Clinical, neuropathological, and immunological short‐ and long‐term feature of a mouse model mimicking human herpes virus encephalitis

1 INTRODUCTION

Neurotropic alphaherpesviruses including herpes simplex viruses 1 and 2 (HSV-1, HSV-2) and Varicella Zoster Virus (VZV) are major human pathogens which can cause devastating neurological diseases [1]. Alphaherpesviruses typically initiate productive infections in mucosal epithelial cells. Subsequently, peripheral sensory nerve endings are infected and viral particles are transported retrogradely within their axons to ganglia of the peripheral nervous system (PNS) to establish reactivatable, life-long latency [2]. HSV-1 infection is the major cause of Herpes Simplex Encephalitis (HSE) [3]. HSE occurs sporadically and is characterized by high mortality of up to 70%, if undiagnosed and untreated, with only a minority of patients returning to normal life [4]. Patients initially present with unspecific clinical signs, which, however, get worse with disease progression and include disorientation, aphasia, changes in mental status, disorders of cranial nerves IV and VI, or seizures [5, 6]. Specifically, behavioral abnormalities occur which include hypomania, Kluver–Bucy syndrome, and memory impairment [7]. Survivors suffer from incriminating life-long sequelae like speech dysfunctions, behavioral, memory and cognitive alterations, and epilepsy [8]. Apart from usually severe affection, few subacute forms of HSE have been described [9]. Further, rare cases of relapsing or chronic CNS inflammation have been reported in both immunocompetent and immunocompromised individuals [10-12].

HSE is characterized by asymmetrical necrotizing inflammation which is mainly restricted to the temporal as well as to the frontal lobe and insular cortex [13, 14]. Macroscopically, brains of surviving HSE patients reveal predominantly unilateral atrophy and yellow-brownish discoloration resulting from necrosis and microhemorrhages of affected brain areas [15, 16]. Histologically, destruction of grey and white matter, necrosis of cortical neurons, glial activation as well as leptomeningeal, and scattered parenchymal infiltration with lymphocytes and histiocytes are present [14, 15, 17, 18]. Extensive edema has been reported [14, 18], whereas glial nodules are frequently detected [17]. Extra temporal involvement of HSE has been described in more than half of the cases and includes lesions in the frontal and parietal cortex, occipital lobe, basal ganglia, brain stem, and pons [19]. Intranuclear inclusion bodies are only inconsistently observed in neurons and astrocytes [14, 17, 20]. Rather rare reports include calcification within necrotic brain areas [21] or granulomatous inflammation with foci of mineralization [22]. Viral antigen containing neurons and astrocytes appear predominantly in amygdaloid nuclei, cortex and white matter of the lateral olfactory striae, entorhinal cortex, subiculum, hippocampus, insula, and cingulate gyrus and to a lesser extent in olfactory bulbs and pons [18].

Despite decades of intensive research, major questions regarding the pathogenesis of HSE remain unanswered. Of central importance is the elucidation of factors enabling HSV-1 to traverse the peripheral nervous system for access to the central nervous system (CNS), specifically to the temporal lobe. Moreover, it is unclear why HSV-1 infection leads to fatal encephalitis in some individuals, but normally results in life-long latency without any clinical signs. In addition, subclinical HSE with recurring reactivation events and associated psychiatric disorders has been discussed decades ago [23, 24]. Valuable knowledge has been gained from a variety of animal models to better understand pathogenesis of the disease, although the data are quite heterogeneous because of the large number of different virus strains, inoculation routes, and variable genetic backgrounds of the animals [25]. In this context, especially HSV-1-infected mice fail to reflect crucial elements of human encephalitis appropriately, because mice develop brain stem or cerebellar encephalitis rather than inflammation associated with the temporal lobe. Although mice are generally highly susceptible to HSV-1 infection, disease outcomes as well as mortality differ widely between inbred mouse strains [26]. Highly susceptible mice usually develop severe neurological signs and succumb to death shortly after infection [27], preventing a more detailed investigation of long-term damage and lesions associated with behavioral abnormalities as seen in human patients.

