Introduction
Increasing evidence indicated that the occurrence and development of neurodegenerative diseases are closely associated with loosed or destructed blood–brain barrier (BBB) which are frequently present in the elderly [
37]. BBB plays a critical role in stabilizing the internal environment of the central nervous system (CNS) by tightly controlling the entry of substances, including plasma proteins, drugs and harmful substances [
31], while destructed BBB allows much more plasma components to enter the brain parenchyma, initiating cascading neuropathology [
32]. One of the main plasma components leaking through destructed BBB is albumin, which is the most abundant protein in the plasma, and plays a vital role in transporting metabolites, nutrition, and maintaining colloid osmotic pressure. BBB disruption increases albumin levels in the cerebrospinal fluid (CSF) and the brains of elderly and the patients with Alzheimer’s disease (AD), frontotemporal dementia (FTD) or other neurodegenerative disease patients [
18,
36]. After entering brain tissue, albumin can activate the microglia, and promote neuroinflammation [
13], and induce excitatory synaptogenesis through astrocytic TGF-β/ALK5 signal, thereby promoting epilepsy occurrence [
2]. However, the association of leaked albumin with neurodegenerative pathogenesis remains largely unknown.
Tau belongs to the family of microtubule-associated proteins, which maintains polymerization and stabilization of microtubules in neuronal axons and dendrites for ensuring essential cargo transportation [
38]. Tau can be physiologically phosphorylated at Ser202/Thr205, Thr181, Thr217, Thr231, Ser396, Ser404, and Ser416 by multiple kinases and phosphatases, such as glycogen synthase kinase-3β (GSK-3β) [
35] and calcium/calmodulin activated protein kinase IIα (CaMKII) [
40]. However, under a pathological state, more tau becomes hyperphosphorylated, aggregates into oligomers and neurofibrillary tangles, causing synaptic impairment [
7] and neuronal loss, and finally resulting in neuronal degeneration and cognitive decline [
29]. Hyperphosphorylation of tau is one of the key pathological hallmarks in several neurodegenerative diseases, such as AD, FTD, Progressive supranuclear palsy, Cortical basal degeneration and other tauopathies [
5]. However, the annotations on the regulation of tau hyperphosphorylation remains far from complete.
Very long-chain saturated fatty acids (VLSFAs) are saturated fatty acids with 20 or more carbons, primarily synthesized through elongation of very-long-chain fatty acids (Elovl) enzymes [
17]. VLSFAs exert a variety of cellular functions and are associated with numerous diseases including cardiovascular disease, diabetes and neurodegenerative disorders [
3,
20]. VLSFAs levels can also act as predictive biomarkers of AD [
8]. In the CNS, A1 astrocytes release VLSFAs that induce neuronal apoptosis and finally lead to memory loss [
12].
In this study, we investigated the effect of MSA on the cell phenotypes and function of microglia, astrocytes and neurons in vitro and in vivo, detected the cellular logic orders by which MSA influenced CNS, explored the association of MSA with tau phosphorylation, neuronal apoptosis, neuroinflammation and mouse cognition, and the corresponding underlying mechanisms.
Materials and methods
Animals
2-month-old C57BL/6J female littermates were purchased from Beijing HFK Bioscience Co.,Ltd. (Beijing, China). All mice were kept in the animal facility of Tsinghua University, and kept in a colony room at 22 ± 2 °C temperature and 45% ± 10% humidity under a 12 h:12 h light/dark cycle. In this study, all animal experiments were performed in accordance with the China Public Health Service Guide for the Care and Use of Laboratory Animals. Experiments involving mice and protocols were approved by the Institutional Animal Care and Use Committee of Tsinghua University.
