J. Neurovirol.
DOI 10.1007/s13365-014-0255-5
Histopathology and immunohistochemistry of tissues
outside central nervous system in bovine rabies
Camila C. Abreu & Priscilla A. Nakayama & Clayton I. Nogueira &
Leonardo P. Mesquita & Priscila F. R. Lopes & Flademir Wouters &
Mary S. Varaschin & Pedro S. Bezerra Jr
Received: 30 January 2014 / Revised: 16 April 2014 / Accepted: 2 May 2014
# Journal of NeuroVirology, Inc. 2014
Abstract We performed a histopathological and immunohis-
tochemical study of tissues outside the central nervous system
in 48 cases of bovine rabies confirmed by direct immunoflu-
orescence and/or immunohistochemistry (IHC) of the central
nervous system. In the bovines of this study, mononuclear
inflammation in all ganglia (trigeminal, spinal, stellate, and
celiac) and adrenal medulla was observed. This injury also
occurred in 85 % of neuro-pituitaries in 55 % of pars inter-
mediate and 15 % of the pars distalis of pituitary evaluated.
IHC was positive in 92.31 % of lumbar spinal ganglia, 90.9 %
of trigeminal ganglia, stellate ganglia of 41.67 and 16.67 % of
the celiac ganglia. One of the evaluated adrenal (1/17) showed
strong immunohistochemical labeling in the cytoplasm of
pheochromocytes. The pituitary IHC was positive in one case
in the neurohypophysis (1/20) and in one case in the pars
intermedia of the adenohypophysis (1/20). Data from this
study indicate that in suspected cases of rabies, besides the
complex pituitary rete mirabile and trigeminal ganglion, the
evaluation of other ganglia, particularly the lumbar spinal, and
adrenal may also contribute to the diagnosis and understand-
ing of the clinical presentation and pathogenesis of the disease
in bovines.
Keywords Adrenals . Bovines . Centrifugal dispersion .
Ganglia . Pituitary . Rabies virus
C. C. Abreu : P. A. Nakayama : C. I. Nogueira : L. P. Mesquita :
P. F. R. Lopes : F. Wouters : M. S. Varaschin
Setor de Patologia Veterinária, Departamento de Medicina
Veterinária, Universidade Federal de Lavras, Lavras,
Minas Gerais, Brazil
P. S. Bezerra Jr (*)
Laboratório de Patologia Animal, Faculdade de Medicina
Veterinária, Universidade Federal do Pará, Campus Castanhal.
Avenida dos Universitários, s/n, Campus I, Jaderlândia,
68.746.360 Castanhal, PA, Brazil
e-mail: psbezerrajunior@gmail.com
Introduction
Rabies is a zoonosis of worldwide distribution, caused by a
virus of the genus Lyssavirus, with neurotropic character that
affects mammals in general (Jackson 2008). In Brazil, the
disease has a significant impact on livestock, and it is one of
the leading causes of death in bovines (Barros et al. 2006). The
bovine rabies is mainly transmitted by vampire bat Desmodus
rotundus (Sota 1993).
The rabies virus, when inoculated by the vampire bat’s
saliva, migrates by motor axonal and sensory pathways until
it reaches the central nervous system (CNS) (Jackson 2010).
Another approach recently recognized is the intravenous in-
oculation (Preuss et al. 2009). It is known that in the vectors,
the centrifugal dispersion of the rabies virus takes place main-
ly in the salivary glands, and saliva is the main form of
transmission of the disease (Jackson 2008). The entry of the
rabies virus by sensory pathway may be important in rabies
transmitted by vampire bats, since these vectors produce more
superficial bites, where the sensory and autonomic innervation
is more abundant (Jackson 2010). The intravenous inoculation
has also been identified as important in disease transmitted by
vampire bats (Preuss et al. 2009).
Studies on rabies lesions outside the CNS have been con-
ducted mainly in humans (Morais and Assis 1985; Jackson
et al. 1999; Jogai et al. 2002; Tobiume et al. 2009) and in
animal species that are vectors (Ajayi et al. 2006;
Balachandram and Charlton 1994). In these species, non-
suppurative inflammation in various tissues such as salivary
glands, nerve plexus of the gastrointestinal tract, nerve gan-
glia, adrenal medulla, and muscle fibers and nervous in the
heart has been described (Ajayi et al. 2006; Balachandram and
Charlton 1994; Van Gehuchten and Nelis 1900). Hyaline
globules have been described in the adrenal medulla
pheochromocytes of human and bovine rabies (Almeida
et al. 1986; Mesquita et al. 2011). In bovines, the studies of

