The phosphoS655 Alzheimer’s Amyloid Precursor Protein (APP) interactome in neuronal differentiation

APP, S655 phosphorylation, SH-SY5Y neuroblastoma cells, neuronal differentiation, neuritogenesis preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 2

The Alzheimer’s Amyloid Precursor Protein (APP) has determinant roles in neuronal development and function, both in its full-length conformation and as some of its proteolytic peptides, particularly secreted (s)APPa. Given that APP phosphorylation tightly regulates its trafficking, proteolysis, and protein-protein binding, it consequently affects several APP functions. The S655 residue, located in the basolateral sorting motif YTSI at APP C-terminus has been observed to be phosphorylated in mature full -length APP and its C -terminal fragments. Previously observed to modify APP’s protein interactions, resulting in altered endolysosomal trafficking , and increased half-life and sAPPa generation, phosphoS655 APP has potential to modulate APP -mediated neuronal differentiation. To study the phosphoS655 differential interactome relevant for neuronal differentiation, SH-SY5Y cells expressing Wt or S655 phosphomutants APP -GFP were differentiated at two time points . APP-GFP and their respective interacting partners were immunoprecipitated using GFP -trap, and interactors identified by mass spectrometry. Both dephospho and phosphoS655 interactomes were generally enriched in similar processes, primarily RNA processing and translation, as well as signal transduction, metabolism, and cytoskeleton remodeling. The smaller phosphoS655 interactome contributes for functional specialization via binding to e.g. FUBP3, ELAVL4, ATXN2, Tubulin, INA. Several of these specific binding partners are known to promot e neurite outgrowth and likely underlie our experimental observation that phosphoS655 APP promotes neuritogenesis, particularly the formation of longer neurit ic extensions. These results are not only important for the body of knowledge on this Alzheimer’s disease core protein, but may also aid in future therapies against this disease. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 3

The Amyloid Precursor Protein (APP) is mainly studied for its association with Alzheimer’s disease (AD). This transmembrane protein is dynamically sorted through the membranes of the cell and intracellular organelles. During trafficking, full -length APP is cleaved by proteases into some functional peptides. Briefly, α-secretases generate sAPPα and a membrane-tethered C-terminal fragment (CTF), whereas cleavage by β-secretase releases the sAPP β along with a CTF. The resulting CTFs are subsequently processed by the γ-secretase complex to produce the p3 or amyloid-β (Aβ) peptides, respectively, and the intracellular AICD fragment (1,2). In addition to the aforementioned classic APP cleavage pathways, other proteolytic processing events have been described for APP (1). Several crucial physiological roles have been attributed to the intricate APP molecule, including most recently in maintaining proteostasis via regulation of TGFβ signaling (3); for comprehensive reviews, see (4,5). Particularly, APP has been described to be determinant for neuronal development and function, with attributed functions in cell migration (6), neuritogenesis (7–9) and synaptic structure, transmission, and plasticity (10–12). Strikingly, loss of APP expression due to a homozygous truncating mutation led to developmental delay, microcephaly, callosal dysgenesis, and seizures in a 20 -month-old infant (13). Several of these functions have been attributed not only to the full-length protein but also to its proteolytic fragments (14–18). However, contradictory

have been reported regarding their role in nervous system development and function (19– 22), and await further investigations. An additional level of complexity is brought by APP regulation through post -translational modifications. As a phosphoprotein, APP phosphorylation tightly regulates its trafficking (23,24), proteolysis (24,25), and protein-protein binding (24,26). APP CTFs and AICD fragments have also been detected in the phosphorylated form (27). One possible APP phosphorylation site is the S655 residue located in the basolateral sorting motif 653YTSI656. Phosphorylation at S655 has been detected in mature APP molecules in rat cortex (28), and in APP-CTFs in the hippocampus of AD patients (29). S655 is described to be phosphorylated by protein kinase C and Rho ‐associated coiled‐coil kinase 1 (28,30,31). Nuclear Magnetic Resonance analyses revealed that APP phosphorylation at S655 induces significant local conformational changes within and downstream of the 653YTSI656 motif (32). These structural alterations likely regulate protein-protein interactions; for instance, the phosphomimetic mutant APP -S655E shows enhanced interaction with the retromer trafficking complex member VPS-35 (23), while APP-CTF-S655E disrupts the interaction with the regulator of lysosomal trafficking AP-3 (33). Ultimately, these alterations in protein-protein interaction directed by S655 phosphorylation lead to enhanced APP half -life and a higher rate of sAPPα secretion, with decreased APP trafficking to lysosomes and reduced A β42 production, when compared to the S655A dephosphomimetic mutant (23,33,34). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 4 The present work sought to identify APP binding partners sensitive to its S655 phosphorylation state, a modification that seems to govern protein-protein interactions essential for APP trafficking and processing, and to examine how these interactions change during neuronal differentiation. To this end, we used the neuritogenic cellular model retinoic acid (RA)-differentiated SH-SY5Y cells transiently overexpressing APP -GFP S655 phosphomutants (S655A and S655E). Cells were collected at differentiation days 3 and 7 given that they represent two distinct stages of RA-induced SH-SY5Y neuronal differentiation. In the first one, associated with an increase in the number of pre-neuritic projections, APP levels rise but holo APP is cleaved at a higher rate into sAPP. Neurites appear more abundantly during the second stage, characterized by neuritic elongation and stabilization of longer neurites in a holo APP-assisted manner (35). The GFPTrap assay was used to immunoprecipitate APP-GFP proteins and their respective interacting partners, which were identified by mass spectrometry (MS). The phophoS655 APP interactome presented a reduced number of protein interactors, when compared to the other categories, denoting a higher degree of functional specificity. In general, the interactors of both S655 dephospho and phosphoAPP were involved in similar biological processes , with some specificities. The phosphoS655 APP interactome presented a special enrichment in some mRNA processing and translation proteins, and in some proteins related to signaling and the cytoskeleton, that are known to interact with each other in a functional network with roles in differentiation, including promotion of neuritogenesis. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 5

SH-SY5Y cell culture, differentiation and transfection The SH-SY5Y human neuroblastoma cell line was maintained in minimum essential media/F12 medium supplemented with 10% fetal bovine serum (FBS). Before differentiation, the culture was enriched in the neuroblastic N -type of cells - immature nerve cells that are differentiated with retinoic acid (RA) into neuronal-like cells - based on their lower substrate adherence comparing to the S-type cells (36). N-type enriched cells were plated at 30,000 cells/cm2 onto 100 mm culture dishes and incubated at day 0 (D0) with complete medium containing 10 µM RA to start neuronal- like differentiation. Cells were kept in a humidified, 37ºC, 5% CO2 incubator and medium was substituted by new medium supplemented with 10 µM RA every other day. Cells were transfected with TurboFectTM, according to the manufacturer’s instructions (Fermentas Life Sciences), at day 2 or day 6 of differentiation with 9 µg cDNA of human APP isoform 695 wild- type (APPWt), S655A dephosphomutant (APPSA), or S655E constitutive phosphomutant (APPSE), fused with GFP (37). The “empty” GFP vector was used as control . The transfection time points were chosen as key regulatory differentiation periods, based on our previous work on time-dependent APP-induced neuritogenic alterations (36). Transfection complexes were kept for 6h on cells, after which medium was changed . After 24h, the efficiency of cell transfections was confirmed by visualization of GFP expressing cells under an Olympus IX81 epifluorescence microscope, and cells collected. GFP Trap IP assays Pull-down of the GFP moiety of the APP-GFP chimeras using GFP-trap® (Chromotek) according to the manufacturer’s instructions. Upon 24h of transfection, SH-SY5Y cells were washed in PBS, and in 1 mL of ice-cold PBS with PMSF (1:100). Cells were collected by scrapping, centrifuged (5 min, 3000 g, 4ºC), and the cells’ pellet lysed for 30 min with 500 µL non-denaturant Lysis buffer (10 mM Tris HCl pH 7.5, 0.5 mM EDTA, 0.5% Gepac -ca-630, 150 mM NaCl, 1 mM PMSF) supplemented with protease inhibitors cocktail (Sigma -Aldrich), 1 mM NaF and 10 mM Sodium Orthovanadate (phosphatase inhibitors). Samples were further centrifuged for 5 min at 20,000 g, an aliquot of supernatant (‘cells lysates’) used to confirm transfection via immunoblot , and the remaining supernatant incubated with previously washed GFP-trap beads (25 µL/sample, 3h with orbital shaking ). Beads were magnetically separated until the supernatant was clear and resuspended in wash buffer. Magnetic separation and washing step were repeated 4 times. Beads were resuspended in 100 µL of 2x SDS-sample buffer and boiled for 10 min at 95ºC to