A Nanobody-Based Toolbox to Probe ApoE4 in the Secretory Pathway and Cytosol

Abstract

Apolipoprotein E4 (apoE4) is the strongest genetic risk factor for late-onset Alzheimer’s disease (AD). Yet the molecular mechanisms underlying its contribution to AD remain to be fully elucidated. Here, we developed and characterized a set of apoE-specific single-domain antibodies (nanobodies) as a molecular toolbox to investigate intracellular apoE4. The nanobodies bind human apoE with nanomolar to sub-nanomolar affinity and recognize both apoE3 and apoE4. Domain-level epitope mapping revealed nanobodies that selectively bind either an N-terminal (residues 1–173) or C-terminal (residues 170–299) apoE4 fragment. Several nanobodies were validated as endoplasmic reticulum-targeted intrabodies that bind apoE4 intracellularly and promote its intracellular retention. These nanobodies constitute a versatile toolbox for probing and manipulating apoE4 in cellular models. As an exploratory application of this nanobody toolbox, we examined cytosolic apoE4, motivated by previous studies suggesting that cytosolic apoE4 fragments may influence AD-related cellular processes. Cytosolic expression of apoE4 resulted in perinuclear protein assemblies and the appearance of a ~25 kDa apoE4 fragment. Using a nanobody-based nuclear relocalization assay, we showed that cytosolic apoE4 remains accessible for nanobody binding but was not relocated to the nucleus by a nuclear localization signal-equipped nanobody. Altogether, this study introduces a nanobody-based toolbox to investigate apoE4 in distinct intracellular contexts, which can be relevant to AD.

1. Introduction

Apolipoprotein E (apoE) is a 299-amino-acid glycoprotein of 34 kDa and the major lipid and cholesterol carrier in the central nervous system [1,2]. Structurally, apoE consists of two main domains connected by a hinge region: an N-terminal domain which harbors the receptor-binding region and the C-terminal domain which contains the lipid-binding region [3]. In the brain, apoE is produced by multiple cell types and is normally secreted. It associates with lipid molecules via its lipid-binding region, while it interacts with members of the low-density lipoprotein (LDL) receptor family via its receptor-binding region to mediate cellular uptake of lipids and cholesterol [4]. The human APOE gene has three common and only slightly divergent alleles, APOE-ε2, APOE-ε3 and APOE-ε4, which encode the apoE2, apoE3 and apoE4 isoforms, respectively [5].

The APOE-ε4 allele is the strongest genetic risk factor for late-onset Alzheimer’s disease (AD), with homozygous carriers of the APOE-ε4 allele having up to a 15-fold greater risk of developing AD than non-carriers [5]. ApoE4 has been implicated in multiple AD-related pathological processes, including amyloid-β (Aβ) accumulation, tau pathology, neuroinflammation, gliosis and neurodegeneration [2,5]. Owing to its central role in AD, apoE4 is considered a potential therapeutic target [5,6,7,8,9]. Besides AD, apoE4 is also associated with Parkinson’s disease, Lewy body diseases, multiple sclerosis, cardiovascular diseases, poor recovery after brain injury and metabolic disorders such as hyperlipidemia and type 2 diabetes [2,6,7].

Despite its involvement in multiple AD-related pathologies, the molecular mechanisms underlying the diverse effects of apoE4 on AD pathology remain incompletely understood [2,10]. Many studies have focused on manipulating APOE expression or genotype [10,11,12,13,14], whereas approaches that directly interrogate apoE4 at the protein level are limited. Molecular tools that enable the investigation of apoE4 biology at the protein level are essential to bridge the gap between genetic associations and mechanistic insight.

Single-domain antibodies, also known as nanobodies (Nbs or VHHs), are versatile molecular tools for studying proteins in disease contexts [15,16,17,18,19]. They are derived from camelid heavy-chain antibodies, consisting of a single antigen-binding domain and possess beneficial properties such as small size (15 kDa), high affinity and stability [17]. Importantly, nanobodies can function intracellularly as intrabodies, allowing direct manipulation of target proteins in living cells [20]. Depending on their design, intrabodies can be used to redirect proteins to defined cellular compartments, modulate protein function, or promote degradation of target proteins [17,20]. These properties make nanobodies particularly well-suited to investigate apoE4 in a cellular context.

Here, we developed and characterized a set of apoE-targeting nanobodies as a novel toolbox. We identified high-affinity binders, and domain-level epitope mapping revealed nanobodies that selectively recognize either an N-terminal or C-terminal apoE4 fragment. Four nanobodies were validated as functional endoplasmic reticulum (ER)-targeted intrabodies, enabling interaction with apoE4 in the secretory pathway, where apoE4 normally resides. This toolbox provides a foundation for future nanobody-based approaches to further dissect apoE4 biology in disease-relevant cellular models.

As an exploratory application of this nanobody toolbox, we investigated cytosolic apoE4. Studies by other researchers have reported that apoE4 fragments interfere with intracellular processes such as mitochondrial function and cytoskeletal organization [6,21,22,23], providing a rationale to examine apoE4 in the cytosolic compartment. Using our nanobody toolbox, we explored the intracellular localization and accessibility of cytosolic apoE4 in mammalian cells.

Altogether, we report the development of a nanobody-based toolbox to investigate (endogenous) apoE4 at the protein level. We further illustrate the use of this toolbox in exploring the behavior of cytosolic apoE4.

2. Materials and Methods

2.1. Immunization, Nanobody Library Generation and Biopanning

The immunization and nanobody selection were performed as described by Vincke et al. [24]. Briefly, three llamas were immunized by four intramuscular injections (days 0, 14, 28 and 35) containing 250 µg of either human recombinant apoE4 (Peprotech, Cranbury, NJ, USA, cat. no. 350-04), human recombinant apoE3 (Peprotech, Cranbury, NJ, USA, cat. no. 350-02), synthetic apoE4 peptide (aa82–131) coupled to the keyhole limpet hemocyanin (KLH) carrier protein (LTPVAEETRARLSKELQAAQARLGADMEDVRGRLVQYRGEVQAMLGQSTEC(KLH)E-NH₂, Caslo, Kongens Lyngby, Denmark) or a combination of the KLH-conjugated synthetic apoE4 peptide and human recombinant apoE4. The llama immunized with recombinant apoE3 was also immunized with the KLH-conjugated apoE4 peptide (aa82–131), but at different periods in time. VHH phage display libraries were constructed from peripheral blood lymphocytes by cloning PCR-amplified VHH fragments into the pMECS phagemid vector. These libraries were subjected to three rounds of phage display biopanning against either recombinant apoE3 (library from apoE3 immunization) or apoE4 (library from apoE4, peptide and apoE4 + peptide immunization).

2.2. Plasmid Construction

Cloning was performed using the In-Fusion HD Cloning Kit (Takara Bio Inc., Shiga, Japan, cat. no. 639650) according to the manufacturer’s instructions, except for the cloning of the nanobody sequences into the pMECS vector, which was executed with T4 DNA ligase (Invitrogen, Carlsbad, CA, USA, cat. no. 15224-041). Correct insertion of each construct was confirmed by DNA sequencing (Mix2Seq kit, Eurofins Genomics, Ebersberg, Germany, cat. no. 3094-0ONMSK).

2.2.1. Bacterial Expression Vectors

Nanobody sequences were cloned into the pMECS vector (as mentioned in Section 2.1) with a C-terminal HA and His₆-tag (pMECS-Nb-HA-His₆). The apoE4 N-terminal fragment containing amino acids (aa) 1 to 173 (apoE4(aa1–173)) was cloned into the pTYB12 vector, resulting in an intein tag fusion protein that allows purification using the Intein-Mediated Purification with an Affinity Chitin-binding Tag (IMPACT) system (pTYB12-apoE4(aa1–173)). The apoE4 C-terminal fragment (apoE4(aa170–299)), containing a C-terminal His₆-tag was cloned into the pMAL-c2X vector to generate a maltose-binding protein (MBP) fusion construct (pMAL-c2X-apoE4(aa170–299)-His₆).

