common clinical problems of PI alpacas included chronic wasting, diarrhea, and respiratory disease. Median survival of PI alpacas that died was 177 days (interquartile range, 555) with a case fatality rate of 50% within 6 months of life. Transplacental infection was confirmed in 82% (9/11) of pregnant females on 1 farm, resulting in the birth of 7 PI crias (7/10 deliveries; 1 animal was aborted). Mean gestation at the beginning and end of BVDV exposure was 64 and 114 days, respectively. CONCLUSIONS AND CLINICAL IMPORTANCE: Natural BVDV type 1b infection during early pregnancy resulted in a high incidence of PI offspring. Although PI alpacas may have distinct clinical characteristics, verification of persistent viremia in the absence of endogenous, neutralizing antibodies is essential to differentiate persistent from chronic infection.

Canine H3N8 influenza virus infection in dogs and mice.

An H3N8 influenza virus closely related to equine influenza virus was identified in racing greyhound dogs with respiratory disease in 2004 and subsequently identified in shelter and pet dogs. Pathologic findings in dogs spontaneously infected with canine influenza virus were compared with lesions induced in beagle and mongrel dogs following experimental inoculation with influenza A/canine/Florida/43/2004. BALB/c mice were inoculated with canine influenza virus to assess their suitability as an experimental model for viral pathogenesis studies. All dogs inoculated with virus developed necrotizing and hyperplastic tracheitis and bronchitis with involvement of submucosal glands as well as mild bronchiolitis and pneumonia. Viral antigen was identified in bronchial and tracheal epithelial cells of all dogs and in alveolar macrophages of several dogs. Many dogs that were spontaneously infected with virus also developed bacterial pneumonia, and greyhound dogs with fatal spontaneous infection developed severe pulmonary hemorrhage with hemothorax. Virus-inoculated BALB/c mice developed tracheitis, bronchitis, bronchiolitis, and mild pneumonia in association with viral antigen in airway epithelial cells and in type 2 alveolar epithelial cells. Virus was not detected in extrarespiratory sites in any animals. The results indicate that canine influenza virus infection consistently induces acute tracheitis and bronchitis in dogs. Mice may be a useful model for some pathogenesis studies on canine influenza virus infection.

Detection of equine herpesvirus-1 in nasal swabs of horses by quantitative real-time PCR.

BACKGROUND: Early identification of inhalation-transmitted equine herpesvirus type 1 (EHV-1) infections has been facilitated by the availability of a number of real-time quantitative PCR (qPCR) tests. A direct comparison between nasal swab qPCR and traditional virus isolation (VI) requires a method for normalizing the qPCR samples and controlling for PCR inhibitors present in some clinical samples. OBJECTIVES: To quantify EHV-1 shedding in viral swabs using an internal control and to compare fast qPCR to VI for the detection of EHV-1 in nasal swabs from horses. ANIMALS: Fifteen horses experimentally infected with EHV-1. METHODS: Experimental study: Nasal swab samples were collected daily after experimental infection for up to 21 days. VI was performed by conventional methods. The DNA was prepared for qPCR with the addition of a known quantity DNA of Marek's disease virus as an internal control. qPCR was performed. RESULTS: The qPCR method detected virus up to day 21 after challenge, whereas VI detected virus only to day 5. The median Kaplan-Meier estimates for EHV-1 detection were 12 days for qPCR and 2 days for VI (P< .0001). When compared with VI, the sensitivity and specificity of qPCR were 97 (95% CI: 86-100) and 27% (95% CI: 20-35). CONCLUSIONS AND CLINICAL IMPORTANCE: We conclude that fast qPCR of nasal swab samples should be chosen for diagnosis and monitoring of herpesvirus-induced disease in horses. Recommended reference ranges of C(T) values are provided as well as justification of a minimum 10-day quarantine period.

Infectious diseases of dogs and cats on Isabela Island, Galapagos.

