Apoptosis and necroptosis of mouse hippocampal and parenchymal astrocytes, microglia and neurons caused by Angiostrongylus cantonensis infection

Background Angiostrongylus cantonensis has been the only parasite among Angiostrongylidae to cause human central nervous system infection characterized by eosinophilic meningitis or meningoencephalitis. The mechanism of the extensive neurological impairments of hosts caused by A. cantonensis larvae remains unclear. The aim of the present study was to investigate apoptosis, necroptosis and autophagy in the brains of mice infected with A. cantonensis, which will be valuable for better understanding the pathogenesis of angiostrongyliasis cantonensis. Methods Functional and histological neurological impairments of brain tissues from mice infected with A. cantonensis were measured by the Morris water maze test and haematoxylin and eosin (H&E) staining, respectively. The transcriptional and translational levels of apoptosis-, necroptosis- and autophagy-related genes were quantified by quantitative real-time polymerase chain reaction (RT-PCR), and assessed by western blot and immunohistochemistry (IHC) analysis. Apoptotic and necroptotic cells and their distributions in infected brain tissues were analysed by flow cytometry and transmission electron microscopy (TEM). Results Inflammatory response in the central nervous system deteriorated as A. cantonensis infection evolved, as characterized by abundant inflammatory cell infiltration underneath the meninges, which peaked at 21 days post-infection (dpi). The learning and memory capacities of the mice were significantly decreased at 14 dpi, indicating prominent impairment of their cognitive functions. Compared with those of the control group, the mRNA levels of caspase-3, -4, -6, and RIP3 and the protein levels of caspase-4, cleaved caspase-3, cleaved caspase-6, RIP3, and pRIP3 were obviously elevated. However, no changes in the mRNA or protein levels of FADD, Beclin-1 or LC3B were evident, indicating that apoptosis and necroptosis, but not autophagy, occurred in the brain tissues of mice infected with A. cantonensis. The quantitative RT-PCR, western blot, IHC, flow cytometry and TEM results further revealed the apoptotic and necroptotic microglia, astrocytes and neurons in the parenchymal and hippocampal regions of infected mice. Conclusions To our knowledge, we showed for the first time that A. cantonensis infection causes the apoptosis and necroptosis of microglia and astrocytes in the parenchymal and hippocampal regions of host brain tissues, further demonstrating the pathogenesis of A. cantonensis infection and providing potential therapeutic targets for the management of angiostrongyliasis.


(Continued from previous page)
Conclusions: To our knowledge, we showed for the first time that A. cantonensis infection causes the apoptosis and necroptosis of microglia and astrocytes in the parenchymal and hippocampal regions of host brain tissues, further demonstrating the pathogenesis of A. cantonensis infection and providing potential therapeutic targets for the management of angiostrongyliasis.
Keywords: Angiostrongylus cantonensis, Pathogenesis, Apoptosis, Necroptosis Background Angiostrongylus cantonensis is the most common cause of eosinophilic meningitis worldwide. As accidental hosts, humans can become infected via the ingestion of undercooked intermediate hosts (Pomacea canaliculata, Achatina fulica, etc.) or contaminated water or food that contains third-stage larvae (L3) [1,2]. After penetrating the intestinal wall, A. cantonensis migrate in the body through the bloodstream, finally passing the blood-brain barrier to further develop in the central nervous system (CNS). Larvae in the brain tissue cause direct mechanical damage and severe inflammation, resulting in eosinophilic meningitis or encephalitis. In addition to the cerebrum and meninges, the cerebellum, brainstem and spinal cord can be affected. Clinical symptoms, manifested as CNS injuries, include severe headache, neck stiffness, convulsions and nausea [3].
Parasitic infections of the CNS include protozoans (Plasmodium, Trypanosoma cruzi, Toxoplasma gondii and amoebae) and metazoans (Schistosoma, Paragonimus, larval stages of Taenia solium, Echinococcus granulosus, Spirometra mansoni, and A. cantonensis), and the pathogenesis of these neuroparasites include the apoptosis and necroptosis of host cells [4]. The major pathogenetic mechanism of cerebral amebiasis lies in the killing of mammalian cells by trophozoite-induced apoptosis [5]. Disruption of the endothelium via apoptosis leads to the leakage of red blood cells, ring haemorrhage, and activated coagulation in adjacent nervous tissue in cerebral malaria [6]. In neurocysticercosis, neurological deficits, particularly learning and memory deficits, are generated by the extensive and significant apoptosis of hippocampal cells [7] and apoptosis of CD3+ T lymphocytes in the brains of hosts induced by T. solium cysticerci may be a mechanism by which the parasite downregulates its host's cellular immune response during early cysticercosis [8]. The most common presentation of cerebral toxoplasmosis in HIV-infected patients is mass lesions consisting of well-defined areas of coagulative necroptosis with or without haemorrhage [9]. The gross appearance of tumoural lesions in neuroschistosomiasis is characterized by the presence of necroticexudative granuloma containing eggs surrounded by necrotic and elongated epithelioid cells [10].
Although A. cantonensis is the most common cause of eosinophilic meningitis worldwide, the major cell populations and cell injuries in the host brain after infection are not entirely clear [3]. To further reveal the pathogenesis of angiostrongyliasis cantonensis in this study, the Morris water maze test was used to test the neural functionality of A. cantonensis-infected hosts. Quantitative RT-PCR, western blot and immunohistochemistry (IHC) analyses were combined to detect apoptosis, necroptosis and autophagy in mouse cerebral parenchymal and hippocampal cells. Flow cytometry and transmission electron microscopy (TEM) were performed to evaluate the cell death of parenchymal and hippocampal astrocytes, neurons and microglia.