dissociate immunocomplexes from GFP-trap beads. After vortexed, beads were magnetically separated, and preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 6 the supernatant transferred to a new microtube. Samples were immediately stored at -20ºC for downstream mass spectrometry. Mass spectrometry (MS) For MS analysis, samples were defrosted, 45 µL of each sample were loaded on a 12% SDS - PAGE gel (Bio-Rad 345-0117) and run for 2h under a constant amperage of 20 mA. Afterwards the gel was fixed for 30 min in 40% MeOH, 10% HAc solution and further stained with Coomassie (Gelcode blue stain, Pierce). Each lane was divided in 3 equal gel bands that were excised and cut into small ( ~1 mm 3) pieces. The gel pieces were washed, in -gel reduced with 6.5 mM dithiothreitol (RT, 60 min), alkylated with 54 mM iodoacetamide (dark, RT, 30 min), and digested by adding trypsin at a concentration of 3 ng/μL (overnight at 37 °C), as described in [23]. Peptides were subsequently extracted with 100% acetonitrile (Biosolve) , dried by vacuum centrifugation, dissolved in 45 µL 10% formic acid, and spun (10 min, 20000 rpm) prior to analysis. 20 µL of each sample was analyzed on a Q exactive plus (Thermo Fisher Scientific, Bremen) connected to Thermo Scientific EASYnLC 1000 (Thermo Fisher Scientific, Bremen). All columns were packed in-house. The trap column was a double fritted 100 µm inner diameter capillary with a length of 20 mm (Dr Maisch Reprosil C18, 3 μm). The analytical column (Agilent Poroshell EC- C18, 2.7 μm) had an id of 75 µm and a length of 50 cm, and was heated to 40°C. Trapping of the sample was performed at a flow rate of 100 n L/min for 10 min in solvent A (0.1 % formic acid in water), and elution performed with a gradient of 7 –38% solvent B (0.1% formic acid 100% acetonitrile) in 75 min, 38–100% B in 3 min, 100% B for 2 min, at a maximum pressure of 800 bar. Nano spray was achieved using a distally coated fused silica emitter (made in-house, o.d. 375 μm; i.d. 20 μm) biased to 1.7kV. The mass spectrometer was operated in the data dependent mode to automatically switch between MS and MS/MS. MS full scan spectra were acquired from m/z 375– 1600 after accumulation to a target value of 3x106. Up to ten most intense precursor ions were selected for fragmentation. HCD fragmentation was performed at normalised collision energy of 25% after the accumulation to a target value of 5x104. MS/MS was acquired at a resolution of 17500. MS Data Processing MS raw data were processed using Proteome Discoverer (PD) 2.5.0.400 software (Thermo Scientific, Bremen, Germany). Protein identification analysis was performed using data from UniProt protein sequence database for the Homo sapiens Reviewed Proteome 2024_0 1 (83,385 entries) along with a common contaminant database from MaxQuant (version 2.6.7.0, Max Planck Institute of Biochemistry, Munich, Germany). FASTA files for the mutant versions of APP were also included to ensure accurate protein identification. Two protein search algorithms were applied: (i) the mass spectrum library search software MSPepSearch, with the NIST human HCD Spectrum preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 7 Library and (ii) the Sequest HT tandem mass spectrometry peptide database search program. Both search nodes used an ion mass tolerance of 10 ppm for precursor ions and 0.02 Da for fragmented ions. The maximum allowed number of missing cleavage sites was set as 2. Cysteine carbamidomethylation was defined as a constant modification, while methionine oxidation, asparagine and glutamine deamidation, peptide N -terminus cyclization (Gln->pyro-Glut), protein N-terminus acetylation, and methionine loss (Met -loss and Met -loss + Acetyl) were defined as variable modifications. Peptide confidence was set to high, the Inferys rescoring node was applied. Data was validated by the Percolator node. Protein-label-free quantitation was performed with the Minora feature detector node at the processing step. Precursor ion quantification was performed at the consensus step with the following parameters: peptides = unique plus razor; precursor abundance ba sed on intensity; normalization mode based on the total peptide amount; and pairwise protein ratio calculation and hypothesis testing based on a background-based t-test (38,39). The dataset of 1,561 proteins identified by the Proteome Discover er across all conditions (six protein lists: APPWt, APPSA, APPSE, either at day 3 or 7 of differentiation), was further filtered in R Studio (version 4.4.3 ) to retain only high -confidence identifications. Only master proteins with ‘high’ FDR confidence were further considered. Proteins with only one unique peptide were discarded, unless they presented a molecular weight (MW)≤20 kDa and a sequence coverage of at least 20% , to minimize bias against smaller proteins in peptide -based identification (40,41). Common contaminant proteins and proteins