2.2.2. Mammalian Expression Vectors

The pCMV4-ApoE4 plasmid, encoding apoE4 with an N-terminal 18-amino-acid secretion signal peptide (SEC1), was obtained from Addgene (Watertown, MA, USA, cat. no. 87087). The sequences of nanobodies 9–74, 17–69, 18–91 and 19–38, as well as a nanobody targeting green fluorescent protein (GFP) (Gulliver Biomed, Ghent, Belgium, cat. no. sdAb-GFP-Nb94), each carrying a C-terminal HA tag, were cloned into the pCMV/myc/ER vector (Addgene, Watertown, MA, USA, not available anymore). Upon expression, this vector directs proteins to the ER via an N-terminal mouse IgH signal sequence (SEC2) and contains a C-terminal myc-tag (will not be used in this study) as well as a C-terminal KDEL ER retention sequence (pCMV/myc/ER-Nb-HA). A DNA fragment containing the nuclear localization signal (NLS) of the SV40 large T antigen (PKKKRKV), followed by mCherry with two restriction sites in between NLS and mCherry was synthesized (Integrated DNA Technologies, Coralville, IA, USA) and cloned into the pcDNA3.1 vector. The vector was subsequently digested at the two restriction sites, and the nanobody sequence of Nb 17–69 (with HA tag) was inserted, yielding the construct pcDNA3.1-NLS-Nb-HA-mCherry. The sequence of apoE4 lacking the 18-amino-acid N-terminal signal peptide was amplified by PCR from the pCMV4-ApoE4 vector as template and cloned into the pEGFP-N1 vector (pEGFP-N1-apoE4(ΔSEC1)), yielding an apoE4(ΔSEC1)-EGFP fusion protein and into the pcDNA3.1 vector (pcDNA3.1-apoE4(ΔSEC1)).

2.3. Protein Expression and Purification

2.3.1. Nanobody

pMECS plasmids containing nanobody cDNAs were heat-shock transformed into E. coli strain WK6 and plated on LB agar with 50 µg/mL ampicillin. A single colony was inoculated in 5 mL LB medium containing 50 µg/mL ampicillin and grown overnight at 37 °C with shaking. The culture was diluted 1:100 in Terrific Broth (TB) medium and incubated at 37 °C with shaking until an OD₆₀₀ of 0.6–0.8 was reached. Protein expression was induced by adding isopropyl-β-D-1-thiogalactopyranoside (IPTG) at a final concentration of 1 mM, and the culture was incubated overnight at 28 °C with shaking. Nanobodies in the pMECS vector are expressed with an N-terminal pelB signal peptide, directing them to the periplasm. Cells were harvested by centrifugation (15 min, 4 °C, 6000 rpm), resuspended in TES buffer (200 mM Tris-HCl, 0.5 mM EDTA, 500 mM sucrose, 2 µg/mL benzamidine, 2 µg/mL leupeptin, 2 µg/mL aprotinin and 1 mM PMSF, pH 8) and incubated on ice for at least 2 h with gentle agitation. A double volume of 4× diluted TES buffer was added, and incubation was continued overnight (on ice, gentle agitation). Via this osmotic shock, periplasmic proteins were released. Cell debris was pelleted by centrifugation (10 min, 4 °C, 14,000 rpm), and the resulting supernatant (periplasmic extract) was either used directly in downstream assays or subjected to purification. When required, His₆-tagged nanobodies were purified from the supernatant by immobilized metal affinity chromatography (IMAC) using Ni High performance Sepharose Beads (GE Healthcare, Chicago, IL, USA, cat. no. 17-5268-01) according to the manufacturer’s instructions. Purity was assessed via SDS-PAGE and Coomassie staining. Purified nanobodies were dialyzed to 20 mM Tris-HCl and 150 mM NaCl buffer (pH 7.5) using Spectrum Spectra/Por 1 RC Dialysis Membrane Tubing 6000 to 8000 Dalton MWCO (Spectrum Laboratories, Rancho Dominguez, CA, USA, cat. no. 132645).

2.3.2. ApoE4(aa1–173)

The pTYB12-apoE4(aa1–173) construct was heat-shock transformed into E. coli strain BL21. Cultivation and induction were performed as described in Section 2.3.1, except that expression was induced with IPTG (1 mM) once the culture reached an OD₆₀₀ > 2 and expression proceeded overnight at 20 °C. Cells were collected by centrifugation (15 min, 4 °C, 6000 rpm), resuspended in chitin column buffer (1 mM EDTA, 50 mM HEPES, 500 mM NaCl, 2 µg/mL benzamidine, 2 µg/mL leupeptin, 2 µg/mL aprotinin and 1 mM PMSF, pH 8), and lysed using a French press followed by sonication. After removal of cell debris by centrifugation (30 min, 4 °C, 10,000 rpm), the supernatant containing the intein tag fusion protein was applied to chitin beads (New England BioLabs, Ipswich, MA, USA, cat. no. S6651S) according to the manufacturer’s instructions for IMPACT purification. Purity was assessed via SDS-PAGE and Coomassie staining.

2.3.3. MBP-ApoE4(aa170–299)

The pMAL-c2X-apoE4(aa170–299) construct was heat-shock transformed into E. coli strain BL21. Expression conditions were as described in Section 2.3.1, except induction was performed at an OD₆₀₀ of ~0.5 with IPTG (1 mM), followed by incubation for 3–5 h at 37 °C. Cells were harvested by centrifugation (15 min, 4 °C, 6000 rpm), resuspended in Ni lysis buffer (50 mM Tris-HCl, 5 mM β-mercaptoethanol, 100 mM KCl, 1% NP-40, 2 µg/mL benzamidine, 2 µg/mL leupeptin, 2 µg/mL aprotinin and 1 mM PMSF, pH 8.25) and lysed using a French press followed by sonication. After removal of cell debris (centrifugation, 30 min, 4 °C, 10,000 rpm), the His₆-tagged MBP-apoE4(aa170–299) fusion protein was purified by IMAC using Ni High Performance Sepharose Beads (GE Healthcare, Chicago, IL, USA, cat. no. 17-5268-01) according to the manufacturer’s instructions. Purity was assessed via SDS-PAGE and Coomassie staining.

2.4. Size Exclusion Chromatography

After IMPACT purification, the N-terminal apoE4 fragment (aa1–173) was further purified by size exclusion chromatography (SEC). The protein sample was loaded onto a Superdex 75 16/600 column connected to the Äkta pure system (Cytiva, Marlborough, MA, USA) with 20 mM Tris-HCl and 150 mM NaCl buffer (pH 7.5). After IMAC purification of the MBP apoE4 fragment (aa170–299) fusion protein, it was further purified by SEC. The eluate was loaded onto a Superdex 75 10/300 column connected to the Äkta pure system (Cytiva, Marlborough, MA, USA) with 20 mM Tris-HCl and 150 mM NaCl buffer (pH 7.5).

2.5. Biolayer Interferometry

Biolayer interferometry (BLI) measurements were performed on the Octet RED96 system (FortéBio, Fremont, CA, USA) at 25 °C. Anti-penta-HIS biosensors (FortéBio, Octet HIS1K) were pre-incubated in PBS (Ca²⁺ and Mg²⁺-free) for 10 min. Subsequently, the biosensors were incubated in PBS with 0.1% bovine serum albumin (BSA, Sigma-Aldrich, St. Louis, MO, USA, cat. no. A2153) for the first baseline assessment. A nanobody solution of 10 µg/mL and an antigen (apoE3 or apoE4) dilution series (50 nM, 100 nM and 200 nM) were prepared in PBS with 0.1% BSA. These solutions, together with a 0 nM antigen solution (PBS + 0.1% BSA), were dispensed at 200 µL per well into a 96-well plate (Greiner Bio-One, Kremsmünster, Austria, cat. no. 655209). Three cycles of the following schedule were performed: biosensor regeneration (5 s in 0.5 M H₂SO₄ followed by 5 s in PBS; repeated 3 times), baseline (60 s in PBS with 0.1% BSA), loading (300 s in nanobody solutions), baseline (60 s in PBS + 0.1% BSA), association (300 s in the antigen solutions) and dissociation (600 s in PBS + 0.1% BSA). The three cycles were followed by a last regeneration step (5 s in H₂SO₄). The flow rate was set to 1000 rpm (shake speed). One replicate per condition was included (4 wells per nanobody, 1 for each antigen concentration). The resulting binding curves were globally fitted to a 1:1 interaction model.

