- Email: [email protected]
Tularaemia in European Brown Hares (Lepus europaeus) and Mountain Hares (Lepus timidus) Characterized by Histopathology and Immunohistochemistry: Organ Lesions and Suggestions of Routes of Infection and Shedding
J. Comp. Path. 2017, Vol. 157, 103e114
Available online at www.sciencedirect.com
ScienceDirect www.elsevier.com/locate/jcpa
DISEASE IN WILDLIFE OR EXOTIC SPECIES
Tularaemia in European Brown Hares (Lepus europaeus) and Mountain Hares (Lepus timidus) Characterized by Histopathology and Immunohistochemistry: Organ Lesions and Suggestions of Routes of Infection and Shedding G. Hestvik*,†, H. Uhlhorn*, F. S€ odersten†, S. Akerstr€ om‡, E. Karlssonx, * E. Westergren and D. Gavier-Wid en*,† *Department of Pathology and Wildlife Diseases, National Veterinary Institute, † Department of Biomedical Sciences and Veterinary Public Health, Swedish University of Agricultural Sciences, ‡ Department of Microbiology, National Veterinary Institute, Uppsala and x Swedish Defence Research Agency, Cementv€agen 20, Ume a, Sweden
Summary Tularaemia is an emerging zoonotic infectious disease caused by the bacterium Francisella tularensis. In Sweden, hares are considered to be key species in the epidemiology of tularaemia. The aim of this study was to characterize the pathology of natural tularaemia infection in European brown hares (EBHs) (Lepus europaeus) and mountain hares (MHs) (Lepus timidus) in Sweden, in order to better understand the presentation of disease and the routes of infection, body dissemination and shedding of F. tularensis. During 2000e2013, 49 EBHs and 37 MHs were diagnosed with tularaemia. Enlargement of the spleen was seen in 80% of EBHs and 62% of MHs. Necrosis was often obvious in the bone marrow, liver, lung and spleen, but 30% of the hares had no lesions or minimal gross lesions. On microscopical examination of tissues from 27 EBHs and three MHs, necrosis was seen in the majority of samples of liver, spleen, bone marrow, lymph node and adrenal glands and was common in the lungs and brain meninges. Immunohistochemistry for Francisella spp. detected bacteria in association with necrosis and inflammation. In several cases, Francisella spp. were also found inside blood vessels, in the renal pelvis, in lactating mammary glands, in bronchioles and in the skin, associated with tick bites. Using quantitative polymerase chain reaction, two genotypes of F. tularensis subsp. holarctica were found; canSNP group B.6, all belonging to subgroup B.7, and canSNP group B.12. There were no differences in pathology between the genotypes. Our results indicate that the urinary tract and mammary glands are important routes for the shedding of F. tularensis. Hunters may not be aware of the risks of contracting tularaemia while handling hares, since infected hares do not always show noticeable gross lesions. Ó 2017 Elsevier Ltd. All rights reserved. Keywords: European brown hare; Francisella tularensis; mountain hare; tularaemia
Introduction Tularaemia is an emerging zoonotic infectious disease caused by the bacterium Francisella tularensis. F. tularCorrespondence to: G. Hestvik (e-mail: [email protected]). 0021-9975/$ - see front matter http://dx.doi.org/10.1016/j.jcpa.2017.06.003
ensis has several subspecies, but only two are considered to cause disease, namely F. tularensis subsp. holarctica, which is found in the Northern hemisphere, and F. tularensis subsp. tularensis, which causes disease in North America only. F. tularensis is a pathogen of great concern because of its extreme infectivity and Ó 2017 Elsevier Ltd. All rights reserved.
104
G. Hestvik et al.
biothreat potential. It is facultatively intracellular and may infect macrophages and various other cell types including hepatocytes, dendritic cells and alveolar type II epithelial cells (Ellis et al., 2002). The bacterium has a wide host range, including mammals, birds and invertebrates. Infection with F. tularensis may be acquired through direct contact with infected tissues and fluids, via mucous membranes, inhalation and ingestion (Bengis et al., 2004). Indirect infection occurs via bites from arthropod vectors, in particular mosquitoes, biting flies and ticks. Little is known about the dissemination of F. tularensis in various organs and tissues after infection, or about the routes of shedding of the bacteria in naturally infected wild animals. This information is fundamental for diagnostic purposes and for better understanding of the epidemiology of tularaemia. In Sweden, tularaemia is a common zoonosis with annual numbers of human cases varying between 14 and 859 over the last 20 years (https:// www.folkhalsomyndigheten.se/folkhalsorapporteringstatistik/statistikdatabaser-och-visualisering/sjukdom sstatistik/harpest/). The disease was first recognized in 1931, when a Mountain hare (MH) (Lepus timidus) sold in a butcher’s shop was presumed to have caused infection in three people that had been in contact with the hare (Granstr€om, 1931). Wildlife hosts are involved in its epidemiology and are a source of infection for man. Tularaemia has historically been restricted to the northern regions of Sweden. In Sweden and Norway, prior to 2002, tularaemia was reported only in MHs and was described as an acute disseminated disease (Morner, 1992; Vikoren et al., 2008). Tularaemia was detected in European brown hares (EBHs) (Lepus europaeus) in Sweden for the first time in € 2002 in Orebro county in central Sweden, and since then cases have been diagnosed in brown hares (EBHs) yearly since 2002. The geographical range for tularaemia has expanded southwards (Tarnvik et al., 2004) affecting EBHs and people. This expansion might be due to a general spread of disease and/or to the involvement of EBHs which, compared with MHs, are more abundant in the southern parts of the country. Both subacute and chronic forms of tularaemia have been described in the EBH in Germany and Hungary (Sterba and Krul, 1985; Gyuranecz et al., 2010). In Sweden, only acute forms have been described in MHs and the pathological presentation in Swedish brown hares (EBHs) has not been studied. The aim of this study was to characterize the pathology of natural tularaemia infection in EBHs in Sweden in order to obtain a better understanding of the routes of infection, body dissemination and shedding of F. tularensis. Furthermore, the lesions were
compared with those seen in the MH. The ultimate goal was to contribute to the understanding of the epidemiological role of the EBH in the southward expansion of tularaemia in Sweden, as the EBH is a key wildlife host involved in the emergence of tularaemia in Sweden.
Materials and Methods Samples
During the period 2000e2013 630 hares (438 EBHs, 183 MHs and nine hares of undetermined species), found dead or killed due to disease, were submitted to the National Veterinary Institute (SVA) in Sweden for post-mortem examination. Out of these, 86 hares, 49 EBHs and 37 MHs, were diagnosed with tularaemia using indirect immunofluorescence (2000e2012) and immunohistochemistry (IHC) or reverse transcriptase polymerase chain reaction (RT-PCR) (2012 onwards). Thirty of the tularaemia-positive hares, 27 EBHs and three MHs, had mild to moderate post-mortem changes and were suitable for histopathological and immunohistochemical examinations. The 49 tularaemia-positive EBHs originated from counties in central Sweden: 18 from Uppsala and the remainder from Dalarna, G€avleborg, J€amtland, Stockholm, S€odermanland, € V€armland, V€astra G€otaland, Orebro and € Osterg€otland. The 37 tularaemia-positive MHs originated from counties in central and northern Sweden, 21 from Dalarna and the rest from G€ avleborg, J€amtland, Norrbotten, Stockholm, V€ armland, V€asterbotten and V€asternorrland. The age of the hares was recorded as adult or leveret. The nutritional state was recorded as normal, poor or cachectic. Polymerase Chain Reaction
For genotyping of F. tularensis subsp. holarctica, qPCR (TaqMan) was used to examine pooled samples of spleen, liver and lung from confirmed tularaemic hares. From the 30 hares chosen for retrospective histopathology, fresh or frozen organ samples were available from 21 EBHs and two MHs. Material from the tissue samples was retrieved using sterile cotton swabs. The swabs were incubated in 380 ml G2 buffer and 20 ml proteinase K solution (EZ1 Tissue DNA Extraction Kit, Qiagen, Sollentuna, Sweden) at 56 C for 15 min under continuous agitation, followed by 5 min incubation at 95 C. DNA was extracted from 200 ml of the resulting lysate using the EZ1 Tissue DNA Extraction Kit and the EZ1 Advanced Instrument (with the Bacteria Card) (Qiagen). The DNA was eluted in 50 ml elution buffer and 1 ml was used as template for each PCR. Francisella tularensis subspp.
Tularaemia in Hares
105
Tissue samples were fixed in 10% neutral buffered formalin, processed routinely and embedded in paraffin wax. For this study, new sections (4 mm) were stained with haematoxylin and eosin (HE). Lesions were scored subjectively as follows. For each section the area affected by necrosis was estimated as: 0e20%, 20e50%, 50e75% or 75e100%. The abundance of each type of inflammatory cell (i.e. lymphocyte, plasma cell, heterophil and macrophage) was graded subjectively as: not observed, few, moderate or numerous.
(RT) with 2% bovine serum albumin (BSA) (SigmaeAldrich, Stockholm, Sweden) for 20 min to block non-specific labelling. The primary antibody was applied at a 1 in 3,500 dilution and the sections were incubated at RT for 45 min. Antibody binding was detected by use of the anti-mouse EnVisionÔ polymer detection system (Dako) and the sections were counterstained with Mayer’s haematoxylin. To assess specificity (i.e. rule out false-positive labelling) a serial section was incubated with 2% BSA instead of the primary antibody. Additionally, for some cases, a serial section was incubated with normal mouse serum and mouse IgG1 (Dako). This was also performed on sections known to be tularaemia negative. To assess sensitivity (i.e. exclude false-negative labelling), known tularaemia-positive control liver or lung from an EBH was included in each run. The location of the labelling was described using the following categories: presence in lesions and/or in healthy tissue, presence in blood vessels, intra- or extracellular location, and in what cell types when located intracellularly. Immunohistochemistry for detection of T lymphocytes (mouse primary antibody CD3 monoclonal; Dako) and B lymphocytes (mouse primary antibody CD79; BioCare Medical, Pacheco, California, USA) was performed on selected cases and organs to identify cell types. Sections were dewaxed and antigen was retrieved in a microwave oven (in Tris EDTA buffer, pH 9.0) at 750 W for 7 min followed by 350 W for14 min. To block endogenous peroxidase activity, sections were incubated with H2O2 in TriseHCl (pH 7.6) for 20 min; thereafter the sections were incubated at RT with 2% BSA for 20 min to block nonspecific labelling. The primary antibody was applied at 1 in 75 (CD3) or 1 in 30 (CD79) dilution and the samples were incubated at RT for 45 min. Antibody binding was detected by use of the anti-mouse EnVisionÔ polymer detection system (Dako) and the sections were counterstained with Mayer’s haematoxylin. A serial section incubated with 2% BSA was used as a negative control.
