Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 https://doi.org/10.1186/s13024-020-00382-8 RESEARCH ARTICLE Open Access Sex-dependent calcium hyperactivity due to lysosomal-related dysfunction in astrocytes from APOE4 versus APOE3 gene targeted replacement mice Raquel Larramona-Arcas1, Candela González-Arias2, Gertrudis Perea2, Antonia Gutiérrez3,4, Javier Vitorica4,5, Tamara García-Barrera6, José Luis Gómez-Ariza6, Raquel Pascua-Maestro7, María Dolores Ganfornina7, Eleanna Kara8,9, Eloise Hudry8, Marta Martinez-Vicente4,10, Miquel Vila1,4,10,11, Elena Galea1,11† and Roser Masgrau1*† Abstract Background: The apolipoprotein E (APOE) gene exists in three isoforms in humans: APOE2, APOE3 and APOE4. APOE4 causes structural and functional alterations in normal brains, and is the strongest genetic risk factor of the sporadic form of Alzheimer’s disease (LOAD). Research on APOE4 has mainly focused on the neuronal damage caused by defective cholesterol transport and exacerbated amyloid-β and Tau pathology. The impact of APOE4 on non-neuronal cell functions has been overlooked. Astrocytes, the main producers of ApoE in the healthy brain, are building blocks of neural circuits, and Ca2+ signaling is the basis of their excitability. Because APOE4 modifies membrane-lipid composition, and lipids regulate Ca2+ channels, we determined whether APOE4 dysregulates Ca2+signaling in astrocytes. Methods: Ca2+ signals were recorded in astrocytes in hippocampal slices from APOE3 and APOE4 gene targeted replacement male and female mice using Ca2+ imaging. Mechanistic analyses were performed in immortalized astrocytes. Ca2+ fluxes were examined with pharmacological tools and Ca2+ probes. APOE3 and APOE4 expression was manipulated with GFP-APOE vectors and APOE siRNA. Lipidomics of lysosomal and whole-membranes were also performed. (Continued on next page) * Correspondence: roser.masgrau@uab.cat † Elena Galea and Roser Masgrau contributed equally to this work. 1 Unitat de Bioquímica de Medicina, Departament de Bioquímica i Biologia Molecular, and, Institut de Neurociències (INc), Facultat de Medicina, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Barcelona, Catalonia, Spain Full list of author information is available at the end of the article © The Author(s). 2020 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. Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 2 of 23 (Continued from previous page) Results: We found potentiation of ATP-elicited Ca2+responses in APOE4 versus APOE3 astrocytes in male, but not female, mice. The immortalized astrocytes modeled the male response, and showed that Ca2+ hyperactivity associated with APOE4 is caused by dysregulation of Ca2+ handling in lysosomal-enriched acidic stores, and is reversed by the expression of APOE3, but not of APOE4, pointing to loss of function due to APOE4 malfunction. Moreover, immortalized APOE4 astrocytes are refractory to control of Ca2+ fluxes by extracellular lipids, and present distinct lipid composition in lysosomal and plasma membranes. Conclusions: Immortalized APOE4 versus APOE3 astrocytes present: increased Ca2+ excitability due to lysosome dysregulation, altered membrane lipidomes and intracellular cholesterol distribution, and impaired modulation of Ca2+ responses upon changes in extracellular lipids. Ca2+ hyperactivity associated with APOE4 is found in astrocytes from male, but not female, targeted replacement mice. The study suggests that, independently of Aβ and Tau pathologies, altered astrocyte excitability might contribute to neural-circuit hyperactivity depending on APOE allele, sex and lipids, and supports lysosome-targeted therapies to rescue APOE4 phenotypes in LOAD. Keywords: APOE4, Astrocytes, Calcium signaling, Sex, Lysosome, Purinergic receptors, Lipidome Background but not according to other authors [18]. In addition, det- Apolipoprotein E is a component of lipoproteins in- rimental actions of APOE4 have been reported in other volved in extracellular lipid transport and cholesterol neurodegenerative disorders such as frontotemporal de- fluxes throughout the body, including the brain [1]. In mentia [19], cerebrovascular disease [20] and traumatic humans, APOE exists in three isoforms: APOE2, 3—the brain injury [21]. most common allele—and 4. Although the three iso- The mechanisms whereby APOE4 is pathological in forms differ only in the amino acids in positions 112 and normal and diseased brain are not totally clarified [15]. 158 at the N-terminal domain of the protein [2], this An outstanding question is whether APOE4 affects the minimal difference results in a structural change in the function of brain cells other than neurons. Although APOE4 isoform that profoundly compromises its func- microglia and neurons secrete ApoE in pathological con- tion [3], as shown by a wealth of studies in mice and ditions, as described in mouse neurons injured with humans that document the impact of APOE genotype kainic acid [22], and in human LOAD brains, where on the structure and function of the healthy brain. Thus, microglia produces and deposits ApoE in senile plaques in humans, APOE4 is associated with reduced memory [23] and appears to mediate Tau pathology [24], it is retention [4], altered neural activity and brain connectiv- often overlooked that APOE is mainly synthesized, se- ity [5], reduced grid-cell like representations [6], reduced creted, and lipidated by astrocytes in physiological and dendritic spine density [7], and hypometabolism mea- pathological conditions [3]. Astrocytes can also take up sured with fluorodeoxyglucose-based PET [8]. Further- lipoproteins, as they express APOE receptors such as the more, APOE4 knock-in mice present alterations of LRP1 receptor [25]. behavior, olfactory memory and neurotransmission, as Astrocytes are building blocks of neural circuits, where well as decreased dendritic arborization and metabolic they modulate neuronal activity [26]. Apart from the im- alterations [9–13], as compared to APOE3 knock-in pact of APOE4 on synaptogenesis [27] and synaptic mice. Not only is normal brain function compromised transmission [28] by lowering the delivery of cholesterol by APOE4, but APOE4 is also the strongest genetic risk from astrocytes to neurons, the question of whether factor in late-onset Alzheimer’s disease (LOAD) [14], the APOE4 compromises the global control of astrocytes principal cause of age-related dementia, affecting over neuronal activities has not been examined. Among millions of people around the world [15]. A complex phenomena modulated by astrocytes there are, to cite a interaction exists among sex, age and APOE4 load. Thus, few, long-term potentiation (LTP), memory consolida- according to a meta-analysis, heterozygous APOE3/ tion, and circadian rhythms [29–32]. Central to APOE4 women present increased risk of LOAD between astrocyte-to-neuron communication is Ca2+ signaling in the ages of 65 and 75 years, and increased risk of mild astrocytes. Thus, although astrocytes are considered cognitive impairment (MCI) between the ages of 55 and non-excitable cells in terms of action potentials, they re- 70, as compared to men [16]. Homozygous APOE4 sub- spond by way of Ca2+ signals to neurotransmitters. In jects show increased risk compared to APOE3/APOE4 turn, such Ca2+ responses promote the release of several heterologous individuals [16–18], with men being at molecules called gliotransmitters, such as ATP, glutam- greater risk as reviewed by Riedel and colleagues [17] ate, D-serine or GABA [29, 30, 32–34]. Ca2+ signaling is Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 3 of 23 thus considered the basis of astrocyte excitability, which APOE3 or APOE4 gene replacing the endogenous mouse is exquisitely precise owing to unique spatiotemporal APOE gene were purchased from Taconic (USA) [42, Ca2+ features resulting from the combination of Ca2+ 43]. All experimental procedures were performed ac- signaling pathways orchestrated by different second cording to the animal research regulations (RD53/2013 messengers, and intracellular organelles [35]. Pathways and 2010/63/UE) of Spain and the European Union, and include inositol 1,4,5-triphosphate (IP3)-mediated Ca2+ with the approval of the Committees of Animal Research release from the endoplasmic reticulum (ER) [36], and from the Institutional Animal Ethics Committee of nicotinic acid adenine dinucleotide phosphate CSIC. Animals were housed in standard laboratory cages (NAADP)-elicited mobilization from acidic organelles, a with ad libitum access to food and water, under a 12:12 heterogeneous population of vesicles highly enriched in h dark-light cycle in temperature controlled rooms. lysosomes [37]. Extracellular Ca2+ entry and mitochon- drial Ca2+ uptake can further shape cytosolic Ca2+ in- Cell culture creases [37, 38]. Of note, dysregulation of Ca2+ signaling APOE3 and APOE4 immortalized astrocytes were a gift in astrocytes has been reported in animal models of from Dr. Holtzman (Washington University) [44]. Im- LOAD [39]. However, Ca2+ signaling has never been mortalized astrocytes were routinely grown at 37 °C in studied in astrocytes in the context of APOE4. 5% CO2 humidified atmosphere air in advanced DMEM APOE genotype confers distinct composition to lipid medium supplemented with 1 mM of Na+ pyruvate, 10% membranes [40], and it is well established that lipid fetal bovine serum (FBS), 100 U/mL of penicillin, composition (e.g., contents of cholesterol and phospho- 100 μg/mL of streptomycin, and 0.2 mg/ml of geneticine. lipids) affects the function of membrane-associated Cell passages were performed with trypsin once per enzymes, receptors, and channels [41]. On the basis of week up to passage 10. this evidence, we posited that APOE4 dysregulates Ca2+ excitability in astrocytes by modifying membrane lipid Ca2+ imaging in astrocytes ex vivo in brain slices composition. To test this hypothesis, we studied ex vivo Animals were sacrificed; the brain was rapidly removed Ca2+ signaling in hippocampal astrocytes of female and and placed in ice-cold modified artificial cerebrospinal male mice with human APOE3 and APOE4 gene fluid medium (ACSF). To reduce swelling and damage targeted replacement of the endogenous mouse APOE in superficial layers, N-methyl-D-glucamine (NMDG)- [42, 43]. We found that Ca2+ responses induced by the ACSF was used during brain sectioning, NMDG being a stimulation of purinergic receptors were upregulated in substitute for sodium ions in a wide range of adult ages APOE4 versus (vs) APOE3 astrocytes in male, but not in and applications [45]. NMDG-ACSF contained [in mM]: female mice. Next, we clarified the underlying mecha- NMDG 93, KCl 2.5, NaH2PO4 1.2, NaHCO3 30, HEPES nisms in cultured immortalized astrocytes expressing 20, MgSO4 10, CaCl2 0.5, glucose 25, sodium ascorbate human APOE3 and APOE4 [44]. The cells reproduced 5, thiourea 2, sodium pyruvate 3, gassed with 95% O2/ the ex vivo model of male mice, as Ca2+ responses were 5% CO2 (pH 7.3–7.4). Hippocampal slices (350 μm) ob- increased in immortalized APOE4 cells as compared to tained with a vibratome (Leica Vibratome VT1200S, APOE3 cells. Cultured APOE4 astrocytes released more Germany) were incubated in NMDG-ACSF for 10 min Ca2+ from their acidic stores upon purinergic stimula- at 34 °C, and then equilibrated for more than 1 h at tion. Manipulation of APOE4 and APOE3 expression re- room temperature (22–24 °C) prior cell loading with vealed that it is the allele, not the reduced synthesis, that fluorochromes in ACSF containing [in mM]: NaCl 124, caused altered Ca2+ signaling in immortalized APOE4 KCl 2.69, KH2PO4 1.25, MgSO4 2, NaHCO3 26, CaCl2 2, astrocytes. Finally, we demonstrated that APOE4 astro- and glucose 10, gassed with 95% O2/5% CO2 (pH 7.3– cytes have distinct lipidome and were refractory to con- 7.4). Slices were loaded with sulforhodamine 101 trol Ca2+ fluxes by extracellular lipids. Taken together, (SR101, 1 μM) for 20 min at 34 °C for astrocyte labeling the data suggest that the APOE genotype modulates [46–48], and washed in ACSF for 10 min at 34 °C. After Ca2+ fluxes in astrocytes in a lipid, lysosome and sex- SR101 labeling, slices were loaded with 2 mM of fluo-4/ dependent manner. Future research will clarify whether AM for 20–30 min at room temperature. After washing dysregulation of astrocyte excitability contributes to the fluo-4/AM overloading in ACSF, slices were kept in increased risk of developing LOAD and other brain ACSF medium supplemented with 10% FBS at the stage pathologies in APOE4 carriers. of a Nikon Eclipse FN1 microscope coupled to a CCD camera (ORCA-R2, Hamamatsu, Japan). Cells were illu- Materials and methods minated for 100–200 ms at 490 nm using LED system Animal model (CoolLED pE-100), and images from stratum radiatum Nine- to 12-week-old male and female APOE3 and astrocytes were acquired every 1 s. The LED system and APOE4 transgenic