Primary microglia and astrocytes cultures
Microglia and astrocyte were purified by immmunopanning from postnatal day 5 mice and cultured as previously described [
4,
10]. Briefly, micro-dissected cortical tissue was digested for 10 min at 37 °C in papain solution (Worthington Biochemical, LS003126), then gently disassociated with a 15 mL pipette to obtain a single-cell suspension. Homogenized cell suspensions were filtered through a 40-micron strainer (BD Bioscience) and the flow-through was centrifuged at 1500 rpm for 10 min to pellet cells. DMEM/F-12 medium (Sigma, D2906) was added to the pellet to resuspend the cells, and cells were placed for 40 min in a 37 °C incubator. The cell suspensions were allowed the CD11b antibodies (ThermoFisher Scientific, 14-0112-82) coated-immunopanning dish for 20 min at room temperature. Unbound cells and debris were removed by washing the dish 10 consecutive times with PBS. Isolated microglia were cultured in a defined serum-free medium containing 49 mL of phenol red free DMEM/F-12 medium containing 500 uL Penicillin–Streptomycin (ThermoFisher Scientific,15,140,122), 2 mM L-glutamine (ThermoFisher Scientific, 25,030,081), 5 μg/mL N-acetyl cysteine (Sigma, A8199), 5 μg/mL insulin (Merck, I9278), 100 μg/mL transferrin (Sigma, T8158), and 100 ng/mL sodium selenite (ThermoFisher Scientific, 11,360,070). Perform a 50% media change every 3 days to maintain the cultures. Cells cultured for 2 weeks were used for the following experiments.
After first immunopanning away microglia, the cell suspensions were added to the O4 antibodies (Sigma, O7139) coated-immunopanning dish for 15 min to remove oligodendrocytes. ITGB5 antibodies (ThermoFisher Scientific,14–0497-82) coated-petri plates was used to isolate astrocytes from remaining cells in suspension. After 40 min of incubation at room temperature, unbound cells and debris were removed by washing the dish 10 consecutive times with PBS. Isolated astrocytes were cultured with a defined serum-free base media including 50% neurobasal, 50% DMEM/F-12, 500 uL Penicillin–Streptomycin, 1mM sodium pyruvate (ThermoFisher Scientific, 11,360,070), 292 μg/mL L-glutamine, 1 μg/mL transferrin, 0.16 μg/mL putrescine (Sigma, P5780), 1 nM progesterone (Merck, P0130), 0.4 ng/mL sodium selenite, 5 ng/ml HBEGF (MedChemExpress, HY-P7400) and 5 μg/ml of N-acetyl cysteine. Perform a 50% media change every 7 days to maintain the cultures. Cells cultured for 2 weeks were used for the following experiments.
Primary neurons cultures
The primary neurons were isolated from cortex and hippocampus of fetal mice as previously described [
33]. Briefly, fetal mice were dissected to isolate the cortical and hippocampal tissue, and carefully peel off the meninges and blood vessels. The fresh brain tissue was cut into small pieces and the minced tissue was incubated in papain at 37 °C for 10 min. After the cells were gently blown into single cells using a pipette, the suspension was filtered with a 40-micron strainer and centrifuged at 1000 rpm for 5 min. Neurons were plated in DMEM medium containing 10% fetal bovine serum (FBS). DMEM medium was replaced with Neurobasal supplemented with B27 (ThermoFisher Scientific, 17,504,044), GluMax (ThermoFisher Scientific, 35,050,061), and pen-strep after 1 h of culture. The medium was replaced every 3 days, and primary neurons were cultured for 2 weeks to use.
Conditional medium preparation
To explore the direct effect of MSA (CUSABIO, CSB-NP000801m) on microglia, astrocytes and neurons, MSA at 7 μM, a similar concentration in patients’ brains, was added to primary microglia, astrocytes and neurons respectively. 24 h later, cells were harvested and used for the detection via RT-qPCR, ICC and Elisa.
To explore the effects of MSA on neurons in a co-culture mode of microglia and astrocytes, we added MSA to co-cultures of mouse primary microglia and astrocytes. After 12 h, replaced culture medium with fresh medium and continue to culture the glia cells for 24 h. Glia medium (GM) was collected and cultured neurons for 48 h. Neurons were then used for subsequent experiments.
To elucidate the effect of MSA-activated microglia on astrocytes and neurons, we firstly added MSA to microglia. After 12 h, replaced culture medium with fresh microglia growth medium and continued to culture the microglia for 24 h. Microglia medium (MM) was collected and used to treat astrocytes and neurons for 24 h, respectively. Astrocytes and neurons were used for subsequent experiments 24 h after treated with MM.
