Neurotropin alleviates cognitive impairment by inhibiting TLR4/MyD88/NF-κB inflammation signaling pathway in mice with vascular dementia

Vascular dementia (VD) is the second most common cause of dementia after Alzheimer's disease. Neuroinflammation contributes to pathogenesis of VD. Neurotropin (NTP) is an analgesic that has been shown to suppress inflammation and neural repair. But its effects on VD are still unclear. Therefore, this study aimed to investigate the therapeutic effects and potential mechanisms of NTP in the VD model mice established by bilateral common carotid artery stenosis method. In VD mice, we found that NTP treatment increased cerebral blood flow by Laser speckle imaging, reduced neuron loss by Nissl, HE and immunochemistry staining, attenuated white matter damage by magnetic resonance imaging and ultrastructural damage by transmission electron microscope, improved cognitive functions by new object recognition test and three-chamber test, Y maze test and Morris water maze test, inhibited significantly glial activation by immunofluorescence methods, reduced the expression of TLR4, down-regulated expression of MyD88 and phosphorylation of NF-κB P65, decreased the levels of pro-inflammatory cytokines IL-1β, IL-6 and TNFα. Further, we showed that administration of a TLR4 inhibitor TAK242 had a similar effect to NTP, while the TLR4 agonist CRX-527 attenuated the effect of NTP in the VD mice. Collectively, our study suggested that NTP alleviates cognitive impairment by inhibiting TLR4/MyD88/NF-κB inflammation signaling pathway in the VD mice. Thus, NTP may be a promising therapeutic approach and a potential TLR4 inhibitor for VD.


Introduction
The prevalence of cerebrovascular diseases increases with age, affecting life quality of patients seriously and putting heavy burden on the society (Rajeev et al., 2022).Vascular cognitive impairment occurs gradually and develops into vascular dementia (VD) in the end.However, there are no effective treatments at present.Investigators are exploring the pathogenesis of VD.Inflammation and apoptosis, as two candidate factors, play an important role in VD (O'Brien and Thomas, 2015).White matter degeneration is one of the most obvious changes in brain structures (Hase et al., 2018).Loss of axons, myelin sheath injury, lacunar infarction (Kalaria, 2018), and changes in vascular parenchyma lead to white matter degeneration (Kalaria, 2018).Chronic ischemic hypoxia, breakdown of the blood-brain barrier and pericyte degeneration are potential factors for vascular cognitive impairment (van der Flier et al., 2018).Destruction of blood-brain barrier may accelerate the process of white matter degeneration.Furthermore, activation of microglia and astrocytes also present in lesions.Studies have shown that inflammation is associated with white matter damage (Iadecola, 2013).Therefore, combating inflammation might be an effective treatment.
Neurotropin (NTP), a nonprotein extract of inflamed rabbit skin inoculated with vaccinia virus.It has been used for nearly 70 years in clinical practice.Also, it is widely used to treat chronic pain in China and Japan.A recent study reported that NTP improved cognitive function and attenuated inflammation via brain-derived neurotrophic factor (BDNF)/nuclear factor-kappa B (NF-κB) pathway in the cortex and the hippocampus APP/PS1 mice.In addition, NTP down-regulated the expression of inflammatory cytokines, such as interleukin-1 beta (IL-1β), interleukin-6 (IL-6) and tumor necrosis factor α(TNFα), both in vivo and in vitro (Fang et al., 2019).In Ts65Dn mice, repeated NTP treatment improved learning and memory (Fukuda et al., 2010).Mutoh et al. demonstrated that NTP prevented early brain injury and improved motor and neurocognitive function after subarachnoid hemorrhage (Mutoh et al., 2021).It was also suggested that NTP inhibited the expression of the NF-κB target gene and JNK phosphorylation induced by IL-1β and TNFα in hepatocytes (Zhang et al., 2014).NTP treatment has been reported to enhance spatial learning accompanied by upregulation of BDNF, in addition to reduce infarct volume in ischemic models.Thus it protected the brain from ischemic stroke and improve brain functions (Nakajo et al., 2015).Zheng et al. investigated anti-inflammatory effects of NTP in BV2-stimulated microglia and its mechanisms.It is reported that NTP inhibited NF-κB activation by inhibiting nuclear NF-κB p65 translocation to cell nucleus, and inhibited mitogen-activated protein kinases (MAPKs) pathway by inhibiting phosphorylation of P38, extracellular regulated protein kinases (ERK), and c-Jun N-terminal kinase (JNK) (Zheng et al., 2018).Moreover, NTP protected hypoxic-ischemic mice, improved survival rate and neurological function, and reduced the expression of pro-inflammatory factors IL-1β, IL-6, and TNFα.Similarly, the expression of three cytokines was also inhibited in the mouse sciatica model (Nishimoto et al., 2016;Hishiyama et al., 2019).To sum up, there is increasing evidence to support the beneficial effects of NTP in suppressing inflammation and improving neural repair.However, the effects of NTP for VD were unknown.Here, we explored the effects of NTP treatment for VD and its underlying mechanisms.
Toll-like receptors (TLRs), a family of pattern recognition receptors, play an important role in inflammation, immune response, pathogen recognition and as a result are involved in the pathogenesis of many diseases (Leitner et al., 2019).Of these TLRs, toll-like receptor 4 (TLR4) is the most well-studied molecule (Shi et al., 2020;Paladugu et al., 2021).TLR4 is mainly expressed in glia, including microglia, astrocytes, and oligodendrocytes (Shi et al., 2020).In the brain, microglia exhibit inflammatory responses through pro-inflammatory mediators; TLR4 expressed in microglia plays a key role (Adhikarla et al., 2021).Binding of the ligand myeloid differentiation factor 88 (MyD88) is not only a key downstream signal transduction ligand of the TLR4 receptor complex, but also an essential adaptor of the NF-κB signaling pathway (Kawasaki and Kawai, 2014).Moreover, TLR4 plays an important role in MyD88-dependent inflammatory responses (Kawasaki and Kawai, 2014).Thus, Binding of MyD88 to the TLR4 domain triggers a complex signaling cascade, ultimately leading to activation of pro-inflammatory factors IL-1β, IL-6 and TNFα (Jack et al., 2005;Miller et al., 2005).The expression level of TLR4 in the VD brain was significantly increased, and was related to the severity of neuron injury (Belkhelfa et al., 2018).Overexpression of TLR4 can lead to activation of downstream signaling pathway MAPKs, and in turn activation of transcription factors (such as NF-κB) and expression of pro-inflammatory factors (Adhikarla et al., 2021).TLR4 activates the MyD88-dependent pathway, resulting in nuclear translocation of NF-κB and producing inflammatory cytokines (IL-1β, IL-6, TNFα) (Kawasaki and Kawai, 2014).Inhibition of TLR4/NF-κB has been shown to protect the brain from injury (Song et al., 2018;Shi et al., 2020;Zhu et al., 2021).Previous studies have shown that neuroinflammation is a key factor in VD (Belkhelfa et al., 2018).Thus, inhibition of TLR4 against inflammation might be an effective VD treatment.
Based on these studies, we proposed that NTP might ameliorate cognitive functions by inhibiting neuroinflammation in a VD mice model, and TLR4/MyD88/NF-κB signaling pathway mediated the effects of NTP.

