Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 https://doi.org/10.1186/s13195-022-01108-2 RESEARCH Open Access Apolipoprotein E imbalance in the cerebrospinal fluid of Alzheimer’s disease patients Matthew Paul Lennol1,2, Irene Sánchez‑Domínguez3,4, Inmaculada Cuchillo‑Ibañez1,2,5, Elena Camporesi6, Gunnar Brinkmalm6, Daniel Alcolea2,7, Juan Fortea2,7,8, Alberto Lleó2,7, Guadalupe Soria4,9, Fernando Aguado3,4, Henrik Zetterberg6,10,11,12,13, Kaj Blennow6,10 and Javier Sáez‑Valero1,2,5* Abstract Objective: The purpose of this study was to examine the levels of cerebrospinal fluid (CSF) apolipoprotein E (apoE) species in Alzheimer’s disease (AD) patients. Methods: We analyzed two CSF cohorts of AD and control individuals expressing different APOE genotypes. Moreo‑ ver, CSF samples from the TgF344-AD rat model were included. Samples were run in native- and SDS-PAGE under reducing or non-reducing conditions (with or without β-mercaptoethanol). Immunoprecipitation combined with mass spectrometry or western blotting analyses served to assess the identity of apoE complexes. Results: In TgF344-AD rats expressing a unique apoE variant resembling human apoE4, a ~35-kDa apoE monomer was identified, increasing at 16.5 months compared with wild-types. In humans, apoE isoforms form disulfide-linked dimers in CSF, except apoE4, which lacks a cysteine residue. Thus, controls showed a decrease in the apoE dimer/ monomer quotient in the APOE ε3/ε4 group compared with ε3/ε3 by native electrophoresis. A major contribution of dimers was found in APOE ε3/ε4 AD cases, and, unexpectedly, dimers were also found in ε4/ε4 AD cases. Under reduc‑ ing conditions, two apoE monomeric glycoforms at 36 kDa and at 34 kDa were found in all human samples. In AD patients, the amount of the 34-kDa species increased, while the 36-kDa/34-kDa quotient was lower compared with controls. Interestingly, under reducing conditions, a ~100-kDa apoE complex, the identity of which was confirmed by mass spectrometry, also appeared in human AD individuals across all APOE genotypes, suggesting the occurrence of aberrantly resistant apoE aggregates. A second independent cohort of CSF samples validated these results. Conclusion: These results indicate that despite the increase in total apoE content the apoE protein is altered in AD CSF, suggesting that function may be compromised. Keywords: Alzheimer’s disease, apoE, Biomarker, Aberrant complexes, Cerebrospinal fluid, Glycoform imbalance Background An important breakthrough in our understanding of Alzheimer’s disease (AD) was the identification of the apolipoprotein E APOE-ɛ4 allele as a risk factor [1]. Apolipoprotein E (apoE) protein is a component of lipo- *Correspondence: j.saez@umh.es protein particles in the plasma, as well as in the cerebro- 1 Instituto de Neurociencias de Alicante, Universidad Miguel Hernández-CSIC, spinal fluid (CSF) [2]. ApoE regulates important signaling Av. Ramón y Cajal s/n, E‑03550 Sant Joan d’Alacant, Spain pathways by interacting with receptors and is present Full list of author information is available at the end of the article © The Author(s) 2022. 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://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​ mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 2 of 19 as sialylated glycoforms [3]. Human apoE lacks the con- residues at both these positions [4]. The amino acid sub- sensus sequence necessary for N-linked glycosylation; stitution of Cys-112 by Arg in apoE4 explains the lower thus, O-linked carbohydrates probably account for gly- number of disulfide-linked dimers in the CSF of APOE cosylation [4]. The impact of apoE glycosylation remains ε3/ε4 subjects compared with APOE ε3/3 subjects, and unclear, but evidence indicates that glycosylation acts as their absence in APOE ε4/ε4 subjects [16, 17], but may an important post-translational mechanism for fine-tun- also explain the reduced ability of apoE4 to mediate ing apoE interaction with receptors and proteins [5]. some of its biological roles, compared with apoE2 or In humans, three versions of the APOE gene exist, ε2 apoE3 [18]. (apoE2), ε3 (apoE3), and ε4 (apoE4) alleles, while other The mature apoE protein has 299 amino acids and a mammals only have one version of the APOE gene, molecular mass of ~35 kDa. However, previous stud- resembling ancestral apoE4 [6]. APOE-ε3 is the most ies performed in the brain [19] and CSF [17] reported common allele (~75%), followed by ε4 (15–20%) and ε2 a ~100-kDa apoE band in non-reducing conditions, as (4–8%) [7]. Compared to the most common APOE ε3/ε3 opposed to the predicted ~70 kDa, which was referred to genotype, each additional copy of the APOE-ε4 allele is as an apoE homodimer. associated with a higher risk of AD and a younger mean Previous studies that considered total CSF apoE levels age of dementia onset. Thus, in individuals with one failed to demonstrate consistent changes when the APOE copy of the APOE-ε4 allele, the risk of AD increases 2–3 genotype was included as a covariate in the models [20– times and 8–12 times in individuals with two copies [8]. 22]. However, other studies associated high CSF apoE Experimental evidence shows the deleterious effect of the concentrations with an increased risk of impaired cogni- apoE4 variant for AD, while the lack of apoE4 appears to tive progression in non-apoE4 carriers [23]. be protective [9]. In contrast, the presence of one or two Anyhow, in order to consider the estimation of apoE copies of the APOE-ε2 allele is associated with a lower levels in CSF as a read-out of AD occurrence or pro- risk of AD and an older mean age of dementia onset gression, in addition to the APOE genotype, the studies [10]; therefore, it has been hypothesized that the apoE2 should also consider changes in the protein conforma- protein could be protective against AD [11]. Indeed, tion/structure that can compromise the biological func- APOE-ε2 homozygotes present an exceptionally low tion of the apoE protein. In this study, we aimed to likelihood of developing AD [12]. The reported effects of characterize the occurrence of different apoE species different APOE genotypes on AD risk vary widely with in AD CSF from individuals with different APOE geno- demographic factors such as gender and ethnicity [7]. types, while considering changes in the balance of apoE Moreover, the percentage of APOE genotypes in cogni- glycoforms and the occurrence of aberrant apoE dimers tively unimpaired people with neuropathological or bio- that could indicate a compromise of apoE function in the marker evidence of preclinical AD, or the percentage of brain. people who meet the criteria for mild cognitive impair- ment with or without biomarker evidence of AD, is not well established (discussed in [12]). Anyhow, despite the Materials and methods 2–3-fold increase in AD prevalence in APOE-ε4 subjects Patients compared to the general population, most of the individ- CSF samples from individuals with known APOE geno- uals with AD are APOE-ε3 homozygotes [13]. types were obtained from two independent cohorts. The Nonetheless, given the important physiological func- CSF samples from both cohorts used for this study were tions of apoE, a malfunctioning of the apoE protein may de-identified aliquots from clinical routine analyses, fol- also contribute to AD pathology in ε4 non-carriers [14]. lowing procedures approved by the Ethics Committees at The differences in the structure of apoE isoforms influ- the University of Gothenburg and the Hospital Sant Pau, ence their ability to bind lipids, receptors, and amyloid-β respectively. Additionally, this study was approved by the (Aβ), which aggregates in plaques within the brain of AD ethics committee at the Miguel Hernandez University, patients [14]. and was carried out in accordance with the Helsinki Dec- Interestingly, apoE forms disulfide-linked homodi- laration regarding research on humans. mers and heterodimers with the apoA-II apolipopro- The CSF samples were obtained by lumbar puncture tein involving the cysteine (Cys) at position 112 [7, 14]. and centrifuged (2000×g, 10 min) and then immedi- Indeed, these apoE homodimers linked by disulfide ately aliquoted and stored in ultrafreezers and kept at bonds could be the native form able to bind to recep- −80°C until analysis. The time between CSF acquisition tors [15]. The three human apoE isoforms differ in the and storage was less than 4 h in all cases. The handling presence of Cys/arginine (Arg) at positions 112 and 158 of the samples was performed following recommended within the receptor binding domain, as apoE4 lacks Cys Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 3 of 19 operating procedures [24]. Freeze-thaw cycles were The second cohort was obtained from the Sant Pau avoided and new aliquots were used for each independ- Initiative on Neurodegeneration (SPIN cohort) [29] ent analysis. from Hospital Sant Pau (Barcelona, Spain). We included The first cohort was from the longitudinal geriatric samples from 29 AD patients (thirteen men and sixteen population study in Piteå, Sweden [25], the Piteå Demen- women, mean age 73±1 years; APOE: 10 ε3/ε3, 10 ε3/ε4, tia Project. The diagnostic evaluation included a clinical 9 ε4/ε4) and ten controls (seven men and three women, examination (detailed medical history and somatic, neu- mean age 69±2 years; APOE: 5 ε3/ε3, 5 ε3/ε4). Typically, ropsychiatric, and neurological status), a neuropsycho- these are patients who present cognitive complaints and logical test battery, routine blood and CSF tests, and a are referred to the specialized memory unit from their CT scan to exclude secondary dementias [26]. All clini- primary care physician. All patients undergo a full neu- cal diagnoses and evaluations were made without knowl- ropsychological evaluation that demonstrates objec- edge of the results of the biochemical analyses and vice tive cognitive impairment. Patients were included in the versa. The cohort consisted of 45 patients with AD (four- cohort when they presented supportive biomarkers of the teen men and thirty-one women, mean age 77±1 years) AD pathophysiological process. Cognitively normal par- and was selected based on the APOE-ε4 status, so that ticipants were volunteers without cognitive complaints fifteen each had APOE ε3/ε3, APOE ε3/ε4, or APOE ε4/ and normal neuropsychological evaluation. More details ε4. In addition, fourteen non-AD controls [seven men about inclusion/exclusion criteria and neuropsychologi- and seven women, mean age (67 ± 3 years); APOE ε3/ cal