2.6. Cell Culture

HEK293T cells (ATCC, Manassas, VA, USA, cat. no. CRL-3216) were maintained in Dulbecco’s Modified Eagle Medium (DMEM; high glucose, GlutaMAX Supplement, pyruvate; Gibco, Waltham, MA, USA, cat. no. 31966047) supplemented with 10% fetal bovine serum (FBS) at 37 °C and 5% CO₂. SH-SY5Y cells (lab of prof. Frank Speleman, Ghent University, Ghent, Belgium) were maintained in RPMI 1640 medium (Gibco, Waltham, MA, USA, cat. no. 11875093) supplemented with 10% FBS at 37 °C and 5% CO₂.

2.7. Transient Transfection

On day 0, HEK293T cells were seeded in 6-well plates at 1.5 × 10⁵ cells per well for downstream immunocytochemistry (ICC) or at 3 × 10⁵ cells per well for co-immunoprecipitation (co-IP) and/or Western blot analysis. SH-SY5Y cells were seeded at 2.5 × 10⁵ cells per well for ICC or at 4 × 10⁵ cells per well for Western blot analysis. For experiments requiring medium collection after transfection, two wells were seeded per condition. Cells were maintained in 1 mL culture medium per well when medium was to be collected and in 2 mL otherwise.

For large-scale apoE4 production (pCMV4-ApoE4 plasmid), HEK293T cells were seeded in T75 flasks at 1.2 × 10⁶ cells per flask in 8 mL culture medium.

On day 1, HEK293T cells were transiently transfected using polyethylenimine (PEI, Polysciences, Warrington, PA, USA, cat. no. 23966-1). For transfections in 6-well plates, plasmid DNA was diluted in 80 µL serum-free DMEM and PEI was diluted separately in 80 µL serum-free DMEM. The PEI solution was added to the DNA solution, mixed briefly by vortexing and incubated for 8 min at room temperature to allow complex formation before addition to the cells. For transfections in T75 flasks, reagent volumes were scaled up proportionally.

DNA:PEI ratios and quantities were optimized for each construct and condition. For apoE4 expression in 6-well plates, 2 µg pCMV4-ApoE4 was transfected with 9 µg PEI, whereas for T75 flasks 16 µg of pCMV4-ApoE4 and 72 µg PEI were used. pCMV/myc/ER-Nb-HA (3 µg), pcDNA3.1-NLS-Nb-HA-mCherry (3 µg), pEGFP-N1-apoE4(ΔSEC1) (3 µg) or pcDNA3.1-apoE4(ΔSEC1) (2 µg) were each transfected with 9 µg PEI. For co-transfection of SEC2-Nb-HA-KDEL and apoE4, 2 µg pCMV/myc/ER-Nb-HA and 2 µg pCMV4-ApoE4 were transfected with 12 µg PEI. For co-transfection experiments combining NLS-Nb-HA-mCherry with apoE4(ΔSEC1) constructs, 2 µg pcDNA3.1-NLS-Nb-HA-mCherry and 1 µg pEGFP-N1-apoE4(ΔSEC1) or 1 µg pcDNA3.1-apoE4(ΔSEC1) were transfected with 9 µg PEI.

SH-SY5Y cells were transiently transfected with pcDNA3.1-apoE4(ΔSEC1) or pCMV4-ApoE4 using the jetPRIME transfection reagent (Satorius, Göttingen, Germany, cat. no. 101000001) according to the manufacturer’s instructions.

2.8. Cell Culture Medium Collection and Cell Extract Preparation

Conditioned cell culture media from two replicate wells of a 6-well plate were pooled one day post-transfection and supplemented with protease inhibitors (2 µg/mL benzamidine, 2 µg/mL leupeptin, 2 µg/mL aprotinin and 1 mM PMSF). For large-scale apoE4 production in T75 flasks, medium was collected two days post-transfection and supplemented with protease inhibitors.

Cells were collected and washed twice with PBS followed by suspension in lysis buffer (20 mM Tris-HCl, 150 mM NaCl, 1% Triton X-100, 2 µg/mL benzamidine, 2 µg/mL leupeptin, 2 µg/mL aprotinin and 1 mM PMSF, pH 7.5) and incubation on ice for 10 min. Cell debris was pelleted by centrifugation (10 min, 4 °C, 14,000 rpm) and the supernatant was collected. When required, total protein concentrations were determined using the Bradford assay (Bio-Rad, Hercules, CA, USA, cat. no. 5000006).

2.9. Co-Immunoprecipitation Assay

Pierce Anti-HA Agarose beads (Thermo Scientific, Waltham, MA, USA, cat. no. 26182) and Amsphere A3 beads (JSR Life Sciences, Sunnyvale, CA, USA) were washed three times with PBS.

2.9.1. Recombinant Proteins

Anti-HA agarose beads (12.5 µg settled beads) were incubated with either 5 µg purified recombinant nanobody or 800 µL periplasmic extract in a total volume of 1 mL buffer (20 mM Tris-HCl, 150 mM NaCl and 0.5% NP-40, pH 7.5) for 1 h at 4 °C with gentle rotation. Beads were washed three times with the same buffer and subsequently incubated with either 5 µg recombinant apoE3 (Peprotech, Cranbury, NJ, USA, cat. no. 350-02), apoE4 (Peprotech, Cranbury, NJ, USA, cat. no. 350-04), apoE4 N-terminal fragment (aa1–173) or MBP-fused C-terminal apoE4 fragment (aa170–299) in 1 mL of the same buffer for 1 h at 4 °C with gentle rotation. After three additional washes in buffer, beads were resuspended in SDS sample buffer and incubated 5–10 min at 95 °C before loading on an SDS-PAGE gel. As a positive control, Amsphere beads were incubated with 2 µg anti-apoE antibody (Novus Biologicals, Centennial, CO, USA, cat. no. NB110-60531), and the co-IP was performed as described above.

2.9.2. Cell Extracts

Anti-HA agarose beads were incubated with cell extract derived from two wells of a 6-well plate in a total volume of 1 mL buffer (20 mM Tris-HCl, 150 mM NaCl and 0.5% NP-40, pH 7.5) for 1 h at 4 °C with gentle rotation. Beads were washed three times in the same buffer, subsequently resuspended in SDS sample buffer, and incubated 5–10 min at 95 °C prior to SDS-PAGE.

2.10. Western Blot Analysis

Cell extract and conditioned medium samples were prepared by adding SDS sample buffer (5×) and incubating them 5–10 min at 95 °C. Equal amounts of cell extract and equal volumes of conditioned medium were electrophoresed on a 12% or 15% SDS-PAGE gel. Pageruler Prestained Protein ladder (10 to 180 kDa; Thermo Scientific, Waltham, MA, USA, cat. no. 26616) was used to determine molecular weight.