the trigeminal ganglia have determined that the trigeminal
ganglioneuritis is an important injury that contributes for
rabies diagnosis in this species (Rech et al. 2006; Summers
et al. 1994).
The distribution of the rabies antigen in peripheral tissues
varies among different species. Most immunohistochemical
studies performed to evaluate this distribution have been
performed in humans (Bagó et al. 2005; Jackson et al. 1999;
Jogai et al. 2002; Tobiume et al. 2009). Similar studies in
animals are scarce, and it was done primerly in the main
vectors like Vulpes vulpes (foxes) and Mephitis mephitis
(skunks) (Balachandram and Charlton 1994), dogs (Ajayi
et al. 2006), and vampire bats (Vieira et al. 2011). In buffaloes
experimentally infected, the rabies virus antigens were detect-
ed ante-mortem in body secretions (nasal, salivary, and rectal)
and in mucosal impression smears (corneal and preputial), by
direct immunofluorescence (DIF) (Singh and Singh 2011).
Furthermore, studies conducted in bovines have shown dif-
ferent results. Delpietro et al. (2001) demonstrated that the
presence of virus in saliva and salivary gland is not significant.
In a study by TaqMan real-time PCR, it was possible to
diagnose rabies in vivo using biopsy samples from bovine
skin (Bansal et al. 2012). Virus antigens are often found in the
trigeminal ganglia of bovines by immunohistochemistry
(IHC) (Pedroso et al. 2009).
However, histological and immunohistochemical wider
studies on the peripheral dispersion of the rabies virus in
bovines were not found. Such studies in other species have
contributed to the understanding of the pathogenesis and
diagnosis of rabies, which highlights the importance of this
study in bovines.
Material and methods
The study was conducted in tissues of 48 cases of bovine
rabies received by the Setor de Patologia Veterinária (SPV) of
the Universidade Federal de Lavras (UFLA), Minas Gerais
State, Brazil, in the period from 2000 to 2011 and confirmed
by the FAT and/or IHC from the CNS, by the Instituto Mineiro
de Agropecuária (IMA) and by SPV-UFLA, respectively. In
29 cases, from the period of 2000–2009, tissues outside the
CNS were used, which were stored in paraffin on the SPV-
UFLA. In the period from 2010 to 2011, a prospective study
was performed of all cases of bovines that died showing
neurological signs. For that study, a wider collection of tissues
was done, and only cases with rabies diagnosis (19 cases)
were included. In these cases, various tissues were collected
for histological processing.
We examined into sentence a total of 48 fragments of
bovines’ brain, 20 pituitary, 33 trigeminal ganglia, 14 lumbar
spinal ganglion, four cervical spinal ganglia, two thoracic
spinal ganglia, 12 celiac and 12 stellate ganglia, 11 cornea,
J. Neurovirol.
10 optic nerve, three lacrimal gland, 11 brachial plexus, 15
nerve ischiadic, 17 adrenal, 15 thyroid, 14 submandibular
salivary gland, 13 tongue, six esophagus, 15 pre-stomachs,
11 abomasum, 19 liver, eight gallbladder, 13 pancreas, 17
intestines, eight tonsil, 14 spleen, 15 lymphnodes, five bone
marrow, seven skeletal muscle, 22 heart, 19 lung, four trachea,
19 kidney, 11 bladder, nine mammary gland, nine uterus, five
ovary, one testicle, and skin of 17 bovines (ear, nape, muzzle,
labial commissure, tactile hair, and eyelid).
The tissues were fixed in 10 % formalin buffered with
phosphates. After fixation for 24–48 h, the material was
cleaved dehydrated in increasing concentrations of ethanol,
diaphanized in xylene and embedded in paraffin. Fragments
were cut at 5 μm and stained with hematoxylin and eosin
(HE). Adrenal cuts were also stained by periodic acid–
Schiff (PAS).
IHC for rabies was performed according to pre-established
protocol (Pedroso et al. 2008). Was employed a longer
blocking endogenous peroxidase, for which the sections were
incubated for 30 min in a solution of hydrogen peroxide at 3 %
in distilled water at room temperature. For blocking of non-
specific binding (background), the sections were treated with
skimmed milk (Molico, Nestlé®) 5 % diluted in phosphate-
buffered saline (PBS) 0.01 M (pH 7, 4) for 1 h. The sections
were incubated with primary antibody (polyclonal anti-rabies,
produced in goats, Millipore, 5199), at a 1:100 dilution in
PBS, incubated for 1 h in a moist chamber at 37 °C. The
revelation of the reaction was performed with the red chromo-
gen (Vector® NovaRED) for 1–2 min, controlling the incuba-
tion time with the appearance of a pale pink color of the cut.
As a positive control, CNS (bulb and cerebellum) sections
of bovine rabies cases previously confirmed in DIF were used.
For negative controls, we used CNS and others tissues from
uninfected bovine, the CNS of which was negative for rabies
in DIF and IHC tests (Fig. 1). Other negative controls were the
conduct of IHC with uncorrelated primary antibody (poly-
clonal anti-Toxoplasma gondii antibodies produced in goats
VMRD, 210-70-TOXO) and the incubation with the omission
of the primary antibody (incubation with PBS).
IHC for neurofilament (monoclonal mouse anti-human
neurofilament protein—clone 2F11) was performed in opti-
cal nerve, brachial plexus, and nerve ischiadic. The protocol
was in accordance with the manufacturer’s instructions
(Dako—M0762).
Results
Spinal ganglia
Of all the bovine rabies cases evaluated, 14 (29.17 %) had the
spinal ganglia collected. Of those cases, 71.43 % of the
ganglia were collected from the lumbar region, 14.29 % from
J. Neurovirol.
Fig. 1 Immunohistochemistry for rabies in control sections from uninfected bovine. Streptavidin–biotin–peroxidase, Vector® NovaRED chromogen. a
Celiac ganglion. b Adrenal medulla. c Pars intermedia of the adenohypophysis. d Neurohypophysis
the lumbar and cervical regions, and 14.29 % from the lumbar,
cervical, and thoracic regions.
In the histopathological evaluation, all spinal ganglia
showed mononuclear inflammation, neuronal necrosis,
neuronophagia (Fig. 2), and Nageotte nodules (Fig. 3), vary-
ing only in intensity of these lesions at different levels of the
spinal cord and from one case to another.
The severest lesions and immunohistochemical markers
were generally more frequent in the lumbar spinal ganglia
(Fig. 4). Of these, 28.57 % showed mild, 14.29 % showed
moderate and 57.14 % showed severe mononuclear inflam-
mation. Neuronal necrosis was characterized by perikaryon
with increased eosinophilia and loss of Nissl substance, asso-
ciated with pyknosis, cariorrexia, and karyolysis. The perikar-
yon of some necrotic neurons was retracted or fragmented.
Other unretracted neurons presented paler center and pyknotic
nuclei when present. When the pyknotic nuclei were present,
they were found on the periphery, resembling axonal reaction.
The neuronophagia was characterized by the presence of
mononuclear cells with moderate amount of cytoplasm
(macrophages) interspersed with fragments of perikaryon
(Fig. 2). The neuronal necrosis and neuronophagia were mild
in 14.29 %, moderate in 71.43 %, and severe in 14.29 %.
Nageotte nodules, characterized by hypertrophy of satellite
cells occupying local loss of the neuronal body (Fig. 3), were
rare in 64.29 % and occasional in 35.71 %. The lumbar spinal
ganglia also showed citoplasmatic vacuolation in 78.57 %,
and this vacuolation was mild in 72.73 % and moderate in
27.27 %. In addition, one case (7.14 %) showed the presence
of Negri bodies. The IHC performed in 13 lumbar spinal
ganglia (Fig. 5) was positive in 92.31 %. The immunolabeling
was mild in 75 %, moderate in 16.67 %, and severe in 8.33 %.
Only in a euthanized bovine, and for that, the clinical course
was shortened, that there was no labeling in the IHC.
Ganglia cervical spinal inflammation was mild in 25 % of
evaluated cases, moderate in 50 %, and severe in 25 %.
Necrosis and neuronophagia in 50 % were mild, moderate in
25 %, and severe in 25 %. The Nageotte nodules were rare in
25 % and occasional in 75 % of the evaluated ganglia. There
was citoplasmatic vacuolation in all the spinal cervical ganglia
evaluated, mild in 50 % of the cases, and moderate in 50 %.
Negri bodies were observed in only one case (25 %). This
same case presented moderate labeling in IHC for rabies in
this ganglion.
The mononuclear inflammation of the thoracic spinal gan-
glia was considered mild in 50 % of the analyzed ganglia and
moderate in 50 %. Necrosis and neuronophagia was mild in
50 % of the evaluated ganglia and moderate in 50 %. Nageotte
 J. Neurovirol.

Fig. 2 Spinal ganglion, bovine,

and rabies. Neuronophagia
characterized by the presence of
the mononuclears cells
interspersed to the fragments of
the pericaryon. HE

nodules were rare in all thoracic spinal ganglia. These showed
mild vacuolation (50 %) or moderate (50 %). IHC was posi-
tive in only one evaluated case (50 %), and this labeling was
mild.
Trigeminal ganglia
Trigeminal ganglion was collected in 33 (68.75 %) of the
bovines with rabies. The mononuclear inflammation was
present in all trigeminal ganglia analyzed (Fig. 4), and it was
considered mild in 27.27 %, moderate in 54.55 %, and severe
in 18.18 %. Neuronophagia and neuronal necrosis were also
present in 100 % of trigeminal ganglia analyzed, and it was
mild in 39.4 %, moderate in 42.42 %, and severe in 18.18 %.
Nageotte nodules were observed in 87.88 % of trigeminal
ganglia evaluated, and it was rare in 72.41 %, occasional in
24.14 %, and frequent in 3.45 %. Negri bodies were present in
54.55 % of these ganglia, and mild vacuolation of the peri-
karyon of the neurons was observed in 72.73 %.
IHC in trigeminal ganglia was positive in 90.9 % of the
cases. In these positive cases, labeling was mild in 26.67 %,
moderate in 36.67 %, and severe in 36.67 %. In two of the

Fig. 3 Spinal ganglion, bovine,

and rabies. Nageotte nodules
characterized by hypertrophy of
the satellite cells in locals of loss
in the neuronal body (arrows).
HE
J. Neurovirol.

Fig. 4 Histology and immunohistochemistry of the ganglia in the bovine rabies. Intensity of the microscopic alterations and the immunodetection of the
rabies antigen

negative cases, there were a few neurons in the analyzed was also present in 33.33 % of these ganglia. IHC was positive
sample, and the other negative case was euthanized shortly in 41.67 % of the stellate ganglia analyzed (Fig. 6b) with mild
after the clinical outcome. labeling in 80 % and moderate in 20 %.

Stellate ganglia Celiac ganglia

In all of the 12 bovines that had stellate ganglia collected
(25 % of total), mononuclear inflammation was observed in
these tissues (Fig. 4), and in the ganglia, the inflammation was
mild in 25 %, moderate in 50 % and severe in 25 % (Fig. 6a).
Neuronophagia and necrosis were observed in 33.33 % of the
evaluated stellate ganglia, and the lesion was considered mild
in 50 % and moderate in 50 %. Mild citoplasmatic vacuolation
All 12 bovines that had celiac ganglia collected (25 % of the
total) had mononuclear inflammation in these tissues (Fig. 4);
these lesions were mild in 41.67 % of the ganglia, moderate in
33.33 %, and severe in 25 %. In 50 % of the evaluated celiac
ganglia, there was also neuronophagia and necrosis, which
were mild in 83.33 % of the cases and moderate in 16.67 %. In
25 % of the celiac ganglia analyzed, there was mild

Fig. 5 Spinal ganglion, bovine,

and rabies. Immunodetection of
the rabies antigen in the
perycarion and in neuron
extension. Streptavidin–biotin–
peroxidase, Vector® NovaRED
chromogen
 J. Neurovirol.

Fig. 6 Stellate ganglion, bovine, and rabies. a Severe mononuclear ganglioneuritis. HE. b Immunodetection of the rabies antigen in the perycarion.
Streptavidin–biotin–peroxidase, Vector® NovaRED chromogen

citoplasmatic vacuolation. IHC was positive in two ganglia
(16.67 %), but the labeling was mild.
Adrenals
Adrenals were collected from 17 (35.42 %) of evaluated
bovines with rabies. At the histopathological examination,
there was mononuclear inflammation surrounding the nerves
and vessels in adrenal medulla in all cases (Fig. 7a). This
inflammation was considered mild in 41.18 % of the cases,
moderate in 47.06 %, and severe in 11.76 %. Hyaline glob-
ules (88.24 %) were also observed in the pheochromocytes’
cytoplasm; most of them were present both in the light and in
the dark zone (86.67 %) of the adrenal medulla. However, in
one case, these were found only in the light zone (6.67 %),
and in another, only one in the dark zone (6.67 %). The
hyaline globules were evidenced by PAS and did not mark
in IHC for rabies. These hyaline globules were rare in
66.67 %, occasional in 20 %, and frequent in 13.33 %. In
both cases (11.77 %) in which they were not observed, these
globules were in an advanced state of autolysis. One of the
evaluated adrenal (5.88 %) showed strong positive immuno-
histochemical markers in the cytoplasm of pheochromocytes
(Fig. 7b).
Pituitaries
The pituitary was collected in 20 (41.67 %) of the evaluated
bovines with rabies. There was mononuclear inflammation in
the neurohypophysis in 85 %. Of these 85 %, the lesion was
considered mild in 29.41 %, moderate in 58.82 %, and severe
in 11.77 %. In the neurohypophysis, IHC was positive in only
one case (5 %) (Fig. 8), with a mild marker probably related to
extensions of neurons. In the pars intermedia of the adenohy-
pophysis, there was inflammation in 55 % of the cases. Of
these 55 %, the lesion was mild in 63.64 % and moderate in
36.36 %. IHC was positive in only one case (5 %) (Fig. 9) with
moderate labeling observed in the cytoplasm of basophil cells.
In the pars distalis of the adenohypophysis, there was inflam-
mation in 15 %; of these 15%, 66.67 % was mild and 33.33 %
was moderate. There was no immunolabeling of this portion
in all cases evaluated.

Fig. 7 Adrenal medulla, bovine, and rabies. a Mononuclear inflammation surrounding the vessels in adrenal medulla. HE. b Immunohistochemical
labeling for rabies in the citoplasm of the feocromocites. Streptavidin–biotin–peroxidase, Vector® NovaRED chromogen
J. Neurovirol.

Fig. 8 Neurohypophysis, bovine,

and rabies. Immunodetection of
the rabies antigen, probably
related to extensions of neurons.
Streptavidin–biotin–peroxidase,
Vector® NovaRED chromogen

Nerves degeneration was moderate in one (7.69 %) case and mild in

Isquiadic nerves were collected from 14 (29.17 %) bo-
vines with rabies. Two of them (14.29 %) had moderate
inflammation and four (28.57 %) had mild inflamma-
tion. There was still wallerian degeneration, evidenced
by IHC for neurofilament, and in two (14.29 %) of
these cases, this change was moderate and in four
(28.57 %) mild.
The brachial plexus was collected from 13 (27.8 %) bo-
vines, and moderate inflammation was observed in one
(7.69 %) case and mild in three (23.08 %). Wallerian
two (15.38 %) cases.
Other tissues evaluated
Injuries or antigens of the rabies virus in the histopathology
and IHC that were performed on other tissues were not found.
Discussion
In all bovines with rabies analyzed in this study, mononuclear
inflammation of variable intensity was observed in the

Fig. 9 Pars intermedia of the

adenohypophysis, bovine, and
rabies. Immunodetection of the
rabies antigen in the cytoplasm of
basophil cells. Streptavidin–
biotin–peroxidase, Vector®
NovaRED chromogen

trigeminal ganglia, spinal (lumbar, cervical, and thoracic),
celiac, and stellate and in the adrenal medulla. Similar inflam-
mation was also observed in 85 % of neurohypophysis ana-
lyzed, in 55 % of the pars intermedia, and in 15 % of the pars
distalis of the adenohypophysis. Inflammation has been fre-
quently associated with the presence of the rabies antigen in
studies in human tissues outside the CNS (Bagó et al. 2005;
Jackson et al. 1999; Jogai et al. 2002; Tobiume et al. 2009),
although the presence of these lesion is not considered specific
(Jackson et al. 1999). In bovines of the present study, this
association was common only in the spinal and trigeminal
ganglia, as described in humans. In stellate and celiac ganglia,
adrenal medulla, neurohypophysis, and pars intermedia of the
adenohypophysis, the presence of antigens in inflamed areas
was less frequent. The inflammation observed in IHC-
negative cases in this study may also be associated with rabies
antigen, but in insufficient quantity for IHC detection. The
shorter clinical evolution of the bovine disease (Barros et al.
2006; Pedroso et al. 2009) related to human (Jackson et al.
1999; Jogai et al. 2002) might be a factor that will justify the
smaller amount of antigen in tissues outside the CNS. There is
no correlation between severity of disease and presence of
rabies antigen or immune infiltration in specific tissues outside
of the CNS. In the majority of bovines in this study, accurate
data of clinical evolution were not informed by the farmers.
Nevertheless, in the present study, a euthanized bovine IHC
was negative even in a local where immunolabeling is fre-
quent (trigeminal ganglia and lumbar spinal), indicating that
the time of disease evolution has been an important factor.
However, some studies in humans found no correlation be-
tween clinical evolution and centrifuge migration of the virus
(Jackson et al. 1999; Jogai et al. 2002).
The collection of the trigeminal ganglion has been recom-
mended in suspected cases of bovine (Pedroso et al. 2009;
Rech et al. 2006) and human rabies (Van Gehuchten and Nelis
1900), and data of the cases presented here support this
statement. In 33 of the analyzed ganglia in this study, inflam-
mation was observed in 100 % and immunodetection of rabies
antigen in 90.9 %. These findings are similar to those de-
scribed in the literature that found inflammation in 95 % (Rech
et al. 2006) and immunohistochemical markers in 100 %
(Pedroso et al. 2009). In this study, only three bovines with
rabies were negative on IHC of trigeminal ganglia; in two of
these cases, the samples showed a small number of neurons.
The other negative bovine had the clinical course of the
disease abbreviated as it was euthanized. In this case, negative
immunohistochemical labeling was probably due to the lower
spread of the virus, due to the interruption of the course of the
disease (Barros et al. 2006).
Almeida et al. (2002) reported that several attacks by
D. rotundus in bovines included muzzle and ears. These
regions receive motor innervation by the facial nerve, which
can serve as a pathway for the rabies virus to trigeminal
J. Neurovirol.
ganglion neurons, once they are involved in reflex activity
in neurons of the nucleus of the facial nerve. Another poten-
tial pathway for the virus to the trigeminal ganglion is by the
centripetal migration from nerve endings of their sensory
neurons in the skin. This sensory pathway has been demon-
strated in some studies (Jackson 2009; Jackson et al. 2010;
Rossiter et al. 2009). The trigeminal ganglion neurons form
the general afferent sensory pathway of the head tissues, so
these injuries can cause reduction or loss of skin sensation
and loss or reduction in reflections on the head (De Lahunta
and Glass 2009). Inflammatory and necrotic lesions observed
in trigeminal ganglia of bovines in this study may justify
these signals in bovine rabies (Barros et al. 2006; Pedroso
et al. 2009).
The spinal ganglia showed similar results to those found in
the trigeminal ganglia, particularly in the lumbar region.
These had inflammation in all cases and immunohistochemi-
cal labeling in 92.31 %, which was not positive only in bovine
euthanized. These data demonstrate that the collection of
lumbar spinal ganglia may contribute to the diagnosis of
bovine rabies, as it was recognized for humans (Van
Gehuchten and Nelis 1900), and it is recommended to collect
the trigeminal ganglion for bovines (Pedroso et al. 2009; Rech
et al. 2006).
One explanation for the lesions observed in spinal ganglia
may be the arrival of the rabies virus in axonal retrograde
pathway. Among the major points of attack of D. rotundus is
the bovines’ dorsum (Almeida et al. 2002; Pereira et al. 2010).
The virus through the neuromuscular junction affects motor
neurons in the spinal cord when inoculated in this region
(Jackson 2008). These connect to the interneurons that synap-
se with the neurons of the spinal ganglia (De Lahunta and
Glass 2009).
Entry of rabies virus into the sensory pathways has also
been demonstrated (Jackson et al. 2010; Rossiter et al. 2009).
In this way, the virus reaches the spinal ganglion initially by
centripetal migration (Rossiter et al. 2009). The neurons of the
spinal ganglia connect to the interneurons involved in reflex
activity or with neurons that project cranially to the brainstem,
as part of the pathway of conscious perception (De Lahunta
and Glass 2009), by this way providing access to the virus to
the CNS. This sensory pathway may be important in rabies
transmitted by vampire bats, since these vectors produce more
superficial bites compared to terrestrial vectors, affecting
mainly the skin and subcutaneous tissue, where the sensory
and autonomic innervation is more abundant (Jackson 2010).
The neurons of the spinal ganglia form the afferent propri-
oceptive general, and thus, the inflammatory and necrotic
lesions most intense in these ganglia observed in bovines in
this study can be correlated with signs of reduction and loss of
skin sensation, ataxia, and incoordination of the hindlimbs
(De Lahunta and Glass 2009). These signs are often observed
in bovine rabies (Barros et al. 2006; Pedroso et al. 2009).
J. Neurovirol.
Immunohistochemical labeling was also observed in
41.67 % of the analyzed stellate ganglia and in 16.67 % of
the celiac ganglia. Lesions in these ganglia were mainly
inflammatory, and necrotic lesions were milder and less fre-
quent than those observed in other ganglia. A pathway for the
virus to these ganglia requires further studies. However, these
ganglia receive innervation in sensory neurons of the spinal
ganglion and sympathetic neurons of the spinal column inter-
mediate (Loukas et al. 2010; Nozdrachev et al. 2003), which
could provide access to the virus. As a reflection of this
innervation, blockages in the stellate ganglia are used, for
example, in the control of cardiac pain, and may have the
effect of reducing the sympathetic inervation in the heart
(Nozdrachev et al. 2003). However, the celiac ganglia block
is used to control pain in abdominal organs such as the
pancreas and may have the additional effect of reducing
sympathetic inervation in the abdominal viscera (Loukas
et al. 2010). Signals related to these blocks have not been
described in bovine rabies, and this may reflect the low
intensity of the lesions in these locations.
All 17 adrenal evaluated showed mononuclear inflamma-
tion in the medullary region surrounding mainly the vessels
and nerves. Similar lesions have been reported in cases of
rabies in humans (Almeida et al. 1986; Jackson et al. 1999;
Jogai et al. 2002; Tobiume et al. 2009). In animals, similar
lesions were also reported in V. vulpes (foxes), Mephitis
mephitis (skunks) (Balachandram and Charlton 1994), and
bovines (Mesquita et al. 2011). These results indicate that the
evaluation of adrenal is also important in assisting the diag-
nosis and understanding of the pathogenesis of the disease.
Hyaline globules in the cytoplasm of pheochromocytes were
observed in 88.24 % of the bovines’ adrenals with rabies in
this study. Similar globules have been described in several
species of laboratory animals (McConnel and Talley 1977),
humans (Hart and Cyrus 1968; Sugie et al. 2005), and bo-
vines (Mesquita et al. 2011). Although it was a nonspecific
injury, a greater number of these inclusions have been linked
to neurological diseases in humans (Hart and Cyrus 1968;
Sugie et al. 2005). These inclusions have been observed in
cases of rabies in humans (Almeida et al. 1986) and bovines
(Mesquita et al. 2011). Data from this study reinforce this
association.
The immunological detection of rabies virus antigens in
adrenal medulla was observed in only one case in this study. In
this case, there was intense labeling of the pheochromocytes’
cytoplasm. Works with IHC in humans also describe the
rabies antigen mainly in adrenal medulla, with additional
marking in nervous plexus of the capsule (Jackson et al.
1999; Jogai et al. 2002; Tobiume et al. 2009). The antigen
has also been detected in pheochromocytes and nerve fibers in
the adrenal medulla in V. vulpes (foxes) and M. mephitis
(skunks), experimentally infected with rabies virus
(Balachandram and Charlton 1994).
The access of the virus to the adrenal medulla is not known.
However, this region receives innervation primarily from
sensory neurons of the spinal ganglion and sympathetic neu-
rons of the middle column of the spinal cord (Zhou et al.
1991). Thus, the possibility of neurons of the dorsal horn of
the spinal cord, infected through motor or sensory pathway,
disseminating the virus in these circuits is feasible. The same
spinal neuron can make connections with sensory neurons of
the skin and visceral sensory neurons. This phenomenon is
called referred pain or cutaneous visceral reflex (De Lahunta
and Glass 2009; Dirksen 2008). Although the exact route of
access of the rabies virus to the adrenal medulla requires
further studies, such connections could justify access of the
virus in bovines in the present study. The fibers innervating
adrenal medulla project themselves through the celiac gangli-
on (Podlech et al. 1996). This observation of lesions and
immunohistochemical markers in both areas demonstrates that
this circuit was also affected in viral spread in these bovines.
In the neurohypophysis, there was mononuclear inflamma-
tion in 85 %, with a positive case in IHC. In the pars
intermedia of the adenohypophysis, there was inflammation
in 55 % of the cases with a positive case in IHC. In the pars
distalis of the adenohypophysis, there was inflammation in
15 %, without immunohistochemical markers. Similar inflam-
mation in the neurohypophysis has been described in cattle
with rabies in 30.77 % of cases (Rech et al. 2006). With
respect to rabies in humans, immunohistochemical labeling
in the anterior lobes (pars distalis) and posterior (pars
intermedia and pars nervosa) in the pituitary can occur
(Tobiume et al. 2009). Literature data associated with the
present study indicate the importance of observing also the
pituitary in cases of suspected rabies.
Conclusions
Data from this study indicate that, in bovines with neurolog-
ical signs, in addition to the observation of the complex
pituitary, rete mirabile, and trigeminal ganglia, the analysis
of other ganglia, particularly the lumbar spinal and adrenal
medulla, can contribute to the diagnosis of rabies and under-
standing of the clinical signs and the pathogenesis of the
disease.
Acknowledgments The authors acknowledged the Fundação de
Amparo à Pesquisa do Estado de Minas Gerais (Fapemig) for funding
this study and the Coordenação de Aperfeiçoamento de Pessoal de
Nível Superior—CAPES by master’s scholarship. Employees of the
Instituto Mineiro de Agropecuária (IMA) are also acknowledged for
the results of direct immunofluorescence for rabies and access to cases
of Perdões—Minas Gerais. The authors also acknowledged the mem-
bers of the examining body of qualification and dissertation, C Hirsch,
PFV Peixoto, DL Raymundo, SC Salvador, and RO Vasconcelos, for
their contributions.

Conflict of interest The authors declare that they have no conflict of
interest.
References
Ajayi BB, Rabo JS, Baba SS (2006) Rabies in apparently healthy dogs:
histological and immunohistochemical studies. Niger Postgrad Med
J 13:128–134
Almeida HO, Teixeira VPA, Oliveira G, Brandão MC, Gobbi H (1986)
Medulite supra-renálica em casos de raiva humana. Mem Inst
Oswaldo Cruz 81:439–442
Almeida EO, Moreira EC, Naveda LAB, Herrmann GP (2002) Combate
ao Desmodus rotundus rotundus (E. Geoffroy, 1810) na região
cárstica de Cordisburgo e Curvelo, Minas Gerais. Arq Bras Med
Vet Zootec 54:1–13
Bagó Z, Revilla-Fernández S, Allerberger F (2005) Value of immunohis-
tochemistry for rapid ante mortem rabies diagnosis. Int J Infect Dis
9:351–352
Balachandram A, Charlton K (1994) Experimental rabies infection of
non-nervous tissues in skunks (Mephitis mephitis) and foxes (Vulpes
vulpes). Vet Pathol 31:93–102
Bansal K, Singh CKK, Ramneek, Sandhu BS, Deka D, Dandale M, Sood
NK (2012) Antemortem diagnosis of rabies from skin: comparison
of nested RT-PCR with TaqMan real time PCR. Braz J Vet Pathol 5:
116–119
Barros CSL, Driemeier D, Dutra IS, Lemos RAA (2006) Doenças do
sistema nervoso de bovinos no Brasil. Coleção Vallée, São Paulo
De Lahunta A, Glass E (2009) Veterinary neuroanatomy and clinical
neurology. Elsevier, St. Louis
Delpietro HA, Larghi OP, Russo RG (2001) Virus isolation from saliva
and salivary glands of cattle naturally infected with paralytic rabies.
Prev Vet Med 48:223–228
Dirksen G (2008) Sistema Digestivo. In: Dirksen G, Gründer H-D, Stöber
M (ed) Rosenberger—Exame clínico dos bovinos, 3rd edn.
Guanabara, Rio de Janeiro, pp 166–228
Hart MN, Cyrus A Jr (1968) Hyaline globules of the adrenal medulla. Am
J Clin Pathol 49:387–391
Jackson AC (2008) Rabies. Neurol Clin 26:717–726
Jackson AC (2009) Update on rabies diagnosis and treatment. Curr Infect
Dis Rep 11:296–301
Jackson AC (2010) Rabies pathogenesis update. Rev Pan-Amaz Saude 1:
167–172
Jackson AC, Ye H, Phelan CC, Ridaura-Sanz C, Zheng Q, Li Z, Wan X,
Lopez-Corella E (1999) Extraneural organ involvement in human
rabies. Lab Investig 79:945–951
Jackson AC, Kammouni W, Zherebitskaya E, Fernyhough P (2010) Role
of oxidative stress in rabies virus infection of adult mouse dorsal root
ganglion neurons. J Virol 84:4697–4705
Jogai S, Radotra BD, Banerjee AK (2002) Rabies viral antigen in extra-
cranial organs: a post-mortem study. Neuropathol Appl Neurobiol
28:34–338
Loukas M, Klaassen Z, Merbs W, Tubbs RS, Gielecki J, Zurada A (2010)
A review of the thoracic splanchnic nerves and celiac ganglia. Clin
Anat 23:512–522
McConnel EE, Talley FA (1977) Intracitoplasmatic hyaline globules in
the adrenal medulla of laboratory animals. Vet Pathol 14:435–440
Mesquita LP, Abreu CC, Wouters F, Varaschin MS, Bezerra PS Jr (2011)
Inclusões intracitoplasmáticas hialinas na medular da adrenal de
bovinos. Arq Bras de Med Vet Zootec 63:247–249
J. Neurovirol.
Morais CF, Assis RVC (1985) Cardiac involvement in human rabies. Rev
Inst Med Trop Sao Paulo 27:145–149
Nozdrachev AD, Fateev MM, Jiménez B, Morales MA (2003) Circuits
and projections of cat stellate ganglion. Arch Med Res 34:106–115
Pedroso PMO, Pescador CA, Bandarra PM, Raymundo DL, Borba MR,
Wouters F, Bezerra PS Jr, Driemeier D (2008) Padronização da
técnica de imuno-histoquímica para raiva em amostras de tecido
do sistema nervoso central de bovinos fixadas em formol e
emblocadas em parafina. Pesqui Vet Bras 28:627–632
Pedroso PMO, Colodel EM, Pescador CA, Arruda LP, Driemeier D
(2009) Aspectos clínicos e patológicos em bovinos afetados por
raiva com especial referência ao mapeamento do antígeno rábico
por imuno-histoquímica. Pesqui Vet Bras 29:899–904
Pereira SN, Gitti CB, Cabral MMO (2010) Análise da distribuição da
região dos ferimentos provocados por morcegos hematófagos
Desmodus rotundus (Geoffroy, 1810) em bovinos sob condições
de campo. Arq Inst Biol 77:203–208
Podlech J, Hengerer F, Fleck M, Walev I, Falke D (1996) Replication of
herpes simplex virus type 1 and 2 in the medulla of the adrenal gland
after vaginal infection of mice. Arch Virol 141:1999–2008
Preuss MA, Faber ML, Tan GS, Bette M, Dietzschold B, Weihe E,
Schnell MJ (2009) Intravenous inoculation of associated rabies
virus causes lethal encephalopathy in mice through invasion of
the brain via neurosecretory hypothalamic fibers. PLoS Pathog
5:1–10
Rech RR, Rissi DR, Silva MC, Inkelmann MA, Barros CSL (2006)
Histomorfologia do gânglio de Gasser, da rete mirabile carotídea e
da hipófise de bovinos: estudo de 199 casos. Pesqui Vet Bras 26:
105–111
Rossiter JP, Hsu L, Jackson (2009) Selective vulnerability of dorsal root
ganglia neurons in experimental rabies after peripheral inoculation
of CVS-11 in adult mice. Acta Neuropathol 118:249–259
Singh H, Singh CK (2011) Ante-mortem diagnosis of rabies from body
secretions and mucosal impression smears in experimentally infect-
ed buffalo calves. Indian J Anim Sci 81:205–208
Sota CA (1993) Control of bovine paralytic rabies in Latin America and
the Caribbean. In: More attention to animal genetic resources. World
Animal Review. http://www.fao.org/docrep/v0600t/v0600T0a.
htm#control. Accessed 15 Jan 2012
Sugie M, Goto J, Kawamura M (2005) Increased norepinephrine-
associated adrenomedullary inclusions in Parkinson’s disease.
Pathol Int 55:130–136
Summers BA, Cummings JF, Lahunta A (1994) Inflammatory diseases of
the central nervous system. In: Summers BA, Cummings JF,
Lahunta A (eds) Veterinary neuropathology, 1st edn. Mosby, St.
Louis, pp 95–188
Tobiume M, Sato Y, Katano H, Nakajima N, Tanaka K, Noguchi A, Inoue
S, Hasegawa H, Iwasa Y, Tanaka J, Hayashi H, Yoshida S, Kurane I,
Sata T (2009) Rabies virus dissemination in neural tissues of autop-
sy cases due to rabies imported into Japan from the Philippines:
Immunohistochemistry. Pathol Int 59:555–566
Van Gehuchten A, Nelis C (1900) Les lesions histologiques de la rage,
chez les animaux et chez l’homme. Bulletin de L’Academie Royale
de Medecine de Belgique, Louvain
Vieira LFP, Pereira FG Sr, Galante AC, Castilho JG, Oliveira RN,
Brandão PE, Kotait I (2011) Detection of rabies virus
nucleoprotein-RNA in several organs outside the central nervous
system in naturally-infected vampire bats. Pesqui Vet Bras 31:
922–925
Zhou XF, Oldfield BJ, Livett BG (1991) Substance P-containing sensory
neurons in the rat dorsal root ganglia innervate the adrenal medulla. J
Auton Nerv Syst 33:247–254