Animals
Female BALB/c mice (6-8 weeks old) were purchased from Guangdong Medical Laboratory Animal Center and housed in a specific pathogen-free environment and had unlimited access to food and water.
One batch of mice was randomly divided into 6 groups (9 mice per group) and orally infected with A. cantonensis third-stage larvae (L3, 30 per mouse), except for the normal controls. Groups were named according to the number of day(s) post-infection (dpi): group G1 (uninfected mice), group G2 (mice infected for 1 day), group G3 (mice infected for 3 days), group G4 (mice infected for 7 days), group G5 (mice infected for 14 days) and group G6 (mice infected for 21 days). Brain samples from each group were prepared for mRNA and protein extraction or fixed for immunohistochemistry analysis. Animals were euthanized under deep anaesthesia by ether inhalation followed by blood-letting.
Another batch of mice was used for studying neurological function, flow cytometry analysis and transmission electron microscopic observation. They were divided into 4 groups and orally infected with 30 L3 for 7, 14 and 21 days or not infected (control group). After neurological function evaluation, flow cytometry analysis of brain tissues was carried out. Brain tissues from the 0 dpi and 21 dpi groups were subjected to transmission electron microscopy.
Infecting BALB/c mice with A. cantonensis larvae All the infectious A. cantonensis L3 used in this study were obtained from Biomphalaria glabrata 21 days after infection of the first-stage larvae (L1) of the parasite. The snails were homogenized and digested in a pepsin-HCl solution (pH 2.0, 500 IU pepsin/g tissue) at 37°C in an incubator for 40 min. Then, infectious L3 were washed in PBS and counted under an anatomical microscope for animal infection [11].

Neurological function evaluation
The Morris water maze test was used as described in a previous study [12] to compare learning and memory skills between normal and infected mice. The water maze consisted of a large, circular tank (180 cm in diameter × 60 cm high) filled with water (22 ± 1°C). A stationary platform was hidden 1.5 cm below the water surface, and white paint was added to make the platform invisible. During cued training, mice were placed into the pool four times each training day (days 1, 2 and 3), once from each of the four quadrants, and then required to find the platform. Animals that located the platform were allowed to rest for 30 s; those that failed to find the platform within 60 s were gently guided to the platform to rest for 30 s. Escape latency, or the time required to reach to the platform, was recorded for each mouse (day 4), and place navigation was used to evaluate their learning and memory skills. The spatial probe test, used to evaluate mouse memory retention, was performed on day 5. The platform was removed, and the probe task was performed in the third quadrant. The swim escape latency, average swim speed, time spent in the target quadrant, and number of times the mice crossed the previous location of the platform were recorded by a video tracking system (Taimeng Tech, Chengdu, China).