associated to hair, skin/epidermis, tongue and gums (UniProt IDs: P02533, P13645, Q5T749, P35527, Q92764, P04264, Q7Z794, P13646, P19013, Q9NSB4) were removed. Following, only proteins present in two biological replicates were considered, and nonspecific APP interactors/proteins likely binding to the GFP tag, were further excluded. For this, the mean of each protein’s abundance in APPWT, APPSA and APPSE conditions were calculated, and their ratios to their respective abundances in control (GFP tag alone) were computed. Only proteins with a resulting fold change ³1.5 relative to control were retained for further analysis. Proteins were considered ‘specific’ of a condition if only identified in that condition (either APPSA or APPSE) ; proteins were considered ‘ enriched’ in a specific APP mutant if their detected abundance was ≥ 2 relative to the other mutant . Total APP interactome includes 233 proteins identified on this work. Bioinformatic analyses A Principal Component Analysis (PCA) was performed to assess clustering and overall variability among the experimental groups. Further, functional classification of the interactors based on the Protein Class ontology was performed with PANTHER (Protein ANalysis THrough Evolutionary Relationships) database. Significantly enriched Gene Ontology (GO) terms, specifically Biological preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 8 Processes (BP), were analysed using the ClueGO plugin (version 2.5.10 ; terms updated on 5 th May 2025) in Cytoscape (version 3.10.3) (42,43), (44,45). The proteins lists corresponding to interactors identified at day 3 and 7 of differentiation for each group here considered – APPSA, APPSE and Total APP interactome (all APP interactors pulled down in the three conditions: Wt, SA, SE), were input into ClueGO simultaneously as separate clusters, and a Functional Clustered analysis was performed to enable comparison of functional enrichment between the two timepoints within each group. The following parameters were applied: a minimum of 3 genes per term or 4% of the term’s genes present; merging of redundant terms was activated, meaning terms with more than 50% overlap were fused, leaving only the most significant term based on p-value. Pathway overrepresentation analyses were performed on g:Profiler (version updated on Jun 16, 2025) (46). For each group and timepoint, the protein lists were input separately, and KEGG, Reactome and WikiPathways were used as pathway sources (Agrawal et al., 2024; Griss et al., 2020; Kanehisa & Goto, 2000; Kolberg et al., 2023). Enrichment s were assessed using the hypergeometric test, with p-values adjusted by the Benjamini-Hochberg correction. Only terms with an adjusted p -value ≤ 0.05 were considered statistically significant, using Homo sapiens [9606] proteins detected in the brain according to The Human Protein Atlas as background (47). Protein-protein interaction (PPI) networks were constructed for the interactors of groups ‘Total’, APPSA and APPSE , at each timepoint. PPIs were constructed usi ng Homo Sapiens data from BioGRID 5.0 (updated on Dec. 25th, 2025) (48), and analysed in Cytoscape (version 3.10.3) (43). The Cytoscape plug-in clusterMaker2 (49) was used to identify densely connected regions within the PPI networks using the MCODE algorithm, with default parameters (Degree Cutoff = 2, Node Score Cutoff = 0.2, K -Core = 2, Max Depth = 100). For the exclusive and enriched APPSE interactors, proteins were grouped into tables based on common cellular processes and reported roles on neuronal differentiation. The Uniprot database was used to manually retrieve each protein’s name, associated GO-BPs, and reported expression and involvement in disease (50). The protein summary information was consulted in NCBI for the individual proteins. Phenotype information was obtained from the Mouse Genomics Informatics (MGI) database, which integrates comprehensive genetic, genomic, and phenotypic data from laboratory m ouse studies (51). Proteins Q96QA5, P68871, A0A0C4DH42, A0A0B4J1Y9 and P0CG12 were excluded from Table 1, because, although expressed in the brain, some are likely contaminants, while others lack sufficient annotation or evidence linking them to neuronal development. APP-GFP/interactor subcellular distribution assays and cell morphology studies SH-SY5Y cells were seeded on poly -D-Lysine (Sigma Aldrich, P6282) precoated coverslips, differentiated, and transfected with APPSE or APP-SA at 2 or 6 DIV, as previously described. At 3 and 7 DIV, cells were fixed with 4% paraformaldehyde for 15–20 min at RT. Samples were further permeabilized with 0.2% Triton X -100 in PBS for 10 min and blocked with 3% bovine serum preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 9 albumin (BSA) in PBS for 1 h. Primary antibodies diluted in blocking solution (1:50) were incubated for 2 h at RT: mouse anti-HuD (alias ELAVL4, E-1, sc-28299), mouse