Immunohistochemistry
Bacterial Culture
To visualize F. tularensis, IHC was performed on formalin-fixed tissues from all histopathologically examined organs using a mouse primary monoclonal antibody FB11 (Meridian Life Science Inc., Nordic Biosite AB, T€ aby, Sweden) directed against F. tularensis sp. lipopolysaccharide antigen. Sections were dewaxed and incubated with H2O2 in TriseHCl (pH 7.6) for 15 min to block endogenous peroxidase activity. Antigen was retrieved by proteinase K (Dako, Agilent, Glostrup, Denmark) treatment for 6 min; thereafter the sections were incubated at room temperature
Aerobic bacterial culture was performed on selected organ samples (seven lungs, two kidneys, nine livers, three spleens, one uterus and two testicles) from 13 hares suspected to be co-infected with additional bacteria. A few samples were submitted for culture in conjunction with the necropsy examination, while the remaining had been stored at 20 C for up to 7 years. The samples were transferred to horse blood agar (in house) and a purple agar plate (in house). Plastic loops were changed between primary, secondary and tertiary streaks. The plates were incubated in
holarctica genotyping was performed on 16 hares from fresh frozen organ samples and on seven hares from freeze-dried organ samples using two indel markers (Ftind49 and Ftind38) and nine canSNP markers (B.13, B.19, B. 23, B.26, B.39, B.40, B.41, B.42 and B.43) for typing of the three major canSNP-groups, B.4, B.6 and B.12, and the subgroups B.7, B.10, B.20, B.23 and B.39 (Svensson et al., 2009; Karlsson et al., 2013) according to the qPCR-based method described previously (Karlsson et al., 2013). Whole genome sequenced F. tularensis subsp. holarctica strains with known genotypes were used as controls, together with no template controls. Necropsy Examination
In hares examined between 2000 and 2009, necropsy findings in the heart, lung, liver, spleen, kidney and bone marrow were recorded routinely. During the period 2010e2013 additional organs examined routinely were brain, heart, airways, intestine, adrenal gland, ovary, testicle, uterus, urinary bladder, intestine and lymph nodes. The skin was scanned for the presence of ectoparasites (i.e. ticks). Descriptions of lesions were retrieved retrospectively from necropsy records and included organ enlargement, hyperaemia, oedema and necrosis/degeneration/inflammation. The lesions were graded as mild, moderate or severe. Histopathology
106
G. Hestvik et al.
an aerobic atmosphere at 37 C and analyzed after 1 and 2 days.
Results Polymerase Chain Reaction
The 21 EBHs and the two MHs examined by qPCR were all positive for F. tularensis subspp. holarctica. Seven samples were genotyped as canSNP group B.6, all of which belonged to subgroup B.7. Sixteen samples were genotyped as canSNP group B.12. These were further divided into subgroups B.20 (n ¼ 4), B.23 (n ¼ 11) and B.39 (n ¼ 1). The genotypes found are consistent with the currently known diversity in Sweden (Karlsson et al., 2013). Necropsy Examination
Gross pathology data was retrieved for the 86 hares subjected to necropsy examination and diagnosed with tularaemia during the period 2000e2013; 49 EBHs and 37 MHs. Table 1 shows the nutritional status in the 74 animals found dead or killed due to disease. The remaining 12, seven EBHs and five MHs died from trauma. Of these latter, three EBHs were in good and three were in poor nutritional status, and one was cachectic. Of the five MHs, four were in good and one was in poor nutritional status. Since tularaemia was not the direct cause of death, these 12 animals were not included in Table 1. The results show that a larger proportion of the EBHs (67%) than the MHs (41%) were cachectic and that a higher percentage of the MHs were in poor or good nutritional state compared with the EBHs. To investigate whether this was a significant difference, indicating that MHs were in a better nutritional state when they died, indicating a shorter course of disease, hares with cachexia and poor nutritional status were grouped together and were compared with hares having a good nutritional status. The association was not significant using a two-sided Fisher’s exact test. Enlargement of, and grossly visible necrosis in, a selection of organs from the 86 tularaemic hares is sumTable 1 Comparison of nutritional status in tularaemic EBHs and MHs found dead or killed due to disease 2000e2013
marized in Table 2. The most common finding, enlargement of the spleen, was seen in 80% of the EBHs and 30% of the MHs. Necrosis was visible as white, pinpoint to a few millimetre diameter foci, often spread diffusely throughout the organ. In bone marrow and spleen, necrotic foci were observed in a larger proportion of the MHs (41% and 62%, respectively) compared with the EBHs (12% and 14%, respectively). In the MHs, the liver was the only additional organ with grossly visible necrosis. In the EBHs, necrosis was also found in the lungs and kidney. The Fisher’s exact test was used to determine if there was a significant difference between EBHs and MHs with respect to the number of individuals having grossly visible necrosis. The difference was significant with P <0.005 for bone marrow and P <0.02 for lung, using a two-sided test. This indicates that grossly detectable necrosis in the bone marrow was more frequent in MHs, while lung necrosis was more common in EBHs. In general, gross lesions were present in a higher number of organs in each individual for the EBHs than in the MHs. In the liver and spleen the differences were not significant. Additional changes observed inconsistently were hyperaemia and haemorrhages, necrosis in lymph nodes and oedema of the lungs. It is noteworthy that five EBHs and nine MHs lacked grossly detectable changes. Moreover, five EBHs and seven MHs showed mild gross lesions, such as slight organ enlargement without visible necrosis, necrosis in the bone marrow and/or nonspecific findings such as oedema and hyperaemia in any organ. In total, 26 of the 86 tularaemic hares (30%) had either no lesions or minimal gross lesions. Nine of these hares had died of trauma. Histopathology and Immunohistochemistry
The histological lesions in the MHs and the EBHs were similar and are described below. Necrosis was a consistent finding, and was associated variably with inflammation. Necrosis was seen in all but one Table 2 Frequently observed gross lesions (necrosis and enlargement in any organ) in tularaemic EBHs and MHs Organ
Cachectic: number Poor nutritional Good nutritional Total number of animals status: number status: number of animals (% per species) of animals of animals (% per species) (% per species) EBH MH Total *
28 (67) 13 (41) 41
10 (24) 13 (41) 23
4 (9) 6 (18) 10
Twelve animals dead from trauma were not included.
42 32 74*
Bone marrow Liver Lung Spleen
EBH (n ¼ 49)
MH (n ¼ 37)
Necrosis: number of cases (%)
Enlargement: number of cases (%)
Necrosis: number of cases (%)
Enlargement: number of cases (%)
6 (12) 9 (18) 7 (14) 7 (14)
NA 3 (6) 0 39 (80)
15 (41) 4 (11) 0 11 (30)
NA 3 (8) 1 (3) 23 (62)
n, total number of examined cases; NA, not applicable.
Tularaemia in Hares
of the examined livers, in a large proportion of the spleens, bone marrow, lymph nodes and adrenal glands, and was common in the lungs and brain meninges (Table 3). Regardless of organ, the overall consistent presentation was multifocal, sometimes coalescing, variably-sized foci of necrosis distributed throughout the organ. Necrosis was multifocal, sometimes coalescing, with variable amounts of karyorrhectic, pyknotic and/or lytic nuclei, and the foci varied in size between 10 and 400 mm. Inflammatory cells associated with the foci necrosis were primarily macrophages and T cells, with occasionally a few plasma cells and heterophils. In the majority of cases this was consistent with an acute or subacute stage of inflammation (Fig. 1). In the liver, the sinusoids contained diffuse and mild to moderate numbers of CD3+ T lymphocytes (Fig. 2). In the spleen, foci of necrosis were located in the red pulp or at the border between red and white pulp (Fig. 1). Subjectively assessed and compared, the red pulp consistently had a lower cell density in most cases compared with the spleen from healthy hares; this change represented mild diffuse necrosis. Inflammation associated with the necrosis was not observed, but a mild to moderate infiltration of inflammatory cells was seen multifocally in the capsule and trabeculae. In five of the spleens there was a mild
Table 3 Number of organs with histological lesions in EBHs and MHs with tularaemia Organ
Adrenal gland Bone marrow Brain Heart Intestine Kidney Liver Lung: conducting airways Lung parenchyma Lymph nodes Mammary gland Ovary Spleen Testis Urinary bladder Uterus
EBH (n ¼ 27) Recorded lesions: *number of organs with lesions/number examined† (%)
MH (n ¼ 3) Recorded lesions: *number of organs with lesions/number examined† (%)
8/11 (72) 19/26 (73) 7/21 (33) 2/19 (11) 2/17 (12) 3/26 (12) 26/27 (96) 4/26 (15) 13/26 (50) 12/13 (92) 2/2 (100) 2/8 (25) 20/26 (77) 1/6 (17) 1/3 (33) 2/5 (40)
1/3 (33) 3/3 (100) 0/2 (0) 0/2 (0) 0/3 (0) 0/2 (0) 3/3 (100) 0/3 (0) 0/3 (0) 2/3 (67) 1/1 (100) 0/1 (0) 2/2 (100) 0/0 () 0/0 () 0/1 (0)
n, total number of hares included in the histopathological investigation. * Types of lesions recorded were necrosis and inflammation. † Number of animals in which respective organ was examined.
107
to moderate increase of inflammatory cells in the red pulp, but necrosis was not observed. The lymph nodes had foci of necrosis in the cortex and/or medulla. Inflammation was not associated with the necrotic areas, but occasionally mononuclear cells were seen in subcapsular and medullary sinuses. The bone marrow had no inflammatory cells associated with the foci of necrosis or in other parts of the sections (Fig. 1). The most common lesion in the lung was multifocal necrosis of the alveolar parenchyma, associated with mild to moderate inflammation (Fig. 1). However, in five EBHs the lung lesions were more severe and chronic. The necrotic foci often coalesced to form large areas, destroying bronchi and bronchioles. The inflammation was moderate to marked, with a mixed population of macrophages, lymphocytes, plasma cells and heterophils. Several hares had mild to marked fibrosis of the lungs (Fig. 3). The trachea was investigated in two of these hares, and in one case there were moderate multifocal areas of necrosis in the lamina propria and submucosa with an associated moderate mixed inflammation. The overlying epithelium showed multifocal erosion and ulceration. Lesions in the leptomeninges were found in seven of 21 brains examined and consisted of necrosis with mild to moderate associated inflammation; occasionally the lesions extended into the adjacent brain tissue (Fig. 4). Three hares had moderate to severe necrosis and/or exudative inflammation of the renal pelvis (Fig. 3). One also had moderate necrosis and mononuclear cell infiltrates in the cortex and the medulla. Most of the affected adrenal glands had small areas of necrosis with associated mild inflammation. In two hares, multifocal epicardial necrosis was evident. The urinary bladder, intestine, ovaries, uteri and testes were rarely examined. Lesions in these organs were few and small, with mild or no associated inflammation, with the exception of suppurative inflammation in one uterus. The mammary gland was examined only in two EBHs and one MH; all were lactating. The MHs had a few foci of necrosis and inflammation in the interalveolar connective tissue. In the two EBHs, necrosis was widespread and inflammation minimal (Fig. 5). In the skin of two EBHs, focal severe necrosis, inflammation and haemorrhage in the dermis, subcutis and cutaneous muscles were associated with still attached ticks (Fig. 6). The pattern of immunolabelling of Francisella spp. was similar in different organs. Immunolabelled bacteria were found both extra- and intracellularly in association with necrosis and inflammation; the most common intracellular location was in macrophages (Fig. 2). In addition, Francisella spp. were found intracellularly in macrophages and T cells in the blood vessels in 13 cases (Fig. 2). To a lesser degree,
108
G. Hestvik et al.
Fig. 1. Acute changes showing necrotic foci with intralesional F. tularensis spp. in the spleen (A, B), bone marrow (C, D) and lung (E, F) on serial sections. HE and IHC. Bars (AeF), 100 mm.
intracellularly-located Francisella spp. were seen in pneumocytes, alveolar macrophages, endothelial cells and megakaryocytes. The presence of Francisella spp. in some locations is worth mentioning because of the relevance for transmission. Bacteria were seen in exudate in the lumen of bronchioles of one hare (Fig. 3) and intracellularly in macrophages and lining epithelial cells in bronchioles in two other hares. In the three kidneys with inflammation in the pelvis, bacteria were detected in the exudate (Fig. 3). In
the three examined mammary glands, Francisella spp. were seen in the milk secretions and in macrophages and epithelial cells (Fig. 5). In the skin of hares with attached ticks, large numbers of Francisella spp. were detected locally in the lesion, both intra- and extracellularly (Fig. 6). Bacterial Culture
In 13 hares with severe inflammation or involvement of organs less commonly presenting with signs of
Tularaemia in Hares
109
Fig. 2. Liver sinusoids containing small cells on HE staining (A), shown by IHC for CD3 to be T-lymphocytes (B). IHC for F. tularensis spp. shows that many of the T-lymphocytes contain cytoplasmic bacteria (C). In addition, F. tularensis spp. can be visualized intraluminally in the central vein (C) and intracellularly in Kupffer cells (D). Bars (AeD), 100 mm.