Sample extraction and cDNA synthesis
Total mouse brain RNA was extracted using TRIzol® Reagent (Thermo Fisher Scientific, Waltham, USA) according to the manufacturer's protocol, and RNA yield and purity were determined by the NanoDropTM spectrophotometer (Thermo Fisher Scientific). The first cDNA strand was synthesized at 42°C for 1 h using the Rever-tAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific).

Quantitative real-time polymerase chain reaction (RT-PCR)
For RT-PCR, primers for detecting cell death genes were designed using AlleleID software. The sequences of the target genes and that of the internal control β-actin gene are listed in Table 1 , and experiments were repeated three times as independent biological replicates. The mRNA levels of these genes were measured by the Ct value (threshold cycle), and the relative expression levels were calculated with the 2 -ΔΔCt method [13].

Immunohistochemistry (IHC)
Mouse brains were isolated, fixed in 4% paraformaldehyde (PFA) overnight at 4°C and embedded in paraffin. Then, the brain sections were de-paraffinized in xylene, rehydrated via a graded ethanol series, and subjected to antigen retrieval by boiling the slices in citrate buffer (pH 6.0) with high heat for 15 min and medium heat for 15 min in a microwave oven. For

Flow cytometry
Mice were euthanized, and their parenchymal and hippocampal tissues were separated, removed and collected into 5 ml of Dounce buffer (Hank's balanced salt solution containing 15 mM HEPES and 0.5 mM Lglutamine). The samples were homogenized on ice using a glass homogenizer, and the suspensions were pipetted onto a 70 μm cell strainer over a 15 ml conical tube and then centrifuged for 1 min at 1000× rpm at 4°C. The pellets were resuspended in FACS buffer (0.5% BSA, 0.1% NaN 3 and 25 mM HEPES in PBS) for single-cell suspension preparation. To increase the microglia yield, myelin was removed using Myelin Removal Beads II (Miltenyi Biotec, Bergisch Gladbach, Germany). Cells were blocked using Fc block prior to surface antibody staining at 4°C for 30 min in the dark. In total, 2.5 μl of

Transmission electron microscopy (TEM)
After perfusion with fixatives, small cubes of the lateroventral hippocampus and parenchymal layers were excised from the brains of normal and 21 dpi mice (n = 3 per group) and immediately fixed in 2.5% glutaraldehyde

Statistical analyses
Statistical analyses were performed using GraphPad Prism 5.0 (GraphPad Software, San Diego, USA) with two-tailed unpaired Student's t-test and SPSS 20.0 (IBM Corp., NY, USA) with one-way analysis of variance (ANOVA) followed by the Tukey-Kramer test. The data are presented as the means ± standard deviation (SD). Differences among comparisons were considered statistically significant for P-values less than 0.05.

Morris water maze test
During the 4 day place navigation test, significantly (F (3,20)  groups. In addition, disparity in the above two parameters between the 14 dpi and 21 dpi groups was statistically significant (F (3,20) = 60.19, P = 0.019 and F (3,20) = 73.31, P < 0.0001, respectively) ( Fig. 1a-c), while no significant differences in average swimming speed were observed between normal and infected mice (P > 0.05) (Fig. 1d). In the spatial probe test on day 5, no significant distinction among the groups in terms of platform area crossing were observed (P > 0.05) (Fig. 1e).
However, mice in the 14 dpi (26.47 ± 9.37 s) and 21 dpi (42.93 ± 18.34 s) post-infection groups significantly extended their time spent in the target quadrant compared with the 0 dpi (6.97 ± 2.48 s) and 7 dpi (11.48 ± 5 s) groups (F (3,20) = 14.25, P < 0.0001), while no significant difference in the time spent in the target quadrant between the 21 dpi and 14 dpi groups was observed (P > 0.05) (Fig. 1f ). The results indicated a clear timedependent reduction in the learning and memory retention capacities of mice infected with A. cantonensis.
Histopathological observation of the brains of mice infected with A. cantonensis H&E staining reflected the histopathological injuries and inflammation of mouse brains infected with A. cantonensis. No histopathological injuries or inflammatory cells were observed in the brains of the 0 dpi and 1 dpi groups. After 3 dpi, a few inflammatory cells infiltrated into the mouse brains, aggregating on and thickening the meninges (black arrows). Worm transection in the infected parenchyma illustrated that this parasite could move into the brain after only 3 days of infection (red arrows). As the infection time extended, more inflammatory cells infiltrated into the mouse brains, and their histopathological statuses became more serious (Fig. 2).