Animals
All male C57BL/6J mice (8-10 weeks old, 23 ± 2 g) were provided by Southern Medical University Animal Experiment Center (Guangzhou, China).The animals were kept in cages access to water and food, under a 12 h; 12 h light-dark cycle.The ambient temperature and humidity were controlled at 21-23 • C and 40-60%, respectively.The animal experiments were determined to minimize the number and suffering of animals.Protocols for our experiments were approved by the Animal Management and Use Committee of Southern Medical University.

Establishment of VD mice model and drug treatment
We established the VD mice model using a bilateral common carotid artery model (BCAS) (Shibata et al., 2004;Shibata et al., 2007;Ihara et al., 2014).After two weeks of adaptation, all mice were randomly divided into five groups: the sham group, the BCAS/normal saline (NS) group, the BCAS/NTP group, the BCAS/TAK242 group and the BCAS/NTP/CRX-527 group (n = 12, each group).
Specifically, the mice were anesthetized with 3.0% isoflurane and placed in the supine position.To expose both common carotid arteries, a small incision was made in the midline of the neck.Microcoils of 0.18 (0.18 mm diameter, 2.5 mm length, Sawane Spring) were placed in both common carotid arteries.Put on one side of the microcoil first, and left the other side about an hour later.The intraoperative and postoperative temperature was kept on a heating pad at 36.5 ± 0.5 • C. When the mice recovered from the anesthesia, they were free to move around the cage and access to water and food.Three days after surgery, both the BCAS/ NS group and the BCAS/NTP group were given saline and NTP (50NU/ Kg, once per day, provided from Nippon Zoki Pharmaceutical Co.Ltd.) for 28 days (Fang et al., 2019), respectively.

New object recognition test
A new object recognition(NOR) experiment was conducted in a black plastic with a size of 29 cm × 47 cm × 30 cm under dim light.The stimulants consisted of two blocks with similar size and different colors.A camera mounted on top of the box was used to record the experiment.The mice were acclimated in the test room for 1 day and in the empty box for 5 min before the experiment began.On the next day of adaption, the mice were placed in the same box with two identical objects to explore for 5 min.One hour later, one object was replaced with a new one.The following behaviors: when the mouse sniffed or touched the new object or placed a front paw on the new object, were defined as exploring.But climbing on the new object was not regarded as exploring.The experiment was terminated when the total exploration time reached 20 s.Exploratory behavior was measured by an experimenter who was blinded to these groups.Discrimination ratio % = new object exploration time/(new object exploration time + old object exploration time) × 100%.A higher discrimination ratio indicates that the animal spent more time to explore new objects, suggesting recognition and memory of familiar objects.
H. Zou et al.

Three-chamber test
We evaluated social behavior and social memory with three-chamber test (3CT) (An et al., 2021;Lee et al., 2022) using plexi -glass boxes with two divided openings (62 cm length × 41 cm width × 21.5 cm height).During the habituation phase, the mice were free to explore three chambers for 10 min.In the test phase, the test mice were given the option of interacting with an unfamiliar mouse in a wire cage (S1) or an empty wire cage.Two cages were placed in the left and right compartments.For social novelty task, the test mice were given the choice of interacting with either a stranger mouse 1 (S1) or a stranger mouse 2 (S2), which were kept in wire cages in a separate chamber.The time spent the interaction with S2 vs. S1 was recorded for the task.High S2 exploration time percentage was considered to be good memory.