Fig. 3. Subacute/chronic lesions in the lung and renal pelvis. HE (A, C) and IHC for F. tularensis spp. (B, D) on serial sections. (A, B) Necrosis with severe inflammation destroying the tissue, and containing large numbers of F. tularensis spp. A bronchiole (*) contains exudate with numerous bacteria. (C, D) Lumen of the renal pelvis is filled with an exudate containing numerous heterophils and bacteria. Bars (AeD), 100 mm.
110
G. Hestvik et al.
Fig. 4. Necrosis and inflammation in the leptomeninges of the brain and arterial cuffing, HE (A). IHC shows numerous F. tularensis spp. in the lesion (B). Bars (A, B), 100 mm.
Fig. 5. Mammary gland from a lactating female. Multifocal areas of necrosis (N) interspersed with milk-containing alveoli (M). F. tularensis spp. are present in both necrotic areas and in the milk (B). HE (A) and IHC F. tularensis spp. (B). Bars (A, B), 100 mm.
inflammation and/or necrosis (e.g. kidney, uterus and testis), a dual infection was considered. Aerobic bacterial culture was performed on a variety of organs (seven lungs, two kidneys, nine livers, three spleens, one uterus and two testes) in the hares from which archived tissues were available. The significance of the bacteria cultured from 11 of the hares was difficult to interpret, including a variable presence of Escherichia coli, Enterococcus spp. and a mixed flora. It was presumed that their presence was probably caused by
post-mortem overgrowth from the intestine or contamination of the sample surfaces during handling at necropsy examination and sampling. Klebsiella pneumoniae was cultured in moderate to large amounts from a hare with severe inflammation in the lungs. In another hare, small numbers of E. coli grew in pure culture in samples from the kidney and uterus. In both of these hares the inflammation was more severe compared with other hares. Both of these hares were considered to have true dual infections.
Fig. 6. Serial section of skin at the site of a tick bite, HE (A) and IHC for F. tularensis spp. (B). (A) Necrosis and inflammation (*) with an underlying haemorrhage. F. tularensis spp. are found in large numbers at the tick bite site (B), suggesting that the tick transmitted the bacteria to the hare. Bars (A, B), 200 mm.
Tularaemia in Hares
Discussion The habitat of MHs is woodland and these animals are spread all over Sweden. The EBH was introduced into southern Sweden during the mid-19th century, and has, during the last decades, expanded north to inhabit the southern half of the country, and to woodlands, in addition to agricultural areas. The two hare species cohabit in many areas and cross-species hybridization is known to occur. In areas where both species are found, the number of MHs has decreased, probably due to less favourable winters for this species, with less snow and higher temperatures (Jansson, 2013). The first cases of tularaemia in Sweden were diagnosed in man and one mountain hare in 1931 and since then tularaemia has been considered to be endemic in the northern regions of Sweden. During the 1970se1980s, MHs that died from tularaemia were found in the northern two thirds of the country (from county Norrbotten in the north to county V€ armland in the central part of the country) and on the island of Stora Karls€o in the southern part of the Baltic Sea (Morner et al., 1988). The MH was the only hare species known to be affected until 2002 when the first case of tularaemia in the EBH was diagnosed. During the past decades, tularaemia has spread south in man and hares, but the reason is not known (Tarnvik et al., 2004). A reason why tularaemia emerged in EBHs in Sweden might be the southward expansion of tularaemia in combination with the northward expansion of populations of EBHs, causing their exposure. Whether EBHs might have a role in spreading the disease further south is not known. Since EBHs had not been tested routinely for tularaemia before 2002, a retrospective study was conducted. Between 1987 and 1999, 909 EBHs were subjected to necropsy examination. Out of these, five had histopathological lesions in the liver or spleen compatible with tularaemia. Archived formalin-fixed and paraffin wax-embedded samples from these five EBHs were tested by IHC for F. tularensis spp. and all were negative. Since the first case in 2002, the disease has been diagnosed regularly in EBHs. Since MHs and EBHs may cohabit, there is a possibility for infection to pass between species by direct contact, as well as by indirect contact through environmental contamination (e.g. water and grass), and possibly also through vectors. During the last decade, tularaemia cases have been diagnosed as far south as the county of J€ onk€ oping, approximately 350 km from the southern country border (SVA post-mortem records). It has been suggested that EBHs are somewhat less susceptible to developing severe tularaemic disease
111
caused by F. tularensis subsp. holarctica than MHs (Gyuranecz, 2012). This is based on findings of subacute/chronic lesions with granulomatous inflammation in EBHs in Hungary and Germany (Sterba and Krul, 1985; Gyuranecz et al., 2010) compared with the acute disseminated lesions in MHs described in northern Europe (Morner et al., 1988; Vikoren et al., 2008). However, in later years there have been descriptions of acutely diseased EBHs in Europe outside Scandinavia, for example in France (Decors et al., 2011). One explanation for the differences could be that the majority of hares with subacute/chronic disease were hunted (i.e. were apparently healthy), while those with acute disease mostly were found dead. In our study, all EBHs and MHs were found dead or were killed due to disease and therefore were suitable for comparison of whether tularaemia at the terminal stage of disease has a more acute course in MHs than in EBHs. Comparison of the nutritional state and inflammatory response could add valuable information, since animals with shorter disease processes are more likely to have a better nutritional status and less inflammatory response. However, comparison of the nutritional status in the EBHs and MHs during the period 2000e2013 showed no significant difference between the two species. Comparison of grossly visible foci of necrosis in the two hare species resulted in a significant difference for only two organs. Necrosis was more frequent in the bone marrow of MHs compared with EBHs and in the lungs of EBHs compared with MHs. Histopathological changes in 21 EBHs and three MHs were consistent with acute disseminated disease. This result is in accordance with the pathology in EBHs found dead in France (Decors et al., 2011) and in MHs in Sweden and Norway (Morner et al., 1988; Vikoren et al., 2008; Josefsen et al., 2012). It was not possible to assess differences in the number of microscopically visible foci of necrosis and in the inflammatory response between EBHs and MHs in our study, since only three MHs were fresh enough to allow good assessment on histopathology. The results of our study, in accordance with additional results from previous studies of MHs, show that tularaemia can have an acute presentation in both hare species, showing necrosis with absent or minimal inflammatory reaction (Morner et al., 1988; Vikoren et al., 2008; Josefsen et al., 2012). It is well known that EBHs develop chronic disease (Sterba and Krul, 1985; Gyuranecz et al., 2010). Information regarding chronic presentations of tularaemia in MHs is lacking, possibly due to absence of investigation of hunted MHs. Six of the EBHs had subacute/chronic changes in the lung and/or kidney and an inflammatory response in the
112
G. Hestvik et al.
spleen in addition to an acute disseminated disease. These may represent chronic infections that were reactivated to give acute disease. The lesions of some investigated organs are worth particular mention. Lesions in the leptomeninges were found in one-third of the examined brains and consisted of necrosis with mild to moderate inflammation; occasionally the lesions extended into the adjacent brain tissue. The only report examining the brain of a hare infected with F. tularensis subsp. holarctica is from a Japanese hare (Lepus brachyurus augustidens), in which encephalitis was recorded (Park et al., 2009). Our findings suggest that meningitis and/or encephalitis is more common than previously appreciated in tularaemia. The mammary gland is another organ not commonly investigated. In our study, the mammary glands of only two EBHs and one MH were examined; all were lactating. All three had necrosis with minimal inflammation in the mammary glands. Mammary lesions were also found in one Hungarian EBH, described as granulomatous inflammation (Gyuranecz et al., 2010). Because all our examined hares had acute and disseminated disease, it is not unexpected that lesions and bacteria were found in many organs that have rarely been investigated in earlier studies, such as brain, mammary gland, testis, ovary and adrenal gland. We postulate that the dissemination in the body was through the haematogenous route, since infection affected multiple organs and lesions of the same type and duration were diffusely distributed throughout each organ. This was also confirmed by IHC, showing F. tularensis spp. within inflammatory cells in the blood vessels in 13 hares. Finding that six EBHs had a partly different disease presentation, with more severe and/or exudative inflammation, we hypothesized that a dual infection with some other bacteria might be the cause. These hares had severe and subacute/chronic lesions in the lung and a more severe inflammatory response in the spleen. In addition, three of them had suppurative inflammation in the renal pelvis. Despite this, the lesions in other affected organs had the same presentation as in the other hares. Most of the cultures showed growth of E. coli and Enterococcus spp., which could have been caused by contamination during handling. Cultures from two hares were assessed to be indicative of true dual infection. Klebsiella pneumoniae was cultured from one lung and E. coli from another hare’s kidney and uterus. Dual infection with K. pneumoniae and F. tularensis spp. has been described in a human case (Givham, 1965). Differences in disease presentation in EBHs as described above could also be due to differences in the pathogenicity of F. tularensis subsp. holarctica geno-