mice homozygous for the human the camera were controlled and synchronized by the Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 4 of 23 NIS-Elements software (Nikon, Japan) that was also used medium without FBS and antibiotics. This solution was for epifluorescence measurements. Astrocyte Ca2+ levels kept at room temperature for 20 min and administered were recorded from the astrocyte cell body, and Ca2+ to the cells. After 5 h, the medium was replaced by variations were quantified as changes in the fluorescence growth medium. The quantities of lipofectamine and signal (F) over the baseline (F0) ((F-F0)/F0). Two proto- DNA or siRNA were: 9 μL of lipofectamine and 2.5 μg/ cols of Ca2+ monitoring were used. Firstly, spontaneous mL GECI plasmids; 12 μL of lipofectamine and 4.5 μg/ events were studied by recording Ca2+ events for 120 s. mL of APOE plasmids (described in [50]), and 4.5 μL of Secondly, ATP-induced responses were studied by re- lipofectamine and 1 μg/mL of APOE siRNA (s194291, cording Ca2+ baseline for 30 s, followed by local applica- Thermo Fisher Scientific) or negative control 1 μg/mL tion of ATP (1 mM; 5 s) to activate purinergic receptors siRNA (Thermo Fisher Scientific). for 30 s, and post ATP period for 60 s. Local application of ATP was delivered by pressure pulses through a Lysosomal pH measurement micropipette (Picospritzer II, Parker Hannifin, Mayfield Lysosomal pH was measured using the ratiometric dye Heights, OH, USA) in the presence of tetrodotoxin LysoSensor Yellow/Blue DND-160 (Thermo Fisher (TTX, 1 μM). Matlab software (MATLAB R2016; Math- Scientific) as described [51] with minor modifications. works Natick) custom-written plugin was used for com- Briefly, cells were incubated with 2 μM of LysoSensor putation of fluorescence values of each region of interest Yellow/Blue in an isotonic solution with addition of 10% (ROI). FBS. Next, cells were incubated with either additional isotonic solution for pH measurement, or with pH cali- Ca2+imaging in astrocytes in vitro in immortalized bration buffers with the corresponding FBS content to astrocytes perform a standard curve for each condition. 15 μM of Intracellular Ca2+ measurement was performed using monensin and 30 μM of nigericin were added to the pH- 2 μM of fura-2/AM, a ratiometric fluorescence indicator calibration solutions to force lysosomal pH to equilibrate whose fluorescence ratio (R340/380) is proportional to with a range of pH values (solutions at 4.0, 4.5, 5.0, 5.5 Ca2+ concentration. Imaging of Ca2+ signaling in organ- and 6.0 pH). Fluorescence was measured with a GENios elles was performed using genetically encoded Ca2+ indi- Pro Fluorometer, and recorded using the XFluor4GEN- cators (GECI). The plasmids for endoplasmic reticulum iosPro software package (TECAN). Lysosomal pH was (pCMV G-CEPIA1er) and mitochondrial (pCMV determined from the ratio of excitation light at 340 nm CEPIA3mt) were a gift from Masamitsu Iono [49]. and 390 nm (collection emission at 535 nm) after Fluorescence was recorded using a TE-2000 U Nikon extrapolation with the standard curve. epifluorescence microscope keeping the cells at 37 °C. Cells were excited with a monochomator (Cairns, UK) Immunocytochemistry and the emitted fluorescence was collected every 2 s by Cells were fixed with 4% paraformaldehyde for 15 min the high sensitivity CCD EG-ORCA camera (Hamama- and permeabilized by adding 0.1% of Triton buffer. tsu Photonics, Japan) using a 40x oil objective (Nikon, Then, 5% of normal goat serum (NGS) was used to Japan). The resulting images were analyzed through block the unspecific unions. Primary antibodies (mouse MetaFluor Software (Universal Imaging, Bedford Hills, monoclonal anti-ApoE (sc53570, Santa Cruz Biotechnol- NY, USA). For experiments using fura-2/AM, the ratio ogy, dilution 1:400) and rat monoclonal anti-Lamp1 between fluorescence after excitation at 340 and 380 nm (1D4B, Hybridoma Bank, dilution 1:200) were incubated was calculated, whereas for experiments using GECI the overnight, followed by one-hour incubation with the sec- fluorescence of the CEPIA indicators at a given time ondary antibody Cy3-donkey anti-rat IgG (712–165-150, point was normalized to the initial fluorescence (F/F0). Jackson Immunoresearch, dilution 1:200) and Alexa Ratiometric fluorescence values were obtained using the fluor 488 goat anti-mouse IgG (A11029, Thermo Fisher MetaFluor software and selecting individual cells (ROI). Scientific, dilution 1:1000). Coverslips were mounted on Two to four coverslips with 15 to 25 cells per coverslip a slide with Fluoromount G. Images were acquired with were analyzed for each condition. Data values were confocal laser scanning microscopy ZEISS LSM 700. further analyzed with GraphPad Prism 6. Analysis of lysosome localization was carried out using the ImarisCell tool of IMARIS software (Bitplane), which Plasmid transfection and siRNA silencing permits manual selection of the nucleus and the mem- Cells were transfected with plasmids using Lipofecta- brane of each cell and computes the distance (μm) from mine 2000 (Thermo Fisher Scientific) while transfection each Lamp1-positive vesicle to the nucleus center. with siRNA to achieve silencing was accomplished with Lamp1-positive vesicles larger than 0.5 μm were consid- Lipofectamine RNAiMAX (Thermo Fisher Scientific). ered. The frequency distribution of vesicles from the nu- Briefly, lipofectamine and DNA or siRNA were added to cleus center to the cellular membrane was represented. Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 5 of 23 Cholesterol staining proceeded to identify metabolites with VIP > 1 compar- Filipin III (Sigma-Aldrich) was used to stain cholesterol. ing their accurate masses with those available in metabo- Cells were fixed with 4% paraformaldehyde for 15 min lomics databases (HMDB, METLIN, KEGG and and incubated with 25 μg/mL Filipin for 30 min at room LIPIDMAPS) [53]. Finally, fold changes of identified me- temperature in the dark. Images were acquired with a tabolites in APOE4 vs APOE3 astrocytes were calculated CCD ORCA-EG monochromatic camera (Hamamatsu) and a multi-t statistical test with corrected probability and the Eclipse TE-2000E (Nikon) epifluorescence (False Discovery Rate –FDR- 5%) was applied. microscope, using a 20x objective. Western blot Lysosomal isolation Cells for protein extraction were lysed with RIPA buffer Lysosomes from astrocytes were isolated as described and extracts were sonicated and centrifuged. Protein ex- [52]. Briefly, collected cells were washed in sucrose 0.25 tracts were quantified with BCA kit (23,225, Thermo M (pH 7.2) and disrupted in a nitrogen cavitation Fisher Scientific) according to the manufacturer’s proto- chamber (Kontes Glass Company), followed by col. 20 μg protein of samples was loaded in 12% poly- homogenization in a Teflon-glass homogenizer and acrylamide gels electrophoresis (PAGE). Electrophoresis centrifuging (2500 xg for 15 min). The mitochondria- was conducted at a constant amperage (30 mA) for ap- lysosomal-enriched faction was collected after 17,000 xg proximately 2 h followed by the protein transference to a centrifugation. The lysosomal-mitochondrial-enriched PVDF membrane at constant voltage (100 mV) for 1.5 h. fraction was then loaded in two subsequent discontinu- 5% non-fat milk was used to block unspecific unions, ous metrizamide/sucrose/percoll density gradients from and the primary antibodies mouse monoclonal anti-V- which lysosomal pure fractions were isolated [52]. Lyso- ATPase subunit V0D1 (ab56441, Abcam, used at 2.5 μg/ somes were broken by 5 consecutive freeze/thaw cycles mL) and monoclonal anti β-actin (a5316, SigmaAldrich, and after centrifugation at 100.000 xg 30 min, lysosomal used at 1/20000), were incubated overnight. The next membrane and intralysosomal content were collected day, the secondary antibody was incubated for 1 h (goat separately. The former pool was used for lipidomic anti-mouse IgG: 31430, Thermo Fisher Scientific, used analysis. at 1/10000). Finally, the membrane was developed using the chemiluminescence kit of BioRad according to the Lipidomics manufacturer’s protocol, and membrane chemilumines- The extraction of metabolites for the untargeted lipido- cence was detected with a Chemidoc MP Image System mic assay was carried out by adding 200 μL CHCl3: (BioRad). Image lab software (BioRad) was used for the MeOH in a proportion of 1 to 2 with 0.1% formic acid quantification of bands. to promote the ionization of molecules. Then, samples were vortexed, centrifuged at 4000 rpm, and analyzed Measurement of mRNA expression with mass spectrometry using the QSTAR XL hybrid Cells were collected adding Trizol Reagent. Then, 0.2 system (Applied Biosystems, Foster City, CA, USA). The mL chloroform was added to the extracts, and these sample was injected at a flux of 15 μL/min through the were centrifuged at 11,500 rpm for 15 min at 4 °C allow- infusion integrated pump. Spectra were acquired during ing the formation of three phases. The RNA-containing positive ionization in a range of m/z from 50 to 1100 phase (superior phase) was isolated by pipetting care- uma. The ionization parameters were: 3300 V of voltage fully, and 0.2 mL of isopropanol was added to precipitate ion spray, 60 V of decluttering potential, and 250 V of of RNA, which was then washed with cold ethanol 75%. focusing potential. Nitrogen was used as a collision gas RNA-sample concentration was determined with a for the spectra acquisition. Markerview™ and SIMCA-P™ Nanodrop 200 spectrophotometer (Thermo Fisher software were used to reduce the results into a two- Scientific). 2 μg of RNA was reverse transcripted to dimensional matrix of peak spectra and intensity of cDNA using 1 μM of oligo DT, 1 μM of hexamers, 0.5 peaks, and for the statistical analysis. Inter-genotype mM dNTPs, 0.45 mM DTT, 10 U RNAse out, RT buffer, comparisons were carried out with the multivariate and 200 U of retrotranscriptase. The PCR program was analysis Partial Least Squares-discriminant analysis divided into four steps: 65 °C for 10 min, 25 °C for 10 (PLS-DA). Next, the Variable Importance in Projection min, 42 °C for 1 h, and 72 °C for 10 min. (VIP) was used to establish which metabolites had more Gene expression was carried out with quantitative impact in the segregation of samples according to real-time PCR (qPCR) using Taqman and SYBR green APOE3 and APOE4 genotypes. ANOVA with a Tukey technology. Fluorescence was detected with the 7500 correction post-test was applied to the group of metabo- Fast Real-Time PCR System. qPCR cycles were the fol- lites with a VIP > 1 to assess, again, inter-genotype lowing: a holding stage of 50 °C (2 min), 10 min at 95 °C differences with the identified metabolites. We then and 40 cycles of 95 °C (15 s), and 60 °C (1 min). A similar Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 6 of 23 protocol was used for SYBR assay but with an extra and in vitro, triggering several physiological functions [34, stage for the melting curve (15 s at 95 °C, 1 min at 60 °C, 55], and because responses elicited by stimulation of 30 s at 95 °C, and 15 s at 60 °C). Data analysis was purinergic receptors are the cause of Ca2+ hyperactivity in performed using Cq value, and the average of the gene AD mouse models [56]. efficiency provided by LinReg PCR software, following In male mice, we detected increased amplitude of the formula 1 + eficienceΛCq of each gene analyzed. spontaneous Ca2+ transients in APOE4 vs APOE3 astro- Expression data were normalized with housekeeping cytes (Fig. 1b). Regarding induced activities, purinergic genes (Gapdh and/or18s), using their geometric mean stimulation caused in both genotypes the increase in calculated according to the geNorm algorithm [54]. magnitude and frequency of Ca2+ transients typically TaqMan primers of APOE (Hs00171168_m1), Trpml seen in astrocytes. The amplitude of Ca2+ responses was (Mm00522550_m1), Gapdh (Mm9999915_g1), and 18S greater in APOE4 cells than in APOE3 cells (Fig. 1c and (Mm03928990) were purchased from Thermo Fisher d). Further, the magnitude of Ca2+ responses decreased Scientific. Primers for SYBR green were designed after removal of ATP (post ATP phase) to basal levels in according to the sequence of the gene: Tpc1 (5′-CTGG APOE3 astrocytes, but remained significantly increased GAGAGATGAATTATCAAGAG-3′; 5′-GTTGTGTAC- over its own basal levels, and with respect to APOE3 GAAGAGGTAGG-3′), Tpc2 (5’GCTGAGCCTTGCTT cells, in APOE4 cells (Fig. 1c and d). The frequency of GTGAGG-3′; 5′-ACACTTCAGGGTCTTCTTCATCA- Ca2+ responses was similar in both genotypes, in both 3′), and Gapdh (5′-AAGCTCATTTCCTGGTATGAC- the ATP and post-ATP phases. It is worth stressing that 3′; 5′-TGGTCCAGGGT TTCTTACTC-3′). the amplitude of spontaneous Ca2+ transients was also statistically increased in APOE4 compared to APOE3 Statistical analysis cells in these set of experiments, confirming