Y-maze test
Short-term spatial memory was assessed using the Y-maze test, consisting of three arms at an angle of 120 • (35 cm length × 10 cm width × 10 cm height).The mice were placed in the center of the Y puzzle and allowed to explore freely for 5 min.Once all limbs entered one arm, the data was recorded and analyzed through free alternate times.Continuous access to all three arms was the definition of alternating behavior.The number of maximum alternations was equal to the total number of arm entries − 2. Alternation percentage = alternation times/(total number of arm entries − 2).Higher alternation percentage suggests better spatial memory.

Morris water maze
After 28 days of NTP treatment, the mice were trained and tested to evaluate spatial learning and memory by Morris water maze (MWM).The MWM test was carried out on a round pool (150 cm in diameter; Height 45 cm).The pool was divided into four quadrants.An underwater platform (10 × 10 cm and 28 cm high) was placed in the middle of target quadrant of the pool; Later, the water was filled to the level of 2 cm above the platform.During acquisition training, the animals underwent training tests four times a day for five consecutive days.The animals were allowed 120 s to find the platform.After failing to find the underwater platform, the animals were guided to the platform and allowed to stay there for 15 s.The escape latency, total distance, and percentage of time spent in the target quadrant were recorded.

Cerebral blood flow by laser speckle imaging
For each recording, the skull surface was wiped with a saline soaked gauze, then covered with a thin gel to prevent drying.Laser speckle imaging (LSI) provides an economical and efficient method to measure blood flow in a wide field of view with high spatial and temporal resolution.LSI has been used for imaging blood flow in tissue perfusion.The relative velocity of blood cells was calculated by integrating the intensity fluctuations of the speckle pattern produced by random interference of scattered coherent light over a finite period.Values of cerebral blood flow (CBF) were measured by LSI (RFLSI, RWD, China), before surgery, 2 h,1d, 3d, 7d, 14d, and 28d after surgery.The mice were anesthetized with isoflurane and placed on a thermostatic heating pad to maintain the rectal temperature between 36.5 • C and 37.5 • C. The skull was exposed through a midline scalp incision.CBF values were expressed as a percentage of the preoperative value.

Magnetic resonance imaging measurements
To determine the treatment of NTP on white and grey matter, Magnetic Resonance Imaging (MRI) was performed (Ihara et al., 2014;Hua et al., 2015;Yang et al., 2020).Diffusion tensor imaging (DTI) data was collected from mouse brains using a 7.0T small animal MRI system (PharmaScan70/16 US, Bruker Biospin MRI GmbH, Germany), and radio-frequency emission and reception were performed using an 86 mm inner diameter independent emission volume coil and a mouse head surface coil.During the scanning process, isoflurane and oxygen were used to anesthetize the mice and maintained the respiratory physiology of the mice.The mice were anesthetized with 3% isoflurane first and then anesthetized with 1-1.5% isoflurane during the scanning process.The mice were fixed on the animal bed with tooth rods and ear rods to avoid the movement of their heads as much as possible.The mice were scanned 28d after surgery or treatment by MRI DTI.
Image analysis: The following regions of interest were included: Corpus Callosum -CC, Internal Capsule-IC, Anterior Commissure -AC, Caudate Putamen -Cpu, Optic Tract-OT.FA (Fractional Anisotropy), MD (Mean Diffusivity), AD (Axial Diffusivity) and RD (Radial Diffusivity) value of the above five regions were calculated.Anisotropy fraction (FA) is one of the main parameters to describe the anisotropy characteristics of MRI scheme including DTI.The axial extension (AD) is often used to reflect the integrity of axons.The vertical diffusion tensor (RD) reflects the diffusion perpendicular to the axonal direction; the change of RD value reflects the pathological changes of demyelination.Mean diffusivity (MD), sensitive to cell, edema, and necrosis, reflects membrane density (Ihara et al., 2014;Hua et al., 2015;Yang et al., 2020).

Transmission electron microscope
Approximately 1 mm 3 hippocampal tissue was taken and immobilized in 2.5% glutaraldehyde solution for 24 h.They were then stained with 1% osmium tetroxide for 1 h.After gradient dehydrated with acetone, embedded with epoxy resin, ultra-thin sections (60 nm thick) were made, stained with 1% uranium acetate citric acid, and observed under projection-electron microscope (Ht7700, Hitachi).The length of the postsynaptic membrane was calculated.

Nissl staining
On the second day after behavioral test, six mice from each group were randomly selected for Nissl staining.The mice were anesthetized with pentobarbital sodium (2%, 50 mg/kg) and injected with NS through the left ventricle until the blood was flushed clean, followed by 4% infusion of paraformaldehyde.After perfusion, the brain was taken out and immersed in 4% paraformaldehyde solution for 24 h.Paraffin sections (4 μm) containing the hippocampus were cut in coronal sequence.Paraffin sections of brain tissue were dewaxed and hydrated, then stained with Nissl solution at 37 • C for 1 h, and washed with deionized water for three times.Slices of the same part of each mouse were examined and then photographed under Olympus microscope.Morphological changes of hippocampal neurons in CA1 and CA3 regions were examined.The number of neurons was calculated in CA1 and CA3 areas.

H&E staining
H&E staining was performed using a standard protocol.Paraffin sections of mouse hippocampus were dewaxed in xylene, hydrated in gradient alcohol and stained with hematoxylin and eosin.Pathological changes in CA1 and CA3 regions were observed under a light microscope, and the percentage of positive cells (ratio to the sham group) was calculated in each image.