tests in this cohort are detailed elsewhere [29]. ε3: 9, APOE ε3/ε4: 5] were included. APOE genotype was In this cohort, the APOE genotype was determined by determined by the solid-phase mini-sequencing method direct DNA sequencing and visual analysis of the result- as previously described [27]. For this study, patients who ing electropherogram performed to identify the two cod- were designated as AD or controls also had typical core ing polymorphisms that encode the three possible apoE CSF biomarker levels [Aβ42 and total tau (T-tau)] using variants [29]. cut-offs that are >90% specific for AD [28], but except for Each center applied their own internally validated cut- CSF Aβ42 and T-tau, all biochemical analyses were made offs, according to their preanalytical and analytical par- without knowledge of the clinical data. The ethics com- ticularities. More details about the cut-offs applied are mittees in Umeå University and University of Gothen- indicated below. Samples were retrospectively selected burg approved the study. from large cohorts to balance age, sex, and APOE sta- tus. Most of the selected cases (92%, 43 of 45 from Table 1 Demographic and biomarker information from the CSF samples obtained from the Gothenburg (Sweden) and Barcelona (Spain) cohorts Cohort: Gothenburg (Sweden) Control Alzheimer’s disease APOE ε3/ε3 ε3/ε4 All ε3/ε3 ε3/ε4 ε4/ε4 All N 9 5 14 15 15 15 45 Age (years) 69±2 62±5 67±3 79±2 78±1 73±1 77±1* Age (range) 60–81 44–75 44–81 62–88 69–84 63–83 62–88 Female/male 5/4 2/3 7/7 11/4 11/4 9/6 31/14 CSF Aβ42 (pg/mL) 845±96 746±121 804±74 470±13* 480±8* 419±21 457±10* CSF tau (pg/mL) 317±53 303±34 312±35 816±88* 917±112* 731±53 840±52* Cohort: Barcelona (Spain) Control Alzheimer’s disease APOE ε3/ε3 ε3/ε4 All ε3/ε3 ε3/ε4 ε4/ε4 All N 5 5 10 10 10 9 29 Age (years) 71±2 67±52 69±2 75±2 73±2 72±2 73±1* Age (range) 66–76 60–72 60–76 64–84 64–83 61–85 61–85 Female/male 1/4 2/3 3/7 7/3 2/8 7/2 16/13 CSF Aβ42 (pg/mL) 1139±248 1010±116 1075±131 607±60* 543±37* 493±62 549±31* CSF tau (pg/mL) 295±49 261±23 278±26 778±94* 624±62* 908±81 765±50* Values are represented as mean ± SEM. *Significantly different (T-test, p< 0.05) from the control group with the same APOE genotype or regardless of the genotype (“All” columns) Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 4 of 19 Gothenburg and 25 of 29 from Barcelona) were catego- Western blotting rized A+T+ according to [30]; thus, subgrouping by the Samples of human or rat CSF (10 μL) were denatured at AT(N) system for analysis was impractical. For full details 98°C for 5 min and resolved by sodium dodecyl sulfate- about the collections, see Table 1. polyacrylamide gel electrophoresis (SDS-PAGE) under reducing or non-reducing conditions (determined by Determination of AD core biomarkers by ELISA the presence or absence of β-mercaptoethanol in the and definition of cut‑offs sample buffer, respectively). Unless specified, the stud- In the cohort from Gothenburg, the levels of the AD core ies presented in the text were performed under reduc- biomarkers T-tau, P-tau, and Aβ42 were measured in the ing conditions. For this study, we used 12% precast CSF using INNOTEST ELISAs (Fujirebio-Europe, Gent, gels (Bio-Rad Laboratories, GmbH, Munich, Germany; Belgium). Patients were designated as AD or controls #4561046). All the samples were analyzed at least in according to CSF biomarker levels using cut-offs that are duplicate (duplicates in separate gels) and distributed in >90% specific for AD: Aβ42 <550 pg/mL and total tau the gels to ensure the comparison by disease condition (T-tau) >400 pg/mL [20]. and APOE genotype. The distribution of the samples in For the cohort from Barcelona, cut-offs for AD bio- the gels was performed by a member of the team and markers measured in the Lumipulse automated platform the experiments were performed by another, the experi- (Fujirebio-Europe) were T-tau > 400 pg/mL, P-tau > 63 menter, in a blind way. pg/mL, and 0.062 for the Aβ42/Aβ40 ratio [29]. Following electrophoresis, proteins were blotted onto All samples were analyzed as part of a clinical routine 0.45-μm nitrocellulose membranes (Bio-Rad Laborato- by board-certified laboratory technicians following strict ries, GmbH, Munich, Germany). Bands of apoE immu- procedures for batch-bridging, analyses, and quality con- noreactivity were detected using either the antibody trol of individual ELISA plates. AB178479 (goat polyclonal; Merck Millipore) or the antibody AB947 (goat polyclonal; Merck Millipore), both common to all apoE isoforms, or alternatively by Transgenic rat CSF an antibody specific to the apoE4 isoform (recognizes The experiments were carried out using a cohort of 107 an internal domain comprising the Arg112 residue rats (53 males and 54 females), including transgenic present exclusively in apoE4 species; mouse monoclo- TgF344-AD rats (n = 52) expressing mutant human APP nal, Novus Biologicals; NBP1-49529). Blots were then (APPsw) and presenilin-1 (PS1ΔE9) genes [31] and wild- probed with the appropriate conjugated secondary type Fischer rats (n = 55). Rats were bred in the animal antibodies (IRDye secondary antibodies, LI-COR Bio- research facilities at the University of Barcelona. Ani- sciences, Lincoln, NE, USA) and imaged on an Odyssey mals were provided with food and water ad libitum and CLx Infrared Imaging System (LI-COR Biosciences). maintained in a temperature-controlled environment in a Band intensities were analyzed using LI-COR software 12/12-h light-dark cycle. CSF samples (50–100 μL) were (ImageStudio Lite). The boxes selected with the Image- collected from ketamine/xylazine-anesthetized animals Quant Studio software for quantification, as well as the by cisternal puncture with a glass capillary in the suboc- completed blots, are shown as supplementary figures. cipital region through the atlanto-occipital membrane, Recombinant apoE3 (Peprotech, ThermoFisher Scien- with a single incision into the subarachnoid space [32]. tific# 350-02) was included into each blot to serve as CSF aliquots from different time points [4 months: 16 a loading reference and for normalizing the immuno- wild-type (8 male, 8 female) and 16 TgF344-AD animals reactivity signal between blots. Specifically, the same (8 male, 8 female); 10.5 months: 17 wild-type (8 male, 9 amount of recombinant apoE3 was always included, female) and 16 TgF344-AD animals (8 male, 8 female); and the immunoreactivity of the apoE bands from each 16.5 months: 22 wild-type (12 male, 10 female) and 20 blot was referred to (divided by) the immunoreactivity TgF344-AD animals (9 male, 11 female)] were analyzed. of recombinant apoE3, thus correcting inter-blot differ- This study was part of a large project assessing various ences and allowing for comparisons across assays. different proteins that included brain analysis at each For blue-native gel electrophoresis, the CSF samples stage; thus, it was not possible to perform longitudinal were not heated (native conditions) and were loaded with measurements in the same animal (repeat sampling) to NuPage LDS 4× Sample Buffer (ThermoFisher Scientific, reduce the number of animals. Animal work was per- NP007) into native-PAGE 4–16% gels (ThermoFisher formed in accordance with the local legislation, with the Scientific, BN1002BOX). Buffers were prepared using approval of the Experimental Animal Ethical Committee native-PAGE Running Buffer (ThermoFisher Scientific, of the University of Barcelona, and in compliance with BN2001) and native-PAGE Cathode Buffer Additive European legislation. (ThermoFisher Scientific, BN2002). Immunoreactivity Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 5 of 19 was detected using the AB178479 antibody and HRP anti- band presence and location. Bands of interest were cut- goat secondary antibody (ThermoFisher). The signal was out from the AD CSF gel lane and recombinant protein visualized by ECL (GE Healthcare Life Science) and ana- lanes and destained using a 1:1 mixture of acetonitrile lyzed using ImageStudio Lite. and 50 mM ammonium bicarbonate solution twice for 15 min. Furthermore, gel pieces were de-hydrated with ApoE immunoprecipitation 100% acetonitrile and dried using a vacuum centrifuge. CSF samples (50 μL) were incubated on a roller over- Samples were subsequently reduced with 10 mM dithio- night with 100 μL PureProteome FlexiBind Magnetic threitol (DTT) for 1 h at 56 °C and alkylated with 25 mM Beads (Merck Millipore, LSKMAGN04) coupled with the iodoacetamide (IAA) for 45 min at room temperature AB178479 apoE antibody (Merck Millipore). The super- in the dark. Gel pieces were further washed with 25mM natant was removed, and the beads were washed and ammonium bicarbonate, de-hydrated with 100% acetoni- then resuspended and boiled at 98 °C for 5 min in SDS- trile, and dried using a vacuum centrifuge once more. PAGE sample buffer and analyzed by western blot with Samples were digested overnight at 37°C using 100 ng/ the AB947 antibody (Merck Millipore) or anti-apoE4 μL trypsin enzyme (Sequencing Grade Modified Trypsin, antibody (Novus Biologicals, NBP1-49529). For a control #V511A, Promega). The next day, digestion was stopped immunoprecipitation, beads were coupled with horse by the addition of 2% trifluoroacetic acid and 75% ace- serum and then incubated with CSF samples. tonitrile solution, and peptides were collected into a new tube (Costar, #3207). Gel pieces were further extracted Enzymatic deglycosylation with the addition of 50% acetonitrile and 0.2% trifluoro- Enzymatic deglycosylation was performed using an acetic acid solution shaking for 30 min. The supernatant Agilent Enzymatic Deglycosylation Kit (Agilent Tech- containing the peptides was transferred to the collec- nologies, GK80110) following the manufacturer’s instruc- tion tube. Pooled extracts for each gel piece were dried tions. Briefly, for each condition, 30 μL of control or AD through vacuum centrifugation and stored at −80 °C CSF was mixed with 10-μl incubation buffer and 2.5-μL pending mass spectrometry (MS) analysis. denaturing buffer and heated at 100 °C for 5 min. The samples were then cooled down to room temperature, Mass spectrometry data analysis and 2.5 μL of detergent (15% NP-40) was added while Dried in-gel digested samples were reconstituted in 7 μL mixing gently. O- (1 μL sialidase and 1 μL O-glycanase) 8% acetonitrile/8% formic acid solution and shaken for or N-linked (1 μL N-glycanase) deglycosylating enzymes 30 min. A total of 6 μL of each sample was investigated were then added according to each different condition using mass