the results Each determination was carried out with cells from at of the 120-s recordings (Fig. 1c and d). least three different passages. A parametric unpaired T- Female mice differed from males in two respects. First, test was used for the comparison of a given variable in the amplitude of spontaneous and ATP-induced events two different conditions or cell types, whereas one-way or was significantly increased (p < 0.001), by at least 2-fold, two-way analysis of variance (ANOVA), with a Tukey’s or in astrocytes from APOE3 females as compared to Dunn’s post hoc test, was used when comparing more APOE3 males (compare Fig. 1b with 2a and 1d with 2c). than two conditions. The software employed was Graph- Second, no differences were observed between APOE3 Pad Prism 6, and data are represented as mean ± SEM and APOE4 astrocytes in females in the magnitude of (standard error of the mean). A p-value < 0.05 was consid- spontaneous and induced events (Fig. 2a-c). This may ered significant (p-value < 0.05 (*), p-value < 0.01 (**), p- suggest that Ca2+ responses in astrocytes from APOE3 value < 0.001 (***) and p-value < 0.0001(****). females represented the maximal Ca2+ response that could not increase further. Altogether, the ex vivo obser- Results vations suggest that expression of human APOE alleles APOE4 expression alters astrocytic excitability modulates Ca2+ transients in astrocytes in a sex-specific To determine whether expression of the allele APOE4 manner, such that expression of the APOE3 allele in alters astrocyte excitability, we recorded Ca2+ in hippo- male mice results in globally decreased Ca2+ transients campal slices of 9–12-week-old male and female mice in in astrocytes—or expression of APOE3 in females in a which the endogenous mouse APOE gene had been re- global increase—as compared to the APOE4 allele. placed with human APOE3 or APOE4 genes. Recordings were made in ACSF medium supplemented with 10% Immortalized APOE4 astrocytes show increased Ca2+ FBS, in order to keep lipoprotein and lipid concentration mobilization from acidic stores as close as possible to physiological conditions. Slices were To gain insight into the mechanism by which expression incubated with the Ca2+ indicator fluo-4/AM and the of different APOE alleles regulates Ca2+ transients in as- astrocytic marker sulforhodamine (SR101) (Fig. 1a, see trocytes, we used immortalized astrocytes that express Materials and methods). We analyzed Ca2+ spontaneous human APOE3 or APOE4 [44] since this in vitro model activity—that is, Ca2+ events at rest conditions—and allows for experimental manipulations that are not feas- neurotransmitter-induced Ca2+ responses in cells co- ible in brain slices. As with slices, cells were supple- labeled with Fluo-4/AM and SR101. To study spontan- mented with 10% FBS, in order to keep lipoprotein and eous activity, we recorded APOE3 and APOE4 astrocytes lipid concentration as close as possible to physiological for 120 s. For receptor-mediated Ca2+ responses, we re- conditions. It is worth stressing that immortalized astro- corded basal peak activity for 30 s, and then we stimulated cytes are aneuploid so the sexual identity is lost. Thus, a slices with 1 mM ATP. We selected purinergic stimulation key question was whether their Ca2+ phenotype is male- because it elicits Ca2+ signals both in astrocytes in vivo or female-like. Our data show that they reproduce a Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 7 of 23 Fig. 1 Enhanced Ca2+ signals in astrocytes from male APOE4 vs APOE3 targeted replacement mice. a Astrocyte from stratum radiatum of hippocampus of male mice labeled with SR101 (red), fluo-4/AM (green), and merged image. Scale bar represents 10 μm. b Spontaneous Ca2+ activity in astrocytes from APOE3 (N = 36 astrocytes) and APOE4 male mice (N = 16 astrocytes). Ca2+ was monitored for 120 s without any stimulation. c Left panels, raster plots of Ca2+ activity in APOE3 (upper panel, N = 146 astrocytes), and APOE4 male mice (lower panel, N = 171 astrocytes). The color code indicates relative fluorescence changes before and after local application of ATP (arrow, 5 s, 1 mM). Right panel, representative traces of Ca2+ signals evoked by an ATP puff in APOE3 (top) and APOE4 (bottom) astrocytes (arrows indicate ATP stimulation). d Quantification of the amplitude and frequency of Ca2+ events for 30 s before (basal peaks), during (ATP) and after local application of stimulus (post-ATP). Statistical significance was established at p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***); One-way ANOVA followed by Dunn’s post hoc test. All the experiments were performed in the presence of TTX (1 μM). N = 4 mice, for both APOE3 and APOE4 male-like Ca2+ signaling phenotype in the presence of 100 μM ATP stimulation resulted in greater Ca2+ re- lipids. First, differences in Ca2+ responses at rest showed sponses in APOE4 than in APOE3 astrocytes (Fig. 3a). the same trend observed in male mice. APOE4 astrocytes Purinergic-induced Ca2+ responses also lasted longer: had significantly different Ca2+ basal levels than APOE3 the response was 64.9 ± 8.2% of the maximum peak sig- astrocytes (p-value = 0.01), the fluorescence ratio being nal after 20 s in APOE4 cells but only 27.7 ± 1.5% in 0.28 ± 0.01 and 0.41 ± 0.03, respectively (Fig. 3a-c). Note APOE3 astrocytes. Moreover, altered Ca2+ signaling was that cultured astrocytes do not show at rest the so-called not restricted to purinergic stimulation, as adrenergic spontaneous Ca2+ oscillations observed ex vivo, but and muscarinic-receptor activation also triggered greater stable basal Ca2+ levels that we could compare thanks to cytosolic Ca2+ responses in APOE4 than in APOE3 as- the ratiometric fura-2/AM Ca2+ indicator. Second, trocytes (Fig. 3b and c). Importantly, the magnitude of Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 8 of 23 Fig. 2 Equal Ca2+ signals in astrocytes from female APOE3 and APOE4 targeted replacement mice. a Spontaneous Ca2+ activity in astrocytes from APOE3 (N = 26 astrocytes) and APOE4 female mice (N = 35 astrocytes). Ca2+ was monitored for 120 s without any stimulation. b Left panels, raster plots of Ca2+ activity in astrocytes from APOE3 (upper panel, N = 165 astrocytes), and APOE4 female mice (lower panel, N = 124 astrocytes). The color code indicates relative fluorescence changes before and after local application of ATP (arrow, 5 s, 1 mM). Right panel, representative traces of Ca2+ signals evoked by an ATP puff in APOE3 (upper panel) and APOE4 (bottom panel) astrocytes (arrows indicate ATP stimulation). c Quantification of the amplitude and frequency of Ca2+ events for 30 s before (basal peaks), during (ATP) and after local stimulus (post-ATP) in APOE3 and APOE4 astrocytes. Statistical significance was established at p < 0.001 (***); One-way ANOVA followed by Dunn’s post hoc test. All the experiments were performed in the presence of TTX (1 μM). N = 4 mice, for both APOE3 and APOE4 purinergic-induced Ca2+ responses was the same with among the principal intracellular Ca2+ sources with two other FBS batches (peak responses after stimulation organelle-specific probes and pharmacological manipula- with 100 μM ATP were 0.36 ± 0.13 in APOE3 and 0.94 ± tions. First, we investigated the mitochondrial Ca2+ up- 0.01 in APOE4; and 0.37 ± 0.04 in APOE3 and 0.91 ± take that characteristically buffers increases in cytosolic 0.05 in APOE4 astrocytes). Taken together, the data Ca2+. Since APOE4 has been described as harming mito- support immortalized APOE3 and APOE4 astrocytes as a chondria in neurons [57], we reasoned that harmed model to study the mechanisms underlying the regula- mitochondria in APOE4 astrocytes could result in defi- tion of Ca2+ responses by APOE alleles in males. cient Ca2+ uptake, and hence in increased intracellular Next, in order to identify which pathways are dysregu- Ca2+ responses. However, expression of the mitochon- lated in APOE4 astrocytes, we examined Ca2+ fluxes drial Ca2+ indicator CEPIA3mt (Fig. 3d) showed higher Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 9 of 23 Fig. 3 Enhanced Ca2+ signals in immortalized astrocytes from APOE4 vs APOE3 targeted replacement mice. Ca2+ responses measured using fura- 2/AM in APOE3 and APOE4 astrocytes after stimulation with (a) 100 μM ATP, (b) 10 μM noradrenaline (NA), and (c) 100 μM acetylcholine (Ach). Representative traces and quantification of the magnitude of the responses are shown (N = 4 for a and c, and N = 3 for b). d Images of astrocytes transfected with the Ca2+ probes for mitochondria (CEPIA3mt) and ER (G-CEPIA1er). Scale bar represents 15 μm. e Representative traces and quantification of mitochondrial Ca2+ in APOE3 and APOE4 cells transfected with CEPIA3mt upon stimulation with ATP (N = 3). f Representative traces and quantification of the decrease in ER Ca2+ upon stimulation of purinergic receptors in APOE3 and APOE4 cells transfected with GCEPIA1er (N = 5). Unpaired parametric T-test was used to compare responses in APOE3 vs APOE4 astrocytes. p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) Ca2+ uptake in APOE4 mitochondria compared to mainly lysosomes and related organelles [58] that we APOE3 astrocytes, consistent with the higher ATP- have shown to be involved in purinergic-induced Ca2+ induced Ca2+ responses in the cytosol (Fig. 3e). Second, responses in astrocytes [37]. Figure 4a shows the main we studied the main Ca2+ signaling pathway in astro- Ca2+ fluxes in lysosomes. We recorded cytosolic Ca2+ cytes, Ca2+ mobilization from the ER through the IP3 re- with fura-2/AM, after inhibiting Ca2+ release from acidic ceptor, by transfecting cells with G-CEPIA1er (Fig. 3d), stores with 100 μM of Ned-19, an inhibitor of NAADP and directly measuring Ca2+ contents inside this organ- receptors responsible for Ca2+ release from these organ- elle. As expected, 100 μM ATP decreased Ca2+ levels in elles [59]. Control cells were treated with DMSO, the the ER of both APOE3 and APOE4 astrocytes, indicative vehicle of Ned-19. As expected, Ned-19 reduced ATP- of Ca2+ being released to the cytosol. Since the process induced Ca2+ responses in APOE3 cells (Fig. 4b), is, although significantly, just slightly reinforced in confirming the contribution of lysosomal Ca2+ to cyto- APOE4 astrocytes (Fig. 3f), the greater cytosolic Ca2+ solic transients [37]. Interestingly, Ned-19 greatly responses in these cells could not rely exclusively on reduced Ca2+ responses in APOE4 astrocytes, such that increased Ca2+ mobilization from the ER. Third, we purinergic-mediated Ca2+ responses in the presence of explored Ca2+ mobilization from acidic stores, which are Ned-19 were of similar magnitude in both cell types. Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 10 of 23 Fig. 4 Dysregulated Ca2+ excitability in APOE4 immortalized astrocytes is due to greater lysosomal V-ATPase activity. a Schematic of the lysosomal Ca2+ mobilization and uptake mechanisms. b Representative traces and quantification of ATP-induced cytosolic Ca2+ responses of cells treated with 0.1% DMSO or 100 μM Ned-19 (diluted in 0.1% DMSO) for 20 min (N = 3). c Quantification by real time qPCR of mRNA expression of Tpc1 and Tpc2 (normalized to Gapdh mRNA contents, N = 5) and Trpml (normalized to Gapdh mRNA contents, N = 4). d Representative traces and quantification of the area under the curve (AUC) of intracellular Ca2+ after lysing the lysosomes with 200 μM GPN (N = 4). e Representative traces and quantification of ATP-induced Ca2+ responses after treating cells with DMSO or 2 μM bafilomycin A1 (Baf) for 20 min (N = 3). Note that the black arrow indicates differences in basal Ca2+. f Quantification of V-ATPase subunit V0D1 with western blot, normalized to β-actin expression (N = 4). g Quantification of the pH of acidic organelles in APOE3 and APOE4 astrocytes (N = 3). Unpaired parametric T-test was used, except in b and e, where one-way ANOVA was used. p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) Hence, Ca2+ release from acidic-stores appears to be re- accepted candidates being two-pore channels 1 and 2 sponsible for the greater purinergic-elicited Ca2+ re- (Tpc1, Tpc2), and transient receptor potential mucolipin sponses in APOE4 astrocytes. Altogether, our results (Trpml) [59]. By contrast, Ca2+ uptake by acidic stores is thus far lend support to the idea that lysosomal-related accomplished through an indirect mechanism whereby Ca2+ release is the main Ca2+ signaling pathway dysregu- V-ATPase pumps H+ into the vesicles, and then H+ are lated in APOE4 astrocytes. exchanged with Ca2+ through the Ca2+/H+ exchanger (CAX) [60] (Fig. 4a). Thus, greater Ca2+ release from Dysregulation of V-ATPase activity contributes to the acidic stores in APOE4 astrocytes might be due to two alteration of Ca2+ responses in APOE4 astrocytes phenomena: greater expression or activation of NAADP Ca2+ is released from acidic