Background
Influenza belongs to orthomyxoviridae, an enveloped segmental negative-strand RNA virus [
1]. According to the antigenicity difference of nuclear protein (NP) and matrix protein (M), influenza viruses can be divided into A, B, C and newly discovered type D influenza virus [
2]. Among them, influenza A virus (IAV) infects a wide range of hosts, which can not only infect a variety of mammals and birds but also infect humans across interspecies barriers [
3]. Due to the lack of proofreading activity of polymerase and segmented genome, the antigenic drift and shift cause the virus diversity and the emergence of novel IAV [
4]. The H7N9 influenza A virus was first isolated in 2013, composed of early H7N9, H7N3 and H9N2 influenza virus gene segments and continued to cause human infections [
5,
6]. The emergence of the highly pathogenic H7N9 influenza virus has seriously threatened poultry production and human health [
7]. In particular, the emergence of highly pathogenic H7N9 virus variants and the ability of limited human-to-human transmission need more attention.
The typical clinical symptoms of influenza are cough, high fever, muscle pain, and general discomfort, but some influenza patients have nervous system symptoms such as febrile seizures, encephalitis/encephalopathy, and myelitis [
8,
9]. Central nervous system complications caused by influenza virus infection have been reported frequently[
8,
10‐
13]. According to previous case reports and studies, Influenza-associated neurological complications often occur in children, and most cases are caused by influenza A and B viruses [
11,
14,
15]. Neurological complications caused by influenza virus infection have gradually attracted more attention, but its pathogenic mechanism remains unclear [
11‐
13,
15,
16]. Whether the influenza virus can invade the central nervous system (CNS) is controversial. However, some laboratory evidence suggested that IVA could infect the CNS. In both ferret and mouse models, H5N1 could reach the CNS through multiple pathways, such as the olfactory, vagus, and vestibulocochlear nerves [
17,
18]. HPAI virus H7N1 could destroy the blood-brain barrier and lead to viremia and pathological changes in the central nervous system of chickens [
19]. Besides, In vivo studies have shown that H7N9 virus RNA was detected in the brains of experimentally infected ferrets and mice [
20,
21], suggesting that avian influenza H7N9 virus might spread to the brains of mammals.
It has been reported that H7N9 virus could cause neurological manifestations in patients [
22,
23]. Besides, human astrocytic and neuronal cells could be infected by H7N9 virus, and viral infection triggered high expression of pro-inflammatory cytokines [
24]. However, there is no enough experimental data to indicate how H7N9 virus enters the CNS. Considering previous studies and viremia in avian influenza virus infection [
25], the blood pathway could be the potential route of the H7N9 avian influenza virus entering the CNS. H7N9 virus needs to pass the blood-brain barrier to reach the CNS. The blood-brain barrier (BBB) is a dynamic regulator of ion balance, a facilitator of nutrient transport and a barrier to potentially harmful molecules that acts as an interface between the central nervous and peripheral circulatory systems [
26,
27]. BBB mainly comprises microvascular endothelial cells, pericytes, astrocytes and basement membrane [
28]. The core element of the BBB is the cerebral vasculature formed by endothelial cells, which build a physical barrier between the blood and the brain. Increased endothelial cell permeability leads to BBB disruption and is a hallmark of CNS infection [
29]. In this study, immortalized human brain microvascular endothelial cells hCMEC/D3 were used to construct an in vitro BBB model to explore whether H7N9 virus could destroy the BBB for the pathogenesis of influenza viral encephalopathy.
Methods
Cell culture
The immortalized human brain capillary endothelial cell line hCMEC/D3 purchased from BeNa Culture Collection was cultured in endothelial cell medium (ECM) supplemented with 5% fetal bovine serum (FBS), 1% endothelial cell growth supplement and 1% penicillin/streptomycin (P/S). The Madin-Darby canine kidney cell line MDCK was cultured in a Dulbecco’s modified eagle medium (DMEM) supplemented with 10% FBS and 1% P/S. Both cells were incubated at 37℃ with 5% CO2 and used from passage 8–20.