type. Of the hares with subacute/chronic lesions in the lung and/or kidney, F. tularensis subsp. holarctica was genotyped from four. Two of these belonged to canSNP group B.12 (subgroups B.20 and B.23, respectively) and the other two belonged to canSNP group B.6 (subgroup B.7). Therefore, the differences found in genotype could not be associated with different disease presentations. Genotyping of F. tularensis subsp. holarctica in tularaemic EBHs in Switzerland resulted in the finding of two canSNP groups, B.6 (subgroup B.10) and B.12 (subgroup B.13) (Origgi and Pilo, 2016). The authors concluded that canSNP group B.6 (subgroup B.10) was associated with splenitis and hepatitis, while canSNP group B.12 (subgroup B.13) was associated with lesions in the pleura, pericardium and kidney, as described in Hungary (Gyuranecz et al., 2010; Origgi and Pilo, 2016). The majority of hares in our study had lesions in the spleen, liver and many other organs, in both canSNP groups. In Sweden, mosquito bites are considered to be the most frequent means of tularaemia infection in man. Another, less frequent route, is entrance through skin erosions or ulcers by direct contact with body secretions from an infected animal. After local multiplication at the infection site, the bacteria are disseminated through the blood and lymphatic vessels to other organs (Gyuranecz, 2012). Mosquito bites are probably also an important route of infection for animals, and a focal skin infection can easily be overlooked at necropsy examination. Various species of ticks are vectors of F. tularensis holarctica in Europe (Hestvik et al., 2014). During the past decades, tularaemia has expanded southwards and ticks northwards in Sweden, therefore ticks may be an increasingly important source of infection. The only clearly proven entry route in our study was through the skin. In two EBHs, severe focal necrosis and inflammation were seen in the skin where ticks were still attached. The ticks were, based on their gross morphology, identified as Ixodes ricinus, the most common tick species in Sweden (http://www.nrm.se/ faktaomnaturenochrymden/djur/insekterochspinde ldjur/spindeldjur/fastingar.12595.html). By IHC, large numbers of Francisella spp. were visualized in the skin lesions, indicating that the tick bite was the site of entry and multiplication. Similar skin lesions associated with a tick bite have been described in a Japanese hare (Lepus brachyurus angustidens), where F. tularensis spp. also were detected by IHC in the intestine of the tick (Park et al., 2009). Infection through inhalation has also been proposed in animals (Gyuranecz et al., 2010; Decors et al., 2011), but this could not be confirmed in our study. The presence of bacteria in bronchiolar epithelial cells and
113
Tularaemia in Hares
macrophages in two hares might indicate signs of entry, but could also be signs of shedding or just the location of the infection. As in other affected organs, the most common lung lesion was multifocal necrosis in the alveolar parenchyma without involvement of conducting airways. This suggests that the infection reaches the lungs through a haematogenous route rather than a respiratory route. However, in one hare, Francisella spp. was also evident in the bronchiolar exudate, indicating that shedding by the respiratory route may occur. The present study has indicated several other possible shedding routes. Francisella spp. was detected in the exudate in the renal pelvis of three hares and in milk in lactating mammary glands of three hares, showing that these are two other ways of possible bacterial shedding. In the intestine and uterus, F. tularensis was detected in superficial mucosal lesions although not in luminal secretions, which at least in theory makes these organs possible routes of shedding. Assuming that the cause of the severe inflammation in the lung, kidney and uterus represented dual bacterial infection, together with the fact that F. tularensis spp. in these cases were only found in exudate, makes it likely that these routes of shedding are not frequent in acutely diseased hares. On the other hand, if these hares represent cases of chronic tularaemia without dual infections, these shedding routes could be important. The majority of hares investigated were found dead or were killed due to disease, so this study describes mostly terminal disease in lethal forms of tularaemia. Little information is available about other presentations of tularaemia; for example, chronic forms or earlier stages of infection in hares in Sweden, which would need to be investigated in apparently healthy hares, for example animals killed by hunting. Twenty-six of the 86 investigated tularaemic hares (30%) had no gross lesions or minimal lesions that could easily be missed. Nine of these 26 hares died of trauma, and the lack of gross lesions in those might be due to early infection at the time of death. Lack of visible gross lesions increases the risk of infection for people who handle wildlife. There are examples of hunters contracting disease after dressing hunted animals with unprotected hands (Munnich and Lakatos, 1979; Wahab and Bjurgren Zelano, 2012; van de Wetering et al., 2015). Hunting dogs are also at risk of infection if they bite the hare or are fed the hare’s internal organs (Nordstoga et al., 2014). Sources of infection for people, predators and scavengers are carcasses of tularaemic animals and water contaminated by diseased animals’ excreta or dead infected animals. Predators and scavengers such as red foxes, wild boars, raccoon dogs, wolves, lynx,
brown bears and certain bird species are considered relatively resistant to tularaemic disease. They may, however, contract the infection and infected hares and other species are a source of infection through predation and scavenging. This has been shown for red foxes, wild boars and raccoon dogs in several studies by the presence of antibodies towards F. tularensis (Hoflechner-Poltl et al., 2000; Hubalek et al., 2002; Al Dahouk et al., 2005; Kuehn et al., 2013) and detection of the bacteria in submandibular lymph nodes of red foxes (Hofer et al., 2010). It is not clear if these animal species can act as reservoirs and contribute to the spread of bacteria. In conclusion, in Sweden over the last 15 years, tularaemia has been expanding its range further south and has been found in a new hare species (EBHs). The present pathological investigation suggests that EBHs and MHs, found dead or killed due to disease, have similar lesions and die of an acute disseminated disease (sepsis). The chronic disease presentation seen in a few hares could reflect a pure chronic disease that has been reactivated and/or a dual infection with an additional bacterium. No differences in pathological presentation between F. tularensis subsp. holarctica genotypes canSNP group B.6 and B.12 were evident. Infection through bite wounds from ticks was seen. Routes of shedding that might be of importance are via the urinary tract, milk, uterine secretions and intestinal content, but since most hares do not show changes in these organs the most important shedding route is probably from dead animals. Animals and people might be infected by exposure to water contaminated by dead tularaemic hares or by ingestion of infected hares. Since tularaemia is not always easily detectable on gross inspection, there is a risk for people handling hares and dogs fed uncooked material from infected hares.
Acknowledgments This work was supported in part by the European Union Seventh Framework Programme (FP7/ 2007e2013), Wild Tech [grant agreement no. 22633].
Conflict of Interest Statement The authors declare that they have no conflicts of interest with respect to their authorship or publication of this article.
References Al Dahouk S, Nockler K, Tomaso H, Splettstoesser WD, Jungersen G et al. (2005) Seroprevalence of brucellosis, tularemia, and yersiniosis in wild boars (Sus scrofa)
114
G. Hestvik et al.
from North-Eastern Germany. Journal of Veterinary Medicine, Series B: Infectious Diseases and Veterinary Public Health, 52, 444e455. Bengis RH, Leighton FA, Fischer JR, Artois M, Morner T et al. (2004) The role of wildlife in emerging and reemerging zoonoses. Revue Scientifique et Technique De L’Office International des Epizooties, 23, 497e511. Decors A, Lesage C, Jourdain E, Giraud P, Houbron P et al. (2011) Outbreak of tularaemia in brown hares (Lepus europaeus) in France, January to March 2011. Eurosurveillance, 16, 3e5. Ellis J, Oyston PCF, Green M, Titball RW (2002) Tularemia. Clinical Microbiology Reviews, 15, 631e646. Givham E (1965) Tularemic pneumonia with Klebsiella superinfection. Journal of the Tennessee Medical Association, 58, 224e226. Granstr€om KO (1931) Tre fall av tularemi (three cases of tularemia). Svenska L€akartidningen, 28, 641e646. Gyuranecz M (2012) Tularaemia. In: Infectious Diseases of Wild Mammals and Birds in Europe, D Gavier-Widen, JP Duff, A Meredith, Eds., Blackwell Publishing Ltd., Chichester, pp. 303e309. Gyuranecz M, Szeredi L, Makrai L, Fodor L, Meszaros A et al. (2010) Tularemia of European brown hare (Lepus europaeus): a pathological, histopathological, and immunohistochemical study. Veterinary Pathology, 47, 958e963. Hestvik G, Warns-Petit E, Smith LA, Fox NJ, Uhlhorn H et al. (2014) The status of tularemia in Europe in a onehealth context: a review. Epidemiology and Infection, 43, 2137e2160. Hofer E, Reisp K, Revilla-Fernandez S, Plicka H, Romanek G et al. (2010) Isolation of Francisella tularensis and Brucella suis from red foxes (Vulpes vulpes). Tieraerztliche Umschau, 65, 229e232. Hoflechner-Poltl A, Hofer E, Awad-Masalmeh M, Muller M, Steineck T (2000) Prevalence of tularaemia and brucellosis in European brown hares (Lepus europaeus) and red foxes (Vulpes vulpes) in Austria. Tierarztliche Umschau, 55, 264e268. Hubalek Z, Treml F, Juricova Z, Hunady M, Halouzka J et al. (2002) Serological survey of the wild boar (Sus scrofa) for tularaemia and brucellosis in South Moravia, Czech Republic. Veterinarni Medicina, 47, 60e66. Jansson G (2013) Mountain hare and brown hare e where and why they hybridize in Sweden? Fauna och Flora (Stockholm), 108, 22e27. Josefsen TD, Mork T, Ytrehus B, Hetta MH, Henrichsen EN et al. (2012) Tularaemia in Finnmark (Tularemi i Finnmark). Norsk Veterinartidsskrift, 124, 173e174. Karlsson E, Svensson K, Lindgren P, Bystrom M, Sjodin A et al. (2013) The phylogeographic pattern of Francisella tularensis in Sweden indicates a Scandinavian origin of Eurosiberian tularaemia. Environmental Microbiology, 15, 634e645. Kuehn A, Schulze C, Kutzer P, Probst C, Hlinak A et al. (2013) Tularaemia seroprevalence of captured and
wild animals in Germany: the fox (Vulpes vulpes) as a biological indicator. Epidemiology and Infection, 141, 833e840. Morner T (1992) The ecology of tularaemia. Revue Scientifique et Technique de L’Office International des Epizooties, 11, 1123e1130. Morner T, Sandstrom G, Mattsson R, Nilsson PO (1988) Infections with Francisella tularensis biovar palaearctica in hares (Lepus timidus, Lepus europaeus) from Sweden. Journal of Wildlife Diseases, 24, 422e433. Munnich D, Lakatos M (1979) Clinical, epidemiological and therapeutic experience with human tularemia e role of hamster hunters. Infection, 7, 61e63. Nordstoga A, Handeland K, Johansen TB, Iversen L, Gavier-Widen D et al. (2014) Tularaemia in Norwegian dogs. Veterinary Microbiology, 173, 318e322. Origgi FC, Pilo P (2016) Francisella tularensis clades B.FTN002-00 and B.13 are associated with distinct pathology in the European brown hare (Lepus europaeus). Veterinary Pathology, 53, 1220e1232. Park C-H, Nakanishi A, Hatai H, Kojima D, Oyamada T et al. (2009) Pathological and microbiological studies of Japanese hare (Lepus brachyurus angustidens) naturally infected with Francisella tularensis subsp. holarctica. Journal of Veterinary Medical Science, 71, 1629e1635. Sterba F, Krul J (1985) Gross and histological studies on the differential diagnosis of brucellosis, tularemia and pseudotuberculosis in hares (Lepus europaeus). In: Transactions of the XVIIth Congress of the International Union of Game Biologists, Brussels, September 17e21, 1985. Part 2, pp. 763e771. Svensson K, Granberg M, Karlsson L, Neubauerova V, Forsman M et al. (2009) A real-time PCR array for hierarchical identification of Francisella isolates. PLoS One, 4, 12. Tarnvik A, Priebe HS, Grunow R (2004) Tularaemia in Europe: an epidemiological overview. Scandinavian Journal of Infectious Diseases, 36, 350e355. Wahab T, Bjurgren Zelano S (2012) J€ agare smittad av harpest vid slakt av hare (Hunter infected by tularemia while dressing a hare). Lakartidningen, 109, 152e153. van de Wetering D, dos Santos CO, Wagelaar M, de Kleuver M, Koene MGJ et al. (2015) A cluster of tularaemia after contact with a dead hare in the Netherlands. Netherlands Journal of Medicine, 73, 481e482. Vikoren T, Handeland K, Djonne B (2008) Fleire tilfelle av tularemi p avist hj a hare i S€ or-Noreg (Several cases of tularemia diagnosed in hares in southern Norway). Norsk Veterinaertidskrift, 120, 583.