stores upon stimulation of receptors, or greater Ca2+ stored in these organelles, due NAADP receptors, which are Ca2+ channels, the most to increased activity of V-ATPase and/or CAX. First, we Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 11 of 23 analyzed the expression of NAADP receptors in astro- it follows that CAX activity must be increased in APOE4 cytes with real time qPCR. Expression of Trpml and astrocytes, too, such that the increased number of H+ Tpc1 channels is the same in APOE3 and APOE4 cells, entering the acidic stores exit through CAX in exchange while expression of Tpc2 channel is lower in APOE4 for Ca2+. As a result, lysosomal Ca2+ concentration and cells (Fig. 4c), perhaps as a result of a negative feedback hence basal Ca2+ and purinergic-induced Ca2+ release mechanism in response to the high Ca2+ signals in from lysosomes, are greater in APOE4 than in APOE3 APOE4 astrocytes. In any event, the increased release of astrocytes. Note that we did not determine nor manipulate Ca2+ from APOE4 lysosomes cannot be attributed to in- CAX expression in APOE3 and APOE4 cells because the creased expression of NAADP receptors. sequence of mouse CAX is unknown. Second, we analyzed Ca2+ contents inside acidic stores with a protocol in which they were osmotically lysed due APOE3 expression in APOE4 astrocytes reduces cytosolic to the accumulation of the peptide Glycyl-L-phenylalanine Ca2+ responses 2-naphthylamide (GPN), followed by its proteolysis by ca- Is the effect of APOE4 due to: (1) loss-of-function due to thepsin C, such that the stored lysosomal Ca2+ is released decreased contents [62], (2) malfunction, (3) gain-of- into the cytoplasm. Then, the area under the curve (AUC) toxic function because of its structure, or (4) misplaced of the cytosolic Ca2+ increase after addition of GPN to the intracellular localization? We first examined ApoE levels astrocytes was used to calculate the total amount of stored in immortalized APOE3 and APOE4 astrocytes by quan- calcium in these organelles. The results showed greater tifying mRNA levels with qPCR. As shown in Fig. 5a, AUC in APOE4 cells (Fig. 4d), supporting a higher con- APOE expression is lower in APOE4 than in APOE3 as- centration of lysosomal-related Ca2+ in APOE4 cells, as trocytes. Therefore, it is plausible that lower expression compared to APOE3. Since there are no pharmacological of APOE4 accounts for Ca2+ signaling alterations. To modulators of CAX, we relied on a pharmacological in- test this possibility, we modulated the quantity of hibitor of V-ATPase, bafilomycin A1, and on measure- APOE3 and APOE4 with the rationale that, if APOE ex- ments of pH in acidic organelles, to establish the pression matters, decreasing ApoE in APOE3 cells would implication of the pump and CAX in the increased lu- increase Ca2+ responses, whereas increasing ApoE in minal Ca2+. Bafilomycin A1 increased basal cytoplasmic APOE4 cells would decrease Ca2+ responses. Upon de- Ca2+ concentration in APOE4 astrocytes to 0.40 ± 0.05 creasing APOE expression in APOE3 astrocytes with an (control values of astrocytes treated with vehicle were APOE siRNA (Fig. 5b), ATP-induced Ca2+ responses 0.30 ± 0.01; p-value = 0.05), whereas there was no effect on were still lower, as compared to APOE4 astrocytes, with non-stimulated intracellular Ca2+ levels in APOE3 (0.44 ± no differences observed between scramble- and siRNA- 0.08 for cells treated with bafilomycin A1 compared to transfected APOE3 cells (Fig. 5c). We also tried the 0.39 ± 0.02 of cells treated with vehicle). This suggests that opposite strategy: we increased APOE expression using the aforementioned lower basal cytoplasmic Ca2+ concen- GFP-APOE3 or GFP-APOE4 plasmids, which overex- tration in APOE4 cells is due to greater V-ATPase- press a GFP-ApoE fusion protein, which then allows one mediated Ca2+ uptake into acidic stores, consistent with to identify cells with increased APOE expression (Fig. the greater intralysosomal Ca2+ levels revealed by the 5d). The GFP signal did not interfere with fura-2/AM GPN experiments. Accordingly, bafilomycin A1 reduced (data not shown). Because transfection efficiencies were the ATP-induced Ca2+ release in APOE4 astrocytes, con- different among experiments (5 to 50% of GFP trans- firming the dependence of these Ca2+ responses on acidic fected cells), we avoided inter-experimental variability by stores (Fig. 4e). In contrast, bafilomycin A1 did not change normalizing peak Ca2+ responses to levels recorded in purinergic-induced Ca2+ responses in APOE3 astrocytes. APOE3 astrocytes transfected with a GFP-expressing Thus, both basal levels and ATP responses in the presence plasmid without APOE constructs. The over-expression of bafilomycin A1 might indicate the existence of low V- of GFP-APOE3, but not of GFP-APOE4, transformed ATPase activity in APOE3 astrocytes. Indeed, the expres- APOE4 into APOE3 astrocytes, in a statistically signifi- sion of V-ATPase subunit V0D1, which is highly expressed cant manner (Fig. 5e), suggesting that Ca2+ hyperactivity in astrocytes [61], was higher in APOE4 cells than in in APOE4 astrocytes is mainly due to the expression of APOE3 cells (Fig. 4f). Because V-ATPase controls the flow this particular isoform, but not to lower APOE expres- of H+ into acidic stores, we analyzed the pH of acidic sion. The data do not support, however, a gain of a toxic stores with the probe lysosensor and fluorometry. We function by APOE4, because, if this were the case, found that both APOE3 and APOE4 astrocytes had stand- APOE4-associated dysfunction would have been potenti- ard lysosomal-related pH, with a minimal but significant ated by GFP-APOE4 over-expression, and would not difference of 0.14 pH units between genotypes, APOE4 ly- have been rescued by GFP-APOE3, since the toxic elem- sosomes being more alkaline (Fig. 4g). If the pH does not ent, APOE4, remained. In agreement, there were no dif- decrease despite increased activity of the V-ATPase pump, ferences between ATP-induced Ca2+ signals in APOE3 Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 12 of 23 Fig. 5 Effects of the APOE allele vs APOE contents on Ca2+-based excitability. a APOE mRNA expression relative to Gapdh and 18S mRNA contents in APOE3 and APOE4 immortalized astrocytes, quantified by real time qPCR (N = 3). b Representative ApoE western blot of APOE3 cells transfected with scramble or APOE siRNA. c Representative traces and quantification of the magnitude of ATP-induced Ca2+ responses in APOE3 astrocytes transfected with lipofectamine (lipo), plus scramble (sc) or APOE siRNA, and APOE4 astrocytes treated only with lipofectamine (N = 4). d Representative images of immunocytochemistry of APOE3 and APOE4 cells transfected with plasmids expressing GFP, GFP-APOE4, and GFP-APOE3, as indicated. Scale bar represents 15 μm. e Representative traces and quantification of the magnitude of 100 μM ATP-elicited Ca2+ responses in APOE3 and APOE4 cells transfected with different plasmids as indicated. Ca2+ peaks after ATP stimulation are relative to the responses of APOE3 cells transfected with GFP. Only GFP fluorescent cells were analyzed (at least 30 cells from 4 independent experiments). f Quantification of the magnitude of Ca2+ responses elicited by 100 μM ATP in APOE3 astrocytes transfected with GFP or GFP-APOE4 plasmids, as indicated. Ca2+ peaks after ATP stimulation are relative to the responses of APOE3 cells transfected with GFP. Only GFP fluorescent cells were analyzed (at least 30 cells from 4 independent experiments). g Representative images of Lamp1 (red) and ApoE (green) immunocytochemistry in APOE3 cells. Merged images and amplifications of the area in the white frame, are displayed. Scale bar represents 15 μm. h Representative images of Lamp1 immunocytochemistry (red) and GFP-ApoE fluorescence (green) in APOE4 cells. Merged images and the amplification of a cell of each image, indicated with a white frame, are displayed. Scale bar represents 15 μm. Unpaired parametric T-test was used in a, and one-way ANOVA in c and e. p < 0.05 (*), p < 0.01 (**), p < 0.001 (***), p < 0.0001 (****) Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 13 of 23 cells transfected with GFP-APOE4 or with GFP (Fig. 5f), supplement without lipoproteins and restricted compos- again ruling out a toxic effect of APOE4. Taken together, ition of lipids: linoleic acid, linolenic acid, progesterone, our data support APOE4 malfunction rather than and corticosteroids. In the absence of lipids, non- toxicity. stimulated basal Ca2+ levels were significantly lower in Finally, because acidic stores accounted for the differ- immortalized APOE4 vs APOE3 astrocytes, as detected ences in Ca2+ responses between genotypes, we studied in the presence of lipids. Thus, in KH, basal levels were whether ApoE was localized in the acidic stores. Because 0.29 ± 0.02 in APOE3 vs 0.23 ± 0.03 in APOE4 (p-value = the quality of staining of such organelles with the fluoro- 0.04); in DMEM supplemented with lipoprotein- phore lysotracker was not optimal for confocal studies deficient FBS, 0.39 ± 0.07 in APOE3 vs 0.29 ± 0.06 in (data not shown), we resorted to using Lamp1, a marker APOE4 (p-value = 0.04); and in DMEM with B27, 0.30 ± for lysosomes, autophagosomes, and different vesicles of 0.02 in APOE3 vs 0.22 ± 0.01 in APOE4 (p-value = the endolysosomal pathway, including late-endosomes 0.007). By contrast, in all three conditions, stimulation of [63]. In APOE3 astrocytes, we used ApoE and Lamp-1 purinergic responses elicited Ca2+ responses of similar immunostaining (Fig. 5g). In APOE4 astrocytes, we re- magnitude in APOE3 and APOE4 astrocytes (Fig. 6a-c, lied on APOE4 overexpression with GFP-APOE4, since f). It was not the case that responses in APOE4 cells had the low amounts of APOE expression in APOE4 cells diminished, but, rather, that responses had increased in precluded immunocytochemical analysis (Fig. 5h). We APOE3 cells. That is, APOE4 astrocytes present the same observed that neither ApoE3 nor ApoE4 from GFP- magnitude of ATP-induced Ca2+ signals regardless of APOE4 colocalized with Lamp-1. These observations the lipids presents, whereas APOE3 astrocytes adapt rule out the likelihood that the alterations in lysosomal- their Ca2+ responses to the concentration of extracellular related Ca2+ handling in APOE4 astrocytes are due to lipids, with responses being low in media rich in lipids loss of a direct interaction of ApoE with channels medi- (presence of FBS) and high in the presence of low lipid ating Ca2+ fluxes, suggesting, instead, indirect actions of concentration, or no lipids. Consistent with the greater the apolipoprotein, perhaps through changes in lipid ATP-triggered Ca2+ responses, the responses also lasted homeostasis. longer in APOE3 astrocytes in the absence of lipids: the response was 36.0 ± 4.0% of the maximum peak signal ATP-induced Ca2+ responses are modulated by after 20 s in KH, 51.6 ± 4.3% in DMEM supplemented extracellular lipids in APOE3 but not APOE4 astrocytes with lipoproteins-deficient FBS and 34.6 ± 3.0% in Considering the wealth of evidence documenting how DMEM with B27, as compared to the previously re- channels and pumps involved in Ca2+ signaling are regu- ported 27.7 ± 1.5% in DMEM with FBS (p < 0.05). Thus, lated by lipids, particularly by membrane lipids [64, 65], the results point to a potentiation of long-lasting ampli- we posited that lipids regulate Ca2+ fluxes in astrocytes, fication Ca2+ pathways, perhaps by Ca2+-induced Ca2+ too, and that aberrant lipid homeostasis accounts for the release from the ER or store-operated Ca2+ entry observed dysregulation of Ca2+ fluxes in APOE4 astro- (SOCE)—that is, Ca2+ influx across the plasma mem- cytes, particularly in lysosome-related organelles. Lipid brane in response to depletion of intracellular Ca2+ modulation of Ca2+ transients in astrocytes is uncharted stores. Still, ATP responses were significantly shorter (p- territory. Thus, to gain insight into the control of astro- value < 0.01) in lipid-depleted APOE3 astrocytes as cyte excitability by lipids in the context of APOE4, we compared to APOE4 astrocytes (72.57 ± 3%, 68.4 ± 6.2% carried out two sets of experiments. First, we aimed to and 81.2 ± 6.8% for APOE4 cells kept in KH, DMEM obtain proof-of-concept that lipids regulate Ca2+ tran- with lipoprotein-deficient FBS and DMEM with B27), sients in astrocytes, by examining the effect of changing indicating that some differences in Ca2+ fluxes remain lipid contents on the excitability of immortalized APOE3 between APOE3 and APOE4 astrocytes. and APOE4 astrocytes. Second, we analyzed cholesterol It is worth stressing that the increased ATP-induced subcellular distribution, and we performed a lipidomic Ca2+ response in APOE3 astrocytes is a fast-onset analysis of lysosomal and whole-membrane to identify process, as it was observed just a few min after lipopro- candidates for