Immunofluorescence staining
After 28 days of NTP treatment, the mice were anesthetized and infused with NS and 4% paraformaldehyde.The brains were removed and left in 4% paraformaldehyde solution for 24 h, followed by dehydration with 30% sucrose.After freezing, coronal sections were continuously cut into 4 μm frozen sections.It was incubated in a blocking solution (0.3% Triton X-100 in PBS,10% normal goat serum) at room temperature for 2 h and then incubated in primary antibodies including microtubule associated protein-2 (MAP2) (Abcam, Cambridge, UK), NeuN (Abcam, Cambridge, UK), microglia IBA1(Abcam, Cambridge, UK) and astroglia GFAP (Abcam, Cambridge, UK).Then sections were incubated in a mixture of fluorescent secondary antibodies (Alexa 488/Alexa 694-conjugated anti-mouse/anti-rabbit IgG).All images were acquired using a fluorescence microscope (Carl Zeiss, Jena, Germany).Quantitative analysis of MAP-2 by immunofluorescence density was performed and the positive cells of IBA1 and GFAP in CA1 and CA3 regions in each image were quantified by Image J (National Institutes of Health, MD, USA).

Western blot
28 days of NTP treatment later, all mice were sacrificed and their brains were taken out quickly.Then the hippocampi were separated quickly.Appropriate amount of cleavage buffer was added, which contained 50 mmol/L Tris-HCl, pH 7.5, 100 mmol/L NaCl, 1% Triton X-100 (RIPA, CST, USA) and phosphatase inhibitors (Cocktail, Sigma-Aldrich, St Louis, MO, USA).The same amount of protein (30 μg/lane) was separated by SDS-PAGE and transferred to PVDF membranes (Millipore, Beijing, China) by electrophoresis.The membranes were blocked with 5% non-fat milk for 1 h, then primary antibodies including TLR4 (Abcam, Cambridge, UK), MyD88 (Abcam, Cambridge, UK), P65 (Cell Signaling Technology, MA, USA) and pP65 (Cell Signaling Technology, MA, USA) were incubated overnight at 4 • C. Washed the membrane with TBST on the next day (10 min three times), then incubated with secondary antibody (Santa Cruz Biotechnology, CA, USA) conjugated with horseradish peroxidase for 1 h at room temperature.Next, washed three times with TBST (10min for each time).Protein bands were detected using ECL + chemiluminescent reagent kits (Millipore, Bedford, MA, USA).Quantification of TLR4, MyD88 and pP65 levels of protein were normalized to the internal control.

Statistics analysis
All the results were expressed as the mean ± standard error of the mean (SEM) or absolute numbers or percentage, unless otherwise indicated.Continuous data that were normally distributed according to the D'Agostino-Pearson normality test were compared using one-way ANOVA with post hoc Bonferroni-Dunn correction where appropriate.For variables that did not pass the normality test, the Mann-Whitney test (for comparisons between two groups) or Kruskal-Wallis test, followed by the Dunn's multiple comparison test (for comparisons among three or more independent groups), was performed.All statistical analyses were performed using SPSS (version 26; IBM, Chicago, IL, USA) and Prism (version 9: GraphPad Software, San Diego, CA, USA).A p-value less than 0.05 was considered statistically significant.

NTP ameliorates cognitive impairment in the VD mice
In our preliminary experiments (data not shown), VD mice were randomly assigned to three groups: BCAS + NTP at 25 NU/kg; BCAS + NTP at 50 NU/kg; BCAS + NTP at 100 NU/kg.Then, behavior tests, including NOR and Y maze, were performed.We found that NTP at NU/kg ameliorated cognitive dysfunctions better compared to NTP at NU/kg in the VD mice, and NTP at 100 NU/kg achieved a comparable effect of 50 NU/kg, consistent with previous study (Mutoh et al., 2021;Yao et al., 2021).Thus, we chose 50 NU/kg NTP for subsequent experiments.
BCAS-induced learning and memory impairment was assessed by NOR and Y-maze test.Our results showed that the NOR index of the BCAS/NS group was lower than the sham group (F (2, 33) = 117.3,P＜0.001), while NTP treatment increased the preference for new objects significantly (P＜0.001) (Fig. 1A).Similar results were obtained by Ymaze test (Fig. 1B).These data suggested that NTP ameliorates shortterm memory impairment induced by BCAS.
To further analyze the changes in social cognitive functions, we measured social cognition ability by 3CT.We found that the BCAS/NS group showed social behavioral deficit (F (2,23) = 49.03,P＜0.001),while NTP treatment reversed the deficit (P＜0.001) (Fig. 1C).
In order to test spatial memory, mice were tested for MWM after NTP treatment.NTP significantly reduced the target quadrant time latency (P ＜0.001) and total distance (P＜0.001), and increased the target quadrant time percentage (P＜0.001)(Fig. 1D,E,and1F).However, VD mice injected with NS did not reverse BCAS-induced cognitive impairment.Taken together, these results suggested that NTP rescues BCAS-induced spatial memory impairment.