We have recently reported on an animal model for HSE which revealed striking analogies to human disease [28]. In our study, 6- to 8-week-old female CD-1 mice developed marked meningoencephalitis after intranasal infection with a mutant of the neurotropic alphaherpesvirus pseudorabies virus (PrV), which is closely related to HSV-1 and usually fatal in all animal species except for pigs. This PrV mutant, designated PrV-ΔUL21/US3Δkin, lacks tegument protein pUL21 and carries a mutation in the active site of the pUS3 protein kinase. While the deletion of most nonessential viral genes had either no or only a slight effect on neuroinvasion, neurovirulence and survival time of infected mice [29], most mice surprisingly survived infection with PrV-ΔUL21/US3Δkin [28]. As in human HSE, lymphohistiocytic inflammation with pronounced neuronal necrosis was predominantly confined to the temporal as well as to the frontal lobes and insular cortex. With progression of the inflammatory reaction, mice revealed behavioral abnormalities such as “star gazing,” which seem to be comparable to behavioral alterations in humans suffering from HSE. Strikingly, only few mice developed severe disease between day 10 and 13 post infectionem (pi), while the majority of infected animals were only moderately affected and able to survive despite extensive neuropathological changes.

In the present study, we analyzed survival as well as clinical and histopathological short- and long-term consequences and compared the inflammatory reaction and associated neuropathological changes to further validate the PrV-mouse model for human HSE.

2 MATERIAL AND METHODS 2.1 Animal experiments

All animal experiments were approved by the State Office for Agriculture, Food Safety and Fishery in Mecklenburg-Western Pomerania (LALFF M-V) with reference number 7221.3-1-064/17. ARRIVE guidelines 2.0 were followed as reported below.

In general, 6- to 8-week-old female CD-1 mice were purchased from Charles River Laboratory and housed in groups of maximum five animals in conventional cages type II L under BSL 2 conditions at a temperature of 20–24°C. Mice were kept under a 12 h light–dark cycle (day light intensity 60%) with free access to food (ssniff Ratte/Maus – Haltung) and clean drinking water. Bedding (ssniff Spezialdiäten Abedd Espen CLASSIC), nesting (PLEXX sizzle nest), and enrichment material (PLEXX Aspen Bricks medium, mouse smart home, mouse tunnel) were provided. An acclimatization period of at least 1 week was allowed prior to inoculation. Animals were anesthetized with 200 µl of a mixture of ketamine (60 mg/kg) and xylazine (3 mg/kg) dissolved in 0.9% sodium chloride which was administered intraperitoneally. Afterwards, a total of 5 µl of PrV-∆UL21/US3∆kin suspension in cell culture media was inoculated in each nostril (1 × 104 plaque forming units [PFU]). Control mice were inoculated with cell culture supernatant from rabbit kidney (RK13) cells (minimum essential medium (MEM) + 5% fetal calf serum [FCS]) accordingly. Mice were monitored 24/7 and scored for clinical signs as described earlier [28]. The animals were sacrificed under deep anesthesia with isoflurane, cardiac bleeding, and final decapitation. In order to allow an unbiased investigation, as well as considering animal welfare conditions, treatment and time points of analysis were determined for each individual animal prior to the experiment by simple randomization. Blinding was performed during allocation of animals and data analysis. The minimum number of animals necessary in the different exploratory experiments were calculated on a disease incidence of 80% of inoculated animals and 0% in mock-infected mice (power = 0.8, α = 0.1).

2.1.1 Long-term investigation

Long-term effects were determined clinically and histopathologically over 6 months in an exploratory study. PrV-∆UL21/US3∆kin-infected animals (n = 5) each were analyzed by histology at 28, 35, 42, 49, 84, and 168 days pi. Mock-inoculated mice (n = 6) were included as control.