Proteins were transferred to an Amersham Protran 0.45 µm Supported nitrocellulose membrane (Cytiva, Marlborough, MA, USA, cat. no. GE10600016). Membranes were blocked for 1 h at room temperature in blocking buffer (5% milk powder and 0.1% Tween-20 (Merck, Darmstadt, Germany) in TBS (20 mM Tris-HCl, 150 mM NaCl, pH 7.5)). Membranes were incubated overnight at 4 °C or for 4 h at room temperature with primary antibodies diluted in blocking buffer. Primary antibodies against the following antigens were used: HA (rabbit anti-HA antibody, Zymed, San Francisco, CA, USA, cat. no. 71-5500 and mouse anti-HA antibody (12CA5), Roche, Basel, Switzerland, cat. no. 11583816001), apoE4 (rabbit anti-apoE antibody [EPR19392], Abcam, Cambridge, UK, cat. no. ab183597; mouse anti-apoE4 (4E4) antibody, Novus Biologicals, Centennial, CO, USA, cat. no. NBP1-49529 and plasma from an apoE-immunized llama (1:2000)) and β-actin (mouse ACTB antibody, Sigma-Aldrich, St. Louis, MO, USA, cat. no. A1978).

After washing three times with TBS containing 0.1% Tween-20, membranes were incubated with secondary antibodies diluted in blocking buffer for 1 h at room temperature. When proteins were detected by fluorescence, the following secondary antibodies were used: Goat Anti-Rabbit IgG (H+L) DyLight 800 conjugated (Thermo Scientific, Waltham, MA, USA, cat. no. 35571), Goat Anti-Rabbit IgG (H+L) DyLight 680 conjugated (Thermo Scientific, Waltham, MA, USA, cat. no. 35568), Goat Anti-Mouse IgG (H+L) DyLight 800 conjugated (Thermo Scientific, Waltham, MA, USA, cat. no. 35521) or Goat Anti-Mouse IgG (H+L) DyLight 680 Conjugated (Thermo Scientific, Waltham, MA, USA, cat. no. 35518). When proteins were detected by enhanced chemiluminescence (ECL), membranes were incubated with the following secondary antibodies: goat anti-llama IgG HRP (Abcam, Cambridge, UK, cat. no. ab112786) or sheep ECL anti-mouse IgG HRP (Cytiva, Marlborough, MA, USA, cat. no. NA931). Membranes were washed three times with TBS containing 0.1% Tween-20.

For membranes incubated with infrared dyes (fluorescent Western blotting), proteins were visualized using the Odyssey DLx Western blot imager (LICORbio, Lincoln, NE, USA). Membranes imaged via ECL were incubated with ECL Prime substrate (Cytiva, Marlborough, MA, USA, cat. no. GERPN2232) according to the manufacturer’s instructions, after which proteins were visualized using the Amersham Imager 680 (Cytiva, Marlborough, MA, USA).

2.11. Immunocytochemistry

Glass coverslips were placed in 6-well plates and HEK293T or SH-SY5Y cells were seeded on the coverslips on day 0. Transfection was performed on day 1, followed by immunostaining on day 2. Cells were fixed with 4% paraformaldehyde (PFA, Merck, Darmstadt, Germany) in PBS for 20 min, permeabilized with 0.2% Triton X-100 (Sigma-Aldrich, St. Louis, MO, USA) in PBS for 10 min, and subsequently blocked with 1% BSA in PBS for 1 h at room temperature.

Primary antibodies against apoE4 (Novus Biologicals, Centennial, CO, USA, cat. no. NBP1-49529), HA (Zymed, San Francisco, CA, USA, cat. no. 71-5500), KDEL (Abcam, Cambridge, UK, cat. no. 176333), Golgin 97 (Proteintech, Rosemont, IL, USA, cat. no. 68648-1-Ig) and/or calnexin (Proteintech, Rosemont, IL, USA, cat. no. 66903-1-Ig) were diluted in 1% BSA in PBS and applied for 1 h at 37 °C.

After washing with 0.1% Tween-20 in PBS, cells were incubated with secondary antibodies for 1 h at room temperature, i.e., Alexa Fluor 488-conjugated anti-mouse antibody (Invitrogen, Carlsbad, CA, USA, cat. no. A-11029), Alexa Fluor 488-conjugated anti-rabbit antibody (Invitrogen, Carlsbad, CA, USA, cat. no. A-11034), Alexa Fluor 594-conjugated anti-mouse antibody (Invitrogen, Carlsbad, CA, USA, cat. no. A-11032), Alexa Fluor 594-conjugated anti-rabbit antibody (Invitrogen, Carlsbad, CA, USA, cat. no. A-11037) and/or Alexa Fluor 647-conjugated anti-mouse antibody (Invitrogen, Carlsbad, CA, USA, cat. no. A-21245). Hoechst was diluted 1:1000 in the secondary antibody solution to stain nuclei. Following secondary antibody incubation, cells were washed again with 0.1% Tween-20 in PBS.

When required, cells were incubated with ProteoStat (Enzo Life Sciences, Farmingdale, NY, USA, cat. no. ENZ-51035, 1:500 dilution) for 30 min at room temperature, after which cells were washed once with PBS. As a positive control for ProteoStat staining, cells were treated with 5 µM MG132 (Sigma-Aldrich, St. Louis, MO, USA, cat. no. M7449) for 18 h. When no protein immunostaining was required, only nuclei were stained by incubating cells for 7 min in Hoechst diluted in PBS.

Finally, coverslips were mounted on glass slides with Vectashield Antifade mounting medium (Vector Laboratories, Newark, CA, USA, cat. no. H-1000-10) and imaged using an Olympus IX81 (Tokyo, Japan), Nikon A1R-MP (Tokyo, Japan), or Zeiss LSM 680 (Oberkochen, Germany) confocal microscope. Image analysis was performed using Fiji (ImageJ2, Version 2.16.0).

2.12. Bio-Informatics

Phylogenetic trees based on alignments of nanobody protein sequences were created using CLC Sequence Viewer (Version 8.0). Trees were constructed with the UPGMA method, and the Jukes–Cantor protein distance measure was used.

Colocalization of two fluorescence channels in microscopy images was quantified by calculating the Pearson’s correlation coefficient (PCC). For each condition, three microscopy slides were prepared (n = 3 biological replicates). Multiple images per slide were acquired, capturing ~45 cells per condition (~15 cells per slide). Each cell was treated as a technical replicate. Cell masks were generated with Cellpose (Version 2.2.2) [25], and using these cell masks, the PCC was calculated per cell in Python (Version 3.10.16; script in Figure S1). The average PCC of all cells analyzed per slide (PCC_averaged) represents the biological replicate value.

2.13. Statistical Analysis

Statistical analyses were performed in RStudio (Version 2024.12.1+563). As sample sizes were limited (n = 3), assessment of normality was not feasible. PCC_averaged values and the cut-off value of 0.5 were Fisher’s Z-transformed prior to hypothesis testing. A one-sided Student’s t-test was performed on the transformed data to evaluate whether the PCC_averaged was significantly greater than 0.5.

2.14. Generative Artificial Intelligence

ChatGPT-5 and GitHub Copilot (Version 0.37.9) were used to assist with Python scripts.

3. Results

3.1. ApoE Nanobodies Bind Recombinant ApoE4 and ApoE3 with High Affinity

To generate a panel of apoE VHHs, three llamas were immunized with either human recombinant apoE4, human recombinant apoE3, synthetic apoE4 peptide (aa82–131), or a combination of the synthetic apoE4 peptide and human recombinant apoE4. This resulted in four distinct VHH libraries, referred to as libraries 9 (apoE4), 17 (apoE3), 18 (peptide) and 19 (peptide + apoE4).

The VHH libraries were subjected to phage display biopanning, yielding a diverse set of apoE-targeting nanobodies. Based on amino acid sequence similarity, these nanobodies were subdivided into families (Figure S2), predominantly defined by their CDR3 sequences. One representative nanobody per family was selected for further characterization. Figure S3 shows partial CDR3 sequences of the selected nanobodies, illustrating their sequence diversity. Nanobodies were named according to their library of origin (e.g., nanobody 9-8 was derived from library 9 and nanobody 17-23 from library 17).