Apoptosis and necroptosis in infected mouse parenchymal and hippocampal cells as detected by immunohistochemistry (IHC)
Expression and localization of caspase-3, cleaved caspase-3, caspase-6, RIP3, LC3B and NeuN were detected in the 0 dpi and 21 dpi groups by IHC. As shown in Figs  caspase-3, the percentage of positive cells increased from 12.67 ± 3.06% to 51 ± 3.61% and 14 ± 4% to 55.33 ± 5.03% in the hippocampus and parenchyma, respectively. The number of RIP3-positive cells increased from 5.67 ± 1.53% to 54.33 ± 4.04% and 9.67 ± 2.08% to 88.33 ± 3.06% in the hippocampus and parenchyma, respectively. A mild elevation in the number of caspase-6-positive cells from 2 ± 1% to 23 ± 4.36% was observed in infected hippocampi, and no substantial changes were observed in the number of caspase-3positive cells. Neither normal nor infected mouse brains exhibited LC3B-positive cells, implying that A. cantonensis infection does not lead to autophagy. Different infectious stages were then measured by similar procedures. The number of cleaved caspase-3-positive cells gradually multiplied from 1 dpi to 21 dpi and peaked at 58 ± 2.65% in parenchymal cells and 56 ± 5.29% in hippocampal cells at 21 dpi (Figs. 9 and 10). Parenchymal RIP3-positive cells peaked at 14 dpi at 82.67 ± 2.31%, while hippocampal RIP3-positive cells peaked at 21 dpi at 53.67 ± 5.51% (Figs. 9 and 10).
The IHC results further proved that post-infectious apoptosis and necroptosis occur in mouse hippocampal and parenchymal cells.