2.1.2 Neurohistopathological analyses of early inflammation

We reused mouse brain tissue sections obtained from the previous experiment [28] to explore pathomorphological changes and the spatial distribution of infiltrating immune cell populations during the first 21 days of infection in detail. Mice sacrificed at 2, 8, 12, 15, and 21 days pi (n = 3) served as positive material. Mock-infected mice were included as control (n = 3).

2.1.3 Immunological analyses of early inflammation

To assess neuroinflammatory response by flow cytometry PrV-∆UL21/US3∆kin-infected (n = 6) as well as control mice (n = 4) were sacrificed at 2, 8, 12, 15, and 21 days pi to identify and quantify inflammatory infiltrates and cytokine levels in the brain.

2.1.4 Neurohistopathological analyses of severely diseased mice

PrV-ΔUL21/US3Δkin-infected animals from the different experiments (n = 6) that show severe clinical signs or found dead were analyzed histopathologically to explore the severe disease outcome. Animals from the first trial to determine the mean time to death and the kinetic study [28] as well as the long-term experiment (this study) were included.

2.2 Virus

PrV-ΔUL21/US3Δkin was generated in a PrV-Kaplan (PrV-Ka) [30] background as described previously [28]. The virus was propagated in RK13 cells grown at 37°C in MEM supplemented with 10% FCS (Invitrogen).

2.3 Histopathological analysis

For histopathological investigation, the skull was removed and the head was fixed in 4% neutral-buffered formalin for at least 1 week followed by decalcification for 3 days in Formical 2000 (Decal, Tallman, NY). From all heads, eight coronal head sections were obtained, embedded in paraffin wax and cut at 3 or 5 µm thick slices, respectively, for further histological and immunohistochemical evaluation [28]. The slices were mounted on Super-Frost-Plus-Slides (Carl Roth GmbH, Karlsruhe, Germany) and stained with hematoxylin–eosin for detailed neuropathological analysis of CNS inflammation.

2.3.1 Special stains

Axonal density was visualized by Bielschowsky's silver impregnation. Dewaxed paraffin sections were treated with 0.25% potassium permanganate solution (3 min) and rinsed in distilled water. Afterwards, 1% potassium sulfate solution was applied to sections (1 min). Sections were rinsed in tap and distilled water before samples were probed with 2% silver nitrate solution overnight. Sections were rinsed in distilled water (3–5 s, 2 times) and incubated with 10% ammoniacal silver nitrate solution (10 min). Sections were dipped in distilled water (5 s) and reduced in 4% formalin.

Myelination was evaluated using Luxol Fast Blue-Cresyl Violet staining. Dewaxed paraffin sections were treated with xylol (2x 2 min), 99.5% (2x 3 min), 95%, 80%, 70%, 50% 1-propanol (3 min each), and distilled water (3 min). After incubation in isopropyl alcohol (15 min), the section were left in luxol fast blue solution (24 h, 57°C), rinsed with distilled water, and differentiated in 0.05% lithium carbonate solution (15 s) and 70% ethyl alcohol (15 s). Sections were rinsed with distilled water and counterstained with 0.1% Cresyl Fast Violet solution, dehydrated in 96% ethyl alcohol (2x 4 min), isopropanol (1x 4 min), and butyl acetate (1x 4 min), and coated with Entellan (Merck, Darmstadt, Germany).

Mineralization was investigated using the von Kossa stain. As described above dewaxed paraffin-embedded section were rehydrated and subsequently incubated with 5% silver nitrate solution (120 min) in the dark. After washing in distilled water, the sections were treated with 1% pyrogallic acid (4 min) and 4% sodium thiosulfate (5 min). The sections were rinsed in tap water (10 min) and counterstained with nuclear fast red (5 min), washed in distilled water, and dehydrated through graded alcohol.