excitability-modulating lipids, and tein removal. It is not transient, because it persisted 12 h changes thereof in the two genotypes. after replacement of DMEM supplemented with FBS for Thus far, all our Ca2+ imaging experiments had been DMEM supplemented with B27 (Fig. 6d and f), and it is performed in medium supplemented with FBS, rich in reversible, because it diminishes if the medium is re- nutrients, including lipoproteins. Here, we replaced this placed again by DMEM supplemented with FBS 5 min medium with three different media with lower lipid prior to Ca2+ recordings (Fig. 6e and f). composition 2–5 min prior to Ca2+ imaging: 1) Krebs Altogether, three conclusions may be drawn from medium (KH), 2) DMEM supplemented with these results: they constitute the first demonstration that lipoproteins-deficient FBS and 3) DMEM with B27, a astrocyte excitability is modulated by lipids and/or Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 14 of 23 Fig. 6 Loss of modulation of Ca2+ signals by lipids in APOE4 astrocytes. Representative traces of 100 μM ATP-induced Ca2+ responses in APOE3 and APOE4 immortalized cells in (a) saline medium or Krebs medium (KH) for 2 to 5 min, (b) DMEM medium supplemented with lipoprotein- deficient serum (Lipoprotein (−)) for 2 to 5 min, (c) DMEM medium supplemented with B27 for 2 to 5 min, and (d) overnight (ON), and (e) DMEM medium supplemented with 10% FBS for 5 min. f Quantification of the ATP-induced intracellular Ca2+ peak in cells kept in the different mediums (N = 3). Two-way ANOVA was used. p < 0.01 (**) as compared to APOE3 cells kept in the same medium lipoproteins, such regulation is reversible but stable in 19, the inhibitor of Ca2+ release from acidic stores APOE3 astrocytes as long as lipids remain present, and (Fig. 7a) might indicate that extracellular lipids decrease is lost in APOE4 astrocytes. Ca2+ levels inside such organelles. Surprisingly, measure- ment of Ca2+ stored inside acidic organelles using the The absence of lipids potentiates Ca2+ release from the GPN-elicited depletion showed no increased Ca2+ load- ER and extracellular Ca2+ entry in APOE3 but not APOE4 ing in APOE3 cells kept in KH (Fig. 7b) compared to astrocytes APOE3 astrocytes kept in the presence of FBS, whereas, Since the core of Ca2+-signaling dysregulation in APOE4 as expected, no difference was observed between the two astrocytes lies in acidic organelles, is Ca2+ homeostasis conditions in APOE4 astrocytes (Fig. 7c). An explanation in these organelles the target of extracellular lipids in is that, in the absence of extracellular lipids, the release APOE3 astrocytes? The finding that the increase in of Ca2+ from lysosomes in APOE3 astrocytes is amplified ATP-induced Ca2+ signals in APOE3 cells in the absence by the activation of other Ca2+ signaling mechanisms. In of lipids/lipoproteins (KH media) was abrogated by Ned- fact, the amplification originates in part, from Ca2+ Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 15 of 23 Fig. 7 Loss of lipid-dependent coordination of Ca2+ fluxes across plasma and lysosomal membranes in APOE4 immortalized astrocytes. a Representative traces and quantification of the magnitude of ATP- triggered intracellular Ca2+ responses in cells treated with DMSO or with 100 μM Ned 19 for 20 min in KH medium (N = 3). b, c Representative traces and quantification of lysosomal Ca2+ content measured as AUC after 200 μM GPN addition to (b) APOE3 and (c) APOE4 astrocytes in KH medium, as compared to DMEM medium supplemented with FBS (FBS) (N = 3) (FBS-related results were presented in Fig. 3); d Representative traces and quantification of Ca2+ ER release in 100 μM ATP-stimulated astrocytes transfected with G-CEPIA1er, and kept in KH medium supplemented or not with FBS (N = 4). e, f Representative traces and quantification of purinergic-induced Ca2+ responses in (e) APOE3 and (f) APOE4 astrocytes in KH medium supplemented with 10% of FBS with or without Ca2+/1 mM EGTA for 10 min, and KH medium with Ca2+ or without Ca2+/500 μM EGTA for 1 min. g Representative images from Lamp1 immunofluorescence of both cell lines and quantification of the percentage of lysosomes in different ranges of distance, the 0 value being the nucleus in each cell. (N = 3, 30 to 40 cells quantified). Scale bar represents 15 μm. One-way ANOVA was used in sections a, d, e and f, unpaired parametric T-test in b and c, and two-way ANOVA in g. p < 0.05 (*), p < 0.001 (***), p < 0.0001 (****) released from the ER, because the Ca2+ content in the EGTA greatly reduced (by 3.3 times) ATP-induced Ca2+ ER was 1.5-fold greater in APOE3 cells kept in KH than responses in APOE3 astrocytes kept in KH without lipo- in DMEM/FBS, as recorded with the Ca2+ probe proteins, but not in APOE3 astrocytes kept in KH with CEPIA1er (Fig. 7d). The same amount of Ca2+ was re- FBS (Fig. 7e). In contrast, EGTA reduced ATP-induced leased from the ER in APOE4 astrocytes kept in the two Ca2+ responses in APOE4 astrocytes to the same extent media, in agreement with the observation that Ca2+ sig- in the presence or absence of FBS (Fig. 7f). Thus, the nals are independent of extracellular lipid concentrations data support the idea that extracellular Ca2+ entry is in this genotype (Fig. 7d). Another source of Ca2+ is the secondary to lysosomal Ca2+ release, and that it is mod- extracellular Ca2+ entry, because the blockage of extra- ulated by extracellular lipids in APOE3 but not in cellular Ca2+ with the cell-impermeable Ca2+ chelator APOE4 astrocytes. We next sought morphological Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 16 of 23 support for this idea by studying the localization of and 0.995 for whole-membranes. The PLS-DA analysis acidic stores inside astrocytes. Again, we resorted to thus confirms that APOE genotype influences membrane Lamp1 immunostaining to be able to use confocal lipid composition in astrocytes. microscopy. Lysosome distribution was abnormal in In order to identify which lipids contributed more to APOE4 cells (Fig. 7g); that is, a greater number of the differential group clustering, we used the Variable Lamp1-positive organelles accumulated near the nucleus Importance in the Projection (VIP), such that lipids with in APOE4 astrocytes than in APOE3 astrocytes. Specific- VIP > 1 were the ones with greater weight on the group ally, 37% of lysosomes are placed at 10 μm from the change. In lysosomal membranes, there were 35 lipids center of the nucleus in APOE4 cells, as compared to with VIP > 1 in the APOE4 vs APOE3. ANOVA analysis 21% in APOE3 astrocytes. Plausibly, the altered with a Tukey correction post-test of the intensity (peak localization may result in changes in the coupling of values) of these 35 metabolites revealed significant dif- such organelles with plasma-membrane channels. ferences due to the APOE phenotype with a p-value < In summary, in APOE3 astrocytes, activation of extra- 0.05, confirming, again, that expression of APOE4 alters cellular Ca2+ entry secondary to intracellular Ca2+ the lipidome of lysosomes. We then proceeded to identi- mobilization underlies the greater Ca2+ responses fication of the particular metabolites and calculation of induced by purinergic receptors in the absence of extra- their fold change (FC) in APOE4 vs APOE3 astrocytes. cellular lipoproteins. This coordination of signaling We could identify 19 lipids: 10 phosphatidylcholines (6 pathways does not take place in the presence of lipopro- increased, 4 decreased), 5 phosphatidylethanolamine (2 teins, suggesting, again, that lipids have the capacity to increased, 3 decreased), 2 lysophosphatidylcholine (2 de- change intracellular Ca2+ fluxes in APOE3 astrocytes. By creased), 1 lysophosphatidylethanolamine (decreased), contrast, APOE4 astrocytes present a higher content of and 1 carnitine (increased) (Fig. 8d). However, multi t- lysosomal Ca2+, but appear to have lost the capacity to test analysis corrected by a false discovery rate (FDR) of have Ca2+ fluxes regulated by lipids. these 19 FC values gave no statistically significant differ- ences (q-value < 0.05). This suggests that joint changes Lipidomics reveals distinct lipid composition in lysosomal in the contents of lipids with VIP > 1 rather than par- and whole-cell membranes from APOE3 and APOE4 ticular lipids account for the segregation of lipidomes astrocytes from APOE3 and APOE4 lysosomes. We posited that lysosomal dysregulation and the refrac- In whole-membranes, 41 metabolites had a VIP > 1, toriness to lipid-based modulation in APOE4 astrocytes comparing APOE4 vs APOE3 astrocytes. ANOVA ana- might be caused by altered lipid trafficking and homeo- lysis with a Tukey correction post-test of the intensities stasis due to APOE4 malfunction—as concluded in a of these 41 metabolites confirmed significant differences previous section. Since ApoE is a major cholesterol due to the APOE phenotype with a p-value < 0.01, in carrier in the brain, we studied cellular cholesterol distri- agreement with the previous PLS-DA analysis. Twety- bution with filipin staining. We found aberrant intracel- one lipids were identified according to their m/z, and lular distribution of cholesterol in APOE4 astrocytes, their FC in APOE4 vs APOE3 calculated (Fig. 8e). A which presented more cholesterol in intracellular multi T-test statistical analysis corrected with FDR, clumps, and less in plasma membrane, than APOE3 cells showed that 11 of these metabolites were significantly (Fig. 8a). This finding points to impaired cholesterol ef- different (q-value < 0.05): 3 lysophosphatidylcholine (1 flux in APOE4 cells. We also carried untargeted lipido- increased, 2 decreased), 5 phosphatidylcholines (3 in- mics because we reasoned that impaired lipid trafficking creased, 2 decreased), 1 phosphatidylserine (decreased) would leave its mark on astrocyte membranes, such that and 2 carnitines (increased). In short, a general trend is the profiling of membrane lipids would provide informa- that carnitines and phosphatidylcholines are more abun- tion about lipid dyshomeostasis in APOE4 astrocytes. To dant in APOE4 astrocytes, whereas APOE3 cells are determine whether APOE4-mediated changes were spe- richer in lysophospholipids. cific to lysosomes, we performed lipidomics in lysosomal Overall, this is the first demonstration that the and whole-membranes of APOE3 and APOE4 immortal- expression of APOE4 changes the lysosomal and ized astrocytes, since, according to a lipid map of the cellular lipidomes in astrocytes, supporting a link be- mammalian cell, organelles present distinct lipid compo- tween altered Ca2+ fluxes and lipid dyshomeostasis. sitions [66]. The multivariate analysis PLS-DA revealed It is worth noting that the different intracellular that the lipids of lysosomal membranes (Fig. 8b) and distribution of cholesterol in APOE3 and APOE4 whole-membranes (Fig. 8c) clustered independently in astrocytes is not due to differences in cholesterol APOE3 and APOE4 genotypes. The predictive accuracy contents between the two genotypes, as the VIP for of the analysis was robust, as the Q2 and R2Y scores cholesterol was consistently lower than 1 in the lipi- were 0.616 and 0.986 for lysosome lipidome, and 0.841 domes (data not shown). Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 17 of 23 Fig. 8 Different lipid profiles in lysosomal and whole-cell membranes from APOE4 vs APOE3 astrocytes. a Cholesterol accumulation visualized by Filipin III staining in immortalized APOE3 and APOE4 astrocytes (N = 2). Scale bar represents 100 μm. The white squares are amplified in the top right images. b, c PLS-DA analysis of lysosomal (b) and whole-cell membranes (c) from APOE3 and APOE4 immortalized astrocytes. Each dot is an individual sample. d, e Representation of changes of lipids with VIP > 1 identified in lysosome (d) and whole-cell (e) membranes as logarithm base 2 of fold changes of intensities in APOE4 vs APOE3. Positive values indicate increase and negative values decrease in APOE4 vs APOE3 cells. q < 0.05 (*), q < 0.01 (**), multi T-test analysis corrected by false discovery rate (FDR) Discussion mice, but not in females, whose astrocytes showed The study has two main general findings. First, in im- increased Ca2+ responses in APOE3 mice, matching mortalized mouse astrocytes expressing human APOE3 those in APOE4 mice. Below we discuss the possible and APOE4 we found that APOE4, in comparison to links between dysregulation of Ca2+ signaling, lipid sig- APOE3, increases receptor-induced Ca2+ responses due naling and lipid homeostasis in astrocytes, and the impli- to increased release of Ca2+ from acidic organelles, cations in neurodegenerative diseases in which APOE4 is which integrate lysosomes and related organelles. a risk factor in both men and women. Further, APOE4-expressing astrocytes present distinct Mechanistically, a key finding is that the pH of acidic lipid profiles and are refractory to Ca2+-signaling regula- organelles is similar in immortalized APOE4 and APOE3 tion by lipids (model in Fig. 9). Second, Ca2+ hyperactiv- astrocytes, despite greater expression and activity of the ity associated with the APOE4 allele was also found lysosomal H+ pump V-ATPase, supporting the idea that ex vivo in astrocytes from targeted replacement male the pH is maintained due to an antiparallel transport of Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 18 of 23 Fig. 9 Summary of the Ca2+ signaling alterations in APOE4 astrocytes. a) Ca2+ signaling pathway of APOE3 (a) and APOE4 (b) astrocytes in the presence (left) or absence (right) of extracellular lipids. The name of channels and receptors that increase cytosolic Ca2+ are in blue rectangles, whereas pathways decreasing cytosolic Ca2+ are in green rectangles. The size of the arrows indicates if the process is increased with respect to APOE3 astrocytes in the presence of extracellular lipids. Plasma-membrane lipids and lipoproteins are shown in different colors in APOE3 and APOE4 astrocytes to reflect their different lipid composition. Organelle and plasmatic membrane graphs were obtained from Smart Servier Medical Art (https://smart.servier.com/) H+ and Ca2+ that extrudes H+ and takes up Ca2+ via we considered this heterogeneous population of acidic CAX. Thus, it is plausible that CAX expression and/or vesicles as a whole, but we cannot rule out specific dif- activity are increased in APOE4 astrocytes, too. In conse- ferential traits in a particular kind of acidic vesicle. For quence, APOE4 acidic stores have greater Ca2+ content, example, a recent report described differences of pH be- and hence release more Ca2+ upon purinergic activation. tween the endolysosomal systems of APOE3 and APOE4 Coordination between intracellular signaling pathways astrocytes [70], suggesting that greater pH differences translates the increased Ca2+ mobilization from acidic than that detected in our study may exist between geno- vesicles to an increased Ca2+ released from ER, and a types, depending on the compartment analyzed. higher Ca2+ uptake into the mitochondria. In accordance It is worth stressing that APOE4 expression not only with most studies in the field [36–38, 51, 58–60, 67–69], alters Ca2+ uptake and mobilization from acidic stores, Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 19 of 23 but also promotes general dysregulation of these organ- endothelial cells [82], whereas phosphoinositides regu- elles in astrocytes, including higher perinuclear lysosome late TRPC3,6,7 channels in numerous cell types [83]. localization and changes in the overall lysosomal lipi- Recently, very-low-density lipoproteins have been shown dome. The intracellular localization of lysosomes, which to inhibit STIM in atrial myocytes [84]; this study and is determined by factors such as phospholipids and ours are the first to report lipoproteins regulating Ca2+ cholesterol, is linked to many of their functions and signaling. It is worth stressing that the all-or-nothing pathology [71, 72], although such studies have only been experimental design consisting in testing the effects of performed in cultured cells, plausibly due to the diffi- media with and without lipoproteins allowed us to culty of tracking the tortuous intracellular distribution obtain proof of concept that lipoproteins modulate Ca2+ of the endolysosomal system in whole tissues. Specific- excitability in astrocytes, but in physiological settings, ally, our study confirms the previous observation that Ca2+ signaling in astrocytes is, plausibly, modulated by aberrant accumulation of cholesterol conditions pro- subtle changes in brain lipid contents. motes perinuclear clustering of lysosomes [71]. On the Importantly, lipoprotein-mediated regulation of puri- other hand, our results showing aberrant intracellular nergic Ca2+ signaling is lost in APOE4 astrocytes. Several cholesterol accumulation, together with alterations of factors could explain this finding. First, the accumulation phospholipid composition in cellular membranes in of lysosomes around the nucleus may uncouple the lyso- APOE4 astrocytes, point to dysregulation of the forma- somal Ca2+ release and SOCE. Second, the entry of tion, internalization, and degradation of lipoproteins, extracellular Ca2+ triggered by low levels of Ca2+ inside processes in which lysosomes and related acidic vesicles acidic vesicles requires the TPC2 channels of acidic vesi- participate [73]. Moreover, alterations of cholesterol traf- cles, which are less expressed in APOE4 than in APOE3 ficking and intracellular cholesterol accumulation have astrocytes. Third, Ca2+ inside the acidic stores is higher been linked to impaired autophagy [74]. Accordingly, in APOE4 astrocytes compared to APOE3 cells; hence impaired autophagy [75] and reduced lysosome- SOCE mechanisms may not be triggered after dependent amyloid degradation [76] have been reported purinergic-induced lysosomal Ca2+ release. Fourth, in immortalized APOE4 astrocytes, and in astrocytes de- APOE4 astrocytes may be devoid of the right concentra- rived from APOE4 iPSC, respectively. Transcriptome tion of lipids to modulate lysosomal and Ca2+ entry analyses of whole brains of aged APOE4 targeted re- channels, as they have different lysosomal and cellular placement mice also revealed dysregulated expression of lipidomes compared to APOE3 cells. Specifically, the de- genes related to the endolysosomal system, although the creased contents of lysophosphatidylcholine species in contributions of the different cellular types were not APOE4 vs APOE3 cellular membranes may explain the studied [77]. Dysregulation of Ca2+ homeostasis might uncoupling of Ca2+ fluxes in APOE4 astrocytes, for these be a cause rather than a consequence of lysosomal dys- lipids activate SOCE [80], and the cation TRPV2 channel function, as there is evidence supporting Ca2+ release [85] in astrocytes. Moreover, the increase in phosphati- from acidic stores controlling endolysosomal trafficking dylcholines in APOE4 lysosomes is consistent with the and autophagy [67, 78]. Moreover, in astrocytes, it has observed potentiation of NAADP-mediated Ca2+ release, been reported that NAADP-induced Ca2+ release from since, as noted, phosphatidylcholines stimulate this path- lysosome-like organelles increases autophagic markers way [65]. Finally, the decrease in plasma membrane [79], and inhibits the fusion of the autophagosome with cholesterol in APOE4 astrocytes compared to APOE3 lysosomes, thus arresting the autophagic fluxes [69]. cells may underlie the reduced SOCE activation, because An unexpected discovery of this study is that lipopro- SOCE requires cholesterol in different cell types [81, 82, teins modulate the magnitude of ATP-induced cytosolic 86]. Cholesterol alterations have been also reported in Ca2+ responses in APOE3 astrocytes by changing the human APOE4-iPSC derived astrocytes [76], and in interplay of Ca2+ signaling and fluxes among organelles brains from 12-month-old APOE4 mice [87]. Interest- and the plasma membrane. The observation that the ingly, cholesterol synthesis is decreased in astrocytes down-regulation of purinergic-induced Ca2+ responses upon aging [88], which might exacerbate APOE4-elicited in the presence of lipids is quick and reversible rules out lipid dyshomeostasis and Ca2+ signaling. the implication of gene expression and down-regulation Remarkably, the differences in Ca2+ signaling between of purinergic receptors. Rather, the phenomenon sup- APOE4 and APOE3 astrocytes were also detected ex vivo ports the emerging notion of lipid-mediated control of in 9–12-week-old male but not in female mice, suggesting Ca2+ channels. Precedents are the activation by lysopho- that the APOE allele affects astrocyte excitability in a sex- sphatidylcholine of astrocytic extracellular Ca2+ entry dependent manner. The present study thus adds to the in- [80], and the regulation by lipids of some of the channels creasing evidence of complex interactions between APOE responsible for SOCE. Specifically, cholesterol regulates genotype and sex, as shown, for example, in lipid-related TRPC1 in neutrophils [81] and STIM in pulmonary metabolic variations [89], cerebrovascular pathology [90], Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 20 of 23 and tau levels in the cerebrospinal fluid [91]. The follow- clearance caused by ApoE, leading to increased accumu- ing scenarios might explain the sex bias in astrocyte excit- lation, and hence aggregation, of Aβ in the brain [70]. In ability in different APOE alleles. First, Ca2+ signals in addition, ApoE4 has recently been reported to potentiate female astrocytes might be by default higher, regardless of Tau-mediated neurodegeneration independently of Aβ the APOE allele. Along these lines, a recent study showed [24]. However, it is increasingly more recognized that al- greater estradiol-induced Ca2+ signals in astrocytes from terations in lysosomal functions [98], lipid homeostasis female mice than from male astrocytes [92]. Second, the [99], and Ca2+ signaling play a role in LOAD, too. The down-regulation of Ca2+-based excitability by lipids ob- Ca2+ hypothesis in LOAD contends that aberrant Ca2+ served in immortalized APOE3 astrocytes might not occur responses in neurons associated with Aβ, Tau, and the in females. Note that ex-vivo Ca2+ responses in female glutamatergic system underlie cognitive impairment mice resemble Ca2+ responses in immortalized astrocytes [100]. In addition, Ca2+ signaling is profoundly dysregu- in the absence of lipoproteins. Emerging evidence indeed lated in astrocytes in animal models of AD [39, 101], points to a distinct impact of APOE genotype on brain mainly due to aberrant activation of purinergic receptors lipid metabolism [76, 87] with a sex bias [93]. In the latter [56]. Our results support the notion that APOE4 may study, lipid clustering by principal component analysis of exacerbate the dysregulation of Ca2+ signaling in male cortical lipidomes unravels sample segregation by sex in astrocytes in LOAD owing to lysosomal dysfunction APOE3 but not in APOE4 16 month-old mice [93]. That caused by lipid dyshomeostasis. Dysregulation of astro- is, sex differences were observed in APOE3 but not in cyte excitability in LOAD may, in turn, contribute to APOE4 mice, as in our study. Specifically, there was a neural-circuit hyperactivity independently of Aβ and trend for greater concentrations of phosphatidylcholines Tau pathologies [102], supporting astrocytic-lysosome and lysophosphatidylcholines in APOE3 male than in targeted therapies in LOAD, at least in male patients. APOE3 female mice [93]. Taking together these and our An outstanding question in APOE4-targeted therapeu- results, it is plausible that the interplay of sex, APOE geno- tics is whether the mutated domain renders ApoE4 type and age differently shapes the composition of lipid toxic, or less efficient, than ApoE3 and ApoE2. Alterna- milieus in male and females through life, resulting in dis- tively, since ApoE is lower in cerebrospinal fluid (CSF) tinct astrocytic Ca2+ responses. of APOE4 individuals [62], and in plasma, CSF, and Whatever the case, our study supports that immortal- brain tissue of APOE3 and APOE4 as compared with ized astrocytes from human APOE replacement mice APOE2 targeted replacement mice [103], is it the prob- may be a model to understand APOE4 pathology in lem that ApoE4 is less abundant? The answers to these males. The fact that Ca2+-based astrocyte excitability questions are important because they will determine controls neural functions [94] lends credence to the hy- whether therapeutic strategies should be aimed to in- pothesis that APOE4-elicited dysregulation of Ca2+ crease or decrease ApoE production, or to replace fluxes in astrocytes contributes to the impairment of APOE4 with APOE3 or APOE2. It seems unlikely that brain activity and metabolism in the healthy brain, as re- the effects of APOE4 on lysosomal-related Ca2+ fluxes peatedly reported in humans [4–6, 8], although, of note, are due to decreased expression of APOE4, because no sex-based stratification existed in these studies. It is APOE4 over-expression did not revert Ca2+ responses to worth noting that the differences in APOE genotype the levels observed in APOE3 cells. It is also unlikely were observed in 10-week old male mice, pointing to that the APOE4 phenotypes are caused by direct inter- early detrimental actions of APOE4 in brain, not surpris- action between ApoE4 and V-ATPase and/or NAADP ingly so, for APOE alleles are acquired at conception. receptors in the acidic organelles, as shown for APOD, This is important because, although metabolic alter- another brain apolipoprotein that participates in vesicle- ations and distinct patterns of brain activity have been mediated astrocyte-to-neuron communication [104] and reported in young