spectrometry (MS) analysis performed with (O-linked, N-linked, or O- and N-linked deglycosyla- a Dionex 3000 nanoflow liquid chromatography system tion) and samples were heated at 37 °C for 3 h. Samples coupled to a Q Exactive (both Thermo Fisher Scientific). were then analyzed by western blot. As for control of the Briefly, a reversed phase Acclaim PepMap C18 (100 Å deglycosylation process, samples exposed to the same pore size, 3 μm particle size, 20 mm length, 75 μm i.d., heating conditions but without deglycosylating enzymes Thermo Fisher Scientific) trap column was used for were included. online desalting and sample clean-up. Separation was performed with a reversed phase Acclaim PepMap RSLC In‑gel digestion C18 (100 Å pore size, 2 μm particle size, 75 μm i.d., 150 In-gel digestion was performed as previously described mm length, Thermo Fisher Scientific) column at a flow [33] in order to investigate the content of western blot rate of 300 nL/min by applying a linear gradient of 0–40% immunoreactive bands of interest using an antibody- B for 50 min. Mobile phase A was 0.1% formic acid in free method. Briefly, 1 mL of a pool of AD CSF (APOE water (v/v) and mobile phase B was 0.1% formic acid and ε3/ε4 and APOE ε3/ε4 cases) was immunoprecipitated 84% acetonitrile in water (v/v/v). with AB178479 antibody and loaded into SDS-poly- Mass spectra were acquired in positive ion mode and acrylamide gel under reducing conditions, as described in a data-dependent manner with a resolution setting of above. ApoE3 and apoE4 recombinant proteins (Pepro- 70,000 for precursor and 17,500 for fragment ion acqui- tech, ThermoFisher Scientific# 350-02 and 350-04) were sitions. Fragmentation was obtained by higher energy also loaded in the gel (10 pmol) and used as a reference collision-induced dissociation (HCD) using a normalized for band excising and positive control. Upon electro- collision energy (NCE) setting of 28. Database searches phoresis, the gel was divided into two parts, one for pro- were made using PEAKS Studio XPRO (Bioinformatic ­ implyBlueTM SafeStain Coomassie tein visualization by S Solutions, Inc., Waterloo, Canada). (ThermoFisher Scientific, cat# LC6060) and one for blotting with the AB947 antibody as confirmation of Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 6 of 19 Statistical analysis Results All the data was analyzed using GraphPad Prism (version CSF apoE in Tg344‑AD rats 7; GraphPad software, San Diego, CA, USA). The test was To determine whether altered CSF apoE levels could used to analyze the distribution of each variable. Firstly, be indicative of pathology-associated changes, we ini- multiple comparisons were performed between groups, tially examined them in a rat transgenic AD model. The ANOVA was used for parametric variables, and the TgF344-AD rat expresses human APP with the Swedish Kruskal-Wallis test for non-parametric variables. A Stu- mutation and human PSEN1 with the Δ exon 9 mutation. dent’s t-test for parametric variables and a Mann-Whit- As mentioned above, while humans have three versions ney U test for non-parametric variables were employed of the APOE gene, other mammals such as rats only have for comparison between two groups and for determining one isoform of the apoE protein, which presents Arg at precise p values. For correlations, the Pearson and Spear- position 112 (https://​web.​expasy.​org/​varia​nt_​pages/​ man tests were used. The results are shown as means ± VAR_​000652.​html) and thus shares the inability to form SEM; the standard deviation (SD) and median values are disulfide-linked dimers with human apoE4. We examined also displayed as indicated in the figure legends. apoE levels in the CSF of 4-, 10.5-, and 16.5-month-old transgenic rats and wild-type littermates by SDS-PAGE Fig. 1 Analysis of CSF apoE in the TgF344-AD rats. A Representative blot of CSF obtained from wild-type (Wt) or transgenic (Tg) rats at 4, 10.5, and 16.5 months. The 100-kDa section of the blot presents enhanced contrast. B Quantification of apoE values obtained from western blots. Data is shown as a percentage with respect to the Wt values obtained at each age. The graphs represent mean ± SEM, and the numbers below represent median ± SD. A significant p value is indicated Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 7 of 19 using the AB178479 antibody. In all the animals and this antibody and subsequent immunoblotting with an at all ages, apoE appeared as a single ~35-kDa band alternative apoE antibody, AB947, confirmed the charac- (Fig. 1A) and no differences were found between males terization of both apoE monomeric species in APOE ε3/ and females (p> 0.05 for the comparison at every age). ε3 samples (Fig. 2B). The same occurred when using an We did not find different glycoforms. Although no dif- anti-apoE4 antibody in APOE ε3/ε4 samples (Fig. 2C). ferences were found at 4 and 10.5 months between wild- An apoE band of ~100 kDa was also observed, almost type and TgF344-AD rats, a trend of apoE increment was exclusively in the AD CSF samples, and this band was observed. At 16.5 months of age, apoE levels were 50% immunoprecipitated similarly to monomers (Fig. 2A–C). higher in TgF344-AD animals than in wild-types (p = ApoE3 and apoE2 isoforms form disulfide-linked dimers 0.003, Fig. 1B). The significant differences for CSF apoE in CSF, but these dimers should be sensitive to the reduc- levels detected between TgF344-AD and controls at 16.5 ing agent β-mercaptoethanol. Additional bands between months of age were maintained when the animals were the monomers and the 100-kDa bands did not appear subgrouped by gender (male: control vs TgF344: p = to follow any specific pattern related with the pathology 0.019; female: control vs TgF344: p = 0.048). condition or the APOE genotype and were not consist- ently represented in the immunoprecipitated fraction; Characterization of apoE in human CSF therefore, they were not considered for further investiga- We examined the presence of apoE species in CSF tions. Immunoprecipitated complexes of 100 kDa, using samples by SDS-PAGE and western blot under reduc- the AB178479 antibody, were dissected after electropho- ing conditions (in presence of the reducing agent resis and examined by MS analysis identifying 14 tryptic β-mercaptoethanol that breaks disulfide bonds) from peptides spanning throughout the sequence of human a cohort of control and AD patients from Gothen- apoE (Uniprot entry P02649_HUMAN), and both apoE3 burg (Sweden; see Table 1) expressing different APOE and apoE4 isoforms were detected. Matching sequences genotypes. are displayed in Table 2. In all CSF samples, using the AB178479 antibody, We first aimed to understand why the monomers apoE appeared as two distinct immunoreactive bands of appeared as two distinct bands with different molecular ~34 and ~36 kDa (Fig. 2A). Immunoprecipitation with masses. We hypothesized that the bands likely represent Fig. 2 Characterization of apoE protein in human control (Ct) and AD CSF samples. A Representative immunoblot of CSF samples immunoblotted with apoE antibody (AB178479). CSF samples immunoprecipitated with the apoE AB178479 antibody (originated in goat) and immunoblotted with B apoE antibody AB947 (originated in goat) common to all apoE variants or C NBP1-49529 (originated in mouse), an antibody specific to the apoE4 isoform. Arrowheads: non-specific immunoglobulins (detected also in beads coupled with AB178479 in absence of human CSF; not shown). The bands detected between 34- and 36-kDa monomers and 100-kDa dimers were not consistently immunoprecipitated across trials. Total, CSF before IP; IP, immunoprecipitated protein; IPc, control immunoprecipitation. D CSF samples immunoblotted with apoE antibody (Ab178479) following O-linked, N-linked, or O- and N-linked deglycosylation. As a control of the deglycosylation process, samples under standard conditions and samples heated to 37 °C in the absence of deglycosylating enzymes were also included. Representative blots of three independent immunoprecipitation or deglycosylation experiments are shown Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 8 of 19 Table 2 Identified peptides from apoE species of human CSF by MS A, B: ApoE peptide chains assessed in monomeric (A) and the 100-kDa (B) species. The 18 aa signal peptide is shown in light gray; numbering is according to the mature protein, and X (marked in green) at position 112 denotes C (Cys) for ε3 and R (Arg) for ε4 isoforms, respectively. C: Data obtained from MS analysis. aStart and end positions refer to the mature protein without signal peptide. bScore as calculated by PEAKS Studio = −10 lg(P), where P is the probability for a false positive as determined by the software. PTM post-translational modifications, ppm mass error of the measured peptide in parts per million different glycoforms of the protein and, thus, an enzy- O-glycosylation could account for the differences in the matic deglycosylation assay was performed. While, as molecular mass between the 36- and 34-kDa apoE spe- expected, N-deglycosylation did not alter the apoE band cies (Fig. 2D). The small differences in the electrophoretic pattern, O-deglycosylation simplified the apoE pat- mobility between glycosylated and deglycosylated apoE tern to a single immunoreactive band, suggesting that monomers are probably related to the slight carbohydrate Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 9 of 19 mass associated to O-glycosylation, but also to changes non-reducing conditions to preserve the disulfide bonds in protein shape that affect their electrophoretic migra- (absence of β-mercaptoethanol). The 100-kDa apoE band tion. Glycosylated proteins are normally more globular in AD samples appeared to be indistinguishable from the than non-glycosylated proteins, because the carbohy- one observed in reducing conditions, and remarkably, drate chains are not linear, even in reducing conditions. this band appeared in the control samples in non-reduc- The apparent levels of the 100-kDa apoE band were not ing conditions, probably representing apoE dimers linked significantly modified following enzymatic deglyco- by disulfide bonds (Fig. 3A). When a specific antibody sylation, suggesting that these species are resistant to for apoE4 was used, NBP1-49529, the 100-kDa immuno- enzymatic deglycosylation (Fig. 2D). We observed the reactivity was also detected in samples from APOE ε3/ occurrence of apoE dimers in the recombinant protein (a ε4 AD subjects under both reducing and non-reducing non-glycosylated species