March 1st, 2017 ½ Received, Accepted, June 9th, 2017
Available online at www.sciencedirect.com
ScienceDirect www.elsevier.com/locate/jcpa
DISEASE IN WILDLIFE OR EXOTIC SPECIES
Tularaemia in European Brown Hares (Lepus europaeus) and Mountain Hares (Lepus timidus) Characterized by Histopathology and Immunohistochemistry: Organ Lesions and Suggestions of Routes of Infection and Shedding G. Hestvik*,†, H. Uhlhorn*, F. S€ odersten†, S. Akerstr€ om‡, E. Karlssonx, * E. Westergren and D. Gavier-Wid en*,† *Department of Pathology and Wildlife Diseases, National Veterinary Institute, † Department of Biomedical Sciences and Veterinary Public Health, Swedish University of Agricultural Sciences, ‡ Department of Microbiology, National Veterinary Institute, Uppsala and x Swedish Defence Research Agency, Cementv€agen 20, Ume a, Sweden
Summary Tularaemia is an emerging zoonotic infectious disease caused by the bacterium Francisella tularensis. In Sweden, hares are considered to be key species in the epidemiology of tularaemia. The aim of this study was to characterize the pathology of natural tularaemia infection in European brown hares (EBHs) (Lepus europaeus) and mountain hares (MHs) (Lepus timidus) in Sweden, in order to better understand the presentation of disease and the routes of infection, body dissemination and shedding of F. tularensis. During 2000e2013, 49 EBHs and 37 MHs were diagnosed with tularaemia. Enlargement of the spleen was seen in 80% of EBHs and 62% of MHs. Necrosis was often obvious in the bone marrow, liver, lung and spleen, but 30% of the hares had no lesions or minimal gross lesions. On microscopical examination of tissues from 27 EBHs and three MHs, necrosis was seen in the majority of samples of liver, spleen, bone marrow, lymph node and adrenal glands and was common in the lungs and brain meninges. Immunohistochemistry for Francisella spp. detected bacteria in association with necrosis and inflammation. In several cases, Francisella spp. were also found inside blood vessels, in the renal pelvis, in lactating mammary glands, in bronchioles and in the skin, associated with tick bites. Using quantitative polymerase chain reaction, two genotypes of F. tularensis subsp. holarctica were found; canSNP group B.6, all belonging to subgroup B.7, and canSNP group B.12. There were no differences in pathology between the genotypes. Our results indicate that the urinary tract and mammary glands are important routes for the shedding of F. tularensis. Hunters may not be aware of the risks of contracting tularaemia while handling hares, since infected hares do not always show noticeable gross lesions. Ó 2017 Elsevier Ltd. All rights reserved. Keywords: European brown hare; Francisella tularensis; mountain hare; tularaemia
Introduction Tularaemia is an emerging zoonotic infectious disease caused by the bacterium Francisella tularensis. F. tularCorrespondence to: G. Hestvik (e-mail: [email protected]). 0021-9975/$ - see front matter http://dx.doi.org/10.1016/j.jcpa.2017.06.003
ensis has several subspecies, but only two are considered to cause disease, namely F. tularensis subsp. holarctica, which is found in the Northern hemisphere, and F. tularensis subsp. tularensis, which causes disease in North America only. F. tularensis is a pathogen of great concern because of its extreme infectivity and Ó 2017 Elsevier Ltd. All rights reserved.
104
G. Hestvik et al.
biothreat potential. It is facultatively intracellular and may infect macrophages and various other cell types including hepatocytes, dendritic cells and alveolar type II epithelial cells (Ellis et al., 2002). The bacterium has a wide host range, including mammals, birds and invertebrates. Infection with F. tularensis may be acquired through direct contact with infected tissues and fluids, via mucous membranes, inhalation and ingestion (Bengis et al., 2004). Indirect infection occurs via bites from arthropod vectors, in particular mosquitoes, biting flies and ticks. Little is known about the dissemination of F. tularensis in various organs and tissues after infection, or about the routes of shedding of the bacteria in naturally infected wild animals. This information is fundamental for diagnostic purposes and for better understanding of the epidemiology of tularaemia. In Sweden, tularaemia is a common zoonosis with annual numbers of human cases varying between 14 and 859 over the last 20 years (https:// www.folkhalsomyndigheten.se/folkhalsorapporteringstatistik/statistikdatabaser-och-visualisering/sjukdom sstatistik/harpest/). The disease was first recognized in 1931, when a Mountain hare (MH) (Lepus timidus) sold in a butcher’s shop was presumed to have caused infection in three people that had been in contact with the hare (Granstr€om, 1931). Wildlife hosts are involved in its epidemiology and are a source of infection for man. Tularaemia has historically been restricted to the northern regions of Sweden. In Sweden and Norway, prior to 2002, tularaemia was reported only in MHs and was described as an acute disseminated disease (Morner, 1992; Vikoren et al., 2008). Tularaemia was detected in European brown hares (EBHs) (Lepus europaeus) in Sweden for the first time in € 2002 in Orebro county in central Sweden, and since then cases have been diagnosed in brown hares (EBHs) yearly since 2002. The geographical range for tularaemia has expanded southwards (Tarnvik et al., 2004) affecting EBHs and people. This expansion might be due to a general spread of disease and/or to the involvement of EBHs which, compared with MHs, are more abundant in the southern parts of the country. Both subacute and chronic forms of tularaemia have been described in the EBH in Germany and Hungary (Sterba and Krul, 1985; Gyuranecz et al., 2010). In Sweden, only acute forms have been described in MHs and the pathological presentation in Swedish brown hares (EBHs) has not been studied. The aim of this study was to characterize the pathology of natural tularaemia infection in EBHs in Sweden in order to obtain a better understanding of the routes of infection, body dissemination and shedding of F. tularensis. Furthermore, the lesions were
compared with those seen in the MH. The ultimate goal was to contribute to the understanding of the epidemiological role of the EBH in the southward expansion of tularaemia in Sweden, as the EBH is a key wildlife host involved in the emergence of tularaemia in Sweden.
Materials and Methods Samples
During the period 2000e2013 630 hares (438 EBHs, 183 MHs and nine hares of undetermined species), found dead or killed due to disease, were submitted to the National Veterinary Institute (SVA) in Sweden for post-mortem examination. Out of these, 86 hares, 49 EBHs and 37 MHs, were diagnosed with tularaemia using indirect immunofluorescence (2000e2012) and immunohistochemistry (IHC) or reverse transcriptase polymerase chain reaction (RT-PCR) (2012 onwards). Thirty of the tularaemia-positive hares, 27 EBHs and three MHs, had mild to moderate post-mortem changes and were suitable for histopathological and immunohistochemical examinations. The 49 tularaemia-positive EBHs originated from counties in central Sweden: 18 from Uppsala and the remainder from Dalarna, G€avleborg, J€amtland, Stockholm, S€odermanland, € V€armland, V€astra G€otaland, Orebro and € Osterg€otland. The 37 tularaemia-positive MHs originated from counties in central and northern Sweden, 21 from Dalarna and the rest from G€ avleborg, J€amtland, Norrbotten, Stockholm, V€ armland, V€asterbotten and V€asternorrland. The age of the hares was recorded as adult or leveret. The nutritional state was recorded as normal, poor or cachectic. Polymerase Chain Reaction
For genotyping of F. tularensis subsp. holarctica, qPCR (TaqMan) was used to examine pooled samples of spleen, liver and lung from confirmed tularaemic hares. From the 30 hares chosen for retrospective histopathology, fresh or frozen organ samples were available from 21 EBHs and two MHs. Material from the tissue samples was retrieved using sterile cotton swabs. The swabs were incubated in 380 ml G2 buffer and 20 ml proteinase K solution (EZ1 Tissue DNA Extraction Kit, Qiagen, Sollentuna, Sweden) at 56 C for 15 min under continuous agitation, followed by 5 min incubation at 95 C. DNA was extracted from 200 ml of the resulting lysate using the EZ1 Tissue DNA Extraction Kit and the EZ1 Advanced Instrument (with the Bacteria Card) (Qiagen). The DNA was eluted in 50 ml elution buffer and 1 ml was used as template for each PCR. Francisella tularensis subspp.
Tularaemia in Hares
105
Tissue samples were fixed in 10% neutral buffered formalin, processed routinely and embedded in paraffin wax. For this study, new sections (4 mm) were stained with haematoxylin and eosin (HE). Lesions were scored subjectively as follows. For each section the area affected by necrosis was estimated as: 0e20%, 20e50%, 50e75% or 75e100%. The abundance of each type of inflammatory cell (i.e. lymphocyte, plasma cell, heterophil and macrophage) was graded subjectively as: not observed, few, moderate or numerous.
(RT) with 2% bovine serum albumin (BSA) (SigmaeAldrich, Stockholm, Sweden) for 20 min to block non-specific labelling. The primary antibody was applied at a 1 in 3,500 dilution and the sections were incubated at RT for 45 min. Antibody binding was detected by use of the anti-mouse EnVisionÔ polymer detection system (Dako) and the sections were counterstained with Mayer’s haematoxylin. To assess specificity (i.e. rule out false-positive labelling) a serial section was incubated with 2% BSA instead of the primary antibody. Additionally, for some cases, a serial section was incubated with normal mouse serum and mouse IgG1 (Dako). This was also performed on sections known to be tularaemia negative. To assess sensitivity (i.e. exclude false-negative labelling), known tularaemia-positive control liver or lung from an EBH was included in each run. The location of the labelling was described using the following categories: presence in lesions and/or in healthy tissue, presence in blood vessels, intra- or extracellular location, and in what cell types when located intracellularly. Immunohistochemistry for detection of T lymphocytes (mouse primary antibody CD3 monoclonal; Dako) and B lymphocytes (mouse primary antibody CD79; BioCare Medical, Pacheco, California, USA) was performed on selected cases and organs to identify cell types. Sections were dewaxed and antigen was retrieved in a microwave oven (in Tris EDTA buffer, pH 9.0) at 750 W for 7 min followed by 350 W for14 min. To block endogenous peroxidase activity, sections were incubated with H2O2 in TriseHCl (pH 7.6) for 20 min; thereafter the sections were incubated at RT with 2% BSA for 20 min to block nonspecific labelling. The primary antibody was applied at 1 in 75 (CD3) or 1 in 30 (CD79) dilution and the samples were incubated at RT for 45 min. Antibody binding was detected by use of the anti-mouse EnVisionÔ polymer detection system (Dako) and the sections were counterstained with Mayer’s haematoxylin. A serial section incubated with 2% BSA was used as a negative control.