Hemosiderosis following hemorrhages was assessed by Prussian Blue staining. Rehydrated paraffin-embedded tissue sections were immersed in 1% hydrochloric acid and 2% potassium ferrocyanide (30 min), rinsed in distilled water, followed by counterstain with nuclear fast red (5 min). Sections were rinsed in distilled water and dehydrated.

2.3.2 Immunohistochemistry

Infiltrating immune cell populations were identified using antibodies against Iba-1 (FUJIFILM Wako, polyclonal rabbit anti-rat, 1:800, for monocytes and macrophages), CD3 (DAKO, polyclonal rabbit anti-human T cell CD3 A452, 1:100, for T cells) and CD79a (DAKO, monoclonal mouse anti-human CD79αcy CloneHM57, 1:50, for B cells). Antibodies against glia-fibrillary-acid-protein (GFAP) (abcam, polyclonal rabbit anti-bovine, 1:100) stained astrocytes. PrV infection was visualized using an in-house-generated rabbit polyclonal antibody against glycoprotein B [31].

Dewaxed and rehydrated paraffin-embedded sections were treated with 3% of hydrogen peroxide (10 min, Merck, Darmstadt, Germany) to block endogenous peroxidases. To demask antigenic sites in tissue (except sections for PrV gB), sections were either treated with 10mM citrate buffer (2x 5 min, microwave, 500W, for GFAP, CD3) or 10mM Tris–EDTA buffer (10mM Tris base, 1mM EDTA solution, 15 min, microwave, 500W, for CD79a) followed by incubation in undiluted normal goat serum (30 min). Primary antibody incubation was followed by biotinylated goat anti-rabbit IgG (1:200; Vector Laboratories, Burlingame, CA, for GFAP) or goat-anti-mouse IgG (1:200, Vector Laboratories, Burlingame, CA, for CD79a) and subsequent avidin–biotin–peroxidase (ABC) complex (Vector Laboratories) for 30 min at room temperature. For CD3 staining, sections were treated with Envison®+ System – HRP (DAKO). Positive antigen–antibody reaction was visualized using AEC-substrate (DAKO, Hamburg, Germany). After rinsing with deionized water, the sections were counterstained with Mayer's hematoxylin for 10 min and mounted with Aquatex (Merck).

2.3.3 Scoring of neurohistopathological changes

In order to characterize PrV-∆UL21/US3∆kin-induced encephalitis during the acute phase of infection [28] in more detail, the following brain regions were stained with hematoxylin and eosin and analyzed histopathologically: brainstem (BS) including medulla oblongata and pons, mesencephalon (MES), diencephalon (DI), temporal lobe (TL) including hippocampus, parietal lobe (PL), and frontal lobe (FL).

The above mentioned brain areas were analyzed for inflammatory changes according to a recently published protocol [32] with slight modifications as given in Table 1. Neuronal necrosis and spongiform changes (Table 1) were scored only in the TL which was the most affected brain region.

TABLE 1. Scoring of histopathological changes assessed on hematoxylin and eosin-stained brain regions Score 0 Score 1 Score 2 Score 3 Inflammation Meninges/perivascular Absent 1–2 cell layers 3–5 cell layers >5 cell layers Neuroparenchymal Absent 1–15 cells 16–30 cells >30 cells Neuronal necrosis Absent Mild, scattered Moderate, multifocal groups of neurons Severe, coalescing groups of neurons Spongiform changes Absent 1%–30% of brain area affected 31%–60% of brain area affected >60% of brain area affected Note The mean value was built from three biological replicates obtained in the kinetic trial [28] and five replicates per indicated time point obtained in the long-term experiment (this study).

Axonal density in the white matter was scored according to a recent protocol [32]. Demyelination was evaluated as published earlier [33]. Parenchymal mineralization and hemosiderosis were determined absent or present. Scoring of all four parameters assessed in the TL during the acute phase of infection is given in Table 2.