since it is produced in bacteria), conditions, while no immunoreactivity was detected in suggesting that sugar residues are not relevant epitopes APOE ε3/ε4 control samples under non-reducing con- for disulfide-linked dimer formation. ditions (Fig. 3B). This could corroborate that apoE4 in As previously mentioned, the ~100-kDa apoE band was AD samples participates in complexes to form 100-kDa observed almost exclusively in AD CSF samples, includ- stable species, in both apoE3/4 or apoE4/4 subjects, that ing APOE ε4/ε4 samples (Fig. 2A). To further character- do not rely on disulfide bonds, due to its lack of Cys112, ize this apoE species, we performed SDS-PAGE studies in thus representing aberrant/anomalous apoE aggregates. Fig. 3 Characterization of CSF apoE species under reducing, non-reducing, and native conditions. A, B Control (Ct) and AD CSF samples immunoblotted under reducing (R) or non-reducing (NR) SDS-PAGE with A apoE antibody (Ab178479) or B apoE4-specific antibody (NBP1-49529). Note that apoE4 is identified as part of the 100-kDa bands despite the inability to form disulfide-linked dimers. C Representative blot of native-PAGE studies, showing the apoE monomers and dimers. The discrepancy in molecular weight is likely due to the differences between native-PAGE and SDS-PAGE electrophoretic conditions which affect the migration of proteins. Rec, recombinant apoE3; Den CSF, denatured CSF. D Quantification of the ratio of apoE dimers compared to monomers across the different APOE genotypes, estimated by native-PAGE (see C). Scatter plots of apoE levels are represented. The graphs represent mean ± SEM, and the numbers below represent median ± SD. Significant p values are indicated Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 10 of 19 In APOE ε3/ε4 control subjects, however, apoE4 does apoE species by SDS-PAGE in reducing conditions and not participate in the 100-kDa species observed under western blotting, using the AB178479 antibody (Fig. 4A). non-reducing conditions, given its complete reliance on The 34-kDa apoE band was significantly increased in disulfide bonds to form functional dimers. AD compared with controls (p = 0.003; Fig. 4B). When To determine the different contributions of the apoE discriminating by APOE genotype, only the APOE ε3/ 100-kDa species in AD and control cases, we first esti- ε3 genotype was significantly elevated in AD compared mated the apoE dimer/monomer balance by native- with control samples (p = 0.02; Fig. 4C). The same anal- PAGE electrophoresis. We included a CSF control sample ysis within the APOE ε3/ε4 genotype exhibited less sta- under fully reducing and denaturing conditions, as well tistical power because the size of the control group is as an apoE3 recombinant protein under native con- small; nonetheless, we observed a trend of 34-kDa apoE ditions, which served to identify the monomeric and increment in ε3/ε4 genotype AD samples with respect to dimeric apoE bands. Two immunoreactive apoE bands controls (p = 0.09; Fig. 4C). The 36-kDa apoE appeared were observed, likely representing apoE monomers and significantly increased in AD compared with controls dimers (Fig. 3C). An immunoreactive band compatible overall (p = 0.002; Fig. 4D), but not among APOE geno- with dimeric complexes was detected in CSF from AD types (Fig. 4E). APOE ε4/ε4 cases, whose lack of Cys112 should elimi- As expected, when we considered the sum of the nate their ability to form disulfide-bond-dependent com- apoE immunoreactivity for the 34- and 36-kDa bands, plexes. Given the difficulty of finding age-matched APOE increased levels were seen in AD patients (27 ± 5%), as ε4/ε4 control subjects (low prevalence of this genotype compared with controls (p = 0.005). Despite the fact in the general and healthy population), we cannot com- that these results indicate a net increase of CSF apoE pare AD APOE ε4/ε4 with control ε4/ε4 cases. We com- in AD samples, when defining a quotient between the pared the apoE dimer/monomer quotient (ratio D/M) apoE monomeric glycoforms (ratio 36 kDa/34 kDa) we between AD CSF samples and controls, subgrouping detected an imbalance in the AD samples, which dis- the samples by APOE genotype (Fig. 3D). In the control played a decreased 36-kDa/34-kDa ratio compared with group, the dimer/monomer quotient was significantly controls (p = 0.007; Fig. 4F). These differences were lower in the APOE ε3/ε4 group (ratio D/M = 0.38) com- maintained when the samples were separated by APOE pared to that of the ε3/ε3 group (ratio D/M = 2.20; p= genotype (ε3/ε3 control: ratio 36 kDa/34 kDa = 1.61 vs 0.006), associated to the inability of the apoE4 isoform ε3/ε3 AD: ratio 36 kDa/34 kDa = 1.46, p = 0.01; ε3/ε4 to form disulfide-linked dimers. The same situation was control: ratio 36 kDa/34 kDa = 1.67 vs ε3/ε4 AD: ratio = found in the AD group, where the dimer/monomer ratio 1.38, p = 0.001; Fig. 4G). Within the AD group, we also decreased as the ε4 allele was present (ε3/ε3: ratio D/M = observed significant differences between the genotypes, 2.27; ε3/ε4: ratio D/M = 1.04; ε4/ε4: ratio D/M = 0.24). as the 36-kDa/34-kDa ratio was significantly lower in For APOE ε3/ε3 subjects, the dimer/monomer ratio in APOE ε3/ε3 (p < 0.0001) and ε3/ε4 (p < 0.0001) samples controls was not significantly different to that found in when compared with ε4/ε4 AD samples (ratio 36 kDa/34 AD subjects; however, for APOE ε3/ε4 subjects, the quo- kDa = 1.64). In each group, CSF apoE levels appeared tient was higher in the AD group compared with controls unaltered when subgrouping between males and females (p = 0.02; Fig. 3D). This may be reflecting the accumula- (p > 0.05 for all the subgroups). There were no clear cor- tion of aberrant aggregates in the AD samples expressing relations between the level of the 34- or 36-kDa apoE or apoE4, in addition to the physiological disulfide-bound the 36-kDa/34-kDa ratio with the age of the subjects, in dimers present in controls. either of the groups considered individually. The immunoreactivity of the 100-kDa apoE species Levels of CSF apoE species in AD was quite faint in control samples, and accordingly, sub- Given the differences in CSF apoE aggregates found stantial differences were found between AD samples and under native conditions between AD and controls, we controls (p < 0.0001; Fig. 4H, I). In the AD group, the evaluated the levels of the 34-kDa, 36-kDa, and 100-kDa 100-kDa apoE species levels were significantly lower in (See figure on next page.) Fig. 4 Analysis of CSF apoE species from the Gothenburg cohort. Control (Ct) and AD CSF samples analyzed by SDS-PAGE. Each individual band was quantified and normalized to the reference value (recombinant apoE). A Representative immunoblot of CSF samples with apoE antibody and legend for graphs. The 100-kDa section of the blot presents enhanced contrast. B, C Statistical analysis of the 34-kDa apoE immunoreactive band in B control and AD and by CAPOE genotype. D, E Statistical analysis of the 36-kDa apoE immunoreactive band in D control and AD and by EAPOE genotype. F, G Statistical analysis of the ratio of 36-kDa/34-kDa immunoreactive bands in F control and AD and by GAPOE genotype. H, I Statistical analysis of the 100-kDa apoE immunoreactive band in H control and AD and by IAPOE genotype. The graphs represent mean ± SEM, and the numbers below represent median ± SD. Significant p values are indicated Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 11 of 19 Fig. 4 (See legend on previous page.) Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 12 of 19 the APOE ε4/ε4 group when compared to both ε3/ε3 (p limit of statistical significance in the APOE ε3/ε4 group < 0.0001) and ε3/ε4 (p < 0.0001) groups. Interestingly, in (p = 0.050) despite the small number of controls. Within the APOE ε4/ε4 AD cases, the samples with the lowest the AD group, the APOE ε4/ε4 samples once again dis- 100-kDa apoE immunoreactivity belonged to the young- played significantly lower 100-kDa apoE levels com- est subjects (n = 5, 68±2 years), as compared to the other pared with ε3/ε3 cases (p = 0.02). The results from this cases (n= 10, 76±1 years; p=0.002). Within the AD cohort corroborate that 100-kDa apoE is associated to APOE ε4/ε4 subgroup, we also observed a statistically AD and that apoE4 has a reduced capacity to form these significant positive correlation between the age of par- complexes. ticipants and the immunoreactivity of the 100-kDa apoE In this cohort, and exclusively in the AD group over- band (r = 0.622, p = 0.013). For the rest of the groups, all, we detected a significant correlation between the we failed to determine a correlation between the 100-kDa age of the subjects and the 100-kDa apoE species (r band and age. =0.645, p= 0.0002), indicating that the appearance of Interestingly, the quotient of 36 kDa/34 kDa mono- the aberrant apoE aggregates may be related to aging in meric glycoforms (r = 0.40, p = 0.007), as well as the lev- AD. This association was maintained within the APOE els of the 34-kDa species (r = 0.32, p = 0.034), correlated ε3/ε3 (r = 0.813, p= 0.004) and the ε3/ε4 (r = 0.799, with Aβ42 in AD individuals. Meanwhile, the levels of the p= 0.006) AD groups. 100-kDa apoE complexes correlated with T-tau levels (r In this cohort, only the levels of the 100-kDa apoE = 0.33, p = 0.028), yet failed to achieve significance with species correlated with Aβ42 levels (r= 0.41, p= 0.027). the Aβ42 levels (r = 0.27, p = 0.070). Thus, despite finding some correlations, none of these significant correlations resulted in consistency between Levels of CSF apoE species in AD in a second cohort the two cohorts. We attempted to validate our results in a second inde- pendent cohort of CSF samples from Barcelona (see Discussion Table 1, Fig. 5A). As in the first cohort, the analyses were Typically, transgenic models produce pathological performed by SDS-PAGE under reducing conditions. changes that partially replicate changes seen in human In this cohort, the 34-kDa apoE levels were significantly patients. In this study, firstly, we have found an increase higher in AD compared with controls (p=0.001), and in CSF apoE in the TgF344-AD rats, with the documented specifically only for those with APOE ε3/ε3 genotype (p occurrence of amyloid pathology around 10 months of = 0.01) (Fig. 5B, C). In contrast to the first cohort, no dif- age [31, 34]. This result can be interpreted as a suggestive ferences in the 36-kDa species were detected (Fig. 5D, gain of function for apoE in AD. In fact, this increase in E). The 36-kDa/34-kDa ratio was again lower in AD CSF apoE content is similar to the one observed in AD compared with control samples (p < 0.001, Fig. 5F), and patients when considering total apoE content. Consid- this difference was maintained when the samples were ering the summation of the apoE immunoreactivity for separated by genotype (AD vs controls for APOE ε3/ε3, 34- and 36-kDa (not including the value for the 100-kDa p=0.02, and for APOE ε3/ε4, p=0.03) (Fig. 5G). ApoE band) species, in samples from AD patients, a significant levels once again appeared to be unaltered when sub- overall increase in total CSF apoE was found in the Goth- grouping between males and females (p> 0.05 for all enburg cohort and a non-significant trend to increase comparisons). In this cohort, we also failed to correlate was seen in the Barcelona cohort compared to controls. 