Discussion
Eosinophilic meningitis or eosinophilic meningoencephalitis in humans is commonly caused by Angiostrongylus cantonensis, a food-borne zoonotic parasite, in southern Asia and the Pacific and Caribbean islands [14,15]. The clinical symptoms of angiostrongyliasis cantonensis are characterized by severe headache, neck stiffness, paraesthesias and cranial nerve palsy [16,17]. The neurological manifestations of neuroangiostrongyliasis include eosinophilic meningitis, encephalitis/encephalomyelitis, radiculitis and cranial nerve abnormalities, ataxia, tremors and paralysis, which have been observed in previous clinical reports [18][19][20] and in studies on infected experimental animal models [11,21]. These characteristics coincide with our results that infected mice exhibited extensive brain injury and neurological disorder, as detected by the Morris water maze test (Fig. 1) and H&E staining (Fig. 2). However, the pathogenesis of A. cantonensis neurological dysfunction remains poorly understood. The tracks and microcavities caused by the migration of larvae in the brains of infected animals, as described in previous studies [21], cannot explain all the clinical neurological injuries in the hosts [20] or the distinct outcomes between permissive and nonpermissive hosts after A. cantonensis infection since the same physical destruction of neural tissue caused by the migration of larvae can be seen within both of their brains [11]. Interestingly, our previous [22] and the current findings indicate the obvious apoptosis and necroptosis of parenchymal and hippocampal astrocytes, neurons and microglia of mice infected with A. cantonensis (Figs. 3-10), which might provide new insights into the pathogenesis of neuroangiostrongyliasis. The cell is the basic structural and functional unit of life. Cell death, as a natural process of life, plays an important role in the development and homeostatic balance of multicellular organisms [23][24][25]. According to morphology and molecular mechanisms, cell death is mainly divided into three categories: apoptosis, autophagy, and necroptosis (or necrotic cell death) [26]. Apoptosis is an automatic and programmed process regulated by signalling pathways, including the death receptor signalling, mitochondria-mediated and endoplasmic reticulum stress pathways [27,28]. All the pathways ultimately activate caspases to complete apoptosis. Caspase-3, -4, and -6 are important members of the caspase family [29], as their function in orchestrating apoptosis has been widely confirmed [30]. In this study, we discovered that compared with those of uninfected mice, the caspase-4 and -6 mRNA levels and cleaved caspase-3, -4, and -6 expression levels of A. cantonensis-infected mice were elevated. As shown in Figs. 3 and 4, A. cantonensis infection could cause the apoptosis of mouse brain cells. Furthermore, qPCR, western blot, immunohistochemistry, flow cytometry and TEM analysis further proved that apoptosis mainly occurs in parenchymal and hippocampal astrocytes, neurons and microglia during A. cantonensis infection (Figs. [5][6][7][8][9][10][11][12][13][14][15]. Although previous studies showed an elevation in the number of apoptotic leukocytes in post-infective parenchyma and subarachnoid space and the induction of apoptosis in endothelial cells and astrocytes after the addition of larvae extracts [31,32], we observed for the first time in this study that A. cantonensis infection caused significant apoptosis amongst parenchymal and hippocampal astrocytes, neurons and microglia, which could partially account for the severe neural damage observed (Fig. 1). In addition, we first confirmed that the levels of RIP3 and pRIP3 increased as the infection time extended (Figs. 3, 4). RIP3 is the key to apoptosis and necroptosis and a specific marker for necroptosis [33]. Thus, the increased RIP3 and pRIP3 levels indicated that A. cantonensis infection could induce the necroptosis of astrocytes, neurons and microglia (Figs. 5-15). Necroptosis is a recently recognized mechanism of programmed cell death that, unlike apoptosis, shows necrotic morphological changes with a patterned regulatory mechanism [34,35]. Necroptosis can be induced by TNF-α, the activator of programmed cell death [36][37][38], and is mediated by its unique RIP3dependent signalling pathway. In most cases, apoptosis and necroptosis co-exist, and a main difference between them is that necroptosis is accompanied by inflammation [39], causing inflammatory diseases, such as neurodegenerative disorders, haemorrhage-induced brain damage, Crohn's disease, multiple sclerosis and certain viral infections [40]. Our study on A. cantonensis infection-induced brain apoptosis and necroptosis is not only consistent with our previous work in that the level of TNF-α, a trigger of apoptosis and necroptosis that is increased in the brains of mice infected with A. cantonensis [41], can be directly induced by the cystatin from A. cantonensis (unpublished data), but also further explains the pathogenesis of severe neural dysfunction accompanied with extensive eosinophilic CNS inflammation during A. cantonensis infection.
As the macrophages of the brain, microglia play both protective and toxic (damaging) roles in non-infectious pathological states, such as brain damage or neurodegenerative disorders [42,43]. As the only immune cells in cerebral parenchyma, microglia are considered the first-line barrier protection for the CNS against pathogens (bacteria, viruses, parasites, etc.) [44][45][46]. Astrocytes are the most widely distributed cells in the mammalian brain and the largest population among glial cells [47]. Thus, astrocytes play significant roles in survival functionality, neuronal development and information transmission, neuronal maturation and synapse formation, CNS metabolism, and damaged blood-brain barrier repair and reconstruction [48,49]. Neurons are the basic structural and functional units of the nervous system, and the key to neural function is the synaptic signalling process, which is partly electrical and partly chemical [50]. The essential function of neurons is to exchange messages through information input, integration, conduction and output. Thus, neurons are crucial to the sensation, learning, memory, cognition and movement of organisms [51,52]. The apoptosis or necroptosis of microglia, astrocytes or neurons can lead to severe neural dysfunction (epilepsy, Alzheimer's disease, cerebrovascular disease, encephalitis, meningitis, etc.), or even death [53,54]. Apoptosis and necroptosis occurred in microglia, astrocytes or neurons after A. cantonensis infection, indicating that neurotrophic agents and inhibitors of apoptosis and necroptosis could be effective in combination with anti-parasitic drugs (albendazole) and steroidal anti-inflammatory drugs (dexamethasone).

Conclusions
In summary, this study is the first to demonstrate that A. cantonensis infection can significantly induce the apoptosis and necroptosis of astrocytes, neurons and microglia in host brain parenchymal and hippocampal tissues. This not only enriches our understanding of A. cantonensis pathogenesis but also provides a clue to novel potential therapeutic strategies against angiostrongyliasis cantonensis.