Immunohistochemistry
Bacterial Culture
To visualize F. tularensis, IHC was performed on formalin-fixed tissues from all histopathologically examined organs using a mouse primary monoclonal antibody FB11 (Meridian Life Science Inc., Nordic Biosite AB, T€ aby, Sweden) directed against F. tularensis sp. lipopolysaccharide antigen. Sections were dewaxed and incubated with H2O2 in TriseHCl (pH 7.6) for 15 min to block endogenous peroxidase activity. Antigen was retrieved by proteinase K (Dako, Agilent, Glostrup, Denmark) treatment for 6 min; thereafter the sections were incubated at room temperature
Aerobic bacterial culture was performed on selected organ samples (seven lungs, two kidneys, nine livers, three spleens, one uterus and two testicles) from 13 hares suspected to be co-infected with additional bacteria. A few samples were submitted for culture in conjunction with the necropsy examination, while the remaining had been stored at 20 C for up to 7 years. The samples were transferred to horse blood agar (in house) and a purple agar plate (in house). Plastic loops were changed between primary, secondary and tertiary streaks. The plates were incubated in
holarctica genotyping was performed on 16 hares from fresh frozen organ samples and on seven hares from freeze-dried organ samples using two indel markers (Ftind49 and Ftind38) and nine canSNP markers (B.13, B.19, B. 23, B.26, B.39, B.40, B.41, B.42 and B.43) for typing of the three major canSNP-groups, B.4, B.6 and B.12, and the subgroups B.7, B.10, B.20, B.23 and B.39 (Svensson et al., 2009; Karlsson et al., 2013) according to the qPCR-based method described previously (Karlsson et al., 2013). Whole genome sequenced F. tularensis subsp. holarctica strains with known genotypes were used as controls, together with no template controls. Necropsy Examination
In hares examined between 2000 and 2009, necropsy findings in the heart, lung, liver, spleen, kidney and bone marrow were recorded routinely. During the period 2010e2013 additional organs examined routinely were brain, heart, airways, intestine, adrenal gland, ovary, testicle, uterus, urinary bladder, intestine and lymph nodes. The skin was scanned for the presence of ectoparasites (i.e. ticks). Descriptions of lesions were retrieved retrospectively from necropsy records and included organ enlargement, hyperaemia, oedema and necrosis/degeneration/inflammation. The lesions were graded as mild, moderate or severe. Histopathology
106
G. Hestvik et al.
an aerobic atmosphere at 37 C and analyzed after 1 and 2 days.
Results Polymerase Chain Reaction
The 21 EBHs and the two MHs examined by qPCR were all positive for F. tularensis subspp. holarctica. Seven samples were genotyped as canSNP group B.6, all of which belonged to subgroup B.7. Sixteen samples were genotyped as canSNP group B.12. These were further divided into subgroups B.20 (n ¼ 4), B.23 (n ¼ 11) and B.39 (n ¼ 1). The genotypes found are consistent with the currently known diversity in Sweden (Karlsson et al., 2013). Necropsy Examination
Gross pathology data was retrieved for the 86 hares subjected to necropsy examination and diagnosed with tularaemia during the period 2000e2013; 49 EBHs and 37 MHs. Table 1 shows the nutritional status in the 74 animals found dead or killed due to disease. The remaining 12, seven EBHs and five MHs died from trauma. Of these latter, three EBHs were in good and three were in poor nutritional status, and one was cachectic. Of the five MHs, four were in good and one was in poor nutritional status. Since tularaemia was not the direct cause of death, these 12 animals were not included in Table 1. The results show that a larger proportion of the EBHs (67%) than the MHs (41%) were cachectic and that a higher percentage of the MHs were in poor or good nutritional state compared with the EBHs. To investigate whether this was a significant difference, indicating that MHs were in a better nutritional state when they died, indicating a shorter course of disease, hares with cachexia and poor nutritional status were grouped together and were compared with hares having a good nutritional status. The association was not significant using a two-sided Fisher’s exact test. Enlargement of, and grossly visible necrosis in, a selection of organs from the 86 tularaemic hares is sumTable 1 Comparison of nutritional status in tularaemic EBHs and MHs found dead or killed due to disease 2000e2013
marized in Table 2. The most common finding, enlargement of the spleen, was seen in 80% of the EBHs and 30% of the MHs. Necrosis was visible as white, pinpoint to a few millimetre diameter foci, often spread diffusely throughout the organ. In bone marrow and spleen, necrotic foci were observed in a larger proportion of the MHs (41% and 62%, respectively) compared with the EBHs (12% and 14%, respectively). In the MHs, the liver was the only additional organ with grossly visible necrosis. In the EBHs, necrosis was also found in the lungs and kidney. The Fisher’s exact test was used to determine if there was a significant difference between EBHs and MHs with respect to the number of individuals having grossly visible necrosis. The difference was significant with P <0.005 for bone marrow and P <0.02 for lung, using a two-sided test. This indicates that grossly detectable necrosis in the bone marrow was more frequent in MHs, while lung necrosis was more common in EBHs. In general, gross lesions were present in a higher number of organs in each individual for the EBHs than in the MHs. In the liver and spleen the differences were not significant. Additional changes observed inconsistently were hyperaemia and haemorrhages, necrosis in lymph nodes and oedema of the lungs. It is noteworthy that five EBHs and nine MHs lacked grossly detectable changes. Moreover, five EBHs and seven MHs showed mild gross lesions, such as slight organ enlargement without visible necrosis, necrosis in the bone marrow and/or nonspecific findings such as oedema and hyperaemia in any organ. In total, 26 of the 86 tularaemic hares (30%) had either no lesions or minimal gross lesions. Nine of these hares had died of trauma. Histopathology and Immunohistochemistry
The histological lesions in the MHs and the EBHs were similar and are described below. Necrosis was a consistent finding, and was associated variably with inflammation. Necrosis was seen in all but one Table 2 Frequently observed gross lesions (necrosis and enlargement in any organ) in tularaemic EBHs and MHs Organ
Cachectic: number Poor nutritional Good nutritional Total number of animals status: number status: number of animals (% per species) of animals of animals (% per species) (% per species) EBH MH Total *
28 (67) 13 (41) 41
10 (24) 13 (41) 23
4 (9) 6 (18) 10
Twelve animals dead from trauma were not included.
42 32 74*
Bone marrow Liver Lung Spleen
EBH (n ¼ 49)
MH (n ¼ 37)
Necrosis: number of cases (%)
Enlargement: number of cases (%)
Necrosis: number of cases (%)
Enlargement: number of cases (%)
6 (12) 9 (18) 7 (14) 7 (14)
NA 3 (6) 0 39 (80)
15 (41) 4 (11) 0 11 (30)
NA 3 (8) 1 (3) 23 (62)
n, total number of examined cases; NA, not applicable.
Tularaemia in Hares
of the examined livers, in a large proportion of the spleens, bone marrow, lymph nodes and adrenal glands, and was common in the lungs and brain meninges (Table 3). Regardless of organ, the overall consistent presentation was multifocal, sometimes coalescing, variably-sized foci of necrosis distributed throughout the organ. Necrosis was multifocal, sometimes coalescing, with variable amounts of karyorrhectic, pyknotic and/or lytic nuclei, and the foci varied in size between 10 and 400 mm. Inflammatory cells associated with the foci necrosis were primarily macrophages and T cells, with occasionally a few plasma cells and heterophils. In the majority of cases this was consistent with an acute or subacute stage of inflammation (Fig. 1). In the liver, the sinusoids contained diffuse and mild to moderate numbers of CD3+ T lymphocytes (Fig. 2). In the spleen, foci of necrosis were located in the red pulp or at the border between red and white pulp (Fig. 1). Subjectively assessed and compared, the red pulp consistently had a lower cell density in most cases compared with the spleen from healthy hares; this change represented mild diffuse necrosis. Inflammation associated with the necrosis was not observed, but a mild to moderate infiltration of inflammatory cells was seen multifocally in the capsule and trabeculae. In five of the spleens there was a mild
Table 3 Number of organs with histological lesions in EBHs and MHs with tularaemia Organ
Adrenal gland Bone marrow Brain Heart Intestine Kidney Liver Lung: conducting airways Lung parenchyma Lymph nodes Mammary gland Ovary Spleen Testis Urinary bladder Uterus
EBH (n ¼ 27) Recorded lesions: *number of organs with lesions/number examined† (%)
MH (n ¼ 3) Recorded lesions: *number of organs with lesions/number examined† (%)
8/11 (72) 19/26 (73) 7/21 (33) 2/19 (11) 2/17 (12) 3/26 (12) 26/27 (96) 4/26 (15) 13/26 (50) 12/13 (92) 2/2 (100) 2/8 (25) 20/26 (77) 1/6 (17) 1/3 (33) 2/5 (40)
1/3 (33) 3/3 (100) 0/2 (0) 0/2 (0) 0/3 (0) 0/2 (0) 3/3 (100) 0/3 (0) 0/3 (0) 2/3 (67) 1/1 (100) 0/1 (0) 2/2 (100) 0/0 () 0/0 () 0/1 (0)
n, total number of hares included in the histopathological investigation. * Types of lesions recorded were necrosis and inflammation. † Number of animals in which respective organ was examined.
107
to moderate increase of inflammatory cells in the red pulp, but necrosis was not observed. The lymph nodes had foci of necrosis in the cortex and/or medulla. Inflammation was not associated with the necrotic areas, but occasionally mononuclear cells were seen in subcapsular and medullary sinuses. The bone marrow had no inflammatory cells associated with the foci of necrosis or in other parts of the sections (Fig. 1). The most common lesion in the lung was multifocal necrosis of the alveolar parenchyma, associated with mild to moderate inflammation (Fig. 1). However, in five EBHs the lung lesions were more severe and chronic. The necrotic foci often coalesced to form large areas, destroying bronchi and bronchioles. The inflammation was moderate to marked, with a mixed population of macrophages, lymphocytes, plasma cells and heterophils. Several hares had mild to marked fibrosis of the lungs (Fig. 3). The trachea was investigated in two of these hares, and in one case there were moderate multifocal areas of necrosis in the lamina propria and submucosa with an associated moderate mixed inflammation. The overlying epithelium showed multifocal erosion and ulceration. Lesions in the leptomeninges were found in seven of 21 brains examined and consisted of necrosis with mild to moderate associated inflammation; occasionally the lesions extended into the adjacent brain tissue (Fig. 4). Three hares had moderate to severe necrosis and/or exudative inflammation of the renal pelvis (Fig. 3). One also had moderate necrosis and mononuclear cell infiltrates in the cortex and the medulla. Most of the affected adrenal glands had small areas of necrosis with associated mild inflammation. In two hares, multifocal epicardial necrosis was evident. The urinary bladder, intestine, ovaries, uteri and testes were rarely examined. Lesions in these organs were few and small, with mild or no associated inflammation, with the exception of suppurative inflammation in one uterus. The mammary gland was examined only in two EBHs and one MH; all were lactating. The MHs had a few foci of necrosis and inflammation in the interalveolar connective tissue. In the two EBHs, necrosis was widespread and inflammation minimal (Fig. 5). In the skin of two EBHs, focal severe necrosis, inflammation and haemorrhage in the dermis, subcutis and cutaneous muscles were associated with still attached ticks (Fig. 6). The pattern of immunolabelling of Francisella spp. was similar in different organs. Immunolabelled bacteria were found both extra- and intracellularly in association with necrosis and inflammation; the most common intracellular location was in macrophages (Fig. 2). In addition, Francisella spp. were found intracellularly in macrophages and T cells in the blood vessels in 13 cases (Fig. 2). To a lesser degree,
108
G. Hestvik et al.
Fig. 1. Acute changes showing necrotic foci with intralesional F. tularensis spp. in the spleen (A, B), bone marrow (C, D) and lung (E, F) on serial sections. HE and IHC. Bars (AeF), 100 mm.
intracellularly-located Francisella spp. were seen in pneumocytes, alveolar macrophages, endothelial cells and megakaryocytes. The presence of Francisella spp. in some locations is worth mentioning because of the relevance for transmission. Bacteria were seen in exudate in the lumen of bronchioles of one hare (Fig. 3) and intracellularly in macrophages and lining epithelial cells in bronchioles in two other hares. In the three kidneys with inflammation in the pelvis, bacteria were detected in the exudate (Fig. 3). In
the three examined mammary glands, Francisella spp. were seen in the milk secretions and in macrophages and epithelial cells (Fig. 5). In the skin of hares with attached ticks, large numbers of Francisella spp. were detected locally in the lesion, both intra- and extracellularly (Fig. 6). Bacterial Culture
In 13 hares with severe inflammation or involvement of organs less commonly presenting with signs of
Tularaemia in Hares
109
Fig. 2. Liver sinusoids containing small cells on HE staining (A), shown by IHC for CD3 to be T-lymphocytes (B). IHC for F. tularensis spp. shows that many of the T-lymphocytes contain cytoplasmic bacteria (C). In addition, F. tularensis spp. can be visualized intraluminally in the central vein (C) and intracellularly in Kupffer cells (D). Bars (AeD), 100 mm.