humans harboring APOE4 [95, 96], prevents lysosomal membrane permeabilization [51], and olfactory-memory impairment exists in 6 month-old given that our immunocytochemistry showed scarce APOE4 mice [13, 89], most studies with human APOE4 ApoE in astrocytic lysosomes. Rather, the data support gene targeted replacement of murine APOE have been malfunction of ApoE4, which could be rescued by over- conducted with aged mice, and occasionally, middle- expression of APOE3, as shown here, and perhaps aged (over 10 months) rodents [97]. APOE2 [50]. Whether drugs designed to increase ApoE4 We emphasize that the APOE4 phenotype observed in lipidation in order to enhance the capacity of the lipo- the aforementioned studies, as well as in our study, oc- protein to carry cholesterol [105] would restore intracel- curs in the absence of LOAD pathology, although it may lular lipid dyshomeostasis in astrocytes is an open render brains more vulnerable to age-dependent question, since the mechanisms whereby the APOE allele ailments. Thus, a wealth of data suggests that APOE4 ex- modifies lipid contents and distribution in astrocyte acerbates the impairment of Aβ processing and membranes remain to be explored. Alternatively, Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 21 of 23 because APOE4 exacerbates ApoE-mediated Aβ aggrega- Authors’ contributions tion [70] and Tau pathology [24], ApoE removal is being RL-A performed calcium imaging experiments in immortalized astrocytes, participated in the rest of in vitro experiments, analyzed data and pursued by immunotherapy [106]. However, considering participated in the writing of the manuscript. CG-A and GP performed exper- the pleiotropic functions of APOE as a lipid carrier be- iments ex vivo. RP-M and MDG performed pH measurements. MM-V and MV tween cells, and plausibly inside cells, chronic removal isolated lysosomes. Lipidomics were done by TG-B, JLG and JV. AG partici- pated in the immunocytochemistry experiments and discussions about lyso- of ApoE may have secondary effects. All in all, develop- somal dysfunction. EH and EK produced and provided the GFP-APOE vectors. ment of APOE4-specific therapeutics is in order. To this EG contributed to the conception of the project, the discussion of results end, better understanding of the basic biology of APOE, and the writing and revising of the article. RM conceived, designed and su- pervised the project, performed some calcium imaging experiments, ana- and of the particulars of APOE4 action and sex bias in- lyzed data and was responsible for data interpretation and writing and dependently of Aβ and Tau, are necessary. Overall, our revising the manuscript. All authors read and approved the final manuscript. findings that APOE4 per se disrupts the lipid-based regulation of Ca+ 2 fluxes in astrocytes due to lysosomal Funding dysregulation supports the notion that astrocyte This research was mainly funded by grants TV3–20141430, TV3–20141432 and TV3–20141431 from La Marató de Televisió de Catalunya (TV3) to EG, AG dysfunction contributes to APOE4 pathology in and JV respectively, and grants 2107 SGR1780 from AGAUR (Generalitat de neurodegeneration. Catalunya) to RM, 2017 SGR547 from AGAUR (Generalitat de Catalunya) to EG, BFU2016–75107-P from Ministerio de Economia, Industria y Competividad (Spanish Government) to GP, BFU2015–68149-R from Ministerio de Ciencia e Innovación (Spanish Government) and co-financed Conclusions by European Regional Development Fund to MDG and PI18/01557 from Taking together ex vivo and in vitro data, we conclude Instituto de Salud Carlos III (ISCiii, Spanish Government) co-financed by that APOE4 malfunction augments Ca2+-based excitabil- FEDER funds from European Union to AG. CG-A was awarded a PhD fellow- ship BES-2017-080303 from Ministerio de Economía, Industria y Competivi- ity in astrocytes due to dysregulation of calcium fluxes dad (Spanish Government). in and out the lysosome, associated with lipid dysho- meostasis, and that the phenomenon might be male- Availability of data and materials associated. One implication of the study is that it Not applicable. supports the use of therapies aimed to restoring lyso- somal dysfunction in astrocytes, including targeted over- Ethics approval expression of APOE3 or APOE2, in order to halt the All experimental procedures were according to the animal research regulations (RD53/2013 and 2010/63/UE) from Spain and European Union, accelerated progression of LOAD in male APOE4 car- and with the approval of the Committees of Animal Research from the riers. Since Ca2+ signaling is central to the regulation of Institutional Animal Ethics Committee of CSIC. neural circuits by astrocytes, another implication of the study is that the therapeutic correction of astrocyte ex- Consent for publication citability might reverse the neural-circuit hyperactivity Not applicable. observed in APOE4-harboring humans and mice in the absence of Aβ and Tau pathologies. Finally, clarifying Competing interests the sex bias in the efficacy of APOE4 targeted therapeu- The authors declare that they have no competing interests. tics is a must, for the mechanisms underlying APOE4 Author details pathology might differ in males and females. 1 Unitat de Bioquímica de Medicina, Departament de Bioquímica i Biologia Molecular, and, Institut de Neurociències (INc), Facultat de Medicina, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Barcelona, Abbreviations Catalonia, Spain. 2Cajal Institute, Consejo Superior de Investigaciones APOE: Apolipoprotein E; ACSF: Artificial cerebrospinal fluid medium; Científicas (CSIC), 28002 Madrid, Spain. 3Departamento de Biología Celular, AUC: Area under the curve; CAX: Ca2+/H+ exchanger; CSF: Cerebrospinal Genética y Fisiología, Facultad de Ciencias, Instituto de Investigación fluid; ER: Endoplasmic reticulum; FBS: Fetal bovine serum; FC: Fold change; Biomedica de Málaga (IBIMA), Universidad de Málaga, 29071 Málaga, Spain. GECI: Genetically encoded Ca2+ indicators; GPN: Glycyl-L-phenylalanine 2- 4 Centro de Investigación Biomédica en Red sobre Enfermedades naphthylamide; h: Hours; IP3: Inositol 1,4,5-triphosphate; KH: Krebs medium; Neurodegenerativas (CIBERNED), 28031 Madrid, Spain. 5Departamento de Lipo: Lipofectamine; LOAD: Late-onset Alzheimer’s disease; Bioquímica y Biología Molecular, Facultad de Farmacia, Universidad de LPC: Lysophosphatidylcholine; LPE: Lysophosphatidylethanolamine; Sevilla, Instituto de Biomedicina de Sevilla (IBiS)-Hospital Universitario Virgen LTP: Long-term potentiation; min: Minutes; NAADP: Nicotinic acid adenine del Rocío/CSIC/Universidad de Sevilla, 41012 Sevilla, Spain. 6Departamento dinucleotide phosphate; NMDG: N-methyl-D-glucamine; de Química, Facultad de Ciencias Experimentales, Campus de El Carmen, PC: Phosphatidylcholine; PE: Phosphatidylethanolamine; Centro de Investigación en Recursos Naturales, Salud y Medio Ambiente PS: Phosphatidylserine; ROI: Region of interest; s: Seconds; Sc: Scramble; (RENSMA), Universidad de Huelva, 21007 Huelva, Spain. 7Departamento de SOCE: Store-operated Ca2+ entry; SR101: Sulforhodamine; Tpc1: Two-pore Bioquímica y Biología Molecular y Fisiología, Instituto de Biología y Genética channels 1; Tpc2: Two-pore channels 2; Trpml: Transient receptor potential Molecular, Universidad de Valladolid-CSIC, 43007 Valladolid, Spain. 8 mucolipin; TTX: Tetrodotoxin; VIP: Variable importance in projection; Alzheimer’s Disease Research Laboratory, MassGeneral Institute for Vs: Versus Neurodegenerative Disease, Massachusetts General Hospital, Harvard Medical School, Charlestown, MA 02129, USA. 9Present Address: Institute of Neuropathology, University Hospital of Zurich, 8091 Zurich, Switzerland. 10 Acknowledgements Neurodegenerative Diseases Research Group, Vall d’Hebron Research We thank Dr. David M. Holtzman for providing us with APOE3 and APOE4 Institute (VHIR), 08035 Barcelona, Spain. 11ICREA, Passeig Lluís Companys 23, immortalized astrocytes. 08010 Barcelona, Catalonia, Spain. Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 22 of 23 Received: 7 August 2019 Accepted: 25 May 2020 26. Kastanenka KV, Moreno-Bote R, De Pittà M, et al. A roadmap to integrate astrocytes into Systems Neuroscience. Glia. 2019;68:23632. 27. Mauch DH, Nägier K, Schumacher S, et al. CNS synaptogenesis promoted by glia-derived cholesterol. Science. 2001;294:1354–7. References 28. Petrov AM, Kasimov MR, Zefirov AL. Brain cholesterol metabolism and its 1. Marais AD. Apolipoprotein E in lipoprotein metabolism, health and defects: linkage to neurodegenerative diseases and synaptic dysfunction. cardiovascular disease. Pathology. 2019;51:165–76. Acta Nat. 2016;8:58–73. 2. Zhao N, Liu C-C, Qiao W, et al. Apolipoprotein E, receptors, and modulation 29. Perea G, Araque A. Astrocytes potentiate transmitter release at single of Alzheimer’s disease. Biol Psychiatry. 2018;83:347–57. hippocampal synapses. Science. 2007;317:1083–6. 3. Liu CC, Liu CC, Kanekiyo T, et al. Apolipoprotein E and Alzheimer disease: 30. Henneberger C, Papouin T, Oliet SHR, et al. Long-term potentiation risk, mechanisms and therapy. Nat Rev Neurol. 2013;9:106–18. depends on release of d-serine from astrocytes. Nature. 2010;463:232–6. 4. Caselli RJ, Reiman EM, Osborne D, et al. Longitudinal changes in cognition 31. Lee HS, Ghetti A, Pinto-Duarte A, et al. Astrocytes contribute to gamma and behavior in asymptomatic carriers of the APOE e4 allele. Neurology. oscillations and recognition memory. Proc Natl Acad Sci. 2014;111:E3343–52. 2004;62:1990–5. 32. Marpegan L, Swanstrom AE, Chung K, et al. Circadian regulation of ATP 5. Stevens BW, DiBattista AM, William Rebeck G, et al. A gene-brain-cognition release in astrocytes. J Neurosci. 2011;31:8342–50. pathway for the effect of an Alzheimer′s risk gene on working memory in 33. Araque A, Carmignoto G, Haydon PG, et al. Gliotransmitters travel in time young adults. Neuropsychologia. 2014;61:143–9. and space. Neuron. 2014;81:728–39. 6. Kunz L, Schröder TN, Lee H, et al. Reduced grid-cell-like representations in 34. Zorec R, Araque A, Carmignoto G, et al. Astroglial excitability and adults at genetic risk for Alzheimer’s disease. Science. 2015;350:430–3. Gliotransmission: an appraisal of Ca 2+ as a Signalling route. ASN Neuro. 2012;4: 7. Ji Y, Gong Y, Gan W, et al. Apolipoprotein E isoform-specific regulation of AN20110061. dendritic spine morphology in apolipoprotein E transgenic mice and 35. Rusakov DA. Disentangling calcium-driven astrocyte physiology. Nat Rev Alzheimer’s disease patients. Neuroscience. 2003;122:305–15. Neurosci. 2015;16:226–33. 8. Reiman EM, Chen K, Alexander GE, et al. From the cover: correlations 36. Raffaello A, Mammucari C, Gherardi G, et al. Calcium at the Center of Cell between apolipoprotein E 4 gene dose and brain-imaging measurements Signaling: interplay between endoplasmic reticulum, mitochondria, and of regional hypometabolism. Proc Natl Acad Sci. 2005;102:8299–302. lysosomes. Trends Biochem Sci. 2016;41:1035–49. 9. Rodriguez GA, Burns MP, Weeber EJ, et al. Young APOE4 targeted replacement 37. Barceló-Torns M, Lewis AM, Gubern A, et al. NAADP mediates ATP-induced mice exhibit poor spatial learning and memory, with reduced dendritic spine Ca2+ signals in astrocytes. FEBS Lett. 2011;585:2300–6. density in the medial entorhinal cortex. Learn Mem. 2013;20:256–66. 38. Eraso-Pichot A, Larramona-Arcas R, Vicario-Orri E, et al. CREB decreases 10. Klein RC, Acheson SK, Mace BE, et al. Altered neurotransmission in the astrocytic excitability by modifying subcellular calcium fluxes via the sigma- lateral amygdala in aged human apoE4 targeted replacement mice. 1 receptor. Cell Mol Life Sci. 2017;74:937–50. Neurobiol Aging. 2014;35:2046–52. 39. Kuchibhotla KV, Lattarulo CR, Hyman BT, et al. Synchronous hyperactivity and 11. Wang C, Wilson WA, Moore SD, et al. Human apoE4-targeted replacement intercellular calcium waves in astrocytes in Alzheimer mice. Science. 2009;323: mice display synaptic deficits in the absence of neuropathology. Neurobiol 1211–5. Dis. 2005;18:390–8. 40. Tambini MD, Pera M, Kanter E, et al. ApoE 4 upregulates the activity of 12. Najm R, Jones EA, Huang Y. Apolipoprotein E4, inhibitory network mitochondria- associated ER membranes. EMBO Rep. 2015;17:1–10. dysfunction, and Alzheimer’s disease. Mol Neurodegener. 2019;14:24. 41. Sunshine H, Iruela-Arispe ML. Membrane lipids and cell signaling. Curr Opin 13. Peng KY, Mathews PM, Levy E, et al. Apolipoprotein E4 causes early Lipidol. 2017;28:408–13. olfactory network abnormalities and short-term olfactory memory 42. Knouff C, Hinsdale ME, Mezdour H, et al. Apo E structure determines VLDL impairments. Neuroscience. 2017;343:364–71. clearance and atherosclerosis risk in mice. J Clin Invest. 1999;103:1579–86. 14. Bertram L, McQueen MB, Mullin K, et al. Systematic meta-analyses of 43. Sullivan PM, Mezdour H, Aratani Y, et al. Targeted replacement of the Alzheimer disease genetic association studies: the AlzGene database. Nat mouse Apolipoprotein E gene with the common human APOE3 allele Genet. 