Selected HA-His₆-tagged nanobodies containing a PelB leader sequence (directing them to the periplasm) were recombinantly expressed in E. coli WK6. Periplasmic extracts containing the nanobodies were used in co-immunoprecipitation assays to assess binding to recombinant apoE4 and apoE3 (produced in E. coli), thereby evaluating antigen recognition and isoform cross-reactivity (Figure 1). All tested nanobodies (9-8, 9-59, 9-74, 9-76, 9-79, 17-4, 17-23, 17-69, 18-24, 18-27, 18-91, 19-38, 19-76, 19-82 and 19-89) bound both apoE4 (Figure 1A) and apoE3 (Figure 1B). In addition, several nanobodies also interacted with recombinant apoE2 (Figure S4).

The binding kinetics of the apoE nanobody set towards apoE4 and/or apoE3 were determined using bio-layer interferometry (Octet system). BLI measures biomolecular interactions in real time by detecting changes in the interference pattern of light reflected from biosensors loaded with His₆-tagged nanobodies as recombinant apoE associates and dissociates. Binding curves were globally fitted to a 1:1 interaction model, from which association rate constants (k_on) were derived from the association phase and dissociation rate constants (k_off) from the dissociation phase. In most cases, the resulting equilibrium dissociation constants (K_d = k_off/k_on) were in the low nanomolar or even sub-nanomolar range, indicating high-affinity binding (Table 1). For several nanobodies (e.g., Nb 17-69), dissociation was extremely slow, and k_off values could not be reliably determined (k_off < 1 × 10⁻⁷ s⁻¹), resulting in K_d values below the limit of detection (LOD; reported as K_d < 1 × 10^–12 M). Although a finite dissociation constant could not be determined for these nanobodies, their slow dissociation rates suggest very strong binding.

The kinetic fits were of high quality, with R² values exceeding 0.96 for all analyses and X² values below 3 for the majority of nanobodies. Representative sensorgrams for four nanobodies together with their corresponding kinetic parameters are shown in Figure 2. Complete kinetic binding data for all 15 nanobodies, including apoE2 measurements for several nanobodies, are provided in Table S1, and the corresponding sensorgrams not shown in Figure 2 are presented in Figure S6. Based on the co-IP results (Figure 1) and the BLI binding kinetics, six nanobodies were selected for further characterization: Nb 9-8, 9-74, 17-69, 18-27, 18-91 and 19-38.

Table 1. Equilibrium dissociation constants (K_d) of apoE nanobodies towards apoE4 and/or apoE3.

	Kd [nM]
Nb	ApoE4	ApoE3
9-8	2.40	1.64
9-59	<0.001 *	<0.001 *
9-74	2.00	<0.001 *
9-76	<0.001 *	<0.001 *
9-79	<0.001 *	<0.001 *
17-4	-	0.01
17-23	-	2.28
17-69	-	<0.001 *
18-24	0.22	-
18-27	1.04	-
18-91	<0.001 *	-
19-38	0.13	-
19-76	1.29	-
19-82	0.45	-
19-89	0.24	-

- Not measured; * Below limit of detection.

Table 1. Equilibrium dissociation constants (K_d) of apoE nanobodies towards apoE4 and/or apoE3.

	Kd [nM]
Nb	ApoE4	ApoE3
9-8	2.40	1.64
9-59	<0.001 *	<0.001 *
9-74	2.00	<0.001 *
9-76	<0.001 *	<0.001 *
9-79	<0.001 *	<0.001 *
17-4	-	0.01
17-23	-	2.28
17-69	-	<0.001 *
18-24	0.22	-
18-27	1.04	-
18-91	<0.001 *	-
19-38	0.13	-
19-76	1.29	-
19-82	0.45	-
19-89	0.24	-

- Not measured; * Below limit of detection.

3.2. Six Selected ApoE Nanobodies Bind ApoE4 Secreted by HEK293T Cells

The six nanobodies (HA-His₆-tagged) selected in the previous section were again expressed in E. coli WK6 and purified from periplasmic extracts using IMAC. SDS-PAGE analysis followed by Coomassie staining revealed clear bands at ~15 kDa for all nanobodies (Figure 3). Nb 18-27 showed an additional band at ~30 kDa, which probably corresponds to a nanobody dimer (Figure S7).

The recombinant apoE antigen used for immunization and in the assays described in the previous section (Section 3.1) was originally purified from bacteria. To ascertain that the six selected nanobodies also recognize apoE4 produced by mammalian cells, apoE4 was transiently expressed in HEK293T cells. As apoE is a secreted protein, it is synthesized with an N-terminal 18-amino-acid signal peptide that directs the protein to the endoplasmic reticulum and into the secretory pathway [4]. ApoE4 displayed a reticular and granular intracellular staining pattern consistent with localization within the ER/Golgi secretory pathway (Figure 4A) [26]. A similar pattern was observed in apoE4-transfected HeLa, Hep3B cells and SH-SY5Y cells (Figure S9). Conditioned cell culture medium containing secreted apoE4 was used in co-IP assays with each of the six selected purified HA-tagged nanobodies. Binding of secreted apoE4 by all six nanobodies (Nb 9-8, 9-74, 17-69, 18-27, 18-91 and 19-38) was confirmed in these assays (Figure 4B).

3.3. Domain-Level Epitope Mapping: Selected Nanobodies Either Bind an N-Terminal or a C-Terminal ApoE4 Fragment

Structurally, apoE consists of two main domains connected by a flexible hinge region (Figure 5A). The N-terminal domain (residues 1–167) adopts a four-helix bundle and harbors the receptor-binding region (residues 135–150), which mediates interaction with members of the LDL receptor family. The C-terminal domain (residues 206–299) consists of amphipathic alpha-helices and contains the lipid-binding region (residues 244–272), which is crucial for lipid interactions [27].

Here, we designed two apoE4 fragments, one encompassing the N-terminal domain and the other the C-terminal domain, and we assessed to which apoE4 fragment each of the six selected nanobodies binds. The N-terminal fragment contains residues 1–173, while the C-terminal fragment contains residues 170–299 (Figure 5A). Both fragments include part of the hinge region to cover the entire protein, with the C-terminal fragment containing the larger part of the hinge. The N-terminal apoE4 fragment (aa1–173) was recombinantly produced in E. coli BL21 and purified using the IMPACT system, followed by size exclusion chromatography (Figure S11A–C; uncropped gels in Figure S12A,B). The C-terminal apoE4 fragment (aa170–299) was expressed in E. coli BL21 as an MPB fusion protein. Fusion to MBP was found to stabilize the small apoE4 fragment (~14 kDa). Following expression, the MBP-apoE4(aa170–299) protein was purified using IMAC followed by SEC (Figure S11D–F; uncropped gels in Figure S12C,D).

The purified apoE4 fragments were subsequently used in co-IP assays with HA-tagged nanobodies 9-8, 9-74, 17-69, 18-27, 18-91 and 19-38. Based on these assays, the nanobodies segregated into two distinct groups binding to one of the two apoE4 fragments. Nanobodies 9-74 and 19-38 interacted with the N-terminal apoE4 fragment, whereas nanobodies 9-8, 17-69, 18-27 and 18-91 bound to the C-terminal apoE4 fragment (Figure 5B,C). None of the nanobodies bound to both the N-terminal and the C-terminal fragment, thereby defining a set of domain-specific nanobodies. From these, two N-terminal domain binders (Nb 9-74 and 19-38) and two C-terminal domain binders (Nb 17-69 and 18-91) were selected for subsequent experiments.