BACKGROUND: Vaccination and importation of dogs and cats are prohibited in the Galapagos, resulting in a uniquely isolated population. The purpose of this study was to determine the prevalence of infectious diseases of dogs and cats that impact their health, could spill over to native wildlife, or sentinel diseases of concern to humans. HYPOTHESIS: The isolation of dogs and cats in the Galapagos protects them from diseases common in mainland populations. ANIMALS: Ninety-five dogs and 52 cats presented during a neutering campaign. METHODS: A prospective cross-sectional study was performed. Blood was collected for serological and DNA evaluation of a panel of infectious diseases. RESULTS: Antibodies against parvovirus (100%), parainfluenza virus (100%), adenovirus 1/2 (66-67%), and distemper virus (22%) were present in dogs. Dirofilaria immitis was also common in dogs (34%), with lower prevalences of Wolbachia pipiens (22%), Bartonella sp. (13%), Ehrlichia/Anaplasma spp. (1%), and Mycoplasma haemocanis (1%) observed. Antibodies against panleukopenia virus (67%), Toxoplasma gondii (63%), calicivirus (44%), and herpesvirus 1 (10%) were detected in cats. Feline leukemia virus antigen, feline immunodeficiency virus antibody, or coronavirus antibodies were not detected. Bartonella sp. (44%) infections were common in cats, but only one was infected with M. haemofelis. CONCLUSIONS AND CLINICAL IMPORTANCE: Despite their relative seclusion from the rest of the world, cats and dogs of Isabela were exposed to many pathogens found in mainland South America. Parasite prophylaxis, neutering, and strict enforcement of animal movement restrictions would control a majority of the diseases. In the absence of vaccination, a reservoir of susceptible animals remains vulnerable to new disease introductions.

Isolation of Salem virus, a novel equine paramyxovirus, and assessment of its etiologic role in a disease outbreak.

Salem virus (SalV) is a recently identified equine virus belonging to the family Paramyxoviridae. The only known isolate was obtained from a horse that was involved in a disease outbreak of undetermined nature and the circumstances of its isolation suggested an etiologic role. However, the experimental infection of a colostrum-deprived foal failed to reproduce the disease; only mild neutropenia and temperature elevation were recorded. An additional attempt to establish an etiological relationship with the disease was made by conducting a retrospective evaluation of the serological profiles of animals involved in the outbreak. Animals reported as being affected by the disease according to a comprehensive United States Department of Agriculture (USDA) database were found to be 48% (n=27) positive for antibodies to SalV, but the percent positive for all horses, affected and unaffected, was actually higher at 56% (n=62). For 15 affected horses for which paired acute and convalescent serum specimens were available, no unequivocal seroconversions to SalV were identified. Furthermore, the horse from which SalV was isolated was not listed as one of the animals affected by the disease. In total, the evidence suggests that SalV was not the etiological agent of the disease and that its isolation was fortuitous.

Clinical, pathologic, immunohistochemical, and virologic findings of eastern equine encephalomyelitis in two horses.

Natural eastern equine encephalitis alphavirus (EEEV) infection was diagnosed in two adult horses with anorexia and colic, changes in sensorium, hyperexcitability, and terminal severe depression. Myocardium, tunica muscularis of stomach, intestine, urinary bladder, and spleen capsule had coagulative necrosis and perivascular lymphocytic infiltrate. Central nervous system (CNS) lesions were diffuse polioencephalomyelitis with leptomeningitis characterized by perivascular T lymphocyte cuffing, marked gliosis, neuronophagia, and multifocal microabscesses. Lesions were more prominent within cerebral cortex, thalamus, hypothalamus, and mesencephalon. EEEV was identified in the cytoplasm of cardiac myocytes and smooth muscle cells of spleen, stomach, intestine, urinary bladder, blood vessels, and dendritic cells. In the CNS, EEEV-positive cells included neurons, astrocytes, oligodendrocytes, microglia, and neutrophils. EEEV was isolated from the CNS of both horses. The detailed description of the encephalic and spinal EEEV localization and the findings of EEEV in extraneural tissues contribute to the understanding of this important multisystemic zoonotic disease.