2007;39:17–23. enhances diet-induced hypercholesterolemia and atherosclerosis. J Biol 15. Yamazaki Y, Painter MM, Bu G, et al. Apolipoprotein E as a therapeutic Chem. 1997;272:17972–80. target in Alzheimer’s disease: a review of basic research and clinical 44. Morikawa M, Fryer JD, Sullivan PM, et al. Production and characterization of evidence. CNS Drugs. 2016;30:773–89. astrocyte-derived human apolipoprotein E isoforms from immortalized 16. Neu SC, Pa J, Kukull W, et al. Apolipoprotein E genotype and sex risk factors astrocytes and their interactions with amyloid-β. Neurobiol Dis. 2005;19:66–76. for Alzheimer disease: a meta-analysis. JAMA Neurol. 2017;74:1178–89. 45. Ting JT, Daigle TL, Chen Q, et al. Acute brain slice methods for adult and 17. Riedel BC, Thompson PM, Brinton RD. Age, APOE and sex: triad of risk of aging animals: application of targeted patch clamp analysis and Alzheimer’s disease. J Steroid Biochem Mol Biol. 2016;160:134–47. optogenetics; 2014. p. 221–42. 18. Rasmussen KL, Tybjærg-Hansen A, Nordestgaard BG, et al. Absolute 10-year 46. Kafitz KW, Meier SD, Stephan J, et al. Developmental profile and properties risk of dementia by age, sex and APOE genotype: a population-based of sulforhodamine 101—labeled glial cells in acute brain slices of rat cohort study. CMAJ. 2018;190:E1033–41. hippocampus. J Neurosci Methods. 2008;169:84–92. 19. Agosta F, Vossel KA, Miller BL, et al. Apolipoprotein E ε4 is associated with 47. Martin R, Bajo-Graneras R, Moratalla R, et al. Circuit-specific signaling in disease-specific effects on brain atrophy in Alzheimer’s disease and astrocyte-neuron networks in basal ganglia pathways. Science. 2015;349:730–4. frontotemporal dementia. Proc Natl Acad Sci. 2009;106:2018–22. 48. Nimmerjahn A, Kirchhoff F, Kerr JND, et al. Sulforhodamine 101 as a specific 20. Baum L, Lam LCW, Kwok T, et al. Apolipoprotein E epsilon4 allele is associated marker of astroglia in the neocortex in vivo. Nat Methods. with vascular dementia. Dement Geriatr Cogn Disord. 2006;22:301–5. 2004;1:31–7. 21. Cao J, El Gaamouch F, Meabon JS, et al. ApoE4-associated phospholipid 49. Suzuki J, Kanemaru K, Ishii K, et al. Imaging intraorganellar Ca2+ at dysregulation contributes to development of tau hyper-phosphorylation subcellular resolution using CEPIA. Nat Commun. 2014;5:4153. after traumatic brain injury. Sci Rep. 2017;7:1–12. 50. Hudry E, Dashkoff J, Roe AD, et al. Gene transfer of human Apoe isoforms 22. Xu Q, Bernardo A, Walker D, et al. Profile and regulation of apolipoprotein E results in differential modulation of amyloid deposition and neurotoxicity in (ApoE) expression in the CNS in mice with targeting of green fluorescent mouse brain. Sci Transl Med. 2013;5:212ra161. protein gene to the ApoE locus. J Neurosci. 2006;26:4985–94. 51. Pascua-Maestro R, Diez-Hermano S, Lillo C, et al. Protecting cells by 23. Parhizkar S, Arzberger T, Brendel M, et al. Loss of TREM2 function increases protecting their vulnerable lysosomes: identification of a new mechanism amyloid seeding but reduces plaque-associated ApoE. Nat Neurosci. 2019;22: for preserving lysosomal functional integrity upon oxidative stress. PLoS 191–204. Genet. 2017;13:1–33. 24. Shi Y, Yamada K, Liddelow SA, et al. ApoE4 markedly exacerbates tau-mediated 52. Storrie B, Amadden E. Isolation of subcellular organelles; 1990. p. 203–25. neurodegeneration in a mouse model of tauopathy. Nature. 2017;549:523–7. 53. González-Domínguez R, García-Barrera T, Gómez-Ariza JL. Metabolomic 25. Holtzman DM, Herz J, Bu G. Apolipoprotein E and apolipoprotein E study of lipids in serum for biomarker discovery in Alzheimer’s disease using receptors: Normal biology and roles in Alzheimer disease. Cold Spring Harb direct infusion mass spectrometry. J Pharm Biomed Anal. Perspect Med. 2012;2:a006312. 2014;98:321–6. Larramona-Arcas et al. Molecular Neurodegeneration (2020) 15:35 Page 23 of 23 54. Vandesompele J, De Preter K, Pattyn F, et al. Accurate normalization of real- 83. Itsuki K, Imai Y, Okamura Y, et al. Voltage-sensing phosphatase reveals time quantitative RT-PCR data by geometric averaging of multiple internal temporal regulation of TRPC3/C6/C7 channels by membrane control genes. Genome Biol. 2002;3:RESEARCH0034. phosphoinositides. Channels. 2012;6:206–9. 55. Khakh BS, Sofroniew MV. Diversity of astrocyte functions and phenotypes in 84. Shiou Y-L, Lin H-T, Ke L-Y, et al. Very low-density lipoproteins of metabolic neural circuits. Nat Neurosci. 2015;18:942–52. syndrome modulates STIM1, suppresses store-operated calcium entry, and 56. Delekate A, Füchtemeier M, Schumacher T, et al. Metabotropic P2Y1 deranges Myofilament proteins in atrial Myocytes. J Clin Med. 2019;8:881. receptor signalling mediates astrocytic hyperactivity in vivo in an 85. Shibasaki K, Ishizaki Y, Mandadi S. Astrocytes express functional TRPV2 ion Alzheimer’s disease mouse model. Nat Commun. 2014;5:5422. channels. Biochem Biophys Res Commun. 2013;441:327–32. 57. Nakamura T, Watanabe A, Fujino T, et al. Apolipoprotein E4 (1–272) 86. Sun Y, Sukumaran P, Varma A, et al. Cholesterol-induced activation of fragment is associated with mitochondrial proteins and affects TRPM7 regulates cell proliferation, migration, and viability of human mitochondrial function in neuronal cells. Mol Neurodegener. 2009;4:35. prostate cells. Biochim Biophys Acta. 2014;1843:1839–50. 58. Patel S, Cai X. Evolution of acidic Ca2+ stores and their resident Ca2+- 87. Peng KY, Pérez-González R, Alldred MJ, et al. Apolipoprotein E4 genotype permeable channels. Cell Calcium. 2015;57:222–30. compromises brain exosome production. Brain. 2019;142:163–75. 59. Patel S, Ramakrishnan L, Rahman T, et al. The endo-lysosomal system as an 88. Boisvert MM, Erikson GA, Shokhirev MN, et al. The aging astrocyte Transcriptome NAADP-sensitive acidic Ca(2+) store: role for the two-pore channels. Cell from multiple regions of the mouse brain. Cell Rep. 2018;22:269–85. Calcium. 2011;50:157–67. 89. Jones NS, Watson KQ, Rebeck GW. Metabolic disturbances of a high-fat diet 60. Melchionda M, Pittman JK, Mayor R, et al. Ca2+/H+exchange by acidic are dependent on apoe genotype and sex. eNeuro. 2019;6:5 ENEURO.0267– organelles regulates cell migration in vivo. J Cell Biol. 2016;212:803–13. 19.2019. 61. Zhang Y, Sloan SA, Clarke LE, et al. Purification and characterization of 90. Finch CE, Shams S. Apolipoprotein E and sex Bias in cerebrovascular aging progenitor and mature human astrocytes reveals transcriptional and of men and mice. Trends Neurosci. 2016;39:625–37. functional differences with mouse. Neuron. 2016;89:37–53. 91. Hohman TJ, Dumitrescu L, Barnes LL, et al. Sex-specific Association of 62. Cruchaga C, Kauwe JSK, Nowotny P, et al. Cerebrospinal fluid APOE levels: Apolipoprotein E with Cerebrospinal Fluid Levels of tau. JAMA Neurol. 2018; an endophenotype for genetic studies for Alzheimer’s disease. Hum Mol 75:989–98. Genet. 2012;21:4558–71. 92. Kuo J, Hamid N, Bondar G, et al. Sex differences in hypothalamic astrocyte 63. Cheng X-T, Xie Y-X, Zhou B, et al. Characterization of LAMP1-labeled response to estradiol stimulation. Biol Sex Differ. 2010;1:7. nondegradative lysosomal and endocytic compartments in neurons. J Cell 93. Shang Y, Mishra A, Wang T, et al. Evidence in support of chromosomal sex Biol. 2018;217:3127–39. influencing plasma based metabolome vs APOE genotype influencing brain 64. Lee AG. Lipid interactions with ion channels. Futur Lipidol. 2006;1:103–13. metabolome profile in humanized APOE male and female mice. PLoS One. 65. Churamani D, Dickinson GD, Ziegler M, et al. Time sensing by NAADP 2020;15:e0225392. receptors. Biochem J. 2006;397:313–20. 94. Shigetomi E, Patel S, Khakh BS. Probing the complexities of astrocyte 66. van Meer G, de Kroon AIPM. Lipid map of the mammalian cell. J Cell Sci. calcium signaling. Trends Cell Biol. 2016;26:300–12. 2011;124:5–8. 95. Perkins M, Wolf AB, Chavira B, et al. Altered energy metabolism pathways in 67. Ruas M, Rietdorf K, Arredouani A, et al. Purified TPC isoforms form NAADP the posterior cingulate in Young adult Apolipoprotein e 4 carriers. J receptors with distinct roles for Ca2+ signaling and Endolysosomal Alzheimers Dis. 2016;53:95–106. trafficking. Curr Biol. 2010;20:703–9. 96. Filippini N, MacIntosh BJ, Hough MG, et al. Distinct patterns of brain activity in 68. Galione A. A primer of NAADP-mediated Ca2+ signalling: from sea urchin young carriers of the APOE- 4 allele. Proc Natl Acad Sci. 2009;106:7209–14. eggs to mammalian cells. Cell Calcium. 2015;58:27–47. 97. Fernandez CG, Hamby ME, McReynolds ML, et al. The role of APOE4 in 69. Lu Y, Hao B, Graeff R, et al. NAADP/TPC2/Ca 2+ signaling inhibits disrupting the homeostatic functions of Astrocytes and Microglia in aging autophagy. Commun Integr Biol. 2013;6:e27595. and Alzheimer’s Disease. Front Aging Neurosci. 2019;11:14. Epub ahead of 70. Prasad H, Rao R. Amyloid clearance defect in ApoE4 astrocytes is reversed by print 11 February 2019. https://doi.org/10.3389/fnagi.2019.00014. epigenetic correction of endosomal pH. Proc Natl Acad Sci. 2018;115:E6640–9. 98. Zare-Shahabadi A, Masliah E, Johnson GVW, et al. Autophagy in Alzheimer’s 71. Pu J, Guardia CM, Keren-Kaplan T, et al. Mechanisms and functions of disease. Rev Neurosci. 2015;26:385–95. lysosome positioning. J Cell Sci. 2016;129:4329–39. 99. Anand S, Barnes JM, Young SA, et al. Discovery and confirmation of diagnostic serum lipid biomarkers for Alzheimer’s disease using direct 72. Ballabio A, Bonifacino JS. Lysosomes as dynamic regulators of cell and infusion mass spectrometry. J Alzheimers Dis. 2017;59:277–90. organismal homeostasis. Nat Rev Mol Cell Biol. 2020;21:101–18. 100. Magi S, Castaldo P, Macrì ML, et al. Intracellular calcium Dysregulation: 73. Thelen AM, Zoncu R. Emerging roles for the lysosome in lipid metabolism. implications for Alzheimer’s disease. Biomed Res Int. 2016;2016:1–14. Trends Cell Biol. 2017;27:833–50. 101. Kelly P, Hudry E, Hou SS, et al. In Vivo two photon imaging of Astrocytic structure 74. Sathyanarayan A, Mashek MT, Mashek DG. ATGL promotes autophagy/ and function in Alzheimer’s disease. Front Aging Neurosci. 2018;10:219. Epub Lipophagy via SIRT1 to control hepatic lipid droplet catabolism. Cell Rep. ahead of print 19 July 2018. https://doi.org/10.3389/fnagi.2018.00219. 2017;19:1–9. 102. Nuriel T, Angulo SL, Khan U, et al. Neuronal hyperactivity due to loss of 75. Simonovitch S, Schmukler E, Bespalko A, et al. Impaired autophagy in inhibitory tone in APOE4 mice lacking Alzheimer’s disease-like pathology. APOE4 astrocytes. J Alzheimers Dis. 2016;51:915–27. Nat Commun. 2017;8:1464. 76. Lin YT, Seo J, Gao F, et al. APOE4 causes widespread molecular and cellular 103. Riddell DR, Zhou H, Atchison K, et al. Impact of Apolipoprotein E (ApoE) alterations associated with Alzheimer’s disease phenotypes in human ipsc- polymorphism on brain ApoE levels. J Neurosci. 2008;28:11445–53. derived brain cell types. Neuron. 2018;98:1141–1154.e7. 104. Pascua-Maestro R, González E, Lillo C, et al. Extracellular vesicles secreted by 77. Nuriel T, Peng KY, Ashok A, et al. The endosomal-lysosomal pathway is astroglial cells transport apolipoprotein D to neurons and mediate neuronal dysregulated by APOE4 expression in vivo. Front Neurosci. 2017;11:1–12. survival upon oxidative stress. Front Cell Neurosci. 2019;12:526. Epub ahead 78. Bootman MD, Chehab T, Bultynck G, et al. The regulation of autophagy by of print 10 January 2019. https://doi.org/10.3389/fncel.2018.00526. calcium signals: do we have a consensus? Cell Calcium. 2018;70:32–46. 105. Safieh M, Korczyn AD, Michaelson DM. ApoE4: an emerging therapeutic 79. Pereira GJS, Hirata H, Fimia GM, et al. Nicotinic acid adenine dinucleotide target for Alzheimer’s disease. BMC Med. 2019;17:64. phosphate (NAADP) regulates autophagy in cultured astrocytes. J Biol 106. Liao F, Hori Y, Hudry E, et al. Anti-ApoE antibody given after plaque onset Chem. 2011;286:27875–81. decreases a accumulation and improves brain function in a mouse model 80. Singaravelu K, Lohr C, Deitmer JW. Regulation of store-operated calcium of a amyloidosis. J Neurosci. 2014;34:7281–92. entry by calcium-independent phospholipase A2 in rat cerebellar astrocytes. J Neurosci. 2006;26:9579–92. 81. Kannan KB, Barlos D, Hauser CJ. Free cholesterol alters lipid raft structure and Publisher’s Note function regulating neutrophil Ca 2+ entry and respiratory burst: correlations Springer Nature remains neutral with regard to jurisdictional claims in with Calcium Channel raft trafficking. J Immunol. 2007;178:5253–61. published maps and institutional affiliations. 82. Zhang B, Naik JS, Jernigan NL, et al. Reduced membrane cholesterol limits pulmonary endothelial Ca 2+ entry after chronic hypoxia. Am J Physiol Circ Physiol. 2017;312:H1176–84.