H7N9 virus infection
The virus A/Shenzhen/13/2013 (H7N9), kindly provided by the Shenzhen Center for Disease Control and Prevention, was propagated in SPF eggs. For virus infection assays, hCMEC/D3 cells were seeded on 6-well plates (5 × 105 cells/well) and incubated until cells were grown to 80-90% confluent monolayer. Cells were then infected with H7N9 virus at different multiplicities of infection (MOI). For the mock group, egg allantoic fluid without virus was added into the medium after cells were grown to 80-90% confluent monolayer. During infection, cell morphology was photographed every 24 h until 48 h after infection and supernatant was collected simultaneously to detect the progeny virus titer at each time point. All infection experiments were carried out in the class III bio-safety lab (BSL-3) at Shenzhen Center for Disease Control and Prevention.
Virus titer assay
The 50% cell culture infectious dose (TCID50) endpoint dilution assay was performed for the virus titer. MDCK cells were inoculated into a 96-well cell plate (2 × 10
4 cells/well) and incubated at 37℃ with 5% CO2 for 24 h, and the virus was diluted in a 10-fold gradient with medium containing 2 µg/mL TPCK-trypsin. When the cells reached 90% density, PBS washed the cells twice. The virus diluted as described above was inoculated on the 96-well plate containing MDCK cells, four wells per dilution, incubated at 37℃ for 1 h. After incubation, the medium containing the virus was discarded, and the cells were washed with PBS once. The culture medium containing 2 µg/mL TPCK-trypsin was added to each well and incubated at 37℃ for 72 h to detect the virus’s erythrocyte agglutination (HA) titer. The TCID50/100µL of the virus was calculated according to Reed-Muench methods (Table
1).
Quantitative real-time PCR
Quantitative real-time PCR (qPCR) was performed to detect the presence of the viral genome in hCMEC/D3 cells and the transcript level of tight junctional complexes. The infected cells were collected at 2 h, 24 h, 48 h, and 72 h after infection, and the total RNA was extracted using the AxyPrep Multisource Total RNA Miniprep Kit (Axygen). PCR amplification was performed using the One Step TB Green PrimeScript RT-PCR Kit II (Takara).
Table 1
The primers used in qPCR
Flu A | GACCAATCCTGTCACCTCTGAC | AGCTGAGTGCGACCTCCTTAG |
Claudin 5 | CTCTGCTGGTTCGCCAACAT | CAGCTCGTACTTCTGCGACA |
Occludin | ACAAGCGGTTTTATCCAGAGTC | GTCATCCACAGGCGAAGTTAAT |
VE-cadherin | TTGGAACCAGATGCACATTGAT | TCTTGCGACTCACGCTTGAC |
β-actin | CTCCATCCTGGCCTCGCTGT | GCTGTCACCTTCACCGTTCC |
Cell viability assay
In order to detect the cell viability of hCMEC/D3 cells after infection, a CCK-8 test was carried out. HCMEC/D3 cells were seeded on a 96-well plate (1.5 × 10
4/well) and incubated until cells were grown to 80% confluent monolayer. PBS washed cells once, and 200µL medium containing 2% FBS was added to each well. Then cells were infected with H7N9 influenza virus at different MOIs. 24 h, 48 and 72 h after infection, the number of living cells was detected by CCK-8 kit (Beyotime). The following formula calculated the cell viability:
$${Cell} \; {viability} = \frac{{OD450}_{H7N9}}{{OD450}_{Mock}}$$
Western blot analysis
To detect the effect of virus infection on tight junction proteins in endothelial cells, cells were lysed in RIPA buffer (Beyotime). Proteins were separated by 10% SDS-PAGE and transferred to polyvinylidene fluoride membranes which were blocked with a blocking buffer (3% bovine serum albumin (BSA) in TBS with 0.05% Tween 20 (TBST)) and incubated with primary antibodies in the TBST in 4℃ overnight. Herein, the anti-ZO-1 antibody (Invitrogen), the anti-Occludin antibody (Invitrogen), the anti-Claudin 5 antibody (Invitrogen), the anti-vascular endothelial (VE)-Cadherin antibody (Abcam), and the anti-beta-Actin antibody (Senta Cruz) were utilized, respectively. After being washed three times with TBST, the membrane was probed with an HRP-conjugated secondary antibody and was developed with SuperSignal West Dura Extended Duration Substrate (Thermo Fisher) and imaged on an Image Quant TM LAS 4000.