Figure 5. Domain-specific binding of apoE nanobodies to apoE4. (A) Schematic representation of apoE structure. The N-terminal domain (residues 1–167) and C-terminal domain (residues 206–299) are connected by a hinge region. The N-terminal domain contains the low-density lipoprotein (LDL) receptor-binding region, while the C-terminal domain contains the lipid-binding region. The three human apoE isoforms differ at residues 112 and 158. Figure adapted from ref. [27], with permission from Springer Nature, 2018. Two apoE4 fragments corresponding to the N-terminal domain (1; aa1–173) and to the C-terminal domain (2; aa170–299) were produced. (B) Co-IP assays showing interaction of the six selected HA-tagged nanobodies with either the N-terminal apoE4 fragment (top; Coomassie-stained SDS-PAGE gels) or the C-terminal apoE4 fragment (bottom; Western blots). Co-IPs were performed using anti-HA agarose beads. The C-terminal apoE4 fragment was produced as an MBP fusion protein to stabilize the small fragment (~14 kDa). Samples were run on multiple gels, indicated by a space between the blocks. On each gel, 3 µg of apoE4(aa1–173) or MBP-apoE4(aa170–299) was loaded as a reference, together with negative control (NC) samples consisting of the respective apoE4 fragment incubated with anti-HA agarose beads in the absence of nanobody. For the N-terminal apoE4 fragment, a positive control (PC) was included in which the fragment was incubated with an anti-apoE antibody and Amsphere beads. The band seen at the nanobody level in the PC lane (top panel, lane 2) is background. For simplicity, one reference and one NC sample are presented. Uncropped gels and Western blots are shown in Figure S13. (C) Subdivision of the six apoE nanobodies according to the apoE4 fragment they bind. Nanobodies 9-74 and 19-38 bind the N-terminal apoE4 fragment (aa1–173), whereas nanobodies 9-8, 17-69, 18-27 and 18-91 bind the C-terminal apoE4 fragment (aa170–299).

Figure 5. Domain-specific binding of apoE nanobodies to apoE4. (A) Schematic representation of apoE structure. The N-terminal domain (residues 1–167) and C-terminal domain (residues 206–299) are connected by a hinge region. The N-terminal domain contains the low-density lipoprotein (LDL) receptor-binding region, while the C-terminal domain contains the lipid-binding region. The three human apoE isoforms differ at residues 112 and 158. Figure adapted from ref. [27], with permission from Springer Nature, 2018. Two apoE4 fragments corresponding to the N-terminal domain (1; aa1–173) and to the C-terminal domain (2; aa170–299) were produced. (B) Co-IP assays showing interaction of the six selected HA-tagged nanobodies with either the N-terminal apoE4 fragment (top; Coomassie-stained SDS-PAGE gels) or the C-terminal apoE4 fragment (bottom; Western blots). Co-IPs were performed using anti-HA agarose beads. The C-terminal apoE4 fragment was produced as an MBP fusion protein to stabilize the small fragment (~14 kDa). Samples were run on multiple gels, indicated by a space between the blocks. On each gel, 3 µg of apoE4(aa1–173) or MBP-apoE4(aa170–299) was loaded as a reference, together with negative control (NC) samples consisting of the respective apoE4 fragment incubated with anti-HA agarose beads in the absence of nanobody. For the N-terminal apoE4 fragment, a positive control (PC) was included in which the fragment was incubated with an anti-apoE antibody and Amsphere beads. The band seen at the nanobody level in the PC lane (top panel, lane 2) is background. For simplicity, one reference and one NC sample are presented. Uncropped gels and Western blots are shown in Figure S13. (C) Subdivision of the six apoE nanobodies according to the apoE4 fragment they bind. Nanobodies 9-74 and 19-38 bind the N-terminal apoE4 fragment (aa1–173), whereas nanobodies 9-8, 17-69, 18-27 and 18-91 bind the C-terminal apoE4 fragment (aa170–299).

3.4. The Four Selected ApoE Nanobodies Are Functional When Targeted to the ER

To evaluate whether the selected apoE nanobodies can function as intracellular tools, we assessed their ability to interact with apoE4 in HEK293T cells. As apoE4 is a secreted protein whose N-terminal signal peptide directs it to the ER [28], the nanobodies were engineered as ER-targeted intrabodies. To this end, each nanobody (Nb 9-74, 19-38, 17-69 and 18-91) was cloned in frame with an N-terminal mouse IgH signal sequence and a C-terminal KDEL ER retention sequence (SEC2-Nb-HA-KDEL; Figure 6A), thereby directing them to the ER lumen and enabling their interaction with intracellular apoE4. KDEL is a well-established ER retention motif that has been used by other researchers to localize nanobodies to the ER lumen [29,30]. Consistent with this, KDEL-equipped apoE nanobodies expressed in HEK293T cells displayed an ER-like reticular staining pattern and overlapped with the ER marker calnexin (Figure S14). Each nanobody construct was subsequently co-transfected with apoE4 (Figure 6B) in HEK293T cells. Figure 6C schematically illustrates the strategy of targeting nanobodies to the ER to allow interaction with apoE4 during its passage through the secretory pathway.

Immunofluorescence analysis revealed overlapping staining patterns of apoE4 and each nanobody, consistent with intracellular interaction (Figure 6D). Similar overlap was observed for Nb 17-69 in co-transfected Hep3B cells (Figure S15). In addition, both apoE4 and the nanobodies exhibited a reticular staining pattern consistent with ER localization, as validated by calnexin staining in Figure S14. Colocalization was quantified using the Pearson’s correlation coefficient, with values above 0.5 and approaching 1 considered indicative of good colocalization [31]. PCC values were calculated per cell and averaged per experiment, reported as PCC_averaged and representing the biological replicate value. All four nanobodies displayed PCC_averaged values that were significantly greater than 0.5 and approaching 1 (0.76, 0.91, 0.79 and 0.94 for nanobodies 9-74, 19-38, 17-69 and 18-91, respectively; Figure 6E). Intracellular interaction between apoE4 and each nanobody was further confirmed by co-IP assays using cell extracts of co-transfected HEK293T cells (Figure 6F).

To assess whether ER-targeted nanobodies affect apoE4 trafficking, cells were co-transfected with apoE4 together with each of the four apoE nanobodies or with a GFP nanobody equipped with the same ER-targeting and retention signals. Compared to the GFP nanobody control, the presence of each apoE nanobody resulted in increased apoE4 levels in cell extracts and reduced apoE4 levels in the conditioned cell culture medium (Figure 6G–I), indicating intracellular retention of apoE4. Notably, KDEL-equipped nanobodies were not detected in the conditioned medium (Figure 6G, right panel), confirming their intracellular retention.

Together, these results validate nanobodies 9-74, 19-38, 17-69 and 18-91 as ER-targeted intracellular binding tools that interact with apoE4 and promote its retention in HEK293T cells. These nanobodies provide a foundation for future strategies aimed at manipulating apoE4 (degradation, block trafficking, etc.) in AD-relevant cell models.

Figure 6. ER-targeted apoE nanobodies 9-74, 19-38, 17-69 and 18-91 bind apoE4 intracellularly and promote its intracellular retention in HEK293T cells. (A) Schematic representation of the ER-targeted nanobody construct containing an N-terminal mouse IgH secretion signal peptide (SEC2), a C-terminal HA-tag, and a KDEL ER retention sequence. (B) ApoE4 expression construct containing an N-terminal 18-amino-acid signal peptide (SEC1). (C) Schematic depiction of cells expressing apoE4 and an ER-targeted apoE nanobody. The KDEL-equipped nanobody is directed to the ER, allowing its interaction with apoE4 as it passes through the secretory pathway. (D) Microscopy images showing colocalization (yellow) of apoE4 (green) and ER-targeted nanobodies (Nb-KDEL; red) in HEK293T cells. Scale bar: 10 µm. (E) Scatter dot plot of the averaged Pearson’s correlation coefficient (PCC_averaged) values quantifying colocalization between apoE4 and each ER-targeted nanobody. Each dot represents the average PCC value of one independent experiment (~15 cells per experiment). Black horizontal lines and error bars indicate the mean ± standard deviation (SD; n = 3). All nanobodies displayed PCC_averaged values significantly greater than 0.5 (red dotted line) and approaching 1, indicative of strong colocalization (Fisher Z-transformation followed by a one-sample t-test against a reference value of 0.5). * p < 0.05; ** p < 0.01; *** p < 0.001. (F) Co-IP assays showing interaction between each of the ER-targeted apoE nanobodies (Nb-KDEL) and apoE4 in cell extracts of co-transfected HEK293T cells (lanes 6). Co-IP was performed using anti-HA agarose beads and analyzed by Western blot. Two negative controls were included: NC1, a co-IP using the extract of cells transfected with apoE4 alone (lanes 4), and NC2, a co-IP using the extract of cells co-transfected with apoE4 and an ER-targeted GFP nanobody (lanes 5). For each condition, the corresponding input (cell extract) is shown (lanes 1–3). Uncropped blots are shown in Figure S16. (G) Western blot analysis showing increased apoE4 levels in cell extracts (left) and reduced apoE4 levels in conditioned medium (right) from HEK293T cells co-transfected with apoE4 and each of the ER-targeted apoE nanobodies, compared to cells co-transfected with apoE4 and an ER-targeted GFP nanobody control. Actin was used as a loading control. Uncropped versions of the shown blots, together with replicate blots (n = 3), are provided in Figure S17. (H) Quantification of intracellular apoE4 levels (cell extract Western blots). ApoE4 levels were normalized to actin and expressed relative to the GFP nanobody control within each blot/replicate. Individual data points are shown as dots (n = 3); bars represent mean ± SD. Raw data are provided in Table S2. (I) Quantification of secreted apoE4 levels (conditioned medium Western blots). ApoE4 levels were expressed relative to the GFP nanobody control within each blot/replicate. Individual data points are shown as dots (n = 3); bars represent mean ± SD. Raw data are provided in Table S2.

Figure 6. ER-targeted apoE nanobodies 9-74, 19-38, 17-69 and 18-91 bind apoE4 intracellularly and promote its intracellular retention in HEK293T cells. (A) Schematic representation of the ER-targeted nanobody construct containing an N-terminal mouse IgH secretion signal peptide (SEC2), a C-terminal HA-tag, and a KDEL ER retention sequence. (B) ApoE4 expression construct containing an N-terminal 18-amino-acid signal peptide (SEC1). (C) Schematic depiction of cells expressing apoE4 and an ER-targeted apoE nanobody. The KDEL-equipped nanobody is directed to the ER, allowing its interaction with apoE4 as it passes through the secretory pathway. (D) Microscopy images showing colocalization (yellow) of apoE4 (green) and ER-targeted nanobodies (Nb-KDEL; red) in HEK293T cells. Scale bar: 10 µm. (E) Scatter dot plot of the averaged Pearson’s correlation coefficient (PCC_averaged) values quantifying colocalization between apoE4 and each ER-targeted nanobody. Each dot represents the average PCC value of one independent experiment (~15 cells per experiment). Black horizontal lines and error bars indicate the mean ± standard deviation (SD; n = 3). All nanobodies displayed PCC_averaged values significantly greater than 0.5 (red dotted line) and approaching 1, indicative of strong colocalization (Fisher Z-transformation followed by a one-sample t-test against a reference value of 0.5). * p < 0.05; ** p < 0.01; *** p < 0.001. (F) Co-IP assays showing interaction between each of the ER-targeted apoE nanobodies (Nb-KDEL) and apoE4 in cell extracts of co-transfected HEK293T cells (lanes 6). Co-IP was performed using anti-HA agarose beads and analyzed by Western blot. Two negative controls were included: NC1, a co-IP using the extract of cells transfected with apoE4 alone (lanes 4), and NC2, a co-IP using the extract of cells co-transfected with apoE4 and an ER-targeted GFP nanobody (lanes 5). For each condition, the corresponding input (cell extract) is shown (lanes 1–3). Uncropped blots are shown in Figure S16. (G) Western blot analysis showing increased apoE4 levels in cell extracts (left) and reduced apoE4 levels in conditioned medium (right) from HEK293T cells co-transfected with apoE4 and each of the ER-targeted apoE nanobodies, compared to cells co-transfected with apoE4 and an ER-targeted GFP nanobody control. Actin was used as a loading control. Uncropped versions of the shown blots, together with replicate blots (n = 3), are provided in Figure S17. (H) Quantification of intracellular apoE4 levels (cell extract Western blots). ApoE4 levels were normalized to actin and expressed relative to the GFP nanobody control within each blot/replicate. Individual data points are shown as dots (n = 3); bars represent mean ± SD. Raw data are provided in Table S2. (I) Quantification of secreted apoE4 levels (conditioned medium Western blots). ApoE4 levels were expressed relative to the GFP nanobody control within each blot/replicate. Individual data points are shown as dots (n = 3); bars represent mean ± SD. Raw data are provided in Table S2.

3.5. ApoE4(ΔSEC1) Localizes in Perinuclear Assemblies, Is Associated with a ~25 kDa ApoE4 Fragment, and Is Not Relocalized to the Nucleus by an NLS-Tagged Nanobody

Studies have reported that apoE4 fragments can cause mitochondrial dysfunction and cytoskeletal abnormalities, suggesting that apoE4 may exert intracellular effects outside the secretory pathway [21,22,23]. To explore the behavior of apoE4 in the cytosol, we generated a cytosolic apoE4 variant lacking its N-terminal secretion signal peptide (apoE4(ΔSEC1)). When expressed in HEK293T and SH-SY5Y cells, apoE4(ΔSEC1) was enriched in the perinuclear region (Figure 7A). Staining with ProteoStat, a fluorescent dye that detects aggregated proteins, suggested that a subset of apoE4-positive cytosolic assemblies corresponds to aggregate-like structures (Figure 7B). In addition, Western blot analysis of apoE4 in HEK293T and SH-SY5Y cells expressing apoE4(ΔSEC1) revealed an additional band of ~25 kDa that was not observed in cells expressing apoE4 containing the secretion signal peptide (Figure 7C). This suggests that apoE4(ΔSEC1) may undergo proteolytic processing in the cytosol.

To assess whether cytosolic apoE4(ΔSEC1) remains accessible for nanobody binding and whether the resulting complex is competent for nuclear relocalization, we employed a nanobody-based intracellular relocalization approach. To this end, HA-tagged Nb 17-69 was fused N-terminally to the SV40 large T antigen nuclear localization signal and C-terminally to mCherry for visualization (NLS-Nb17-69-mCherry; Figure 8A). ApoE4(ΔSEC1) was fused to EGFP at its C-terminus for visualization (Figure 8B). Figure 8C schematically illustrates the NLS-nanobody relocalization strategy used to probe nuclear import competence of cytosolic apoE4(ΔSEC1).

Intracellular binding of NLS-equipped nanobody 17-69 to apoE4(ΔSEC1) was confirmed by co-IP using extracts of HEK293T cells that co-express both proteins (Figure 8D). When expressed alone, the NLS-equipped nanobody 17-69 showed strong nuclear localization in HEK293T cells (Figure 8E). However, upon co-expression with apoE4(ΔSEC1)-EGFP, neither nuclear relocalization of apoE4(ΔSEC1) nor nuclear enrichment of the NLS-equipped nanobody was observed (Figure 8F). Instead, both proteins accumulated in the cytosol.

Together, these observations indicate that cytosolic apoE4(ΔSEC1) accumulates in perinuclear assemblies with aggregate-like properties, is associated with the appearance of a ~25 kDa apoE4 fragment and is not relocalized to the nucleus by an NLS-tagged nanobody.

Figure 8. SV40 nuclear localization signal (NLS)-equipped nanobody 17-69 binds apoE4(ΔSEC1), but apoE4(ΔSEC1) is not directed to the nucleus. (A) Schematic representation of an HA-tagged nanobody N-terminally fused to the SV40 large T antigen NLS and C-terminally fused to mCherry. (B) Schematic representation of apoE4(ΔSEC1) fused to EGFP at its C-terminus. (C) Schematic representation of the nanobody-based intracellular relocalization strategy in HEK293T cells. Upon co-expression of the NLS-equipped nanobody (fused to mCherry) and apoE4(ΔSEC1) (fused to EGFP), both are present in the cytosol, where the nanobody binds apoE4(ΔSEC1) and is intended to direct it to the nucleus. (D) Co-IP assay showing intracellular interaction between NLS-Nb17-69-mCherry and apoE4(ΔSEC1), co-expressed in HEK293T cells, using anti-HA agarose beads (lane 4). As a negative control (NC), co-IP was performed using the cell extract of HEK293T cells expressing apoE4(ΔSEC1) alone (lane 3). The input samples (cell extracts) are shown in lanes 1 and 2. The uncropped blot is shown in Figure S19. (E) Microscopy images of NLS-Nb17-69-mCherry (red) expressed in HEK293T cells, showing nuclear localization of the nanobody. Scale bar: 10 µm. (F) Microscopy images of apoE4(ΔSEC1)-EGFP (green) co-expressed with NLS-Nb17-69-HA-mCherry (red) in HEK293T cells. Both apoE4(ΔSEC1) and NLS-equipped Nb 17-69 remain in the cytosol. Scale bar: 10 µm.

Figure 8. SV40 nuclear localization signal (NLS)-equipped nanobody 17-69 binds apoE4(ΔSEC1), but apoE4(ΔSEC1) is not directed to the nucleus. (A) Schematic representation of an HA-tagged nanobody N-terminally fused to the SV40 large T antigen NLS and C-terminally fused to mCherry. (B) Schematic representation of apoE4(ΔSEC1) fused to EGFP at its C-terminus. (C) Schematic representation of the nanobody-based intracellular relocalization strategy in HEK293T cells. Upon co-expression of the NLS-equipped nanobody (fused to mCherry) and apoE4(ΔSEC1) (fused to EGFP), both are present in the cytosol, where the nanobody binds apoE4(ΔSEC1) and is intended to direct it to the nucleus. (D) Co-IP assay showing intracellular interaction between NLS-Nb17-69-mCherry and apoE4(ΔSEC1), co-expressed in HEK293T cells, using anti-HA agarose beads (lane 4). As a negative control (NC), co-IP was performed using the cell extract of HEK293T cells expressing apoE4(ΔSEC1) alone (lane 3). The input samples (cell extracts) are shown in lanes 1 and 2. The uncropped blot is shown in Figure S19. (E) Microscopy images of NLS-Nb17-69-mCherry (red) expressed in HEK293T cells, showing nuclear localization of the nanobody. Scale bar: 10 µm. (F) Microscopy images of apoE4(ΔSEC1)-EGFP (green) co-expressed with NLS-Nb17-69-HA-mCherry (red) in HEK293T cells. Both apoE4(ΔSEC1) and NLS-equipped Nb 17-69 remain in the cytosol. Scale bar: 10 µm.

4. Discussion

Apolipoprotein E4 is the strongest genetic risk factor for late-onset Alzheimer’s disease and has been implicated in multiple AD-related pathologies, including Aβ plaque deposition, tau pathology, neuroinflammation, synaptic dysfunction and mitochondrial impairment [2,5,6,7,32]. However, the molecular mechanisms by which apoE4 contributes to these AD-related pathologies remain incompletely understood. In this study, we developed a set of apoE-targeting nanobodies as high-affinity research tools to investigate apoE4 at the protein level in cellular models. Such research tools are important to further dissect apoE4 biology.

Fifteen apoE-targeting nanobodies were generated that bind recombinant apoE4 and apoE3, purified from bacteria, with high affinity. Six of these nanobodies were further validated to recognize apoE4 produced and secreted by mammalian cells, underscoring their relevance for cellular applications. Fragment-based epitope mapping revealed that the nanobodies segregate into two groups, recognizing either the 20 kDa N-terminal (Nb 9-74 and 19-38) or the 14 kDa C-terminal apoE4 fragment (Nb 9-8, 17-69, 18-27 and 18-91). This domain-level epitope mapping is of particular interest given the modular architecture of apoE, in which the N-terminal and C-terminal domains are associated with distinct structural and functional properties [33]. Domain-selective nanobodies therefore provide a means to interrogate domain-specific functions of apoE4, such as receptor interaction or lipid-binding-dependent processes. The availability of nanobodies targeting different apoE4 domains expands the versatility of the toolbox and enables future studies to selectively probe domain-specific aspects of apoE4 biology in vitro and in vivo.

The N- and C-terminal fragment binders were validated as ER-targeted intrabodies. The KDEL-equipped nanobodies colocalized with apoE4, bound apoE4 intracellularly and reduced apoE4 secretion, demonstrating robust interaction with apoE4 within the ER. These findings establish a foundation for future nanobody-based approaches, such as the manipulation of apoE4 trafficking or targeted apoE4 degradation, to further investigate apoE4’s role in AD within disease-relevant cellular models.

In their current ER-targeted format, these nanobodies can be expressed in disease-relevant systems such as APOE-ε4 homozygous induced pluripotent stem cell (iPSC)-derived neurons to retain apoE4 within the ER and assess downstream AD-relevant readouts, including Aβ peptides or phosphorylated tau (p-tau), without implying therapeutic intervention. This strategy, in which an ER-targeted intrabody partially blocks protein trafficking, has been applied to other proteins in the ER/Golgi pathway to study protein function [34,35,36], underscoring their utility as research tools.

ApoE4 fragments have been shown to trigger intracellular p-tau accumulation and mitochondrial dysfunction in neuronal cells [21,23], supporting the hypothesis that proteolytically cleaved apoE4 fragments may exert deleterious effects in the cytosol that contribute to AD pathogenesis. Therefore, we investigated the behavior of cytosolic apoE4 using our nanobody toolbox. When expressing a cytosolic apoE4 variant lacking its secretion signal peptide (apoE4(ΔSEC1)) in mammalian cells, perinuclear apoE4 assemblies with aggregate-like properties were observed. Using a nanobody-based relocalization assay, we demonstrated binding of an NLS-equipped nanobody (Nb 17-69) to apoE4(ΔSEC1). This is consistent with our own findings [37,38] and those of others [39,40,41,42] showing that nanobodies are functional in the cytoplasm of mammalian cells. Interestingly, cytosolic apoE4 failed to redirect to the nucleus after binding of the NLS-equipped nanobody, while the NLS-tagged nanobody in the absence of apoE4 strongly localized in the nucleus. This behavior contrasts with previous observations using a gelsolin-specific nanobody fused to the same NLS (gelsolin Nb11-NLS), which efficiently relocalized its antigen to the nucleus [37]. We hypothesize that apoE4 may aggregate in the cytosol, thereby preventing its relocalization to the nucleus by an NLS-equipped nanobody. However, future studies will have to confirm this.

In addition, cytosolic expression of apoE4 was associated with the appearance of a lower molecular weight apoE4 fragment of approximately 25 kDa, which was not detected when apoE4 was expressed via the secretory pathway. ApoE4 is known to be structurally less stable than apoE2 and apoE3 and is thereby more susceptible to proteolytic cleavage [33,43]. The hinge region between the N-terminal and C-terminal domains contains numerous protease-sensitive sites and is more exposed in apoE4, making this region especially vulnerable for cleavage [44]. This is concomitant with the ~25 kDa fragment observed here, which could arise from proteolysis within the hinge region. Furthermore, apoE fragments, including N-terminal fragments of ~25 kDa, have been detected in human AD brains [45,46,47]. Our data suggest that the cytosolic localization of apoE4 may promote its susceptibility to proteolytic cleavage.

Together, we present a novel, domain-selective apoE-targeting nanobody toolbox to investigate intracellular apoE4 biology. As an exploratory application, this toolbox was used to probe the behavior of cytosolic apoE4. Our observations are compatible with a model in which cytosolic apoE4 undergoes proteolytic processing and accumulates in aggregates. Future studies will be required to validate this hypothesis and to determine whether such cytosolic behavior of apoE4 contributes to cellular processes relevant to AD.