34- or 36-kDa levels or the 36-kDa/34-kDa ratio with the However, the biochemical discrimination of different age of the subjects. human CSF apoE species and the altered balance of these As in the first cohort, the 100-kDa apoE levels were species lead us to believe that, despite the increase in higher in AD than in controls (p = 0.005; Fig. 5H). When total CSF apoE levels determined in the AD transgenic the samples were stratified by APOE genotype (Fig. 5I), model and AD patients, the imbalance between apoE the significant differences between AD and controls were species should be interpreted as indicative of a potential maintained in the ε3/ε3 group (p = 0.01) and were in the impairment in apoE function in the brain. Thus, higher (See figure on next page.) Fig. 5 Analysis of CSF apoE species from the Barcelona cohort. Control (Ct) and AD CSF samples analyzed by SDS-PAGE. Each individual band was quantified and normalized to the reference value (recombinant apoE). A Representative immunoblot of CSF samples with apoE antibody and legend for graphs. The 100-kDa section of the blot presents enhanced contrast. B, C Statistical analysis of the 34-kDa apoE immunoreactive band in B control and AD and by CAPOE genotype. D, E Statistical analysis of the 36-kDa apoE immunoreactive band in D control and AD and by EAPOE genotype. F, G Statistical analysis of the ratio of 36-kDa/34-kDa immunoreactive bands in F control and AD and by GAPOE genotype. H, I Statistical analysis of the 100-kDa apoE immunoreactive band in H control and AD and by IAPOE genotype. The graphs represent mean ± SEM, and the numbers below represent median ± SD. Significant p values are indicated Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 13 of 19 Fig. 5 (See legend on previous page.) Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 14 of 19 levels of apoE could paradoxically result in less func- ε3 and ε3/ε4, while in the Barcelona cohort, the ratios tionality if the increase is represented by complexes and were at a similar level among AD individuals with dif- immature glycoforms. ferent APOE genotypes. Additional studies will serve to Indeed, we have identified two monomeric apoE spe- determine if the imbalance between apoE glycoforms cies in human CSF and demonstrated that the balance is a common feature for AD APOE-ε4 homozygote between these species in AD patients differs compared subjects. to that of controls. Some previous studies that did not Indeed, the changes observed in this study are less distinguish the contribution of particular apoE species obvious in ε4/ε4 samples. This discrepancy may be due have indicated that CSF apoE levels in AD patients are to the fact that small changes in apoE levels for ε4/ε4 sub- increased [35], also at follow-up [36], but many stud- jects could be more detrimental than in the rest of APOE ies addressing total CSF apoE levels are inconclusive genotypes, perhaps caused by the basal compromise and found no clear association with the AD condition in some of the biological functions of apoE in the brain or APOE genotype [20–22]. In addition to recurrent related with the inability of the apoE4 isoform to form confounding factors such as the handling of the sam- dimers. ples, and also considering differences in the diagnostic APOE-ε4 is the strongest risk factor gene for AD, accuracy between cohorts, the inconsistencies found in although inheriting APOE-ε4 does not mean a person these previous reports could be associated mostly with will definitely develop the disease. Thus, the opportunity the determination method used, as some are based in to analyze the subset of APOE ε3/ε4 control individuals MS [17, 18], while others use immunoassays [16], both with no AD-like symptoms is very interesting. As stated, of which fail to discriminate between apoE species. all the cases were retrospectively selected from large Even if an immunoassay is the most available and desir- cohorts and based on the determination of AD core bio- able approach for quantitative analysis of altered lev- markers. The diagnostic uncertainty is inherent in this els of a biomarker, this method does not easily detect type of studies, but the control individuals with APOE subtle changes in specific species (imbalance in glyco- ε3/ε4 genotype displayed similar apoE values as the ones forms) and/or does not detect particular species suffer- obtained in APOE ε3/ε3 individuals. ing conformational changes (aberrant dimers). Correct apoE glycosylation is fundamental for its func- The 34- and 36-kDa species are likely different O-gly- tion and lipoprotein binding capacity. ApoE glycosyla- coforms, and the difference in electrophoretic mobility tion can modulate receptor affinity, lipid-binding ability, of the apoE glycoforms could be a consequence of its lipid transportation, and metabolic functions [41–43]. sialylation [37]. ApoE is exclusively O-glycosylated and Furthermore, apoE deglycosylation reduces its binding to can be capped with one or two sialic acids [5]. In CSF, Aβ42 [44] and may induce Aβ42 accumulation [45]. Our the existence of two glycans per molecule of apoE has results suggest that the imbalance between the differ- been demonstrated [38], and previous studies indicate ent glycoforms of apoE monomers observed in AD may that astrocytes secrete two differential glycoforms of interfere with its biological function, contributing to the apoE [39] and that the sialo and asialo forms of apoE progression of the disease. Interestingly, apoE glycosyla- can both be secreted into the medium [40]. Our results tion also plays a key role in the protection against self- indicate that the 34-kDa apoE monomers, which appear association and spontaneous aggregation [46]. to be less sialylated than 36-kDa apoE monomers [3], As mentioned, the apoE isoforms encoded by APOE are present at a higher proportion in AD subjects com- ε3 or ε2 are able to form disulfide-linked hetero- and pared with controls, in both independent cohorts. homodimers through the Cys residue at position 112, Whether or not these 34-kDa species can participate while APOE ε4 (which presents Arg at position 112) in disulfide-linked apoE dimers or pathological com- and apoE from non-human mammals are unable to plexes, as described here, requires further study. form these oligomeric species. However, in our studies, Moreover, the altered balance between apoE glyco- apoE4 isoforms were present in 100-kDa aggregates in forms should be validated in external cohorts. Here, APOE ε3/ε4 AD cases, and these aggregates were iden- most of the results obtained in the Gothenburg cohort tified in most of the APOE ε4/ε4 AD patients. These were validated in a second independent cohort from 100-kDa complexes are compatible in molecular mass Barcelona, despite the small size of the groups in this to disulfide-linked apoE dimers, which exist as a major cohort. Nonetheless, some inconsistent results were portion of apoE in human CSF of APOE ε3 or ε2 carriers observed between cohorts regarding the ratio of the [16]. The existence of SDS-resistant dimers of apoE4 was 36-kDa/34-kDa species. In the Gothenburg cohort, this suggested when studying the in vitro formation of SDS- ratio was significantly higher in AD individuals with resistant Aβ-apoE complexes [47]; but, to our knowledge, an APOE ε4/ε4 genotype compared with APOE ε3/ it has never been demonstrated in vitro or in vivo. The Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 15 of 19 definitive identity of the 100-kDa species was confirmed However, despite the fact that Aβ can contribute to the by the diverse immunoprecipitation analyses combining formation of stable apoE dimers as a crosslinking agent, antibodies originated from diverse animal species and the behavior of the resulting species may differ from the MS studies. Rats express a unique apoE variant most other apoE/Aβ aggregates. We favor the hypothesis that closely related to the human ε4-type haplotype. How- the stable apoE complexes may have compromised bio- ever, in the transgenic rat model of AD, we were not able logical activity, regardless of the presence of Aβ. to observe the 100-kDa resistant apoE species that we It is also interesting to note that apoE binds Aβ in observed in AD APOE ε4/4 cases. Likewise, the possibil- an isoform-specific manner. Thus, monomeric apoE4 ity that inactive monomers of apoE occur in this animal binds to Aβ peptide more rapidly than monomeric model requires further study; however, models in which apoE3 or apoE2, and so it appears that the efficiency of the amyloid condition results in an increase of apoE binding correlates inversely with the risk of develop- expression should consider this possibility. ing AD pathology [62]. Moreover, soluble SDS-stable ApoE dimers or multimers may be the biologically complexes of apoE4/Aβ precipitate more rapidly than important species, particularly in receptor binding [15]. apoE3/Aβ complexes [63]. Whether these monomeric In a previous study, the levels of apoE dimers in the CSF apoE/Aβ complexes trigger the formation of oligomeric from AD subjects were not different from those in con- complexes, and the potential compromise of the apoE trols [48], although in this study they did not assess the peptides involved in these complexes on Aβ clearance nature of the aberrant β-mercaptoethanol resistant com- in vivo, require analysis. plexes. In our AD samples, the 100-kDa apoE complexes The aberrant apoE complexes may also influence the are aberrantly resistant to reducing conditions; thus, they role of apoE on lipid metabolism and transport. It is may represent a different species compared to the biolog- assumed that unlipidated apoE monomers are the spe- ically active disulfide-bound