Microvascular endothelial cells monolayer resistance measurement
Cell resistance can reflect the function of the physiological barrier. In order to detect whether H7N9 virus infection could affect the function of BBB in vitro, Electric cell-substrate impedance spectroscopy (ECIS) technology was used to detect the resistance changes of hCMEC/D3 cells after infection. ECIS experimental procedure involved a pretreating 8W10E ECIS plate with 10 mM L-cysteine prior to coating with 3 µg/cm2 collagen I (Gibco). Epithelial cells were seeded on wells at a density of 4 × 104/well, and the volume of culture medium for each well was 400ul. Cells were then infected H7N9 virus with different MOIs until a barrier had formed, typically ~ 60 h post-seeding. The endothelial barrier resistance was then monitored, at which point multi-frequency (ranging from 62.5 to 64,000 Hz) data was collected and modelled using ECIS software (Applied Biophysics).
In vitro BBB permeability assay
The hCMEC/D3 cells were seeded at a density of 5 × 104 per well on a Transwell insert (0.4 μm pore size, Corning) coated with 3 µg/cm2 collagen I in 100 µl complete ECM and 500 µl of the same medium were added to the basal chamber. Usually, confluent monolayer formation was assessed 48 h post-plating. To test whether the permeability of the barrier model was affected by H7N9 virus (MOI = 1), FITC-dextran (wt4000, Sigma) permeability was determined during the BBB exposure to H7N9 virus. The inserts were placed onto companion plates containing 500 µl of ECM. Before FITC-dextran was added to the media in the apical insert, the inserts were washed twice with prewarmed PBS. Then FITC-dextran was diluted to 100 µg/ml with ECM and added to the apical insert to incubate 30 min at 37℃, and 100 µl of ECM was collected from the basal companion wells. The fluorescence intensity of FITC-dextran was measured by a fluorescent plate reader (excitation 492 nm and emission 518 nm).
Evans blue-BSA permeability assay
The Evans Blue (EB) was dissolved in PBS into 0.5% (w/v) solution, and then BSA was added to the EB solution to a concentration of 1% (w/v). After complete vortex mixing, let the solution stand for 30 min at room temperature and filter it with a 0.22 μm filter for later use. The hCMEC/D3 cells were seeded on a Transwell insert and infected by H7N9 virus as above. 2 h, 24 and 48 h after infection, EB-BSA was added to the apical insert to incubate at 37℃, and 50 µl of ECM was collected from the basal companion wells every 10 min until 50 min. After collecting the ECM, the basal was replenished every time. 0.5% EB solution was double-diluted to plot a standard curve simultaneously. The concentration of EB was measured by a microplate reader (OD620), and taking the time as abscissa and OD620 as the ordinate, the standard curve and the experimental group curve were plotted, respectively. The permeability coefficient (Pe) of EB-BSA was calculated by the following formula:
$$Pe=\frac{1}{1/{m}_{e}-1/{m}_{s}}\times \frac{1}{s}$$
where ms is the slope of the standard curve, and me is the slope of the curve in the experimental group. Moreover, s is the surface area of the inserts (1.12 cm
2).
Statistics analysis
Prism 8 (GraphPad) was used for data analysis and chart drawing. If the data of the two groups were compared, the student’s t-test was use to for statistical analyses. If more than two groups were compared, One-way ANOVA was used for statistical analyses. If p < 0.05, there is a statistical difference. Statistical significance is defined as, n.s., not significant, *p < 0.05, **p < 0.01, ***p < 0.001.
Discussion
Human infections with HPAI H7N9 virus cause primarily respiratory disease, but some populations, such as children, the elderly, and gravidas, are at risk of influenza-related central nervous system complications. However, the pathogenic mechanism of influenza viral encephalopathy has not been illustrated. In this study, we found that H7N9 influenza virus could disrupt the BBB function in vitro and increase its permeability by down-regulating the expression of junctional complex proteins. The present study provides new insight into the pathogenesis of H7N9 viral encephalopathy.
If H7N9 virus enters the CNS through the bloodstream, it must cross the BBB, a barrier between blood and brain tissue. Changes in the barrier characteristics are an essential reason for the pathology and progression of various neurological diseases [
26‐
28,
38]. Here we used hCMEC/D3 as endothelial cells in the BBB to determine the infectivity and virulence of H7N9 virus on human brain microvessels endothelial cells, which is an important component of the BBB. The infection experiment has shown H7N9 virus could infect brain endothelial cells and produce infectious progeny viruses that did not cause cell death but affected the cell morphology during infection, which provides preliminary evidence that the H7N9 virus has the potential to infect the BBB.