NTP increases CBF in the VD mice
The reduction of CBF is observed in the VD (Shibata et al., 2004).Thus, increasing CBF might be a therapeutic strategy for VD.To observe the effects of NTP on CBF, we detected the change of CBF following NTP by LSI.In the BCAS/NS group, LSI showed that CBF dropped to about 70% 2 h after surgery and began to recover on the first day and remained low until day 28 (Fig. 2A and B), consistent with previous studies (Shibata et al., 2004).CBF in BCAS/NS group was significantly decreased when compared to the sham group (F (2,6) = 204.506,P = 0.000).However, compared with the BCAS/NS group, NTP improved CBF after 7 days of postoperative administration and maintained to the 28th day (F (2,6) = 71.936,P = 0.001) (Fig. 2A and B).These results suggested that NTP increased CBF in the VD mice model.

NTP alleviates neuronal ultrastructure damage in the VD mice
Synaptic structural plasticity, like functional plasticity, plays a key role in cognitive function and interacts with functional plasticity (Vierk et al., 2015).To further explore whether NTP attenuated neuronal ultrastructure damages induced by BCAS, transmission electron microscope was used to observe the ultrastructure of myelin sheath, synapses, mitochondria and neurons in the hippocampus among the three groups (Fig. 4A).In the sham group, the myelin sheath structure was clear and showed lamellar structure surrounding the axon.In the BCAS/NS group, the lamellar structure and normal structure were disorganized and vacuolar, even destroyed.Conversely, the extent and area of demyelination were clearly reduced in the BCAS/NTP group (black arrows).As shown in the quantization chart, the percentage of demyelinating parts in the BCAS/NS group was higher than that in the sham group(t (8) = 12.52, P＜0.001).However, it was improved after NTP treatment (t (8) = 4.303, P = 0.0026, Fig. 4B).Next, we examined the synaptic structure, and observed that BCAS decreased length of the postsynaptic density (PSD) (t (47) = 2.256, P = 0.0288,compared to the sham group), while NTP prevented the change (t (33) = 3.941,P＜0.001,compared to the BCAS/NS group) (Fig. 4C).Synapses in the sham group showed clear synaptic membranes with synaptic vesicles and postsynaptic density.However, mice in the BCAS/NS group showed membrane rupture, synaptic cleft fusion and even loss of synaptic vesicles at presynaptic and postsynaptic densities.In contrast, NTP treatment clearly restored visible synaptic structures compared to those observed in BCAS mice (red arrows).These data suggested that NTP reduces damage to synaptic ultrastructure, thereby enhancing synaptic plasticity.No serious damage was observed in the mitochondrial crest and morphology in the sham group.The BCAS/NS group showed more mitochondrial crest fracture, outer membrane destruction and cavitation phenomenon.The BCAS/NTP group exhibited less damage than the BCAS/NS group (green arrows).As shown in Fig. 4A yellow arrows, no swelling and degeneration occurred in the sham group, but the shrinkage of the nuclei and the irregular nuclear membranes were presented in the BCAS/NS group.Clear nuclear membrane, less edema and numerous organelles were still visible in the BCAS/NTP group.Collectively, NTP treatment attenuates the BCAS-induced damage in ultrastructure.

NTP restores hippocampal tissue damage in the VD mice
To investigate BCAS-induced hippocampal damage and the effects of NTP on this damage, we used Nissl and HE staining and immunofluorescence staining.Pathological observation showed significant differences among the three groups.As shown in Fig. 5A, the hippocampal neurons in the sham group were full in shape, regular in arrangement and normal in density.The number of cells in the hippocampal CA1 area (F (4,47) = 27.15,P= 0.000)and CA3 (F (2, 15) = 62.19,P < 0.001) area drastically reduced in the BCAS/NS group compared with the sham group, with irregular arrangement, cell disorder and abnormal morphology of neurons.In the BCAS/NTP group, most hippocampal neurons were regular in the morphology, with less loss of neurons.The statistical results presented that the neuronal number in CA1 (P＜0.001) and CA3 (P＜0.001) regions of the BCAS/NS group were much less relative to the sham group, while the number of neurons in CA1 (P＜ 0.001) and CA3 (P＜0.001)regions of the BCAS/NTP group was increased markedly compared to the BCAS/NS group (Fig. 5B).HE staining showed neat and compact arrangements of hippocampal neurons in the sham group.In the BCAS/NS group, some nuclei of neurons in the hippocampal CA1 (F (2, 13) = 46.72,P < 0.001) and CA3 (F (2, 12) = 37.48, P < 0.001) regions were hyperchromatic, pyknotic or even disappeared.However, these irregularities were significantly improved in the BCAS/NTP group (Fig. 5C).The HE statistical results showed that the pyramidal cell number in CA1 (P＜0.001) and CA3 (P＜0.001) of the BCAS/NS group were much less than the sham group.And the pyramidal cells in CA1(P = 0.0014) and CA3 (P = 0.0199) of the BCAS/NTP group increased comparing to the BCAS/NS group (Fig. 5D).
Through co-localization of NeuN and MAP-2 in CA1 (F (2, 15) = 8.609, P = 0.0032) and CA3(F (2, 15) = 11.996,P = 0.001) regions, neuronal nuclei in the model group were blurred and MAP-2 was discontinuous expression related to the sham group, indicating possible neuron damage.In contrast, the BCAS/NTP group showed mitigated irregular changes and was more continuous expression (Fig. 5E and F).Quantitative analysis showed that the fluorescence density of the BCAS/NS group was significantly lower than the sham group in CA1(P = 0.0072) and CA3 (P＜0.001) regions and was increased markedly in the BCAS/ NTP group comparing to the BCAS/NS group in CA1(P = 0.0021) and CA3(P = 0.0011) regions (Fig. 5G and H).To sum up, NTP prevents hippocampal damage in the VD mice model.