Fig. 3. Subacute/chronic lesions in the lung and renal pelvis. HE (A, C) and IHC for F. tularensis spp. (B, D) on serial sections. (A, B) Necrosis with severe inflammation destroying the tissue, and containing large numbers of F. tularensis spp. A bronchiole (*) contains exudate with numerous bacteria. (C, D) Lumen of the renal pelvis is filled with an exudate containing numerous heterophils and bacteria. Bars (AeD), 100 mm.
110
G. Hestvik et al.
Fig. 4. Necrosis and inflammation in the leptomeninges of the brain and arterial cuffing, HE (A). IHC shows numerous F. tularensis spp. in the lesion (B). Bars (A, B), 100 mm.
Fig. 5. Mammary gland from a lactating female. Multifocal areas of necrosis (N) interspersed with milk-containing alveoli (M). F. tularensis spp. are present in both necrotic areas and in the milk (B). HE (A) and IHC F. tularensis spp. (B). Bars (A, B), 100 mm.
inflammation and/or necrosis (e.g. kidney, uterus and testis), a dual infection was considered. Aerobic bacterial culture was performed on a variety of organs (seven lungs, two kidneys, nine livers, three spleens, one uterus and two testes) in the hares from which archived tissues were available. The significance of the bacteria cultured from 11 of the hares was difficult to interpret, including a variable presence of Escherichia coli, Enterococcus spp. and a mixed flora. It was presumed that their presence was probably caused by
post-mortem overgrowth from the intestine or contamination of the sample surfaces during handling at necropsy examination and sampling. Klebsiella pneumoniae was cultured in moderate to large amounts from a hare with severe inflammation in the lungs. In another hare, small numbers of E. coli grew in pure culture in samples from the kidney and uterus. In both of these hares the inflammation was more severe compared with other hares. Both of these hares were considered to have true dual infections.
Fig. 6. Serial section of skin at the site of a tick bite, HE (A) and IHC for F. tularensis spp. (B). (A) Necrosis and inflammation (*) with an underlying haemorrhage. F. tularensis spp. are found in large numbers at the tick bite site (B), suggesting that the tick transmitted the bacteria to the hare. Bars (A, B), 200 mm.
Tularaemia in Hares
Discussion The habitat of MHs is woodland and these animals are spread all over Sweden. The EBH was introduced into southern Sweden during the mid-19th century, and has, during the last decades, expanded north to inhabit the southern half of the country, and to woodlands, in addition to agricultural areas. The two hare species cohabit in many areas and cross-species hybridization is known to occur. In areas where both species are found, the number of MHs has decreased, probably due to less favourable winters for this species, with less snow and higher temperatures (Jansson, 2013). The first cases of tularaemia in Sweden were diagnosed in man and one mountain hare in 1931 and since then tularaemia has been considered to be endemic in the northern regions of Sweden. During the 1970se1980s, MHs that died from tularaemia were found in the northern two thirds of the country (from county Norrbotten in the north to county V€ armland in the central part of the country) and on the island of Stora Karls€o in the southern part of the Baltic Sea (Morner et al., 1988). The MH was the only hare species known to be affected until 2002 when the first case of tularaemia in the EBH was diagnosed. During the past decades, tularaemia has spread south in man and hares, but the reason is not known (Tarnvik et al., 2004). A reason why tularaemia emerged in EBHs in Sweden might be the southward expansion of tularaemia in combination with the northward expansion of populations of EBHs, causing their exposure. Whether EBHs might have a role in spreading the disease further south is not known. Since EBHs had not been tested routinely for tularaemia before 2002, a retrospective study was conducted. Between 1987 and 1999, 909 EBHs were subjected to necropsy examination. Out of these, five had histopathological lesions in the liver or spleen compatible with tularaemia. Archived formalin-fixed and paraffin wax-embedded samples from these five EBHs were tested by IHC for F. tularensis spp. and all were negative. Since the first case in 2002, the disease has been diagnosed regularly in EBHs. Since MHs and EBHs may cohabit, there is a possibility for infection to pass between species by direct contact, as well as by indirect contact through environmental contamination (e.g. water and grass), and possibly also through vectors. During the last decade, tularaemia cases have been diagnosed as far south as the county of J€ onk€ oping, approximately 350 km from the southern country border (SVA post-mortem records). It has been suggested that EBHs are somewhat less susceptible to developing severe tularaemic disease
111
caused by F. tularensis subsp. holarctica than MHs (Gyuranecz, 2012). This is based on findings of subacute/chronic lesions with granulomatous inflammation in EBHs in Hungary and Germany (Sterba and Krul, 1985; Gyuranecz et al., 2010) compared with the acute disseminated lesions in MHs described in northern Europe (Morner et al., 1988; Vikoren et al., 2008). However, in later years there have been descriptions of acutely diseased EBHs in Europe outside Scandinavia, for example in France (Decors et al., 2011). One explanation for the differences could be that the majority of hares with subacute/chronic disease were hunted (i.e. were apparently healthy), while those with acute disease mostly were found dead. In our study, all EBHs and MHs were found dead or were killed due to disease and therefore were suitable for comparison of whether tularaemia at the terminal stage of disease has a more acute course in MHs than in EBHs. Comparison of the nutritional state and inflammatory response could add valuable information, since animals with shorter disease processes are more likely to have a better nutritional status and less inflammatory response. However, comparison of the nutritional status in the EBHs and MHs during the period 2000e2013 showed no significant difference between the two species. Comparison of grossly visible foci of necrosis in the two hare species resulted in a significant difference for only two organs. Necrosis was more frequent in the bone marrow of MHs compared with EBHs and in the lungs of EBHs compared with MHs. Histopathological changes in 21 EBHs and three MHs were consistent with acute disseminated disease. This result is in accordance with the pathology in EBHs found dead in France (Decors et al., 2011) and in MHs in Sweden and Norway (Morner et al., 1988; Vikoren et al., 2008; Josefsen et al., 2012). It was not possible to assess differences in the number of microscopically visible foci of necrosis and in the inflammatory response between EBHs and MHs in our study, since only three MHs were fresh enough to allow good assessment on histopathology. The results of our study, in accordance with additional results from previous studies of MHs, show that tularaemia can have an acute presentation in both hare species, showing necrosis with absent or minimal inflammatory reaction (Morner et al., 1988; Vikoren et al., 2008; Josefsen et al., 2012). It is well known that EBHs develop chronic disease (Sterba and Krul, 1985; Gyuranecz et al., 2010). Information regarding chronic presentations of tularaemia in MHs is lacking, possibly due to absence of investigation of hunted MHs. Six of the EBHs had subacute/chronic changes in the lung and/or kidney and an inflammatory response in the
112
G. Hestvik et al.
spleen in addition to an acute disseminated disease. These may represent chronic infections that were reactivated to give acute disease. The lesions of some investigated organs are worth particular mention. Lesions in the leptomeninges were found in one-third of the examined brains and consisted of necrosis with mild to moderate inflammation; occasionally the lesions extended into the adjacent brain tissue. The only report examining the brain of a hare infected with F. tularensis subsp. holarctica is from a Japanese hare (Lepus brachyurus augustidens), in which encephalitis was recorded (Park et al., 2009). Our findings suggest that meningitis and/or encephalitis is more common than previously appreciated in tularaemia. The mammary gland is another organ not commonly investigated. In our study, the mammary glands of only two EBHs and one MH were examined; all were lactating. All three had necrosis with minimal inflammation in the mammary glands. Mammary lesions were also found in one Hungarian EBH, described as granulomatous inflammation (Gyuranecz et al., 2010). Because all our examined hares had acute and disseminated disease, it is not unexpected that lesions and bacteria were found in many organs that have rarely been investigated in earlier studies, such as brain, mammary gland, testis, ovary and adrenal gland. We postulate that the dissemination in the body was through the haematogenous route, since infection affected multiple organs and lesions of the same type and duration were diffusely distributed throughout each organ. This was also confirmed by IHC, showing F. tularensis spp. within inflammatory cells in the blood vessels in 13 hares. Finding that six EBHs had a partly different disease presentation, with more severe and/or exudative inflammation, we hypothesized that a dual infection with some other bacteria might be the cause. These hares had severe and subacute/chronic lesions in the lung and a more severe inflammatory response in the spleen. In addition, three of them had suppurative inflammation in the renal pelvis. Despite this, the lesions in other affected organs had the same presentation as in the other hares. Most of the cultures showed growth of E. coli and Enterococcus spp., which could have been caused by contamination during handling. Cultures from two hares were assessed to be indicative of true dual infection. Klebsiella pneumoniae was cultured from one lung and E. coli from another hare’s kidney and uterus. Dual infection with K. pneumoniae and F. tularensis spp. has been described in a human case (Givham, 1965). Differences in disease presentation in EBHs as described above could also be due to differences in the pathogenicity of F. tularensis subsp. holarctica geno-
type. Of the hares with subacute/chronic lesions in the lung and/or kidney, F. tularensis subsp. holarctica was genotyped from four. Two of these belonged to canSNP group B.12 (subgroups B.20 and B.23, respectively) and the other two belonged to canSNP group B.6 (subgroup B.7). Therefore, the differences found in genotype could not be associated with different disease presentations. Genotyping of F. tularensis subsp. holarctica in tularaemic EBHs in Switzerland resulted in the finding of two canSNP groups, B.6 (subgroup B.10) and B.12 (subgroup B.13) (Origgi and Pilo, 2016). The authors concluded that canSNP group B.6 (subgroup B.10) was associated with splenitis and hepatitis, while canSNP group B.12 (subgroup B.13) was associated with lesions in the pleura, pericardium and kidney, as described in Hungary (Gyuranecz et al., 2010; Origgi and Pilo, 2016). The majority of hares in our study had lesions in the spleen, liver and many other organs, in both canSNP groups. In Sweden, mosquito bites are considered to be the most frequent means of tularaemia infection in man. Another, less frequent route, is entrance through skin erosions or ulcers by direct contact with body secretions from an infected animal. After local multiplication at the infection site, the bacteria are disseminated through the blood and lymphatic vessels to other organs (Gyuranecz, 2012). Mosquito bites are probably also an important route of infection for animals, and a focal skin infection can easily be overlooked at necropsy examination. Various species of ticks are vectors of F. tularensis holarctica in Europe (Hestvik et al., 2014). During the past decades, tularaemia has expanded southwards and ticks northwards in Sweden, therefore ticks may be an increasingly important source of infection. The only clearly proven entry route in our study was through the skin. In two EBHs, severe focal necrosis and inflammation were seen in the skin where ticks were still attached. The ticks were, based on their gross morphology, identified as Ixodes ricinus, the most common tick species in Sweden (http://www.nrm.se/ faktaomnaturenochrymden/djur/insekterochspinde ldjur/spindeldjur/fastingar.12595.html). By IHC, large numbers of Francisella spp. were visualized in the skin lesions, indicating that the tick bite was the site of entry and multiplication. Similar skin lesions associated with a tick bite have been described in a Japanese hare (Lepus brachyurus angustidens), where F. tularensis spp. also were detected by IHC in the intestine of the tick (Park et al., 2009). Infection through inhalation has also been proposed in animals (Gyuranecz et al., 2010; Decors et al., 2011), but this could not be confirmed in our study. The presence of bacteria in bronchiolar epithelial cells and