dimers. The relevance of an cies that form disulfide-linked dimers; however, it is also apoE dimer/monomer profile in AD was also addressed believed that apoE must be properly lipidated to partici- previously in plasma, with the identification of dimers pate in cholesterol and lipid transport. Aberrant dimers only in APOE-ɛ3 carrier subjects, the levels of which are not linked by disulfide bonds, but we can only specu- decreased in the demented group [49]. A recent report late whether these species are lipidated or not, and if the using two-dimensional gel electrophoresis indicated that occurrence of these aberrant dimers could compromise plasma apoE is elevated in AD with respect to controls the role of apoE regulating lipid homeostasis by mediat- [50]. However, it is worth noting that apoE does not cross ing lipid metabolism and transport. ApoE4 is poorly lipi- the blood-CSF barrier [51]. dated compared with apoE2 and apoE3 [64], and reduced ApoE can form heteromeric complexes with other binding affinity of apoE4 for HDL results in a greater apolipoproteins [17] and with proteins such as the ciliary proportion of unlipidated apoE, hence forming aggre- neurotrophic factor [52] or APP [53], among others, but gates that can be more toxic for neurons than apoE2 and principally with Aβ. Indeed, apoE can form in vitro SDS- apoE3 aggregates [65]. Since lipidation of apoE impedes stable complexes with Aβ [1, 54, 55], but the interaction aggregate formation [66], we presume that these aber- with exogenous Aβ does not induce drastic changes to rant dimers are not lipidated; nonetheless, this possibility the overall size of the Aβ/apoE-containing lipoprotein should be tested. particles [55]. The formation of noncovalent apoE/Aβ Finally, we found a correlation between the 100-kDa complexes (1:1) is implicated in both Aβ clearance and apoE levels and age in AD samples, which suggests that fibrillization, and the three isoforms of apoE are able to during pathological aging, apoE could be more likely to form these complexes [56]. Complexes of apoE and Aβ form non-disulfide-bound aggregates in the CSF. In the have been demonstrated in non-pathological human TgF344 rats, only the older animals showed statistically CSF [55] and in AD brain [57, 58]. Thus, Aβ may act as significant high apoE levels; accordingly, these AD mod- a triggering driver for the crosslinking and stabilization els show an age-dependent increase of the levels of Aβ40 of aberrant apoE complexes. In the AD brain, the balance and Aβ42 from 6 months of age [31]. between soluble to insoluble apoE/Aβ aggregates has The imbalance of apoE glycoforms and the existence of been associated with impaired apoE activity in Aβ clear- aberrant apoE aggregates in the CSF from AD individu- ance, as apoE is responsible for the accumulation and als could be considered as a read-out of alterations of the fibrillization of Aβ [59]. The effects of apoE on Aβ aggre- biological activity of apoE in the brain of AD individuals. gation may be restricted to HDL-like particle-bound The possibility that CSF levels of apoE are under strong apoE [60]. Other studies have demonstrated that apoE genetic influence by the APOE polymorphism is plau- influences Aβ clearance despite minimal interaction [61]. sible; however, the relevance of these changes in CSF Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 16 of 19 apoE levels on AD pathology remains elusive. The net Acknowledgements Not applicable. increase of apoE levels in the CSF from AD individuals could be favored by aging. This increment, mainly due Authors’ contributions to the 34-kDa glycoform of apoE, which is likely hypo- MPL, IC, and JSV were involved in the design of the study. MPL, IC, JSV, and ISD contributed to data treatment. HZ, KB, DA, JF, and AL all provided CSF sialylated, and the appearance of a β-mercaptoethanol- samples. FA and GS provided rat samples. MPL and ISD were responsible for resistant 100-kDa apoE species, could indicate that the the experimental work performed in the study. EC and GB performed the ability of apoE in AD to achieve its biological functions in-gel digestion and the MS analysis. The authors read and approved the final manuscript. may be compromised. In conclusion, while apoE levels tend to increase in Funding AD CSF, this increase is more noticeable in certain gly- This work was supported by grants from the Fondo de Investigaciones Sanitarias (FIS; PI15/00665 and PI19-01359, co-funded by the Fondo Europeo coforms of monomers and aberrant complexes that may de Desarrollo Regional, FEDER “Investing in your future”), from the CIBERNED hinder its biological activity. A specific description of (Instituto de Salud Carlos III, Spain) and from the Direcció General de Ciència how these species affect apoE signaling and Aβ clearance I Investigació, Generalitat Valenciana (AICO/2021/308). We also acknowledge financial support from the Spanish Ministerio de Economía y Competitividad, should improve our understanding of the role of apoE in through the “Severo Ochoa” Programme for Centres of Excellence in R&D the AD pathology. (SEV-2017-0723). MPL is supported by a BEFPI fellowship from the Generalitat Valenciana. This study was also supported by the FIS, Contratos para la intensificación Conclusions de la actividad investigadora en el SNS from Instituto de Salud Carlos III The imbalance of apoE glycoforms and the existence of (PI14/01126, PI17/01019, PI20/01473, and INT16/00171 to JF; PI18/00435 to DA; PI14/1561, PI17/01896, and PI20/1330 to AL). This work was also sup‑ aberrant apoE aggregates in the CSF from AD individu- ported by the CIBERNED program (Program 1, Alzheimer Disease to AL and als could be considered as a read-out of alterations of the SIGNAL study, www.​signa​lstudy.​es), the National Institutes of Health (NIA biological activity of apoE in the brain of AD individuals. grants 1R01AG056850-01A1, R21AG056974, and R01AG061566 to JF), and Departament de Salut de la Generalitat de Catalunya, Pla Estratègic de Recerca The possibility that CSF levels of apoE are under strong i Innovació en Salut (SLT002/16/00408 to AL). This work was also supported by genetic influence by the APOE polymorphism is plau- Generalitat de Catalunya (SLT006/17/00119 to JF and SLT006/17/00125 to DA). sible; however, the relevance of these changes in CSF This study was also supported by the Spanish Ministry of Science, Innovation and Universities MICINN/FEDER (PID2019-107738RB-I00, to FA, a fellowship to apoE levels on AD pathology remains elusive. The net ISD and the “Maria Maeztu Unit of Excellence Program”) and Instituto de Salud increase of apoE levels in the CSF from AD individuals Carlos III (PI18/00893 to GS, co-funded by ERDF, “A way to make Europe”). could be favored by aging. This increment, mainly due HZ is a Wallenberg Scholar supported by grants from the Swedish Research Council (#2018-02532); the European Research Council (#681712); the Swedish to the 34-kDa glycoform of apoE, which is likely hypo- State Support for Clinical Research (#ALFGBG-720931); the Alzheimer Drug sialylated, and the appearance of a β-mercaptoethanol- Discovery Foundation (ADDF), USA (#201809-2016862); the AD Strategic Fund resistant 100-kDa apoE species, could indicate that the and the Alzheimer’s Association (#ADSF-21-831376-C, #ADSF-21-831381-C, and #ADSF-21-831377-C); the Olav Thon Foundation; the Erling-Persson ability of apoE in AD to achieve its biological functions Family Foundation; Stiftelsen för Gamla Tjänarinnor; Hjärnfonden, Sweden may be compromised. (#FO2019-0228); the European Union’s Horizon 2020 research and innova‑ In conclusion, while apoE levels tend to increase in tion programme under the Marie Skłodowska-Curie grant agreement No 860197 (MIRIADE); the European Union Joint Programme – Neurodegenera‑ AD CSF, this increase is more noticeable in certain gly- tive Disease Research (JPND2021-00694); and the UK Dementia Research coforms of monomers and aberrant complexes that may Institute at UCL (UKDRI-1003). KB is supported by the Swedish Research hinder its biological activity. A specific description of Council (#2017-00915); the Alzheimer Drug Discovery Foundation (ADDF), USA (#RDAPB-201809-2016615); the Swedish Alzheimer Foundation (#AF- how these species affect apoE signaling and Aβ clearance 742881); Hjärnfonden, Sweden (#FO2017-0243); the Swedish state under the should improve our understanding of the role of apoE in agreement between the Swedish government and the County Councils, the the AD pathology. ALF-agreement (#ALFGBG-715986); the European Union Joint Program for Neurodegenerative Disorders (JPND2019-466-236); the National Institute of Health (NIH), USA (grant #1R01AG068398-01); and the Alzheimer’s Association 2021 Zenith Award (ZEN-21-848495). Abbreviations Aβ: Amyloid-beta; AD: Alzheimer’s disease; APP: Amyloid precursor protein; Availability of data and materials ApoE: Apolipoprotein E; CSF: Cerebrospinal fluid; CNS: Central nervous system; The datasets used and/or analyzed during the current study are available from HDL: High-density lipoprotein; MS: Mass spectrometry; SDS: Sodium dodecyl the corresponding author on reasonable request. sulfate; PAGE: Polyacrylamide gel electrophoresis; Tg: Transgenic; Wt: Wild-type. Declarations Supplementary Information The online version contains supplementary material available at https://​doi.​ Ethics approval and consent to participate org/​10.​1186/​s13195-​022-​01108-2. This study was approved by the ethics committee at the Miguel Hernandez University (ref# UMH.iNJS.01.18), and it was carried out in accordance with the Additional file 1. Images of complete blots from which figures were Helsinki Declaration regarding research on humans. The present assays were obtained, and also the boxes selected with the ImageQuant Studio soft‑ performed on de-identified left-over aliquots from clinical diagnostic CSF ware for quantification. samples; thus, no consent for participation was required. The samples were obtained following procedures approved by the Ethics Committees at the University of Gothenburg and the Hospital Sant Pau, respectively. Animal work Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 17 of 19 was performed in accordance with the local legislation, with the approval of Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​35041​991/. [Cited 2022 the Experimental Animal Ethical Committee of the University of Barcelona, Feb 22]. and in compliance with European legislation. 