To elucidate the effect of MSA-activated astrocytes on microglia and neurons, we added MSA to astrocytes. After 12 h, replaced culture medium with fresh astrocyte growth medium and continued to culture the astrocytes for 24 h. Astrocyte medium (AM) was collected and cultured microglia and neurons for 24 h, respectively. Microglia and neurons were used for subsequent experiments 24 h after treated with AM.
MAM preparation: MSA firstly stimulated microglia, after 12 h, the medium was replaced with fresh microglia growth medium. After 24 h, supernatant was collected and continued to culture astrocytes for 12 h. Then, the culture medium was replaced with fresh astrocyte growth medium and continued to culture for 24 h. Microglia-Astrocyte Medium (MAM) was collected and added to neurons for 48 h. Neurons were collected for subsequent experiments.
AMM preparation: MSA firstly stimulated astrocytes, after 12 h, the medium was replaced with fresh astrocyte growth medium. After 24 h, supernatant was collected and continued to culture microglia for 12 h. Then, the culture medium was replaced with fresh microglia growth medium and continue to culture for 24 h. Astrocyte-Microglia Medium (AMM) were collected and added to neurons for 48 h. Neurons were collected for subsequent experiments.
Standard RT-qPCR
Gene expression of inflammatory factors, phenotypic markers, phagocytosis receptors and neurotrophic factors in glia and neurons were detected by RT-qPCR. Total RNA was extracted by using the RNeasy Lipid Tissue kit (Qiagen, #74,804) according to the manufacturer’s instructions. cDNA was prepared from total RNA using the PrimeScript RT-PCR kit (Takara, #RR037Q). Relative gene expression of the cDNA was assayed using a 7500 Fast real-time PCR instrument (Applied Biosystems) with SYBR Select Master Mix (Applied Biosystems, 4,472,908). qPCR data were analyzed by the ΔΔCT method by normalizing the expression of each gene to housekeeping gene β-Actin and then to the control groups. Primer sequences for mouse are listed in Additional file
1: Table S1.
Neurotransmitter detection by biochemical kits
The levels of glutamic acid in neurons, both intracellularly and extracellularly, were detected with Glutamic Acid (Glu) Content Assay Kit (Solarbio, BC1585). Supernatant and cell were collected and processed for the detection of glutamate release follow the manufacturer’s protocol. Neuronal glutamic acid contents were determined by comparing the absorbance value with the calibration plot for standard solutions. The absorbance values were measured at 340 nm.
General Gamma-Aminobutyric Acid (GABA) ELISA Kit (Jonln, T0731) was used to measure GABA concentrations in neurons, both intracellularly and extracellularly. Supernatant and cells were collected and processed for the detection of GABA release follow the manufacturer’s protocol. Neuronal GABA contents were determined by comparing the absorbance value with the calibration plot for standard solutions. The absorbance values were measured at 645 nm.
Western Blot (WB)
Total sample protein from cell and mice brain were extracted from RIPA protein lysate (Beyotime, P0013B) according to the manufacturer’s instructions. The concentration of protein was determined by the classical BCA protein determination method (Beyotime, P0010S). SurePAGE™ precast sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gel (Mellun, MA0456) was used for electrophoresis. After electrophoresis, protein samples were transferred to PVDF membrane at 300 mA constant current for 0.5–2 h. The membrane was sealed at room temperature for 1 h with 5% nonfat milk. Primary antibody was applied overnight at 4 °C. In the next day, after washed thrice for 5 min each with TBST, the membranes were coped with secondary antibody for 1 h. Then membranes were washed thrice again and imaged in Amersham Imager system (GE Healthcare). The images were analyzed by ImageJ software. The primary antibodies used in this study were described in Additional file
1: Table S2.
Targeted lipidomics
When conditional medium prepared, MAM was collected, quickly frozen in liquid nitrogen, stored in dry ice, and sent to the Beijing Bio-Tech Pack Technology Company Ltd. The samples were analyzed by liquid chromatography-mass spectrometry (LC–MS) analysis.