113
Tularaemia in Hares
macrophages in two hares might indicate signs of entry, but could also be signs of shedding or just the location of the infection. As in other affected organs, the most common lung lesion was multifocal necrosis in the alveolar parenchyma without involvement of conducting airways. This suggests that the infection reaches the lungs through a haematogenous route rather than a respiratory route. However, in one hare, Francisella spp. was also evident in the bronchiolar exudate, indicating that shedding by the respiratory route may occur. The present study has indicated several other possible shedding routes. Francisella spp. was detected in the exudate in the renal pelvis of three hares and in milk in lactating mammary glands of three hares, showing that these are two other ways of possible bacterial shedding. In the intestine and uterus, F. tularensis was detected in superficial mucosal lesions although not in luminal secretions, which at least in theory makes these organs possible routes of shedding. Assuming that the cause of the severe inflammation in the lung, kidney and uterus represented dual bacterial infection, together with the fact that F. tularensis spp. in these cases were only found in exudate, makes it likely that these routes of shedding are not frequent in acutely diseased hares. On the other hand, if these hares represent cases of chronic tularaemia without dual infections, these shedding routes could be important. The majority of hares investigated were found dead or were killed due to disease, so this study describes mostly terminal disease in lethal forms of tularaemia. Little information is available about other presentations of tularaemia; for example, chronic forms or earlier stages of infection in hares in Sweden, which would need to be investigated in apparently healthy hares, for example animals killed by hunting. Twenty-six of the 86 investigated tularaemic hares (30%) had no gross lesions or minimal lesions that could easily be missed. Nine of these 26 hares died of trauma, and the lack of gross lesions in those might be due to early infection at the time of death. Lack of visible gross lesions increases the risk of infection for people who handle wildlife. There are examples of hunters contracting disease after dressing hunted animals with unprotected hands (Munnich and Lakatos, 1979; Wahab and Bjurgren Zelano, 2012; van de Wetering et al., 2015). Hunting dogs are also at risk of infection if they bite the hare or are fed the hare’s internal organs (Nordstoga et al., 2014). Sources of infection for people, predators and scavengers are carcasses of tularaemic animals and water contaminated by diseased animals’ excreta or dead infected animals. Predators and scavengers such as red foxes, wild boars, raccoon dogs, wolves, lynx,
brown bears and certain bird species are considered relatively resistant to tularaemic disease. They may, however, contract the infection and infected hares and other species are a source of infection through predation and scavenging. This has been shown for red foxes, wild boars and raccoon dogs in several studies by the presence of antibodies towards F. tularensis (Hoflechner-Poltl et al., 2000; Hubalek et al., 2002; Al Dahouk et al., 2005; Kuehn et al., 2013) and detection of the bacteria in submandibular lymph nodes of red foxes (Hofer et al., 2010). It is not clear if these animal species can act as reservoirs and contribute to the spread of bacteria. In conclusion, in Sweden over the last 15 years, tularaemia has been expanding its range further south and has been found in a new hare species (EBHs). The present pathological investigation suggests that EBHs and MHs, found dead or killed due to disease, have similar lesions and die of an acute disseminated disease (sepsis). The chronic disease presentation seen in a few hares could reflect a pure chronic disease that has been reactivated and/or a dual infection with an additional bacterium. No differences in pathological presentation between F. tularensis subsp. holarctica genotypes canSNP group B.6 and B.12 were evident. Infection through bite wounds from ticks was seen. Routes of shedding that might be of importance are via the urinary tract, milk, uterine secretions and intestinal content, but since most hares do not show changes in these organs the most important shedding route is probably from dead animals. Animals and people might be infected by exposure to water contaminated by dead tularaemic hares or by ingestion of infected hares. Since tularaemia is not always easily detectable on gross inspection, there is a risk for people handling hares and dogs fed uncooked material from infected hares.
Acknowledgments This work was supported in part by the European Union Seventh Framework Programme (FP7/ 2007e2013), Wild Tech [grant agreement no. 22633].
Conflict of Interest Statement The authors declare that they have no conflicts of interest with respect to their authorship or publication of this article.
References Al Dahouk S, Nockler K, Tomaso H, Splettstoesser WD, Jungersen G et al. (2005) Seroprevalence of brucellosis, tularemia, and yersiniosis in wild boars (Sus scrofa)
114
G. Hestvik et al.
from North-Eastern Germany. Journal of Veterinary Medicine, Series B: Infectious Diseases and Veterinary Public Health, 52, 444e455. Bengis RH, Leighton FA, Fischer JR, Artois M, Morner T et al. (2004) The role of wildlife in emerging and reemerging zoonoses. Revue Scientifique et Technique De L’Office International des Epizooties, 23, 497e511. Decors A, Lesage C, Jourdain E, Giraud P, Houbron P et al. (2011) Outbreak of tularaemia in brown hares (Lepus europaeus) in France, January to March 2011. Eurosurveillance, 16, 3e5. Ellis J, Oyston PCF, Green M, Titball RW (2002) Tularemia. Clinical Microbiology Reviews, 15, 631e646. Givham E (1965) Tularemic pneumonia with Klebsiella superinfection. Journal of the Tennessee Medical Association, 58, 224e226. Granstr€om KO (1931) Tre fall av tularemi (three cases of tularemia). Svenska L€akartidningen, 28, 641e646. Gyuranecz M (2012) Tularaemia. In: Infectious Diseases of Wild Mammals and Birds in Europe, D Gavier-Widen, JP Duff, A Meredith, Eds., Blackwell Publishing Ltd., Chichester, pp. 303e309. Gyuranecz M, Szeredi L, Makrai L, Fodor L, Meszaros A et al. (2010) Tularemia of European brown hare (Lepus europaeus): a pathological, histopathological, and immunohistochemical study. Veterinary Pathology, 47, 958e963. Hestvik G, Warns-Petit E, Smith LA, Fox NJ, Uhlhorn H et al. (2014) The status of tularemia in Europe in a onehealth context: a review. Epidemiology and Infection, 43, 2137e2160. Hofer E, Reisp K, Revilla-Fernandez S, Plicka H, Romanek G et al. (2010) Isolation of Francisella tularensis and Brucella suis from red foxes (Vulpes vulpes). Tieraerztliche Umschau, 65, 229e232. Hoflechner-Poltl A, Hofer E, Awad-Masalmeh M, Muller M, Steineck T (2000) Prevalence of tularaemia and brucellosis in European brown hares (Lepus europaeus) and red foxes (Vulpes vulpes) in Austria. Tierarztliche Umschau, 55, 264e268. Hubalek Z, Treml F, Juricova Z, Hunady M, Halouzka J et al. (2002) Serological survey of the wild boar (Sus scrofa) for tularaemia and brucellosis in South Moravia, Czech Republic. Veterinarni Medicina, 47, 60e66. Jansson G (2013) Mountain hare and brown hare e where and why they hybridize in Sweden? Fauna och Flora (Stockholm), 108, 22e27. Josefsen TD, Mork T, Ytrehus B, Hetta MH, Henrichsen EN et al. (2012) Tularaemia in Finnmark (Tularemi i Finnmark). Norsk Veterinartidsskrift, 124, 173e174. Karlsson E, Svensson K, Lindgren P, Bystrom M, Sjodin A et al. (2013) The phylogeographic pattern of Francisella tularensis in Sweden indicates a Scandinavian origin of Eurosiberian tularaemia. Environmental Microbiology, 15, 634e645. Kuehn A, Schulze C, Kutzer P, Probst C, Hlinak A et al. (2013) Tularaemia seroprevalence of captured and
wild animals in Germany: the fox (Vulpes vulpes) as a biological indicator. Epidemiology and Infection, 141, 833e840. Morner T (1992) The ecology of tularaemia. Revue Scientifique et Technique de L’Office International des Epizooties, 11, 1123e1130. Morner T, Sandstrom G, Mattsson R, Nilsson PO (1988) Infections with Francisella tularensis biovar palaearctica in hares (Lepus timidus, Lepus europaeus) from Sweden. Journal of Wildlife Diseases, 24, 422e433. Munnich D, Lakatos M (1979) Clinical, epidemiological and therapeutic experience with human tularemia e role of hamster hunters. Infection, 7, 61e63. Nordstoga A, Handeland K, Johansen TB, Iversen L, Gavier-Widen D et al. (2014) Tularaemia in Norwegian dogs. Veterinary Microbiology, 173, 318e322. Origgi FC, Pilo P (2016) Francisella tularensis clades B.FTN002-00 and B.13 are associated with distinct pathology in the European brown hare (Lepus europaeus). Veterinary Pathology, 53, 1220e1232. Park C-H, Nakanishi A, Hatai H, Kojima D, Oyamada T et al. (2009) Pathological and microbiological studies of Japanese hare (Lepus brachyurus angustidens) naturally infected with Francisella tularensis subsp. holarctica. Journal of Veterinary Medical Science, 71, 1629e1635. Sterba F, Krul J (1985) Gross and histological studies on the differential diagnosis of brucellosis, tularemia and pseudotuberculosis in hares (Lepus europaeus). In: Transactions of the XVIIth Congress of the International Union of Game Biologists, Brussels, September 17e21, 1985. Part 2, pp. 763e771. Svensson K, Granberg M, Karlsson L, Neubauerova V, Forsman M et al. (2009) A real-time PCR array for hierarchical identification of Francisella isolates. PLoS One, 4, 12. Tarnvik A, Priebe HS, Grunow R (2004) Tularaemia in Europe: an epidemiological overview. Scandinavian Journal of Infectious Diseases, 36, 350e355. Wahab T, Bjurgren Zelano S (2012) J€ agare smittad av harpest vid slakt av hare (Hunter infected by tularemia while dressing a hare). Lakartidningen, 109, 152e153. van de Wetering D, dos Santos CO, Wagelaar M, de Kleuver M, Koene MGJ et al. (2015) A cluster of tularaemia after contact with a dead hare in the Netherlands. Netherlands Journal of Medicine, 73, 481e482. Vikoren T, Handeland K, Djonne B (2008) Fleire tilfelle av tularemi p avist hj a hare i S€ or-Noreg (Several cases of tularemia diagnosed in hares in southern Norway). Norsk Veterinaertidskrift, 120, 583.
March 1st, 2017 ½ Received, Accepted, June 9th, 2017