NTP decreases glia proliferation induced by BCAS
Glial cells, especially astrocytes and microglia, play an important role in neuroinflammation.Previous studies have detected the activation of glial cells in the brains of animals with chronic hypoperfusion (Shibata et al., 2004;Choi et al., 2011).To investigate whether NTP affected activation of microglia and astrocytes in the hippocampus of VD mice, IBA1 and GFAP were used as markers of microglia and astrocytes, respectively.In this study, the number of microglia in the BCAS/NS group was significantly increased in CA1(F (4, 15) = 27.456,P = 0.000) region and these microglia had larger cell bodies as compared to the sham group (Fig. 6A and B).After NTP treatment, a less number of activated cells was observed compared to the BCAS/NS group in CA1 region (P＜0.001).Microglia had less dendritic branching, reduced    H. Zou et al. branch length and decreased complexity of dendritic trees in the hippocampal CA1 region of the BCAS/NTP group (Fig. 6A and B).
Similarly, reactive astrocytes in CA1 region were observed an increased number in the BCAS/NS group (F (4, 25) = 15.787,P＜0.001),which manifested swelling, an increase in astrocyte volume indicative of both hypertrophy and activation(Fig.6C and D).On the contrast, reactive astrocytes were observed a decrease in the BCAS/NTP group, with lower density and decreased volume compared to the BCAS/NS group (P ＜0.001) (Fig. 6C and D).These findings suggested that NTP attenuates BCAS-induced glia responses.

TLR4 mediates NTP anti-inflammation effect in the VD mice
Given that the neuroprotective effects of NTP may be mediated at least in part by TLR4, we hypothesized that TLR4 antagonists might have similar effects to NTP in inhibiting inflammation and ameliorating memory, whereas TLR4 agonists might reverse the protective effect of NTP.To further elucidate whether the anti-inflammatory effect of NTP was at least in part mediated by TLR4, the specific TLR4 antagonist TAK242 and the specific agonist CRX-527 were applied.Mice were randomly assigned to the sham group, the BCAS/normal saline (NS) group, the BCAS/NTP group, the BCAS/TAK242 group and the BCAS/ NTP/CRX-527 group (each group, n = 12).
Next, we observed the effects of the two inhibitors on glial cells.More IBA-1 positive cells were observed in the BCAS/NTP/CRX-527 group than the BCAS/NTP group in the CA1 and CA3 regions of the hippocampus (CA1: P＜0.001; CA3: P = 0.004) (Fig. 6A and B).Additionally, GFAP fluorescence staining showed a great increase of the positive cell number in the BCAS/NTP/CRX-527 group than the BCAS/ NTP group in CA1 and CA3 regions(CA1: P = 0.0037; CA3: P＜0.001) (Fig. 6C and D).
To assess whether improved memory function was associated with TLR4 in the BCAS/NTP group, mice were subjected to the Y-maze test after treatment.In comparison with the sham group, the BCAS/NS group showed a significant decrease in spontaneous alternation rate in Y-maze test (F (4,47) = 27.15,P= 0.000).NTP and TAK242 treatment increased significantly % correct alternation compared to the BCAS/NS group.However, the therapeutic effect of NTP on % correct alternation of the BCAS mice was apparently abolished by CRX-527 administration (P＜ 0.001, compared to the BCAS/NTP group).Additionally, there was no difference between BCAS/NTP and BCAS/TAK242 group(P＞0.5)(Fig. 8), suggesting that over-activation of TLR4 reverses the effects of NTP on memory impairment, and further TLR4 mediated memory improvement of NTP in the VD mice.
Together, these results indicated that NTP exerts its antiinflammatory effect by inhibiting TLR4/MyD88/NF-κB signaling pathway in the VD mice model.