Immunohistochemistry (IHC)
Mice were anesthetized with chloral hydrate and transcardially perfused with cold PBS. The dissected mouse brains were fixed in 4% paraformaldehyde at 4°C overnight, dehydrated and embedded in paraffin. Making 5 μm paraffin sections, after dewaxing and hydration, antigen retrieval was carried out. After permeabilized with 0.3% Triton X-100 for 10 min, the sections were blocked in 10% donkey serum (Solarbio, SL050) for 1 h at room temperature. Subsequently, the sections were incubated with the primary antibodies overnight at 4 °C, and followed by corresponding fluorescently conjugated secondary antibodies, respectively, and imaged on a Leica TCS SP8 confocal microscope. For DAB staining, sections were incubated with appropriate secondary antibodies, and the staining was developed by incubating with DAB. Images of DAB-stained sections were captured using an Olympus BX61 microscope. The primary antibodies used in this study were described in Additional file
1: Table S2. All images were analyzed by ImageJ software.
Immunocytochemistry (ICC)
The cell slides in the 24-well plate were washed three times with PBS and fixed with 4% paraformaldehyde for 15 min. After permeabilized with 0.3% Triton X-100 for 10 min, the slides were blocked in 10% donkey serum for 1 h at room temperature. Subsequently, the sections were incubated with the primary antibodies overnight at 4 °C, and followed by corresponding fluorescently conjugated secondary antibodies, respectively, and imaged on a Leica TCS SP8 confocal microscope. The primary antibodies used in this study were described in Additional file
1: Table S2. All images were analyzed by ImageJ software.
Enzyme-linked immunosorbent assay
Levels of the inflammatory factors (IL-1β, TNF-α, IL-6) in samples from brain lysates of mice and cell culture supernatant were detected by ELISA kits (Neobioscience technology), according to the manufacturer’s protocols. The absorbance at 450 nm was measured using a SpectraMax M5 microplate reader.
Levels of Aβ in soluble or insoluble extractions of mouse hippocampal tissues were determined with Aβ 1–38 (Aβ38), Aβ 1–40 (Aβ40), and Aβ 1–42 (Aβ42) MSD Triplex assay kit (Meso Scale Discovery, Rockwilly, MA, USA, N45199A-1) according to the manufacturer’s instructions.
Stereotaxic injection
2-month-old C57BL/6J female mice were anesthetized with 1.2% tribromoethanol and placed in a stereotaxic device. The skull was exposed by a midline scalp incision, and a craniotomy was drilled unilaterally or bilaterally above each cannula implantation or injection site.
For the injection of MSA, the cannulas were placed into the lateral ventricles ( − 0.2 mm anteroposterior, 1.4 mm mediolateral, 1.8 mm dorsoventral). Insert the infusion needle attached to the 10 μL micro-syringe into the guide tube, 1 μL of MSA solution is delivered within 2 min. The injections were given every 4 day for 16 days (total of 5 injections) and every 4 day for 60 days (total of 16 injections) respectively.
In order to reduce the expression of Elovl1, adeno-associated virus (AAV) carrying Elovl1 shRNA was provided by OBiO Technology (Shanghai) Corp., Ltd. The sequence of Elovl1 shRNA was GAATCATGGCTAATCGGAAGC. The serotype of AAV was 2/5. For AAV injections, 3 μL (6.03 × 1012 vg/mL) of shElovl1 was injected bilaterally into the cortex (+ 1.3 mm anteroposterior, 0.7 mm mediolateral, 0.8 mm dorsoventral) and hippocampus ( − 2 mm anteroposterior, 1.7 mm mediolateral, 1.6 mm dorsoventral). Mice were injected with AAV one month before MSA treatment. After each injection, the needle was left in place for 2 min and then slowly withdrawn. The surgical site was cleaned with sterile saline and the incision sutured. After surgery, animals were monitored and provided post-surgical care.
Behavioral tests
Novel object recognition (NOR) test, Y-maze test and Morris water maze (MWM) test were applied to detect the memory and cognitive function of mice. NOR, a method for exploring animal recognition and memory of new objects, is based on the instinct of mice to explore the characteristics of new objects. In the adaptation phase, each mouse was allowed to freely explore the open-field area (a white box 40 cm wide × 40 cm deep × 40 cm high) during 5 min. In the training phase, mice were exposed to two identical objects, which they were allowed to freely explore for 5 min. Recognition memory was tested after 6 h by exposing the mice to one familiar and one novel object. The times that the mice explored the novel and the old object were recorded independently. The discrimination index was determined by performing the following calculation: (Timenovel-Timeold) / (Timenovel + Timeold).