4. Rall SC, Weisgraber KH, Innerarity TL, Mahley RW. Structural basis for receptor binding heterogeneity of apolipoprotein E from type III hyper‑ Consent for publication lipoproteinemic subjects. Proc Natl Acad Sci U S A. 1982;79:4696–700 Not applicable. Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​62893​14/. [Cited 2022 Feb 22]. Competing interests 5. Kockx M, Traini M, Kritharides L. Cell-specific production, secretion, and HZ has served at scientific advisory boards and/or as a consultant for Abbvie, function of apolipoprotein E. J Mol Med (Berl). 2018;96:361–71 Available Alector, Annexon, Artery Therapeutics, AZTherapies, CogRx, Denali, Eisai, from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​29516​132/. [Cited 2022 Feb 22]. Nervgen, Novo Nordisk, Pinteon Therapeutics, Red Abbey Labs, Passage Bio, 6. McIntosh AM, Bennett C, Dickson D, Anestis SF, Watts DP, Webster TH, Roche, Samumed, Siemens Healthineers, Triplet Therapeutics, and Wave; has et al. The apolipoprotein E (APOE) gene appears functionally monomor‑ given lectures in symposia sponsored by Cellectricon, Fujirebio, Alzecure, phic in chimpanzees (Pan troglodytes). PLoS One. 2012;7 Available from: Biogen, and Roche; and is a co-founder of Brain Biomarker Solutions in https://​pubmed.​ncbi.​nlm.​nih.​gov/​23112​842/. [Cited 2022 Feb 22]. Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program 7. Heffernan AL, Chidgey C, Peng P, Masters CL, Roberts BR. The neurobiol‑ (outside submitted work). KB has served as a consultant, at advisory boards, ogy and age-related prevalence of the ε4 allele of apolipoprotein E in or at data monitoring committees for Abcam, Axon, Biogen, JOMDD/Shi‑ Alzheimer’s disease cohorts. J Mol Neurosci. 2016;60:316–24 Available madzu, Julius Clinical, Lilly, MagQu, Novartis, Prothena, Roche Diagnostics, from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​27498​201/. [Cited 2022 Feb 22]. and Siemens Healthineers and is a co-founder of Brain Biomarker Solutions in 8. Roses AD. Apolipoprotein E alleles as risk factors in Alzheimer’s disease. Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program, Annu Rev Med. 1996;47:387–400 Available from: https://​pubmed.​ncbi.​ all unrelated to the work presented in this paper. JF has served as a consultant nlm.​nih.​gov/​87127​90/. [Cited 2022 Feb 22]. for Novartis and Lundbeck; has received honoraria for lectures from Roche, 9. Uddin MS, Kabir MT, Al Mamun A, Abdel-Daim MM, Barreto GE, Ashraf NovoNordisk, Nestle, Esteve, and Biogen; and served at advisory boards for AC GM. APOE and Alzheimer’s disease: evidence mounts that target‑ Immune, Zambon, and Lundbeck. D.A. participated in advisory boards from ing APOE4 may combat Alzheimer’s pathogenesis. Mol Neurobiol. Fujirebio-Europe and Roche Diagnostics and received speaker honoraria from 2019;56:2450–65. Fujirebio-Europe, Roche Diagnostics, Nutricia, Krka Farmacéutica S.L., Zambon 10. Corder EH, Saunders AM, Risch NJ, Strittmatter WJ, Schmechel DE, Gaskell S.A.U., and Esteve Pharmaceuticals S.A. AL has served at scientific advisory PC, et al. Protective effect of apolipoprotein E type 2 allele for late onset boards from Fujirebio-Europe, Nutricia, Roche-Genentech, Biogen, Grifols, and Alzheimer disease. Nat Genet. 1994;7:180–4. Roche Diagnostics and has filed a patent application of synaptic markers in 11. Li Z, Shue F, Zhao N, Shinohara M, Bu G. APOE2: protective mechanism neurodegenerative diseases. and therapeutic implications for Alzheimer’s disease. Mol Neurodegener. 2020;15:63. Author details 12. Reiman EM, Arboleda-Velasquez JF, Quiroz YT, Huentelman MJ, Beach TG, 1 Instituto de Neurociencias de Alicante, Universidad Miguel Hernández- Caselli RJ, et al. Exceptionally low likelihood of Alzheimer’s dementia in CSIC, Av. Ramón y Cajal s/n, E‑03550 Sant Joan d’Alacant, Spain. 2 Centro de APOE2 homozygotes from a 5,000-person neuropathological study. Nat Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas Commun. 2020;11:667. (CIBERNED), Sant Joan d’Alacant, Spain. 3 Department of Cell Biology, Physiol‑ 13. Saddiki H, Fayosse A, Cognat E, Sabia S, Engelborghs S, Wallon D, et al. ogy and Immunology, Faculty of Biology, University of Barcelona, Barcelona, Age and the association between apolipoprotein E genotype and Alzhei‑ Spain. 4 Institute of Neurosciences, University of Barcelona, Barcelona, Spain. mer disease: a cerebrospinal fluid biomarker-based case-control study. 5 Instituto de Investigación Sanitaria y Biomédica de Alicante (ISABIAL), PLoS Med. 2020;17 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​ Alicante, Spain. 6 Institute of Neuroscience and Physiology, Department 32817​639/. [Cited 2022 Feb 22]. of Psychiatry and Neurochemistry, the Sahlgrenska Academy at the University 14. Martens YA, Zhao N, Liu C-C, Kanekiyo T, Yang AJ, Goate AM, et al. ApoE of Gothenburg, Mölndal, Sweden. 7 Sant Pau Memory Unit, Neurology Depart‑ Cascade Hypothesis in the pathogenesis of Alzheimer’s disease and ment, Hospital de la Santa Creu i Sant Pau, Biomedical Research Institute related dementias. Neuron. 2022;110(8):1304–17. Sant Pau, Universitat Autònoma de Barcelona, Barcelona, Spain. 8 Barcelona 15. Dyer CA, Smith RS, Curtiss LK. Only multimers of a synthetic pep‑ Down Medical Center, Fundació Catalana Síndrome de Down, Barcelona, tide of human apolipoprotein E are biologically active. J Biol Chem. Spain. 9 Laboratory of Surgical Neuroanatomy, School of Medicine and Health 1991;266:15009–15. Sciences, University of Barcelona, Barcelona, Spain. 10 Clinical Neurochemistry 16. Weisgraber KH, Shinto LH. Identification of the disulfide-linked homodi‑ Laboratory, Sahlgrenska University Hospital, Mölndal, Sweden. 11 Department mer of apolipoprotein E3 in plasma. Impact on receptor binding activity. of Neurodegenerative Disease, Institute of Neurology, University College J Biol Chem. 1991;266:12029–34. London, London, UK. 12 UK Dementia Research Institute at UCL, London, UK. 17. Rebeck GW, Alonzo NC, Berezovska O, Harr SD, Knowles RB, Growdon JH, 13 Hong Kong Center for Neurodegenerative Diseases, Hong Kong, China. et al. Structure and functions of human cerebrospinal fluid lipoproteins from individuals of different APOE genotypes. Exp Neurol. 1998;149:175– Received: 3 May 2022 Accepted: 16 October 2022 82 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​94546​26/. [Cited 2022 Feb 22]. 18. Minagawa H, Gong JS, Jung CG, Watanabe A, Lund-Katz S, Phillips MC, et al. Mechanism underlying apolipoprotein E (ApoE) isoform-dependent lipid efflux from neural cells in culture. J Neurosci Res. 2009;87:2498–508 References Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​19326​444/. [Cited 2022 1. Strittmatter WJ, Weisgraber KH, Huang DY, Dong LM, Salvesen GS, Pericak- Feb 22]. Vance M, et al. Binding of human apolipoprotein E to synthetic amyloid 19. Elliott DA, Halliday GM, Garner B. Apolipoprotein-E forms dimers in beta peptide: isoform-specific effects and implications for late-onset human frontal cortex and hippocampus. BMC Neurosci. 2010;11 Avail‑ Alzheimer disease. Proc Natl Acad Sci U S A. 1993;90:8098–102 Available able from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​20170​526/. [Cited 2022 Feb from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​83674​70/. [Cited 2022 Feb 22]. 22]. 2. Husain MA, Laurent B, Plourde M. APOE and Alzheimer’s disease: from 20. Cruchaga C, Kauwe JSK, Nowotny P, Bales K, Pickering EH, Mayo K, et al. lipid transport to physiopathology and therapeutics. Front Neurosci. Cerebrospinal fluid APOE levels: an endophenotype for genetic studies 2021;15 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​33679​311/. for Alzheimer’s disease. Hum Mol Genet. 2012;21:4558–71 Available from: [Cited 2022 Feb 22]. https://​pubmed.​ncbi.​nlm.​nih.​gov/​22821​396/. [Cited 2022 Feb 22]. 3. Moon HJ, Haroutunian V, Zhao L. Human apolipoprotein E isoforms 21. Martínez-Morillo E, Hansson O, Atagi Y, Bu G, Minthon L, Diamandis EP, are differentially sialylated and the sialic acid moiety in ApoE2 attenu‑ et al. Total apolipoprotein E levels and specific isoform composition in ates ApoE2-Aβ interaction and Aβ fibrillation. Neurobiol Dis. 2022;164 cerebrospinal fluid and plasma from Alzheimer’s disease patients and Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 18 of 19 controls. Acta Neuropathol. 2014;127:633–43 Available from: https://​ 38. Hu Y, Meuret C, Go S, Yassine HN, Nedelkov D. Simple and fast assay for pubmed.​ncbi.​nlm.​nih.​gov/​24633​805/. [Cited 2022 Feb 22]. apolipoprotein E phenotyping and glycotyping: discovering isoform- 22. Minta K, Brinkmalm G, Janelidze S, Sjödin S, Portelius E, Stomrud E, et al. specific glycosylation in plasma and cerebrospinal fluid. J Alzheimers Dis. Quantification of total apolipoprotein E and its isoforms in cerebrospinal 2020;76:883–93 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​32568​ fluid from patients with neurodegenerative diseases. Alzheimers Res Ther. 201/. [Cited 2022 Feb 22]. 2020;12 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​32054​532/. 39. Lanfranco MF, Sepulveda J, Kopetsky G, Rebeck GW. Expression and secre‑ [Cited 2022 Feb 22]. tion of apoE isoforms in astrocytes and microglia during inflammation. 23. van Harten AC, Jongbloed W, Teunissen CE, Scheltens P, Veerhuis R, van Glia. 2021;69:1478–93 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​ der Flier WM. CSF ApoE predicts clinical progression in nondemented 33556​209/. [Cited 2022 Feb 22]. APOEε4 carriers. Neurobiol Aging. 2017;57:186–94 Available from: https://​ 40. Wernette-Hammond ME, Lauer SJ, Corsini A, Walker D, Taylor JM, Rall SC. pubmed.​ncbi.​nlm.​nih.​gov/​28571​653/. [Cited 2022 Feb 22]. Glycosylation of human apolipoprotein E. The carbohydrate attachment 24. del Campo M, Mollenhauer B, Bertolotto A, Engelborghs S, Hampel H, site is threonine 194. J Biol Chem. 1989;264:9094–101. Simonsen AH, et al. Recommendations to standardize preanalytical 41. Flowers SA, Grant OC, Woods RJ, Rebeck GW. O-glycosylation on cer‑ confounding factors in Alzheimer’s and Parkinson’s disease cerebrospinal ebrospinal fluid and plasma apolipoprotein E differs in the lipid-binding fluid biomarkers: an update. Biomark Med. 2012;6:419–30. domain. Glycobiology. 2020;30:74–85 Available from: https://​pubmed.​ 25. Andreasen N, Blennow K, Sjödin C, Winblad B, Svärdsudd K. Prevalence ncbi.​nlm.​nih.​gov/​31616​924/. [Cited 2022 Feb 22]. and incidence of clinically diagnosed memory impairments in a geo‑ 42. Ke LY, Chan HC, Chen CC, Chang CF, Lu PL, Chu CS, et al. Increased APOE graphically defined general population in Sweden. The Piteå Dementia glycosylation plays a key role in the atherogenicity of L5 low-density Project. Neuroepidemiology. 