Discussion
With the increase in incidence, investigators are paying more attention to the mechanisms of VD.Long-term cerebral hypoperfusion is the main cause.Inadequate blood flow activates a cascade of cellular and molecular reactions, leading to a breakdown of the blood-brain barrier and neurodegeneration (Rajeev et al., 2022).Chronic cerebral hypoperfusion with reduced CBF can result in brain damage through oxidative stress (Toyama et al., 2014), apoptosis (Choi et al., 2014), neuroinflammation (Wang et al., 2020), and abnormal energy metabolisms (Li et al., 2003).
Although NTP has been used in clinical practice for more than 70 years, its therapeutic efficacy in VD has not been explored, and the underlying mechanisms of NTP are still understood poorly.In this study, we reported for the first time that NTP significantly ameliorated memory impairment and promoted white matter remodeling and neuroprotection in the VD mice by inhibiting inflammatory pathways.
The main symptoms of VD are reduced cognitive functions and impaired learning and memory (Tsivgoulis et al., 2014).In our study, NOR test, Y-maze test, 3CT and MWM test showed reduced memory in the BCAS/NS group, consistent with previous studies in animal models of chronic cerebral hypoperfusion (Farkas et al., 2007;Balucani et al., 2012;Washida et al., 2019;Han et al., 2020).Importantly, in our study, NTP administration alleviated BCAS-induced memory impairment.Consistently, recent studies have shown that NTP improves cognitive function through the BDNF/NF-κB pathway in APP/PS1 double transgenic mice (Fang et al., 2019).It has also been reported that oral NTP at a daily clinical dose enhanced spatial memory in mice in a temporary focal ischemia model, protected the brain from ischemic stroke, and was accompanied by upregulations of BDNF (Nakajo et al., 2015).Similarly, Fukuda et al. showed that repeated oral NTP for three months improved spatial cognition in Ts65Dn mice and prevented the down-regulation of BDNF, which proved to be a beneficial treatment for degenerative diseases (Fukuda et al., 2010).These studies suggested that NTP improves cognitive functions in the different disease models.Abnormal neuroinflammation is one of the preconditions leading to endothelial and neuronal cell damage in VD.Neuroinflammation exacerbates the process of cognitive dysfunction in chronic cerebral hypoperfusion (Wang et al., 2020).As the first line of defense against inflammation in the brain, microglia promotes the release of cytokines (TNFα and IL-1β) and MCP-1, ultimately leading to neuroinflammation and apoptosis (Feng et al., 2017;Rosenberg, 2017;Alam et al., 2018).Some studies have shown that neuroinflammation is associated with cognitive impairment.For example, the research of Bradburn et al. supported the cognitive impairment caused by neuroinflammation in mild cognitive impairment and Alzheimer's disease (Bradburn et al., 2019).NTP has been shown to reduce neuroinflammation and alleviate cognitive deficits in APP/PS1 mice through the BDNF/NF-κB pathway (Fang et al., 2019).In LPS-stimulated BV2 cells, NTP reduced NF-κB P65 translocation to the nucleus and the expression of TNFα and IL-6 through NF-κB and MAPKs signaling pathways (Zheng et al., 2018).Inhibition of NF-κB and JNK in hepatocytes attenuates IL-1β and TNFα -mediated inflammatory cytokine expression and hepatocyte death (Zhang et al., 2014).NTP has also been reported in clinical studies to help treat cerebral edema associated with acute ischemic stroke (De Reuck et al., 1994).Consistent with these publications, we reported that NTP reduces neuroinflammation and the release of pro-inflammatory cytokines IL-1β, IL-6, and TNFα in mice treated with VD.These data suggested that NTP alleviated cognitive dysfunction in VD model mice through anti-inflammatory effects.
It is well known that TLR4 contributes to triggering inflammation, and TLR4 promotes the release of inflammatory cytokines IL-1β, IL-6, and TNFα (Tsivgoulis et al., 2014;Leitner et al., 2019).Several studies have shown that TLR4 is a therapeutic target for cerebral ischemia (Han et al., 2020).Previous studies have shown that TLR4 inhibition reduces oxidative stress-induced injury, decreases ferroptosis activation, and alleviates neuroinflammation after hypoxic ischemic brain damage (Feng et al., 2017).Fukuda et al. proposed that TLR4 was involved in the neuroprotective action of NTP and Trk neurotorphin receptors (Fukuda et al., 2020).Experiments conducted by Fukuda et al. showed new membrane fractions contained TLR4.Inhibition of TLR4 function by TAK242 prevented the formation of these unidentified membrane fractions, TLR4 activity is required for the formation of unidentified raft-like fractions with NTP treatment.In addition, TAK242 strongly blocked neuroprotective effects of NTP.In our experiment, NTP was treated in the VD model in the pathological state of inflammation activation.Once inflammation was over-activated, NTP could inhibit the over-expressed TLR4 receptor.Therefore, blocking TLR4 inhibits the membrane fractions.The MyD88-dependent pathway is the first and most important way (Zhu et al., 2021).Activation of innate immune receptors (such as TLR4) activates NF-κB via MyD88, thereby promoting NF-κB nuclear translocation (Shibata et al., 2004;Song et al., 2018).TLR4 knockout mice reduced neuroinflammation in the brain and may play a protective role in mice with traumatic brain injury (Ahmad et al., 2013).In LPS-induced neuroinflammation models, hesperidin treatment decreased reactive glial cell activation, inhibited the release of pro-inflammatory factors, protected neurons, and improved cognitive function by inhibiting TLR4/NF-κB signaling (Ahmad et al., 2013).Effects of intrathecal epigallocatechin gallate (EGCG), an inhibitor of TLR4, EGCG can reduce the expression of IL-1β and TNFα.Moreover, IL-10 expression could be increased to improve hyperalgesia in sciatic nerve rats (Kuang et al., 2012).These reports suggested that inhibition of TLR4 attenuates inflammation, protects neurons from damage and improves cognitive function, serving as a promising therapeutic target.Correspondingly, our study showed that BCAS increased the expression of TLR4 and its adaptor MyD88 and its downstream protein NF-κB P65, suggesting activation of TLR4 in the VD mice.To further verify whether the effects of NTP treatment were mediated by TLR4/MyD88/NF-κB signaling pathway, VD mice were treated with TLR4 inhibitor TAK242.TAK-242 selectively binds to TLR4 and subsequently disrupts the interaction between TLR4 and the adapter molecule, thereby inhibiting TLR4 signal transduction and its downstream signaling pathways (Matsunaga et al., 2011).TAK242 has been reported to inhibit inflammatory injury and restore neural function in a mouse model of cerebral hemorrhage by inhibiting the expression of downstream signaling molecules (Wang et al., 2013).In our study, TAK242 inhibited neuroinflammation induced by BCAS, similar to the effects of NTP.Finally, we administered the TLR4-specific agonist CRX-527 to elucidate the underlying mechanisms of NTP effects.CRX-527, a TLR4-specific agonist, induced significant NF-κB promoter activation at 24 h after stimulation (Bowen et al., 2012).CRX-527 induces activation of downstream MyD88 signaling cascades leading to early activation and nuclear translocation of the NF-κB, promoting inflammation (Zhang et al., 2022).We found that CRX-527 abolished the effect of NTP, worsened neuroinflammation, and increased memory impairment, suggesting that TLR4 plays a major role in VD inflammation.The anti-inflammatory effect of NTP in VD mice is through regulation of TLR4.This suggested that NTP blocks the activation of sensor molecules in a neural pathway that could be a treatment for neuroinflammatory nervous system diseases.To sum up, we showed that NTP suppresses inflammation through TLR4/MyD88/NF-κB pathway.

Conclusions
Taken together, we showed that NTP treatment alleviates cognitive impairment by inhibiting TLR4/MyD88/NF-κB inflammatory signaling pathway in the VD mice.NTP might be a potential TLR4 inhibitor for VD.This study provides a new direction for the treatment of VD by elucidating inflammatory mechanisms of VD.

Fig. 1 .
Fig. 1.NTP ameliorates memory impairment induced by BCAS.(A)The discrimination ratio % in new object recognition experiments.(B)% correct alterations in the Y-maze test, and(C) time spent in stranger mouse 2 (S2) shown as a measure of social behavior in the social novelty task.Quantification of escape latency (D), total distance (E) for reaching the hidden platform during 5 days, and time spent in the target quadrant (F) were measured using MWM task in the sham group, the BCAS/ NS group and the BCAS/NTP group (n = 12).***P＜0.001BCAS/NS vs. sham group; ### P＜0.001 BCAS/NTP vs. BCAS/NS group.[one-way ANOVA followed by Student-Newman-Keuls test and two-way ANOVA followed by Bonferroni test].

Fig. 2 .
Fig. 2. NTP increases cerebrovascular blood flow after BCAS.(A) Representative laser speckle images showing cerebral blood flow (CBF) measured at the indicated times in the sham group and the BCAS/NS group and BCAS/NTP group.(B) The data represent mean ± SEM expressed as a percentage of the preoperative value (n = 3).***P＜0.001BCAS/NS vs. sham group; ### P＜0.001 BCAS/NTP vs. BCAS/NS group.[two-way ANOVA followed by Bonferroni test].

Fig. 4 .
Fig. 4. NTP alleviates neuronal ultrastructure damage after BCAS.(A)Electron microscope analysis indicates effects of BCAS and NTP on myelin (black arrows), synapsis (red arrows), mitochondria (green arrows) and neuron (yellow arrows) in the hippocampus in the sham group, the BCAS/NS group and the BCAS/NTP group.(B) Quantitative analysis show the percentage of demyelination in each group (n = 5).(C)Quantitative analysis shows the length of postsynaptic density in the three groups (n = 17-20).Scale bar: 200 nm *P < 0.01, **P < 0.01, ***P＜0.001,compared as indicated.[one-way ANOVA followed by Student-Newman-Keuls test].

Fig. 5 .
Fig. 5. NTP reduces hippocampal tissue damage after BCAS.(A)Representative pathological changes of CA1 and CA3 pyramidal neurons in the hippocampus in the sham group, the BCAS/NS group and the BCAS/NTP group.Scale bar: 50 μm.(B)Quantitative analysis of the number of CA1 and CA3 neurons in hippocampus in the three groups for A (n = 6).(C)Representative HE staining images of CA1and CA3 neurons in the hippocampus in the three groups.(D) Quantitative analysis of the number of CA1and CA3 pyramidal cells in hippocampus in the three groups for C (n = 5-7).Representative immunofluorescence images show colocalization of MAP-2 (red) and NeuN (green) in CA1 (E) and CA3 (F) regions of the hippocampus in the three groups.DAPI (blue) indicates the nucleus.Scale bar:100 plex.Quantitative analysis shows immunofluorescence intensity of MAP-2 of CA1(G) and CA3(H) regions, respectively (n = 6) **P < 0.01 and ***P＜0.001BCAS/NS vs. sham group; ## P < 0.01 and ### P＜0.001 BCAS/NTP vs. BCAS/NS group.[one-way ANOVA followed by Student-Newman-Keuls test].

Fig. 6 .
Fig. 6.NTP decreases BCAS-induced glia proliferation mediated by TLR4.(A)Representative immunofluorescent images indicate IBA1 + staining (red) in the CA1 region of the hippocampus in the sham group, the BCAS/NS group, the BCAS/NTP group, the BCAS/TAK242 group and the BCAS/NTP/CRX-527 group.Scale bar: 20 plex.(B)Quantitative analysis of immunofluorescence intensity of IBA1 staining for A. (C)Representative immunofluorescent images indicate GFAP + staining (green) in the CA1 region of the hippocampus in the five groups.Scale bar: 20 plex.(D)Quantitative analysis shows immunofluorescence intensity of GFAP staining for C. *P < 0.01, **P < 0.01 ***P＜0.001,compared as indicated.[one-way ANOVA followed by Student-Newman-Keuls test].