Y‐maze test was used to evaluate the spatial working memory of mice. The Y-maze test was performed in the Y-shaped maze having three arms. Three arms were randomly defined as novel arm, the starting arm and other arm. The arm with the dotted line represents the randomly chosen starting arm; the blocked off arm represents the novel arm. During the first training trial, each mouse was allowed to freely explore the starting arm and the other arm for 10 min. After a one-hour interval, the mice were allowed to freely explore all three arms for 5 min. A camera mounted above the maze automatically records the distance traveled, arm entries, and the time spent in each arm.
Mice were tested for their spatial learning and memory abilities using the MWM test. The water maze consisted of a water pool (1222 cm in diameter) containing opaque water and a platform (102 cm in diameter) submerged 1 cm below the water surface. During learning trials for 5 days, mice freely swam for 60 s to find the hidden platform. Mice that failed to find the platform within 60 s were guided to and remained on the platform for 10 s. The mice were trained twice a day with a 4 h interval between training sessions. To evaluate spatial memory retention, the platform was removed for a probe trial 24 h after the last day of hidden platform training. The swimming path of the mouse was recorded by video camera and analyzed by Ethovision XT version 16.0 (Noldus software).
Adult microglia isolation
Adult microglia were isolated from the mouse brain as previously described [
1]. In brief, the brains of mice were minced in the Hibernate A (Thermo Fisher Scientific, A1247501)/B27 medium and dissociated for 15 min at 37 °C with 0.25% Trypsin–EDTA containing DNase I, after neutralized with DMEM supplemented with 10% FBS, cells were separated by Optiprep (Merck, D1556) density gradient centrifugation. Fractionated microglia from the mice treated with or without MSA were obtained for further RNA sequence analysis.
Adult astrocyte isolation
For adult astrocyte isolation, mice were perfused with cold PBS under isoflurane anesthesia, and then the brain was removed, dissected, and rinsed in HBSS. Enzymatic cell dissociation was then performed using an Adult Brain Dissociation Kit (Miltenyi Biotec, 130-107-677), according to the manufacturer’s instructions. The resulting single cell suspension was centrifuged at 300 g for 10 min at room temperature, resuspended in 40% Percoll and centrifuged at 800 g for 20 min with breaks off. After Fc receptors were blocked using anti-mouse CD16/CD32 (eBioscience, 14-0160-82) for 10 min, cells were incubated with ACSA-2-PE for 30 min on ice, and washed with 1 ml of blocking buffer and sorted using a FACSAria II cell sorter. Fractionated astrocytes from the mice treated with or without MSA were obtained for further RNA sequence analysis.
Adult neuron isolation
Adult neurons were isolated in an enriched population from the mice treated with or without MSA using the Mitlenyi MACS system, the Adult Brain Dissociation Kit (Miltenyi Biotec, 130-107-677), and Adult Neuron Isolation Kit (Miltenyi Biotec, 130-126-603). Mice were perfused with cold PBS under isoflurane anesthesia, and then the brain was removed, dissected, and rinsed in HBSS. The cortex and hippocampus tissues were isolated and transferred to a MACS C-tube. Samples was then dissociated using the appropriate pre-set protocol on the Miltenyi gentleMACS Octo Dissociator instrument with heaters attached. Fractionated neurons from the mice treated with or without MSA were obtained for further RNA sequence analysis.
RNA sequence and data analysis
The microglia, astrocytes and neurons isolated from the mice treated with or without MSA were sent to OE Biotech, Inc., (Shanghai, China) Company for RNA extraction and eukaryotic transcriptome sequencing. The RNA libraries were sequenced by OE Biotech, Inc., Shanghai, China. We are grateful to OE Biotech, Inc., (Shanghai, China) for assisting in sequencing and/or bioinformatics analysis. Bioinformatic analysis was performed using the OECloud tools at
https://cloud.oebiotech.com/task/. The volcano map (or other graphics) was drawn based on the R (
https://www.r-project.org/) on the OECloud platform (
https://cloud.oebiotech.com/task/).