TABLE 2. Scoring of temporal lobe tissue sections for axonal density, demyelination, mineralization, and hemosiderosis Score 0 Score 1 Score 2 Score 3 Axonal density No reduction 1/3 loss 1/3-2/3 loss >2/3 loss Demyelination Absent Scattered Multifocal Coalescing Mineralization Absent or present Hemosiderosis Absent or present 2.3.4 Scoring of inflammatory cells (immunohistochemistry)

Temporal lobe infiltration by CD3+ T cells, CD79+ B cells, and Iba-1+ microglia/macrophages was determined during the acute phase of infection and scored as illustrated in Table 3 based on a recent protocol [32] with few adaptions. Astrogliosis based on GFAP immunoreactivity was assessed as absent or present. Infiltration of cells was evaluated in 20x or 40x magnification (high power filed = HPF).

TABLE 3. Scoring details of CD3, CD79, Iba-1, and GFAP signals in temporal lobe sections Score 0 Score 1 Score 2 Score 3 CD3 Meningeal/perivascular (40x) Absent 2 layers 3–5 layers >5 layers Neuroparenchymal (40x) <10 cells 11–20 cells >20 cells CD79 Meningeal/perivascular (20x) Absent 1 layer 2 layers >2 layers Neuroparenchymal (20x) <10 cells 11–20 cells >20 cells Iba-1 Meningeal/perivascular (40x) Absent 2 layers 3–5 layers >5 layers Neuroparenchymal (40x) <10 cells 11–20 cells >20 cells GFAP Meningeal/perivascular (20x) Absent or present Neuroparenchymal (20x) 2.4 Cell preparation and antibody staining for flow cytometric analysis

Brain samples were prepared for single-cell isolation according to a recently published protocol [34] with slight modifications. Briefly, after removing from the scull, brains were immediately transferred to ice-cold PBS and kept on ice. The cerebellum was removed and the remaining brain cut into small pieces on ice. For cell isolation, brain pieces were pressed through a cell strainer (70 µm, BD Biosciences, Heidelberg, Germany), homogenized, and taken up in 2 ml cOmpleteTM Mini EDTA-free protease inhibitor cocktail (Roche, Basel, Switzerland). Half of the homogenate was used for the analysis of the infiltrating immune cells or cytokines. The homogenate was centrifuged (286 × g, 4 °C, 5 min), and the supernatant was discarded. The cell pellet was resuspended in 1 ml digestion buffer (Liberase with low thermolysin concentration to a concentration of 2 U/ml in Hanks Balanced Salt Solution [HBSS] containing calcium [Ca] and magnesium [Mg]) and incubated for 30 min at 37°C with gentle agitation. The suspension was pressed through a cell strainer (70 μm), washed with 10 ml of DNAse-free washing buffer (HBSS (Ca/Mg free) containing 10% of FCS), and centrifuged (286 × g, 18°C, 5 min). The supernatant was discarded, and the cell pellet was carefully resuspended in 5 ml density gradient medium (25%, room temperature), and centrifuged (521 × g, 18°C, 30 min, acceleration/deceleration = 0). The myelin layer and the supernatant were aspirated, and the cell pellet was resuspended in 10ml DNAse-free washing buffer and centrifuged again (286 × g, 10°C, 5 min). The supernatant was discarded, and the cells were resuspended in 100µl of cold washing buffer. Cell counting and assessment of cell viability were achieved using trypan blue staining (dilution 1:10).

For flow cytometric antibody staining, the cell pellet of 1 ml brain homogenate was suspended in FACS-buffer (PBS containing 0.1% Sodium azide and 0.1% BSA) and treated with CD16/CD32 Fc-Receptor blocking reagent (2.5 μg/ml). Cells were stained with primary antibodies listed in Table 4 for 15 min at 4°C in the dark. For staining of whole blood, erythrocytes were lysed after surface staining by conventional lysis buffer (1.55 M NH4Cl, 100 mM KHCO3, 12.7 mM Na4EDTA, pH 7.4, in distilled water). Gating is shown in SI 1.