Serologic survey of selected viral agents in recently captured wild North American river otters (Lontra canadensis).

Blood samples were collected from 64 wild North American river otters (Lontra [Lutra] canadensis) from northern and eastern New York State and analyzed for serologic evidence of exposure to selected viral agents during a 1995 1996 translocation program. No clinical signs of disease nor lesions suggestive of prior viral exposure were seen. Titers were detected for antibodies against canine distemper virus, canine herpesvirus-1, and canine parvovirus-2 but not for antibodies against canine adenovirus-1, canine coronavirus, canine parainfluenza virus, rabies virus, feline herpesvirus-1, feline calicivirus, or feline coronavirus. This is the first report of titers for antibodies against canine herpesvirus-1 in North American river otters, and it suggests a low prevalence of antibody titers against most canine viruses in otter populations in northern and eastern New York. Confounding variables in this study could include exposure to domestic dogs associated with the project, prolonged time spent in captivity, and concurrent bacterial or parasitic infection. Stress-associated humoral immune suppression could have altered serologic profiles, especially in otters exposed to dogs after trapping but before venipuncture.

Identification and phylogenetic comparison of Salem virus, a novel paramyxovirus of horses.

A virus that could not be identified as a previously known equine virus was isolated from the mononuclear cells of a horse. Electron microscopy revealed enveloped virions with nucleocapsid structures characteristic of viruses in the Paramyxoviridae family. The virus failed to hemabsorb chicken or guinea pig red blood cells and lacked neuraminidase activity. Two viral genes were isolated from a cDNA expression library. Multiple sequence alignments of one gene indicated an average identity of 45% as compared to Morbillivirus N protein sequences. A weaker relationship was found with Tupaia paramyxovirus (TPMV) and Hendra virus (HeV) N proteins. In the second gene, multiple open reading frames (ORFs) were identified, corresponding to the arrangement of the P, V, and C ORFs in the Morbillivirus and Respirovirus viruses. Short stretches in the C-terminal regions of the P and C proteins showed limited homologies to viruses in the Morbillivirus genus but no obvious relationship to viruses in other genera. The V ORF translation product contained a highly conserved, cysteine-rich domain that is common to most viruses in the Paramyxovirinae subfamily. Sequencing of P gene cDNA clones confirmed the use of a cotranscriptional editing mechanism for the regulation of P/V expression. Based on the location of its origin it has been named Salem virus (SalV).

Comparison of virus isolation and reverse transcription polymerase chain reaction assay for detection of bovine viral diarrhea virus in bulk milk tank samples.

The use of a reverse transcriptase polymerase chain reaction (RT-PCR) assay to screen bulk milk tank samples for bovine viral diarrhea virus (BVDV) has proven to be a sensitive and economical means to evaluate the lactating animals in a herd. The assay is capable of detecting the presence of a single persistently infected animal within a group of several hundred cows. Over a 3-year period, 144 samples from 97 farms were tested for BVDV using an RT-PCR assay in conjunction with a classical virus isolation (VI) procedure to measure the relative effectiveness of the techniques. Virus could be detected with both methods when the milk from a single persistently infected animal was diluted 1:600 with the milk from a herd of BVDV-negative animals. Based on individual farms, there was an overall prevalence of 12.4% BVDV infection, and the correlation between the 2 assays was 95.9%. In terms of sensitivity, specificity, and turnaround time, RT-PCR was superior to VI. However, of the 17 samples that were VI positive, 4 were RT-PCR negative. RT-PCR may not detect all naturally occurring BVDV isolates because they may contain minor sequence variations in the primer regions. VI and RT-PCR are both suitable for detection of BVDV in bulk milk samples when used independently, but to increase the probability of successful detection and to provide cross-checks against assay contamination, it is desirable to utilize both methods in parallel.

Strategies for evaluation of enveloped virus inactivation in red cell concentrates using hypericin.

Photodynamically induced virus inactivation appears promising in preventing transmission of enveloped virus infections in transfusible blood products. The potential for utilizing hypericin as a photosensitizer to inactivate key enveloped viruses in packed red cell concentrates (PRC) was evaluated. In addition to inactivating effectively > or = 10(6) TCID50 of human immunodeficiency virus (HIV), inactivation of bovine viral diarrhea virus (BVDV) in PRC was used as a model for hepatitis C virus to overcome the deficiency in reliable experimental systems for hepatitis C virus (HCV) inactivation. BVDV was two orders of magnitude more sensitive to inactivation by hypericin than HIV. As part of the virucidal efficacy analyses, the effects of photosensitization on hemopoietic cell lines carrying quiescent integrated HIV provirus were studied as models for evaluating virus inactivation in latently infected cells. Phorbol ester-induced virus production by these cells was effectively prevented by photosensitization with hypericin. A refinement of the illumination conditions, incorporating a monochromatic sodium light source with an emission spectrum coinciding with the absorption peak of hypericin, was highly virucidal, however, caused unacceptable levels of hemolysis. Red blood cells could be protected from phototoxic cellular damage by complexing hypericin with human serum albumin (albumin-hypericin), but the decrease in hemolysis was at the expense of virucidal efficacy. Thus, excitation of hypericin with a fluorescent source appears to be useful potentially for virus inactivation in PRC.

Evaluation of a new sandwich ELISA kit that uses serum for detection of cattle persistently infected with BVD virus.

Cattle immunotolerant to and persistently infected (PI) with bovine viral diarrhea (BVD) virus (BVDV) constitute the mechanism by which BVDV persists in and spreads among cattle herds. Detection and elimination of PI cattle are necessary for control of BVD. Serum is an excellent specimen for BVD PI testing because of high survivability of BVDV in serum and ease of collection, storage, and handling. Currently, microtiter virus isolation (VI) employing serum and sandwich ELISAs (S-ELISA) on tissues or leukocytes are used for BVDV PI screening. This paper evaluates a new S-ELISA kit that uses serum as the diagnostic sample. Cattle sera (n = 408) were tested using VI and the S-ELISA. The VI detected 172 BVDV-positive sera. Of these, 18 were confirmed PI cattle. The S-ELISA was positive on all PI samples. Considering only the PI animals, and using VI as the gold standard, the relative sensitivity of S-ELISA was 100%. The overall relative sensitivity was 93.6% and the agreement quotient (kappa) was 0.94. The relative specificity of the kit, based on 236 VI-negative sera, was 100%. These data indicate that the new kit is very adequate for detection of BVDV PI cattle.

Response to modified live and killed multivalent viral vaccine in regularly vaccinated, fresh dairy cows.

Vaccination programs for viral pathogens in the dairy industry span the full spectrum of possibilities even though few of these have been evaluated in field situations. One such program is the vaccination of fresh cows 30 to 60 days postpartum with modified live viral (MLV) vaccines. The purpose of this study was to evaluate the antibody response to booster vaccinations during this period. The impact of vaccinations on milk production and reproductive performance was also examined. The response of cattle boosted with MLV bovine viral diarrhea virus (BVDV) was greatly enhanced compared with the saline controls and the killed vaccine test group. Similar increases were not seen with the MLV infectious bovine rhinotracheitis virus (IBR) and bovine respiratory syncytial virus (BRSV). Changes in milk production were not detected. There was a positive effect on the rate of conception with the MLV group even though there was no evidence of the presence of the three viruses in the herd at the time of study. Although this was a single field trial, and thus limited in scope and repeatability, the results indicate that the vaccines used had a positive effect.

Detection of bovine immunodeficiency virus DNA in the blood and semen of experimentally infected bulls.

Five 18- to 24-month-old bulls were inoculated with either a cell suspension containing bovine immunodeficiency virus (BIV-FL112; 3 bulls) or a BIV-free cell suspension (2 bulls). Blood and semen specimens were collected once a week for 14 weeks, and seroconversion was confirmed by indirect immunofluorescent antibody (IFA) testing. The presence of BIV in blood and semen was determined by virus isolation and/or polymerase chain reaction (PCR) assays. Antibodies to BIV were detected in the 3 experimentally infected bulls as early as day post inoculation (DPI) 17, and levels peaked at DPI 37-58. BIV was isolated from the peripheral blood mononuclear cells (MNCs) of the infected bulls at DPI 9 (2 bulls) and DPI 23 (1 bull), and could be isolated from one animal up to DPI 65. PCR analysis of MNC DNA, using BIV pol gene primers, detected virus in all three of the experimentally infected bulls from DPI 9 until the termination of the experiment at DPI 98. Efforts to isolate a significant number of non-spermatozoal cells (NSC) by gradient separation from the semen of the experimentally infected bulls were unsuccessful. Two methods for the extraction of total NSC DNA from up to 2 ml of non-extended semen were employed; however, no BIV pol fragment was amplified from these DNA preparations. Additionally, 30 bulls from artificial insemination (AI) centers were evaluated for BIV infection by PCR. No amplification products were obtained from MNC DNA from the AI submissions using primer sets for both the BIV pol and env genes.

Experimental infection of calves with bovine viral diarrhea virus genotype II (NY-93).

To ascertain the virulence of bovine viral diarrhea virus (BVDV) genotype II, isolate NY-93 was inoculated intranasally into 3 calves, 2 of which were treated with a synthetic glucocorticoid prior to and after virus inoculation. Anorexia, fever (up to 42 C), dyspnea, and hemorrhagic diarrhea developed 6 days after intranasal inoculation with BVDV NY-93. The condition of all calves deteriorated further until the end of the study on day 14 postinoculation. The most significant postmortem macroscopic changes in all calves were limited to the gastrointestinal tract and consisted of moderate to severe congestion of the mucosa with multifocal hemorrhages. Microscopic lesions found in the gastrointestinal tract were similar to those observed in mucosal disease, including degeneration and necrosis of crypt epithelium and necrosis of lymphoid tissue throughout the ileum, colon, and rectum. The basal stratum of the epithelium of tongue, esophagus, and rumen had scattered individual necrotic cells. Spleen and lymph nodes had lymphocytolysis and severe lymphoid depletion. Severe acute fibrinous bronchopneumonia was present in dexamethasone-treated calves. Abundant viral antigen was detected by immunohistochemistry in the squamous epithelium of tongue, esophagus, and forestomachs. BVDV antigen was prominent in cells of the media of small arteries and endothelial cells. The presence of infectious virus in tissues correlated with an absence of circulating neutralizing antibodies. These findings highlight the potential of BVDV genotype II to cause severe disease in normal and stressed cattle.

Experimental infection of the human granulocytic ehrlichiosis agent in horses.

Human blood collected from two patients from Westchester County, New York with human granulocytic ehrlichia (HGE) infection was inoculated into two ponies. Inoculated ponies developed clinical signs similar to a previous report (Madigan et al., 1995). Histopathological changes involved follicular hyperplasia of lymphoid tissues. HGE DNA was detected by PCR in muscle, fascia, peritoneum, and adrenal gland after the ponies produced a high level of antibodies to HGE. We suggest that HGE may reside in poorly vascularized connective tissues, where the antibodies may have some difficulties to penetrate, resulting in persistent infection. Since HGE and E. equi cause very similar diseases in both humans and horses, they may be the same organism with minor genetic differences.