Statistical analysis
Data were analyzed with GraphPad Prism v.9.5.1. Each figure legend denotes the statistical analysis used. All data are represented as mean ± s.e.m. For comparisons between groups, first, it was determined whether the data were normally distributed using the Shapiro–Wilk test (Sigma-Plot). If data were normally distributed, one-way ANOVA was used with post hoc Holm–Sidak test for pairwise comparisons or an unpaired t-test with two-tailed p values. If not, Mann–Whitney rank sum test (two groups) or Kruskal–Wallis one-way ANOVA on ranks (three or more groups) with post hoc Dunn’s test was used. In all cases, statistical difference was considered significant at *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.
Discussion
Elderly people often present with impaired BBB, and more attention has been paid to the correlation between disrupted BBB and neurodegenerative diseases [
27,
28]. Many blood-derived protein such as thrombin, fibrinogen and albumin are prohibited from entering the brain by the intact BBB, but they may enter the brain when BBB is impaired [
9,
24]. Albumin, the most abundant protein in plasma, has remained largely unexplored in its role in CNS. Our present study found that IL-1α, TNF-α and C1q secreted by MSA-activated microglia induced astrocytes to their A1 phenotype to generate VLSFAs, the released VLSFAs subsequently triggered neuron apoptosis through PERK endoplasmic reticulum stress response pathway. The injected MSA in mouse brains induced tau phosphorylation via microglial NLRP3 inflammasome pathway, caused neuronal apoptosis, resulting in the decline in learning and spatial memory ability.
Neuron death is one of the most common pathological features in neurodegenerative disorders. Recent evidence showed that VLSFAs played a critical role in neuron apoptosis [
12], while VLSFAs generation in astrocytes was dependent on Elovl1. We here found that the levels of VLSFAs and Elovl1 were significantly promoted in astrocytes by C1q, TNF-α and IL-1α in the MSA-induced microglial conditional medium, which was consistent that previous report that neurotoxic reactive astrocytes were induced by activated microglia [
23]. Moreover, MSA-induced astrocytic conditional medium failing to induce neuron apoptosis. Thus, MSA-induced neuron apoptosis follows a logic order from microglia to astrocytes to neurons.
Various cells in the brain live in the same complex environment, interact with each other and are subjected to direct or indirect effects of leaked MSA. MSA induced microglia and astrocytes to MGnD and A1 phenotype, respectively, and AM and MM also correspondingly further activated microglia and astrocytes, forming a hazardous microenvironment containing various inflammatory factors, cytokines and VLSFAs in CNS. It was reported that long chain saturated fatty acids promoted inflammation [
14], and that there was a positive feedback-loop regulation between VLSFAs and inflammatory factors. We further found that MSA-induced detrimental microenvironment incurred tau phosphorylation and neuron apoptosis, and significantly decreased learning, memory abilities in mice, while Elovl1 knockdown broke this detrimental effect.
Hyperphosphorylated tau, the main component of neurofibrillary tangles in neurons, is a major pathological feature of several neurodegenerative diseases [
22]. Hyperphosphorylated tau is apt to aggregate, forming aggregates such as oligomers and fibrils which are toxic to neurons, leading to neuronal death [
5]. The present injected MSA in mouse brains induced microgliosis and generation of NLRP3, IL-1β and IL-18. Consistent with previous study [
16], these inflammatory factors promoted the generation of tau phosphorylation-related kinases GSK3β and CaMKIIα, leading to obvious tau phosphorylation at multiple sites including Thr181, Ser199, Thr217 and Thr231 (Fig.
3). It has been extensively reported that tau181, tau 217 and tau 231 are desired biomarkers for early diagnosis of AD [
25,
26], and brain-derived tau aggregate seeds can spread tauopathy throughout the brain, further inducing the mis-folding of endogenous tau and neuronal degeneration [
6,
11]. Therefore, leaked MSA through impaired BBB would be a primary cause for tauopathies. In contrast, MSA did not induce the production of Aβ and α-synuclein, suggesting that MSA mainly contribute to tauopathies rather than other amyloid related pathology.
Conclusions
In summary, our study here revealed that MSA induced tau phosphorylation and neuron apoptosis based on MSA-activated microglia and astrocytes, respectively, showing the important role of MSA in initiating tauopathies and cognitive decline, and providing a potential therapeutic target for tauopathies.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.