Endothelial cell-cell connections are critical to barrier function, and two substructures of the junctional complexes that exist in the BBB participate in the cell connection, namely adherens junctions and tight junctions [
27,
31]. The adherens junctions mediate cell-cell adhesion, and the tight junctions establish cell polarity and regulate the paracellular transport of ions and small molecules [
34]. In this study, we found that H7N9 virus could infect human brain microvascular endothelial cells hCMEC/D3 and reduce the expression of cell junctional complexes proteins in cells. The transmembrane molecules TJ proteins, including claudins and occludins, are closely related to the intercellular junction. Claudin-5 is the most enriched TJ protein at the BBB [
39], and mice that lack claudin-5 have a size-selective leak of the BBB [
40]. Occludin is a tetraspanin strongly expressed at the interface of CNS endothelial cells [
41], an in vitro culture experiment disrupting occludin homotypic interactions suggests that it is important for barrier resistance [
26]. In infected hCMEC/D3 cells, the decreased expression of both claudin-5 and occludin indicated the effect of H7N9 virus infection on intercellular junctions of endothelial cells. Besides TJs, AJs are other major cell-to-cell connecting structures that sense and respond to tensile forces at the intercellular contact interface. The TJs interact with basal AJs, including the transmembrane protein VE-cadherin, which has five extracellular cadherin repeat domains mediating cell junctions [
42]. VE-cadherin expression is restricted to the vascular system playing a crucial role in establishing AJs [
43]. Consistent with TJ proteins results, VE-cadherin expression was down-regulated at 48 h after infection, resulting in disruption of cell junctions. As transmembrane adhesion complexes, the ZO proteins are cytoplasmic scaffolding proteins located in the cytoplasmic domain of TJs [
31,
38]. Among ZO proteins family, ZO-1 plays a critical role in endothelial cell junctions by linking TJ proteins to the actin cytoskeleton and as a linkage between the AJs and TJs [
26,
32,
44]. However, we did not find H7N9 influenza virus down-regulated the expression of ZO-1 in human brain microvascular endothelial cells, which indicated H7N9 virus disturbed the cell-cell contacts by directly down-regulating the transmembrane part of junctional complexes, rather than affecting the interaction with AJs and TJs components and the link between TJs and the cytoskeleton. We concluded H7N9 virus caused down-regulation of the expression of cell junction proteins, allowing the virus to disrupt the cell barrier. In addition, we detected the mRNA level of three down-regulated proteins, VE-cadherin, occludin, and claudin-5. The results showed that all proteins were lower than the mock control at the transcriptional level. This indicated that H7N9 virus infection directly reduced expression by affecting tight junction proteins’ transcription.
Based on the results above, we speculated that the infection of H7N9 virus would affect the normal physiological function of the BBB. In order to verify the phenomenon, we detected the cell resistance to see if the barrier was intact. ECIS has been applied to study BBB biophysical and biomedical functions including cell motility, wound healing and cell-cell adhesion [
45,
46]. In this study, ECIS was used to detect cell resistance to reflect the integrity of the barrier function. With the infection of H7N9 virus, the resistance of monolayer hCMEC/D3 cells decreased gradually, which was consistent with the previous results. That is, the down-regulation of intercellular junction-related proteins affected the cell barrier function. In addition to the resistance detection of endothelial cells, we constructed an in vitro BBB model using Transwell inserts to verify the ECIS result. FITC-dextran is a bio-marker of the BBB permeability of high molecular mass molecules [
47,
48], and the EB dye, also as a permeability marker, is often used to evaluate BBB integrity [
49,
50]. Here they were used as indicators of whether the BBB function was impaired. Consistent with the electrical resistance test results, increased FITC-dextran and EB-BSA permeability in the infected model also demonstrated that H7N9 virus infection could cause in vitro BBB destruction. The destruction of the BBB increases the permeability of potentially neurotoxic molecules that may be harmful to the central nervous system and provides a way for H7N9 virus to enter the CNS.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.