TABLE 4. Antibodies used in flow cytometric analyses Antigen Host Isotype Conjugate Clone Manufacturer B220 Rat IgG2a PerCP-Cy5.5 RA3-6B2 eBioscience CD3 Armenian hamster IgG1 BUV395 145-2C11 BDBiosciences CD4 Rat IgG2b BV510 GK1.5 BioLegend CD8 Human IgG1 PE-Cy7 REA601 Miltenyi CD11b Rat IgG2b BV711 M1/70 BioLegend CD45 Rat IgG2b FITC 30-F11 BioLegend Ly6G Rat IgG2a AF700 1A4 BDBiosciences NK1.1 Mouse IgG2a BV786 PK136 BioLegend 2.5 Cytokine assay

For cytokine analysis in the brain, the LegendPlex™ Mouse Anti-Virus Response Panel was used to quantify 13 mouse cytokines including interferons IFN-α, IFN-β, and IFN-γ; interleukins IL-1β, IL-6, IL-10, and IL-12 as well as chemokines CCL2, CCL5, CXCL1, CXCL10, TNF-α, and GM-CSF, according to manufacturer’s instructions (BioLegend, Koblenz, Germany).

2.6 Statistical analysis

Statistical analyses and graphical visualization of data were performed using Graph Pad Prism (Version 8.4.2). To analyze brain immune cell infiltration and cytokines, ordinary one-way ANOVA with Holm-Sidak’s post hoc test was performed to compare infected animals from 2, 8, 12, 15, and 21 days pi to all control mice. Values with p ≤ 0.05 were considered significant and are indicated by asterisks (*).

3 RESULTS 3.1 Long-term dynamics and clinical signs after PrV-∆UL21/US3∆kin infection

In our first study [28], we monitored PrV-∆UL21/US3∆kin-infected mice for 21 days and investigated viral spread and inflammatory reaction in a detailed kinetic experiment. The animals developed meningoencephalitis starting at day 9 pi. Interestingly, the majority of mice showed only mild-to-moderate clinical signs or remained completely asymptomatic despite of an extensive inflammatory reaction, which could be detected until the end of the study. The animals were able to survive, with the exception of three mice, which died or had to be euthanized between 9 and 13 days pi. Notably, localization of viral antigen, which was detectable until day 15, and the very pronounced inflammatory response localized to the temporal lobe were largely comparable to human HSE. Mice also developed behavioral alterations including star gazing which may resemble abnormalities observed in human patients. Based on these data, we aimed to investigate the further course of infection including clinical alterations as well as central nervous lesions beyond 21 days in a long-term experiment. To this end, clinical signs of PrV-∆UL21/US3∆kin-infected mice were recorded until day 168 pi. For this experiment, 47 PrV-∆UL21/US3∆kin-infected mice and six control mice were used. Starting from day 28 pi, five infected animals and one mock-infected mouse each were sacrificed on days 35, 42, 49, 84, and 168 pi for neurohistopathological examination.

As observed previously [28], at day 5 pi few mice (6%) started to show clinical signs typical for PrV-∆UL21/US3∆kin infection. Subsequently, the incidence of clinical signs increased and reached almost 50% on day 8 pi. On day 10 pi, 87% of animal showed clinical signs which was the highest incidence detected in this experiment. On day 19, it decreased to 54%. Thereafter the number of mice showing clinical signs increased again to 77% on day 24 pi, thereafter decreasing continuously to 25% on day 46 pi. This was followed by a slight increase to 48% on day 53, followed by another decrease to ca. 30% on day 96 pi. The incidence from then on was low, at a maximum of around 24%. However, the incidence slightly increased again to around 30% from day 133 pi and decreased from day 146 pi to almost 18%. However, starting at day 153 pi, a new wave of clinical signs was noted which reached almost 77% at day 165 pi (Figure 1). In summary, we detected an essentially biphasic course of infection. The acute, first phase of the infection with two peaks at around days 10 and 20 slowly subsided about 3 weeks after the infection. The disease rate then remained at a low level, but increased again after about 6 months defined as the second phase of disease.

image

Number of diseased animals in proportion to the total number of infected animals over 168 days. The percentage of healthy (beige), mildly affected (green), moderately affected (yellow) and severely affected (red) animals is shown on the left y-axis. The incidence at each time point is indicated by a dotted line and shown on the right y-axis. Red arrows indicate the time points (28, 35, 42, 49, 84, and 168 days pi) at which five infected mice and one control mouse each were sacrificed and analyzed histopathologically

Early after infection, animals showed nonspecific clinical signs including ruffled fur or hunching, while two out of 47 animals remained clinically inapparent until day 28 pi. Several mice had mild pruritus and conjunctivitis and developed a nasal bridge edema. Out of 47 mice (72%), 34 animals developed alopecic skin erosions in various body regions including the head, limbs, abdomen or back, mainly occurring between day 8 and 14 pi. Some animals even developed multiple or recurring skin lesions. Typically, lesions were not hemorrhagic and healed within a week of onset. Out of 47 mice (32%), 15 animals started to show behavioral alterations including reduced activity levels and “star gazing” within the first two weeks pi while further 19 animals (40%) started to show clinical signs in the fourth week pi. Rather mild nervous signs were observed in 12 animals (25%) and included slight facial fasciculations mainly occurring between day 10 and 13 pi. Four mice showed mild ataxia mainly between day 12 and 14 pi. In total, three mice had to be euthanized because of severe convulsions and excitations at 11, 12, and 13 days pi.

Beyond 28 days pi, 29 out of the remaining 39 mice (74%) revealed recurring clinical signs mainly characterized by behavioral abnormalities, but also ruffled fur, hunched back or alopecia. Notably, not until day 159 pi, two mice (designated as mouse 1 and 2) developed seizures interrupted by phases with normal behavior or reduced activity levels. These two animals did not show any particular abnormalities in the acute phase of the infection. More specifically, mouse 1 showed first clinical signs as early as 5 days pi such as nasal bridge edema, ruffled fur, blepharospasm, and reduced activity. In the following, but only for a short time, the animal developed photophobia and signs of nervousness. Until the end of the study period, the mouse showed alternating phases of staring, nervousness, and hunching. Mouse 2 showed sickness from day 8 pi on with hunching, ruffled fur, and mild itch. From day 23 onwards, the animal had reduced activity levels which improved on day 40 pi. The animal was then clinically normal until day 156.

3.2 Long-term CNS lesions after PrV-∆UL21/US3∆kin infection

In the previous study, all mice investigated at 21 days pi showed marked meningoencephalitis [28], therefore we intended to investigate neuropathological changes also at later time points of infection. For this, five infected animals each were sacrificed at 28, 35, 42, 49, 84, and 168 days pi. The severity and location of inflammation was determined on hematoxylin and eosin stained tissue sections as described earlier [28]. Inflammatory cell infiltrates and reactive changes were differentiated by immunohistochemistry targeting CD3, Iba-1, and GFAP on three animals each at each time point as illustrated in Figure 2.

image

Representative images of long-term temporal lobe lesions after PrV-∆UL21/US3∆kin infection. At all investigated time points between day 28 and 168 pi focal mild meningoencephalitis (H&E stain) composed of CD3+ and Iba+ infiltrates (immunohistochemistry, ABC method) were detectable. GFAP+ glial cells (immunohistochemistry, ABC method) were generally mildly increased except in a single animal 49 days pi. Magnification 20x

On day 28 pi, four out of five animals revealed inflammation confined to the FL and TL. Immunohistochemistry identified mainly CD3+ and Iba-1+ infiltrates while CD3+ T cells predominated in two mice. One mouse had a severe temporal meningoencephalitis with extensive necrosis of hippocampal neurons, perivascular infiltration, and mild edema. Mild astrogliosis was present in all animals. Results of hematoxylin and eosin, CD3, Iba-1, and GFAP immunohistochemistry staining are demonstrated in Figure 2. At 35 days pi, meningoencephalitis composed of CD3+ and Iba-1+ cells was found in all animals similar to those investigated 28 days pi. One animal showed moderate spongy changes in the BS. Likewise, all animals sacrificed on day 42 pi showed inflammatory reaction in the CNS. In one mouse multifocal gliosis was observed, while all other animals revealed mainly mild lymphohistiocytic meningoencephalitis in the TL and FL, confirmed by CD3+ and Iba-1+ immunohistochemistry. As observed earlier, mild spongiform changes were located to the TL as well as to the BS in two out of five animals. In mice investigated on day 49 pi, mild lymphohistiocytic temporofrontal inflammation and gliosis were found in three out of five animals, which were comparable to samples of 42 days pi. Later on day 84 pi, four out five animals suffered from CNS inflammation. However, in contrast to earlier time points mild, but mixed cellular inflammatory response was found consisting of moderate numbers of lymphocytes and histiocytes, admixed with few neutrophils mainly affecting the TL and FL. At the end of the experiment, on day 168 pi, two out of five animals that were scheduled to be sacrificed that day as well as one mouse which was further euthanized because of seizures showed histopathological abnormalities including mild mixed-cellular meningoencephalitis, as seen on day 84 pi, as well as gliosis. Severe and widespread spongiform changes, but no inflammatory reactions were present in the mouse with seizures, particularly in the TL.

Despite of long-term CNS inflammation detected in majority of animals, immunohistochemistry against PrV gB was negative except for one animal examined 49 days pi, showing few positive neurons located in the PL and TL as well as another animal 84 days pi with a single positive neuronal signal in TL (SI 2).

3.3 Kinetics and distribution of the CNS inflammatory response in the acute phase

The first term of our long-term study reflected the results of the previous experiment during the first 21 days [28]. Both indicate a biphasic course of disease with an acute phase of infection with two peaks. Mice developed meningoencephalitis from day 9. As the degree of inflammation progresses, the animals developed predominantly moderate clinical signs, including behavioral alterations, but mostly survived the infection. Only six mice died during this critical phase in three independent experiments (mean time to death and kinetic study [28] and long-term experiment [this study]). We therefore aimed to analyze this acute and critical phase in more histopathological detail.

Eight coronal head sections were stained with hematoxylin–eosin. Different parts of the brain including the BS, MES, DI, TL, PL, and FL of three animals each sacrificed at 2, 8, 12, 15, and 21 days pi were evaluated for meningeal/perivascular and parenchymal inflammatory infiltration based on a score from 0 to 3 (Table 1). Figure 3 illustrates the dynamics and score per animal and brain area at the indicated time points.

image

Semiquantitative scoring of meningeal/perivascular (red) and parenchymal inflammation (blue) affecting different brain regions. H&E stained sections of three mice per time point including the temporal lobe (TL), brain stem (BS), frontal lobe (FL), parietal lobe (PL), DI (diencephalon), and MES (mesencephalon) were scored

While infected animals showed no histopathological changes at 2 days pi, meningeal/perivascular and parenchymal infiltration, respectively, affecting the BS and TL were mildly present in single PrV-ΔUL21/US3Δkin-infected mice sacrificed at 8 days pi. However, severe meningoencephalitis was observed in all animals sacrificed 12 days pi mainly affecting the TL. While moderate inflammation was present in BS, moderate to only mild inflammatory response was observed in DI, PL, FL, and MES. However, up to severe inflamma

留言 (0)

沒有登入
gif