1999;18:144–55. lipoprotein. FASEB J. 2020;34:9802–13 Available from: https://​pubmed.​ 26. Andreasen N, Minthon L, Davidsson P, Vanmechelen E, Vanderstichele ncbi.​nlm.​nih.​gov/​32501​643/. [Cited 2022 Feb 22]. H, Winblad B, et al. Evaluation of CSF-tau and CSF-Abeta42 as diag‑ 43. Kacperczyk M, Kmieciak A, Kratz EM. The role of ApoE expression and nostic markers for Alzheimer disease in clinical practice. Arch Neurol. variability of its glycosylation in human reproductive health in the light 2001;58:373–9. of current information. Int J Mol Sci. 2021;22 Available from: https://​ 27. Blennow K, Ricksten A, Prince JA, Brookes AJ, Emahazion T, Wasslavik C, pubmed.​ncbi.​nlm.​nih.​gov/​34281​251/. [Cited 2022 Feb 22]. et al. No association between the alpha2-macroglobulin (A2M) deletion 44. Sugano M, Yamauchi K, Kawasaki K, Tozuka M, Fujita K, Okumura N, et al. and Alzheimer’s disease, and no change in A2M mRNA, protein, or pro‑ Sialic acid moiety of apolipoprotein E3 at Thr(194) affects its interaction tein expression. J Neural Transm (Vienna). 2000;107:1065–79. with beta-amyloid(1-42) peptides. Clin Chim Acta. 2008;388:123–9 Avail‑ 28. Hansson O, Zetterberg H, Buchhave P, Londos E, Blennow K, Minthon L. able from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​18023​277/. [Cited 2022 Feb Association between CSF biomarkers and incipient Alzheimer’s disease 22]. in patients with mild cognitive impairment: a follow-up study. Lancet 45. Chua CC, Lim ML, Wong BS. Altered apolipoprotein E glycosylation is Neurol. 2006;5(3):228–34. S1474442206703556. https://​doi.​org/​10.​1016/​ associated with Abeta(42) accumulation in an animal model of Niemann- S1474-​4422(06)​70355-6. Pick Type C disease. J Neurochem. 2010;112:1619–26 Available from: 29. Alcolea D, Clarimón J, Carmona-Iragui M, Illán-Gala I, Morenas-Rodríguez https://​pubmed.​ncbi.​nlm.​nih.​gov/​20070​866/. [Cited 2022 Feb 22]. E, Barroeta I, et al. The Sant Pau Initiative on Neurodegeneration (SPIN) 46. Lee Y, Kockx M, Raftery MJ, Jessup W, Griffith R, Kritharides L. Glycosyla‑ cohort: a data set for biomarker discovery and validation in neurodegen‑ tion and sialylation of macrophage-derived human apolipoprotein E erative disorders. Alzheimers Dement (N Y). 2019;5:597–609. analyzed by SDS-PAGE and mass spectrometry: evidence for a novel site 30. Jack CR, Bennett DA, Blennow K, Carrillo MC, Dunn B, Haeberlein SB, et al. of glycosylation on Ser290. Mol Cell Proteomics. 2010;9:1968–81 Available NIA-AA Research Framework: toward a biological definition of Alzhei‑ from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​20511​397/. [Cited 2022 Feb 22]. mer’s disease. Alzheimers Dement. 2018;14:535–62. 47. Martel CL, Mackic JB, Matsubara E, Governale S, Miguel C, Miao W, et al. 31. Cohen RM, Rezai-Zadeh K, Weitz TM, Rentsendorj A, Gate D, Spivak I, Isoform-specific effects of apolipoproteins E2, E3, and E4 on cerebral et al. A transgenic Alzheimer rat with plaques, tau pathology, behav‑ capillary sequestration and blood-brain barrier transport of circulating ioral impairment, oligomeric aβ, and frank neuronal loss. J Neurosci. Alzheimer’s amyloid beta. J Neurochem. 1997;69:1995–2004 Available 2013;33:6245–56 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​23575​ from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​93495​44/. [Cited 2022 Feb 22]. 824/. [Cited 2022 Feb 22]. 48. Montine KS, Bassett CN, Ou JJ, Markesbery WR, Swift LL, Montine TJ. 32. Sáez-Valero J, Pérez De Gracia JA, Lockridge O. Intraperitoneal admin‑ Apolipoprotein E allelic influence on human cerebrospinal fluid apolipo‑ istration of 340 kDa human plasma butyrylcholinesterase increases proteins. J Lipid Res. 1998;39:2443–51. the level of the enzyme in the cerebrospinal fluid of rats. Neurosci Lett. 49. Patra K, Giannisis A, Edlund AK, Sando SB, Lauridsen C, Berge G, et al. 2005;383:93–8 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​15936​ Plasma apolipoprotein E monomer and dimer profile and relevance to 518/. [Cited 2022 Feb 22]. Alzheimer’s disease. J Alzheimers Dis. 2019;71:1217–31 Available from: 33. Camporesi E, Lashley T, Gobom J, Lantero-Rodriguez J, Hansson O, https://​pubmed.​ncbi.​nlm.​nih.​gov/​31524​156/. [Cited 2022 Feb 22]. Zetterberg H, et al. Neuroligin-1 in brain and CSF of neurodegenerative 50. Laffoon SB, Doecke JD, Roberts AM, Vance JA, Reeves BD, Pertile KK, et al. disorders: investigation for synaptic biomarkers. Acta Neuropathol Com‑ Analysis of plasma proteins using 2D gels and novel fluorescent probes: mun. 2021;9:19. in search of blood based biomarkers for Alzheimer’s disease. Proteome 34. Chaney AM, Lopez-Picon FR, Serrière S, Wang R, Bochicchio D, Webb SD, Sci. 2022;20:2 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​35081​ et al. Prodromal neuroinflammatory, cholinergic and metabolite dysfunc‑ 972/. [Cited 2022 Feb 22]. tion detected by PET and MRS in the TgF344-AD transgenic rat model of 51. Liu M, Kuhel DG, Shen L, Hui DY, Woods SC. Apolipoprotein E does not AD: a collaborative multi-modal study. Theranostics. 2021;11:6644–67. cross the blood-cerebrospinal fluid barrier, as revealed by an improved 35. Heywood WE, Galimberti D, Bliss E, Sirka E, Paterson RW, Magdalinou NK, technique for sampling CSF from mice. Am J Phys Regul Integr Comp et al. Identification of novel CSF biomarkers for neurodegeneration and Phys. 2012;303 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​22933​ their validation by a high-throughput multiplexed targeted proteomic 021/. [Cited 2022 Feb 22]. assay. Mol Neurodegener. 2015;10 Available from: https://​pubmed.​ncbi.​ 52. Gutman CR, Strittmatter WJ, Weisgraber KH, Matthew WD. Apolipoprotein nlm.​nih.​gov/​26627​638/. [Cited 2022 Feb 22]. E binds to and potentiates the biological activity of ciliary neurotrophic 36. Lindh M, Blomberg M, Jensen M, Basun H, Lannfelt L, Engvall B, et al. factor. J Neurosci. 1997;17:6114–21 Available from: https://​pubmed.​ncbi.​ Cerebrospinal fluid apolipoprotein E (apoE) levels in Alzheimer’s disease nlm.​nih.​gov/​92362​23/. [Cited 2022 Feb 22]. patients are increased at follow up and show a correlation with levels of 53. Haas C, Cazorla P, de Miguel C, Valdivieso F, Vázquez J. Apolipoprotein tau protein. Neurosci Lett. 1997;229:85–8 Available from: https://​pubmed.​ E forms stable complexes with recombinant Alzheimer’s disease beta- ncbi.​nlm.​nih.​gov/​92235​97/. [Cited 2022 Feb 22]. amyloid precursor protein. Biochem J. 1997;325(Pt 1):169–75 Available 37. Halim A, Rüetschi U, Larson G, Nilsson J. LC-MS/MS characterization of from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​92246​43/. [Cited 2022 Feb 22]. O-glycosylation sites and glycan structures of human cerebrospinal fluid 54. Wisniewski T, Golabek A, Matsubara E, Ghiso J, Frangione B. Apolipopro‑ glycoproteins. J Proteome Res. J Proteome Res. 2013;12:573–84 Available tein E: binding to soluble Alzheimer’s beta-amyloid. Biochem Biophys Res from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​23234​360/. [Cited 2022 Feb 22]. Lennol et al. Alzheimer’s Research & Therapy (2022) 14:161 Page 19 of 19 Commun. 1993;192:359–65 Available from: https://​pubmed.​ncbi.​nlm.​nih.​ gov/​84847​48/. [Cited 2022 Feb 22]. 55. Ladu MJ, Munson GW, Jungbauer L, Getz GS, Reardon CA, Tai LM, et al. Preferential interactions between ApoE-containing lipoproteins and Aβ revealed by a detection method that combines size exclusion chromatography with non-reducing gel-shift. Biochim Biophys Acta. 2012;1821:295–302 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​ 22138​302/. [Cited 2022 Feb 22]. 56. Deroo S, Stengel F, Mohammadi A, Henry N, Hubin E, Krammer EM, et al. Chemical cross-linking/mass spectrometry maps the amyloid β peptide binding region on both apolipoprotein E domains. ACS Chem Biol. 2015;10:1010–6 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​25546​ 376/. [Cited 2022 Feb 22]. 57. Näslund J, Thyberg J, Tjernberg LO, Wernstedt C, Karlström AR, Bogda‑ novic N, et al. Characterization of stable complexes involving apolipopro‑ tein E and the amyloid beta peptide in Alzheimer’s disease brain. Neuron. 1995;15:219–28 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​76195​ 25/. [Cited 2022 Feb 22]. 58. Permanne B, Perez C, Soto C, Frangione B, Wisniewski T. Detection of apolipoprotein E/dimeric soluble amyloid beta complexes in Alzhei‑ mer’s disease brain supernatants. Biochem Biophys Res Commun. 1997;240:715–20 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​93986​ 32/. [Cited 2022 Feb 22]. 59. Russo C, Angelini G, Dapino D, Piccini A, Piombo G, Schettini G, et al. Opposite roles of apolipoprotein E in normal brains and in Alzheimer’s disease. Proc Natl Acad Sci U S A. 1998;95:15598–602 Available from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​98610​15/. [Cited 2022 Feb 22]. 60. Padayachee ER, Zetterberg H, Portelius E, Borén J, Molinuevo JL, Andreasen N, et al. Cerebrospinal fluid-induced retardation of amyloid β aggregation correlates with Alzheimer’s disease and the APOE ε4 allele. Brain Res. 2016;1651:11–6 Available from: https://​pubmed.​ncbi.​nlm.​nih.​ gov/​27653​981/. [Cited 2022 Feb 22]. 61. Verghese PB, Castellano JM, Garai K, Wang Y, Jiang H, Shah A, et al. ApoE influences amyloid-β (Aβ) clearance despite minimal apoE/Aβ associa‑ tion in physiological conditions. Proc Natl Acad Sci U S A. 2013;110 Avail‑ able from: https://​pubmed.​ncbi.​nlm.​nih.​gov/​23620​513/. [Cited 2022 Feb 22]. 62. Aleshkov S, Abraham CR, Zannis VI. Interaction of nascent ApoE2, ApoE3, and ApoE4 isoforms expressed in mammalian cells with amyloid peptide beta (1-40). Relevance to Alzheimer’s disease. Biochemistry. 1997;36:10571–80. 63. Sanan DA, Weisgraber KH, Russell SJ, Mahley RW, Huang D, Saunders A, et al. Apolipoprotein E associates with beta amyloid peptide of Alzhei‑ mer’s disease to form novel monofibrils. Isoform apoE4 associates more efficiently than apoE3. J Clin Invest. 1994;94:860–9. 64. Kanekiyo T, Xu H, Bu G. ApoE and Aβ in Alzheimer’s disease: accidental encounters or partners? Neuron. 2014;81:740–54. 65. Hatters DM, Peters-Libeu CA, Weisgraber KH. Apolipoprotein E structure: insights into function. Trends Biochem Sci. 2006;31:445–54. 66. Hubin E, Verghese PB, van Nuland N, Broersen K. Apolipoprotein E associated with reconstituted high-density lipoprotein-like particles is protected from aggregation. FEBS Lett. 2019;593:1144–53. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions