Álvarez et al. BMC Veterinary Research (2021) 17:29 https://doi.org/10.1186/s12917-021-02745-9 RESEARCH ARTICLE Open Access Aquaporin-4 protein expression in normal canine brains Patricia Álvarez1, Ester Blasco2, Martí Pumarola2,3 and Annette Wessmann1* Abstract Background: Aquaporin-4 (AQP4) is in growing recognition as potential marker for cancer progression, differentiation and therapeutic intervention. No information is available about AQP4 expression in the normal canine brain. The aim of this histopathological study is to confirm the presence of AQP4 by immunohistochemistry technique in a group of non-pathological canine brains and to describe its expression and distribution across the brain. Results: Twelve non-pathological canine brains of various ages (ranging from 21 days to 17 years) and breeds were included in the study. Immunohistochemical expression of AQP4 was analyzed using formalin-fixed paraffin-embedded brain tissue sections. The findings were correlated between AQP4 expressing cells and astrocytes using glial fibrillary acidic protein (GFAP). AQP4 expression was more marked in the astrocyte foot processes of subpial, perivascular and periventricular surfaces in all specimens. The majority of the canine brain sections (9/12) presented with an AQP4 predilection for white matter tracts. Interestingly, the two youngest dogs (21 days and 3 months old) were characterized by diffuse AQP4 labelling in both grey and white matter tracts. This result may suggest that brain development and ageing may play a role in the AQP4 distribution throughout the canine brain. Conclusions: This is the first study to describe immunohistochemical distribution of AQP4 in normal canine brains. The AQP4 expression and distribution in non-pathological canine brains was comparable to other species. Larger studies are needed to substantiate the influence of breed and ageing on AQP4 expression in the normal canine brain. Keywords: Aquaporin-4, Brain, Dog, Grey matter, Immunohistochemistry, Non‐pathological, White matter, Brain development Background such as neuromyelitis optica spectrum disorders, tumor types Aquaporins (AQPs) are a family of transmembrane water- and grades, tumor proliferation, migration, angiogenesis and channel proteins distributed in membranes of various bio- tumor-associated edema in people [2–6]. logical cells of animals and humans mainly facilitating trans- Most information about CNS AQP4 is available in port of water between cells [1]. One of the most prominent humans, rats and mice [7, 8]. In normal circumstances, aquaporins in the central nervous system (CNS) is AQP4 is expressed more abundantly in the astrocyte Aquaporin-4 protein (AQP4) [1]. AQP4 plays a crucial role foot processes of the perivascular, subpial and subepen- in maintaining water homeostasis, cell migration and neu- dymal areas and in the basolateral membrane of epen- roexcitation in the brain [1]. The AQP4 expression in the dymal cells. These locations are special sites of major CNS has been associated with neuroinflammatory conditions fluid transport where AQP4 is responsible for water bal- ance regulation in and out the CNS [1]. Similar locations * Correspondence: annette.wessmann@scarsdalevets.com have been demonstrated in other species including 1 Neurology and Neurosurgery Service, Pride Veterinary Centre, Derby, UK Full list of author information is available at the end of the article Macaca Fascicularis and Psittacus erithacus [9, 10]. © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Álvarez et al. BMC Veterinary Research (2021) 17:29 Page 2 of 8 Only scarce information is available about AQP4 dis- periventricular surfaces also displayed a strong positive tribution in diseased canine brains. Klemens et al. [11] immunoreaction in the basolateral membrane of the described the expression of AQP4 in canine cerebellar ependymal cells and the associated bands of the sube- samples of canine distemper virus infected dogs. This pendymal astrocyte foot processes in a characteristic study showed that there was a more severe loss of AQP4 branching pattern. The intensity of the labeling was in acute distemper lesions when compared to subacute gradually fading towards the neuropil. APQ4 immunore- or chronic lesions in these dogs. Spitzbarth et al. [12] action was absent in choroid plexus cells, neurons or analyzed AQP4 expression using tissue microarray tech- other glial cells aside from astrocytes. The common nique searching for AQP4 expression patterns in differ- areas of greater AQP4 expression in the canine brain are ent types of canine CNS neoplasms and highlighted illustrated in Fig. 1. AQP4 as a novel marker helping to discriminate be- Nine out of 12 dogs (N03-N11) showed a clear distinc- tween canine astrocytoma and oligodendroglioma. Con- tion between the grey and white matter. A predilection for sidering the growing recognition of AQP4 in the CNS AQP4 immunoreaction was concentrated in the white and its potential diagnostic and therapeutic implications, matter tracts and the subcortical cerebral transition area knowledge about the distribution of AQP4 in normal between the grey and white matter. A longitudinal pattern canine brain is urgently needed as basis for future stud- alongside nerve fibers involving mostly the white matter ies. The purpose of this study is first, to confirm the was identified in the subcortical white matter, corona presence of AQP4 in the normal canine brain using im- radiata, internal capsule and corpus callosum. The basal munohistochemistry technique and secondly, to describe nuclei showed variable expression from absent to mild the AQP4 distribution in non-pathological brains in a patchy and punctuated immunoreaction surrounding the group of dogs of different age and breed (Table 1). vascularization. Interestingly, the 2 youngest dogs (21 days old N01, 3 months old N02, 2/12) showed a slightly differ- ent presentation. All encephalic structures examined for Results N01 and N02 (2/12) (Table 2) were characterized by a AQP4-expressing cells were confirmed using immuno- generalized homogenous strong immunoreaction involv- histochemistry in all canine brains included in this study. ing both grey and white matter. In the neocortex of these The expression of AQP4 in our canine population was 2 dogs, AQP4 expressing cells were subjectively more con- examined by semiquantitative descriptive analysis. The centrated in the deeper laminae of the cortical grey matter complete data is contained in Table 2. AQP4 immuno- and located primarily surrounding blood vessels of any histochemistry was co-localized with GFAP immunohis- size in a branching-fibrillar pattern. To a lesser extent, im- tochemistry confirming the common astrocytic origin. munoreaction was also present in the perineuronal astro- All twelve canine brains showed the greatest AQP4 cyte processes. The canine brain N12 (17 years old) immunoreaction in the plasma membrane of the peri- showed a well-distinguished variation compared to the vascular astrocyte foot processes in a branching-fibrillar majority of the population (N03-N11). A greater and dif- pattern. The glia limitans externa, located between the fuse AQP4 immunoreaction was identified in the grey CNS parenchyma and the pia mater, showed a well- matter of the neocortex of the frontal and temporal areas. defined homogenous monolayer labeling. The The archicortex was largely positive for AQP4 expres- sion in all canine brains (N01-N12). The labeling was Table 1 Canine brains included in this study denser in the inner margin of the dentate gyrus and ex- Case Number Breed Age ternal margin towards the meninges in the Cornu N01 Cross breed 21 days Ammonis surrounding the neuronal bodies. Interest- N02 Bichon Maltese 3 months ingly, only in the canine brain N12 (17 years old) the de- N03 White Swiss Shepherd dog 3 months gree of AQP4 immunostaining around the Cornu N04 Cocker Spaniel 5 months Ammonis was increased. N05 Pitbull 3 years The paleocortex (piriform lobe) and parahippocampal gyrus presented a moderate diffuse and homogeneous la- N06 Cocker Spaniel 4 years belling in all cases (N01-N12). N07 Miniature Schnauzer 5 years The brainstem showed a similar pattern in 9 canine N08 Bull Terrier 5 years brains (N03-N12) with a clear predilection for white N09 Cross breed 13 years matter tracts. The mesencephalon, the substantia nigra N10 Yorkshire Terrier 14 years and interpeduncular nuclei had partial immunoreaction. N11 Scottish Terrier 15 years The red nuclei and oculomotor nuclei were clearly less positive. At the pons and medulla oblongata, immunore- N12 Cross breed 17 years action appeared to be denser in the white matter tracts Álvarez et al. BMC Veterinary Research (2021) 17:29 Page 3 of 8 Table 2 Expression of AQP4 immunohistochemistry in the 12 canine brains following the semiquantitative classification system N01 N02 N03 N04 N05 N06 N07 N08 N09 N10 N11 N12 Neocortexa GM ++ ++ +/++ -/+ -/+ n/ai -/+ -/+ -/+ -/+ -/+ +/++ WM +++ +++ +++ +++ +++ n/a ++ +++ +++ +++ +++ +++ b Basal nuclei GM ++ ++ -/+ -/+ -/+ n/a +/++ +/++ -/+ + + ++ WM ++ ++ ++ ++ +++ n/a ++ ++ ++ ++ ++ ++ c Archicortex GM ++ ++ + n/a +/++ -/+ -/+ +/++ + n/a + ++ WM +++ +++ ++ n/a ++ - - -/+ ++ n/a +++ +++ Paleocortexd ++ ++ n/a + n/a n/a ++ ++ +/++ +++ ++ +/++ Diencephalone GM ++ ++ -/+ n/a -/+ n/a - - n/a - + + WM +++ +++ ++ n/a ++ n/a ++ ++ n/a + +++ +++ Midbrain Ponsf GM ++ ++ n/a +/++ + + -/+ - -/+ - - - WM +++ +++ n/a +++ ++ ++ ++ ++ ++ +/++ ++ +/++ Medulla oblongatag GM ++ ++ n/a +/++ + -/+ - -/+ -/+ - -/+ - WM +++ +++ n/a +++ +++ ++ ++ ++ +++ +++ ++ ++ Cerebellumh GM ++ ++ n/a ++ +/++ -/+ - -/+ -/+ - - +/++ WM +++ +++ n/a +++ +++ ++ ++ +++ +++ +++ +++ +++ Abbreviations: GM Grey matter, WM White matter, n/a Not applicable; -, negative; +, mild staining; ++, moderate staining; +++, strong staining; Neocortexa: frontal, parietal and temporal areas and subcortical white matter and corpus callosum; Basal nucleib: Striatum body and internal capsule; Archicortexc: hippocampal formation, fimbria and fornix; Paleocortexd: piriform lobe; Diencephalone: epithalamus (habenula nuclei), thalamic nuclei and white matter tracts; Midbrain/ponsf: oculomotor nuclei, red nuclei, substantia nigra, periaqueductal grey matter and white matter tracts (crus cerebri, cerebellar tracts, transverse pontine fibers); Medulla oblongatag: caudal cranial nerves nuclei, white matter tracts (pyramids, medial longitudinal fasciculus, trapezoid body); Cerebellumh: cerebellar cortex and subcortical white matter; n/ai: Not applicable, corresponds to an incomplete evaluation due to the lack of tissue or low immunoreaction at the level of the cerebellar peduncles, crus cerebri, brains a crucial role in maintaining water homeostasis, transverse pontine fibers and pyramids. cell migration and neuroexcitation [1]. The cerebellar cortex showed a similar immunoreac- The distribution of AQP4 showed many similarities to tion in 9 specimens (N03-N11). AQP4 expressing cells reports of other species. All examined canine brains exhibited a predilection for the cerebellar white matter. were characterized by the strongest immunoreaction in A low AQP4 immunoreaction was sporadically accumu- the plasma membrane of astrocytes, particularly in the lated in the Purkinje layer in a perineuronal pattern and astrocytic foot processes of the perivascular space, the in the glomeruli of the granular layer. In the canine subpial and subependymal areas, and in the basolateral brain N12, a linear labeling of the molecular layer of the membrane of the ependymal cells, which are specific re- cerebellar cortical grey matter was present contrary to gions of major fluid transport. This is a characteristic the common findings. distribution comparable to other species and supports The habenular nuclei was only examined in three the assumption that AQP4 plays also in the canine a specimens of varied age including a young (N02, 3 major role as a water-channel protein maintaining the months old), adult (N08, 3 years old) and senior dog perivascular volume and cerebral blood perfusion as pre- (N11, 15 years old) and showed an equally strong AQP4 viously reported in other species [7, 9, 10, 13, 14]. expression and diffuse distribution. A specific polarized distribution of AQP4 in astrocytes A comparison of the AQP4 immunohistochemistry dis- towards the astrocytic foot processes was seen in our ca- tributions in the above described representative encephalic nine population. A specific anchoring mechanism is structures of dogs of varied ages is illustrated in Fig. 2. found in this location which is called the dystrophin- associated protein complex [15]. The dystroglycan was Discussion found to interact with laminin and dynamin, which is This is the first study to confirm the expression and dis- necessary for the regulation of AQP4 internalization and tribution of AQP4 using immunohistochemistry in nor- therefore explaining the asymmetric enrichment towards mal canine brains. Previous veterinary publications have perivascular astrocyte foot processes [16]. exclusively investigated the presence of CNS AQP4 im- AQP4 immunoreaction was further absent in neurons, munohistochemistry in pathological conditions [11, 12], other glial cells and choroid plexus in our canine popula- but no information was available so far in non- tion. This was observed similarly in other species too [7– pathological canine brains. This finding supports the as- 9, 14], apart from a reported low expression of AQP4 in sumption that AQP4 plays also in the normal canine the choroid plexus in rat and human brain tissue [4, 13]. Álvarez et al. BMC Veterinary Research (2021) 17:29 Page 4 of 8 Fig. 1 The immunohistochemistry of the main areas of AQP4 expression are represented with a schematic illustration (modified from Papadopoulos et al [1]): Brain surface-glia limitans externa, Ventricular lining-glia limitans interna and blood brain barrier A moderate diffuse homogeneous AQP4 immuno- same regions [18]. Thus, the results of our study sug- staining was seen in the paleocortex throughout our ca- gests a similar significance in canine brains. nine population, whereas a more concentrated labelling The distribution of AQP4 towards the white matter in the neocortex towards the subcortical white matter was evident in the majority of dogs (N03-N11, aged be- and the transition area between the grey and white mat- tween 3 months and 15 years). Specifically, the expres- ter was seen in the majority of the specimens (9/12). It sion was located in the subcortical white matter of both can be suspected that this AQP4 distribution is related neocortex and cerebellum. The immunoreaction pattern to the primitive origin of the paleocortex compared to was parallel to the nerve fibers. It is suspected that this the neocortex. The paleocortex has 3 cortical grey mat- particular distribution might play a role in maintaining ter layers with its white mater located in an external dis- intercellular junctions at the nodes of Ranvier [7]. Fur- position, whereas the neocortex comprises 6 layers of ther, the white matter is characterized by a considerable cortical grey matter with the white matter located in a concentration of fibrillary astrocytes compared to the subcortical disposition [17]. grey matter [8]. Fibrillar astrocytes have highly branched The hippocampal formation presented with increased processes compared to protoplasmic astrocytes [19]. It is AQP4 immunostaining in the dentate gyrus and Cornu considered that this distinct characteristic may account Ammonis in all the canine brains of this study. This is for the higher concentration of positive labeling within not unexpected considering published findings in the the white matter. On the contrary, monkeys and adult mice brain where AQP4, was shown to be associated rats have been described to have a greater AQP4 expres- with the facilitation of rapid water fluxes required for sion in the grey matter instead [9, 13]. One of their ex- maintaining K + homeostasis during electrical activity in planations was its higher concentration of blood vessels Álvarez et al. BMC Veterinary Research (2021) 17:29 Page 5 of 8 Fig. 2 CNS AQP4 immunoreactivity comparison between dogs of varied ages at a different representative encephalic structure in transverse sections. Parietal/temporal cortex (1) (0.5 times magnification), Midbrain/pons (2) (0.6-1x), Archicortex with hippocampal formation (3) (2-2.5 times magnification) and Cerebellar cortex (4) (10-15 times magnification). The youngest dogs (a: N01 and N02) showed a generalized AQP4 labeling diffusively involving the neuroparenchyma. The adult and senior dogs (b and c: N05,N06, N10 and N11) presented greater expression of AQP4 in the subcortical WM and at the interface between major fluid compartments (periventricular and subpial surfaces). The hippocampal distribution in b and c was concentrated in the GM of the dentate gyrus and Amon’s horn compared to the white matter [9]. Although we did not area is needed in the formation of the blood brain bar- find increased AQP4 expression in the grey matter com- rier during the brain development and the control of the pared to the white matter in our study, increased AQP4 perivascular volume [21, 22]. As seen in the developing labelling was additionally concentrated in the grey/white mouse brain, AQP4 immunoreactivity varies between matter transition in most of the canine brains (9/12). brain regions according to neuronal differentiation [21]. The latter could be explained by a higher concentration This suggests that the homogeneous distribution of of capillaries and perivascular astrocyte foot processes AQP4 along the brain grey and white matter structures expressing AQP4 in this transition zone [20]. Moreover, in two of the youngest canine dogs (N01, N02) may be the grey and white matter transition area is where a related with postnatal brain development, neuronal mi- clearer distinction between protoplasmic versus fibrillary gration, blood brain barrier formation and the need of astrocytes is present, which may account for the pres- an adequate hydrated milieu to favor the neuronal shifts. ence of AQP4 immunostaining yet less strong compared Whereas the distribution of AQP4 towards the white to the white matter alone. matter seen in the majority of the canine population, There was an AQP4 distribution tendency towards the could be associated with AQP4’s primary role as water white matter with increasing age in our study popula- channel protein and therefore might be more limited to tion. The 2 youngest canine brains (21 days old N01, 3 regions of actual fluid exchange, such as perivascular months old N02) showed a generalized and homogenous (blood), subpial and periventricular (CSF) areas. AQP4 immunostaining in both, the grey and white mat- A different distribution pattern was seen in the canine ter, across the brain. The reasoning behind this particu- brain N12 when compared to the majority of the studied lar finding is uncertain. Previous studies in rats have canine brains (9/12). The AQP4 immunoreaction was explored whether increased AQP4 expression in this increased in the grey matter (cerebral frontal, parietal Álvarez et al. BMC Veterinary Research (2021) 17:29 Page 6 of 8 and cerebellar cortices and Cornu Ammonis) compared astrocytes of other canine circumventricular organs such to the white matter. One possible explanation may be as the area postrema, and thus supporting the important contemplated based on findings about dementia in contribution of this water-channel in the maintenance of people [23]. Owasil et al. [23] concluded that the expres- the homeostasis [30]. sion of AQP4 was associated with astrocytes in the Limitations of this study are the limited number of ca- brain, and that the distribution of these astrocytes in the nine brains of various ages and breeds. The aim of the grey and white matter was correlated with the patient’s current study was to confirm the presence of AQP4 in age and the severity of cerebral amyloid angiopathy. non-pathological brains and to identify tendencies of Here, demented patients seemed to have a higher con- AQP4 distribution in the canine brain, which was centration of AQP4 labeling in the cortical grey matter achieved. However, the limited number of cases can only compared to non-demented patients [23]. Interestingly, give an idea about the wider dog population. Larger the case N12 was the oldest dog with 17 years of age. In scale studies would be ideal to focus on particular as- dogs, the diagnosis of canine cognitive dysfunction syn- pects of AQP4 expression in normal canine brains such drome is based on behavioral signs and exclusion of as particular ages (young or senile group) or possible other medical conditions [24]. Further, it is known clin- breed specific findings. Another limitation of this study ically that the prevalence and severity of canine cognitive is the lack of clinical data. Inclusion criteria were brains dysfunction increases with the age [25]. Different publi- defined as ‘non-pathological’ based on histopathological cations have studied that dogs affected by progressive examination only. More clinical information would have cognitive impairment share certain histopathological been ideal to investigate our hypothesis that the unusual changes including alterations related to the structure of AQP4 distribution in the oldest dog (N12, 17 years old) certain proteins as in Alzheimer disease: neuronal loss, was caused by early or preclinical canine cognitive dys- astrocytosis, amyloid-β deposition and rarely neurofibril- function syndrome. lary tangles [26, 27]. As AQP4 is more abundantly expressed in astrocytes, AQP4 concentration is likely in- Conclusions creased in diseases resulting in astrocytosis, which is a This is the first study to confirm the presence AQP4 in hallmark of neuroinflammation. Thus, the AQP4 overex- non-pathological canine brains and to describe its distri- pression seen in this dog (N12) raises the suspicion that bution in normal canine brains using immunohisto- preclinical cognitive dysfunction might have been chemistry. APQ4 expression and distribution in the present. Further studies would be needed to investigate canine brain was comparable to other species including the influence of canine cognitive dysfunction syndrome humans. AQP4 was widely expressed in the astrocyte on the AQP4 distribution in the brain compared to se- cell plasma membrane, particularly in the foot processes nile brains and further, to possibly investigate AQP4 as a of the blood brain barrier and at the border between the potential immunomarker for early detection of canine CNS and the cerebrospinal fluid. A particular redistribu- cognitive dysfunction syndrome. We cannot exclude the tion tendency was observed for AQP4 expression with possibility that the increased grey matter AQP4 labelling ageing from grey matter to white mater in most dogs. noted in this particular brain (N12) could have also been Brain development as well as ageing may affect AQP4 an individual variant. Publications report a greater distribution throughout the canine brain. AQP4 expression in the grey matter of rats and mon- Knowledge about AQP4 expression and distribution in keys, yet this is not described for the dog [9, 13]. normal canine brains will aid the understanding of the The habenula was examined in 3 cases in which AQP4 importance of abnormal AQP4 expression in canine immunostaining was found to be homogeneous despite brains with different pathological conditions. the varied age (3 months to 15 years). The habenula is located in the dorsomedial border of the thalamus close Methods to the pineal gland, which is forming part of a group of Animals and specimen collection specialized structures defined as circumventricular or- A total of 12 canine brains were retrospectively evaluated. gans [28, 29]. The circumventricular organs are located The dogs died or were euthanized due to causes unrelated to mainly at the midsagittal line around the third and the CNS and donated to the Unit of Murine and Compara- fourth ventricles, often protruding into the lumen [30]. tive Pathology, Autonomous University of Barcelona, Spain. The singular characteristic of the circumventricular or- Owner agreement was obtained at the time of the donation gans is the lack of blood brain barrier due to the pres- of the canine body. Post-mortem examination and harvesting ence of fenestrated capillaries [30]. They are recognized of the brains for histopathological analysis were performed to play important integrative roles in the regulation of within 24 hours after death. Throughout this time, the ani- fluid and mineral balance [31]. This is not completely mal bodies were refrigerated. A complete post-mortem new. AQP4 expression was already demonstrated in examination including gross and histopathological analysis of Álvarez et al. BMC Veterinary Research (2021) 17:29 Page 7 of 8 the bodies was performed. Thorough examination of staining intensity (+++) to the subpial and subependymal formalin-fixed paraffin-embedded transverse sections of the areas. The staining intensity became progressively fainter brains, using a rotatory microtome, followed by (++,+, -) the further away from the periventricular and hematoxylin-eosin staining confirmed the lack of significant perivascular regions. Thereafter, the above-named other abnormalities. The age ranged between 21 days to 17 years representative encephalic structures were compared to (mean 6.4 years). Different canine breeds were included. De- the staining intensity seen in the subpial and subependy- tailed information about the dogs is listed in Table 1. mal areas and the staining intensity was attributed ac- Representative sections of the brain were evaluated at cordingly. This pattern was found to be repeatable in all the level of the frontal cortex/corpus striatum, parieto- examined canine brains. The analysis was performed temporal cortex/diencephalon, mesencephalon, cerebel- within the same day and was repeated up to three times lum/pons and medulla oblongata. to ensure consistency of the results. Abbreviations Histology procedure AQP4: Aquaporin-4; AQPs: Aquaporins; CNS: Central nervous system; Samples of all tissues were fixed in 10% buffered forma- GFAP: Glial fibrillary acidic protein; GM: Grey matter; PBS: Phosphate-buffered lin and paraffin-embedded. saline; RT: Room temperature; WM: White matter Sections 3-µm thick of paraffin-embedded brain tissue Acknowledgements of different selected areas were stained using standard We thank the technical staff at the UPMiC for their assistance with this research hematoxylin-eosin dye. Primary antibodies specific for and the owners for the donation of the cases. AQP4 and GFAP were used to characterize the distribu- Authors’ contributions tion of the AQP4 and the astrocytes within the brain. All authors conceived and designed the study. EB performed the Sections 3‐µm thick were mounted on capillary glass immunohistochemical analysis. PA analyzed all the canine brain samples. slides, dewaxed, and rinsed with water and treated with MP and EB supervised the experiments. PA drafted the manuscript, which was supervised and modified by AW. All authors have critically 3% peroxide hydrogen for 35 minutes to inhibit en- reviewed the article content and approved it in its final version. dogenous peroxidase activity. When antigen retrieval was necessary (GFAP), sections were heated for 20 mi- Funding nutes in a bain-marie at 98ºC with 10 mM citrate buffer This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. pH 6.0, cooled for 30 minutes at room temperature (RT), and rinsed in phosphate‐buffered saline (PBS). Availability of data and materials Nonspecific binding was blocked with 30% normal goat The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. serum diluted in PBS for 1 hour at RT. The samples were incubated overnight at 4 °C with primary anti- Ethics approval and consent to participate bodies: rabbit anti-AQP4 antibody (1:800, Chemicon, Owner agreement was obtained at the time of the donation of the canine body to use the data freely for research purpose. The approval for conducting Temecula, CA), and rabbit anti-GFAP antibody (1:1000, this study was also granted by the Unit of Murine and Comparative Pathology Dako, Denmark). The sections were then rinsed with without a request to ethical committee as the samples were collected after PBS and incubated for 40 minutes at RT with a labeled euthanasia elected by the owner or natural death. None of the animals was sacrificed for this study. polymer according to the manufacturer’s instructions (Labeled Polymer—Dako REAL Envision-HRP K011, Consent for publication Dako, Denmark). Staining was completed by 10-minute Not applicable. incubation with 3, 3′-diaminobenzidine and counter- Competing interests staining in hematoxylin for 3 seconds. For the negative The authors have no personal interest, direct or indirect, in any matter that control, an isotype-specific immunoglobulin was used as raises or may raise a conflict and declare that they have no financial a substitute for the primary antibody in all experiments; competing interests. no immunostaining was detected in these sections. Author details 1 Neurology and Neurosurgery Service, Pride Veterinary Centre, Derby, UK. 2 Analysis Unit of Murine and Comparative Pathology (UPMiC), Department of Animal Medicine and Surgery, Veterinary Faculty, Autonomous University of Microscopic description and semiquantitative evaluation Barcelona, Barcelona, Spain. 3Center on Bioengineering, Biomaterials and of the expression and intensity of AQP4 immunoreac- Nanomedicine (CIBER-BBN) Networking Research, Universitat Autònoma de tion was performed. The observer (PA) was not blinded Barcelona, Bellaterra, Barcelona, Spain. to the signalment and clinical history of the canine study Received: 23 October 2020 Accepted: 1 January 2021 population. The AQP expression was distributed accord- ing to the following scale: (-) absence of detectable stain- ing, (+) mild staining, (++) moderate staining in focal or References 1. Papadopoulos MC, Verkman AS. Aquaporin water channels in the nervous diffuse patterns and (+++) strong staining (Table 2). system. Natural Reviews Neuroscience. 2013;14(4):265–77. https://doi.org/10. This scale was achieved by assigning the strongest 1038/nrn3468. Álvarez et al. BMC Veterinary Research (2021) 17:29 Page 8 of 8 2. Fang W, Zheng Y, Yang F, Meng-Ting C, Chun-Hong S, Zhi-Rong L, et al. 23. Owasil R, O’Neill R, Keable A, Nimmo J, Sharp MM, Kelly ML, et al. The Short segment myelitis as the initial and only manifestation of aquaporin-4 Pattern of AQP4 Expression in the Ageing Human Brain and in Cerebral immunoglobulin G-positive neuromyelitis optica spectrum disorders. Ther Amyloid Angiopathy. Int J Mol Sci. 2020;21(4):1225. https://doi.org/10.3390/ Adv Neurol Disord. 2020;13:1–10. https://doi.org/10.1177/ ijms21041225. 1756286419898594. 24. Landsberg GM, Nichol J, Araujo JA. Cognitive dysfunction syndrome: a 3. Nataf S. The Demonstration of an Aqp4/Tgf-beta 1 Pathway in Murine disease of canine and feline brain aging. Vet Clin North Am Small Anim Astrocytes Holds Implications for Both Neuromyelitis Optica and Progressive Pract. 2012;42(4):749–68,vii. https://doi.org/10.1016/j.cvsm.2012.04.003. Multiple Sclerosis. Int J Mol Sci. 2020;21(3):1035. https://doi.org/10.3390/ 25. Azkona G, García-Belenguer S, Chacón G, Rosado B, León M, Palacio J. ijms21031035. Prevalence and risk factors of behavioural changes associated with age- 4. Maugeri R, Schiera G, Di Liegro CM, Fricano A, Lacopino DG, Di Liegro I. related cognitive impairment in geriatric dogs. J Small Anim Pract. 2009; Aquaporins and Brain Tumors. Int J Mol Sci. 2016;17(7):1029. https://doi.org/ 50(2):87–91. https://doi.org/10.1111/j.1748-5827.2008.00718.x. 10.3390/ijms17071029. 26. Yu CH, Song GS, Yhee JY, Kim JH, Im SH, Nho WG, et al. Histopathological 5. Nico B, Ribatti D. Role of aquaporins in cell migration and edema formation and immunohistochemical comparison of the brain of human patients with in human brain tumors. Exp Cell Res. 2011;317(17):2391–6. https://doi.org/ Alzheimer’s disease and the brain of aged dogs with cognitive dysfunction. 10.1016/j.yexcr.2011.07.006. J Comp Pathol. 2011;145(1):45–58. https://doi.org/10.1016/j.jcpa.2010.11.004. 6. Papadopoulos MC, Saadoun S. Key roles of aquaporins in tumor biology. 27. Ozawa M, Chambers JK, Uchida K, Nakayama H. The Relation between Biochim Biophys Acta Biomembr. 2015;1848(10PtB):2576–83. https://doi.org/ canine cognitive dysfunction and age-related brain lesions. The Journal of 10.1016/j.bbamem.2014.09.001. veterinary medical science. 2016;78(6):997–1006. https://doi.org/10.1292/ 7. Badaut J, Lasbennes F, Magistretti PJ, Regli L. Aquaporins in brain: jvms.15-0624. distribution, physiology, and pathophysiology. J Cereb Blood Flow Metab. 28. Cottrell GT, Ferguson AV. Sensory circumventricular organs: central roles in 2002;22(4):367–78. https://doi.org/10.1097/00004647-200204000-00001. integrated autonomic regulation. Regul Pept. 2004;117:11–3. https://doi.org/ 8. Satoh J, Tabunoki H, Yamamura T, Arima K, Konno H. Human astrocytes 10.1016/j.regpep.2003.09.004. express aquaporin-1 and aquaporin-4 in vitro and in vivo. Neuropathology. 29. Rønnekleiv OK, Møller M. Brain-pineal nervous connections in the rat: an 2007;27(3):245–56. https://doi.org/10.1111/j.1440-1789.2007.00774.x. ultrastructure study following habenular lesion. Exp Brain Res. 1979;37(3): 9. Arciénega II, Brunet JF, Bloch J, Badaut J. Cell locations for AQP1, AQP4 and 551–62. https://doi.org/10.1007/BF00236823. 9 in the non-human primate brain. Neuroscience. 2010;167(4):1103–14. 30. Oliveira M, Fernández F, Solé J, Pumarola M. Morphological, histological and https://doi.org/10.1016/j.neuroscience.2010.02.059. immunohistochemical study of the area postrema in the dog. Anat Sci Int. 10. Blasco E, Martorell J, De la Fuente C, Pumarola M. Immunohistochemical 2018;93(2):188–96. https://doi.org/10.1007/s12565-016-0388-0. Study of Aquaporins in an African Grey Parrot (Psittacus erithacus) With 31. Fry WM, Ferguson AV. Circumventricular Organs. Encyclopedia of Hydrocephalus. J Avian Med Surg. 2014;28(4):309–15. https://doi.org/10. Neuroscience. Boston: Academic, 2009: pp. 997–1002. https://doi.org/10. 1647/2013-059. 1016/B978-008045046-9.00462-9. 11. Klemens J, Ciurkiewicz M, Chludzinski E, Iseringhausen M, Klotz D, Pfankuche VM, et al. Neurotoxic potential of reactive astrocytes in canine distemper Publisher’s Note demyelinating leukoencephalitis. Scientific Reports. 2019;9(1):11689. https:// Springer Nature remains neutral with regard to jurisdictional claims in doi.org/10.1038/s41598-019-48146-9. published maps and institutional affiliations. 12. Spitzbarth I, Heinrich F, Herder V, Recker T, Wohlsein P, Baumgärtner W. Canine Central Nervous System Neoplasm Phenotyping Using Tissue Microarray Technique. Vet Pathol. 2017;54(3):369–79. https://doi.org/10.1177/ 0300985816688745. 13. Mobasheri A, Marples D, Young IS, Floyd RV, Moskaluk CA, Frigeri A. Distribution of the AQP4 water channel in normal human tissues: protein and tissue microarrays reveal expression in several new anatomical locations, including the prostate gland and seminal vesicles. Channels. 2007; 1(1):29–38. https://doi.org/10.4161/chan.3735. 14. Rash JE, Yasumura T, Hudson CS, Agre P, Nielsen S. Direct immunogold labeling of aquaporin-4 in square arrays of astrocyte and ependymocyte plasma membranes in rat brain and spinal cord. Proc Natl Acad Sci. 1998;95: 11981–6. https://doi.org/10.1073/pnas.95.20.11981. 15. Nagelhus EA, Ottersen OP. Physiological roles of aquaporin-4 in brain. Physiol Rev. 2013;93(4):1543–62. https://doi.org/10.1152/physrev.00011.2013. 16. Tham DK, Joshi B, Moukhles H. Aquaporin-4 Cell-Surface Expression and Turnover Are Regulated by Dystroglycan, Dynamin, and the Extracellular Matrix in Astrocytes. PLoS One. 2016;11(10):e0165439. https://doi.org/10. 1371/journal.pone.0165439. 17. Brat DJ. Normal Brain Histopathology. Practical Surgical Neuropathology: A Diagnostic Approach. 2nd ed. Philadelphia: Elsevier, 2017: pp. 19–37. 18. Binder DK, Yao X, Zador Z, Sick TJ, Verkman AS, Manley GT. Increased seizure duration and slowed potassium kinetics in mice lacking aquaporin-4 water channels. Glia. 2006;53(6):631–6. https://doi.org/10.1002/glia.20318. 19. Taft JR, Vertes RP, Perry GW. Distribution of GFAP + astrocytes in adult and neonatal rat brain. Int J Neurosci. 2005;115(9):1333–43. https://doi.org/10. 1080/00207450590934570. 20. Kalaria R, Ferrer I, Love S. Vascular Disease, Hypoxia and Related Conditions. Greenfield’s Neuropathology. 9th ed. Florida: CRC Press, 2015: pp. 59–209. 21. Wen H, Nagelhus EA, Amiry-Moghadda M, Agre P, Ottersen OP, Nielsen S. Ontogeny of water transport in rat brain: postnatal expression of the aquaporin-4 water channel. Eur J Neurosci. 1999;11(3):935–45. https://doi. org/10.1046/j.1460-9568.1999.00502.x. 22. Fallier-Becker P, Vollmer JP, Bauer HC, Noell S, Wolburg H, Mack AF. Onset of Aquaporin-4 Expression in the Developing Mouse Brain. Int J Dev Neurosci. 2014;36:81–9. https://doi.org/10.1016/j.ijdevneu.2014.06.001.