anti-FXR2 (1G2, sc-32266), mouse anti-α-internexin (G-9, sc-271302), anti-FBP3 (alias for FBUP3, E -8, sc-398466) (Santa Cruz Biotechnology), rabbit anti -ATXN2 (GeneTex, GTX130329), rabbit anti-α-tubulin (TU-01, Novus Biologicals, NB500-333). Secondary Alexa Fluor 594-conjugated goat anti-mouse (A11005) and anti -rabbit (A11012) antibodies (Inv itrogen), diluted in blocking solution (1:300), were incubated for 1 h at RT. Coverslips were mounted with DAPI -containing Vectashield antifading mounting medium (Vector, H-1200) and images were acquired using a Zeiss LSM 880 Airyscan confocal microscope (100x oil objective). Morphometric analysis was performed on differentiated SH-SY5Y cells, transfected with the three APP-GFP cDNAs (Wt, SE or SA) at 6 DIV and fixed at 8 DIV, for neurites to have more time to develop under a specific APP overexpression background. 30 randomly selected digitized images were analyzed per sample, with an average of 40 transfected cells (GFP expressing) per biological replica of each condition (Wt, SE, or SA). Measurements were performed using ImageJ XXX on matching PhC microphotographs. The number and length of processes per cell were quantified and categorized as follows: <20 µm, 20-35 µm, 35-50 µm, and ≥50 µm. Processes ≥35μm were already considered neurites. Processes mean length only includes processes ≥20 µm. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 10

S655- and differentiation time-dependent APP interactors and their functional classes This study aimed to gain insight on the role of APP S655 phosphorylation in neuronal differentiation, by identifying APP protein interactors dependent on its S655 phosphorylation state. Neuronal-like differentiated SH-SY5Y cells, overexpressing Wt or S655 phosphomutants APP - GFP cDNAs for 24h prior to collection, were used as the protein pool. The GFP Trap assay was used to immunoprecipitate the APP proteins (APPWt, APPSA, and APPSE) and their respective interacting partners, which were subsequently identified by mass spectrometry. In total, 233 proteins were identified across all groups and timepoints, of which 64 were already described APP interactors, while 169 represent newly identified APP interactors. The number of interactors identified for each group and timepoint, along with their intersections, are shown in Venn diagrams (Figure 1A). From differentiation day 3 (D3) to day 7 (D7) , APPWt and APPSE groups present a relatively stable number of interactors (n=110à108 and n=78à72, respectively), whereas the number of APPSA interactors increased substantially (n=97à163). Of note, only proteins appearing in at least two replicates were here considered ; if all interactors with a fold change >1.5 relative to the GFP control were considered, all groups would exhibit an increase in interactor number over the course of differentiation (Supplementary Figure 1A), suggesting that the amount of APP interactors increases as the cell specializes into a neuron. The number of APPWt interactors is intermediate to the ones of the phosphomutant groups, consistent with the fact that APPWt can be either S655 phosphorylated or not (Figure 1A). As such, both APPSE and APPSA share most of their interactors with APPWt on D3: 83.3% for APPSE and 86.6% for APPSA (with 60 interactors common to the two phosphomutants). This overlap decreases to 50.3% (APPSE) and 58.3% (APPSA) at D7, suggesting that S655 phosphorylation imposes more specification as the cells become more neuron al-like, negatively regulating a higher number of APP interactions at D7 . Consistent with a specialization of the APP interactome upon S655 phosphorylation, Principal Component Analysis (PCA) of all replicates showed APPSE samples tightly clustered, APPSA samples widely dispersed, and APPWT samples cluster ed either with APPSE or with APPSA (Supplementary Figure 1A). The main proteins contributing for groups separation at each day are presented in Supplementary Figure 1B . APPSE maintained ~20% exclusive interactors at both D3 (18, 23.1%) and D7 (16, 22.2%), while the number of APPSA exclusive interactors was higher and markedly increased from D3 (37, 38.1%) to D7 (107, 65.6%). Figure 1B shows that the identity of APP interactors partially change s as the cell s acquires neuronal traits, such as a more neuronal -like proteome and morphology. For all the subsequent analyses, the identified interactors were subgrouped into three categories: 1) Total APP interactors, which includes all APP interactors identified in this work (131 at D3; 199 at D7); 2) APPSA interactors (97 at D3; 163 at D7); and 3) APPSE interactors (78 at D3; 72 at D7). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 11 Figure 1. APP protein interactors change with S655 phosphorylation state and differentiation time. A. Venn diagrams presenting the numbers of interactors experimentally detected in the pull -downs of APPWt and the S655 phosphomimetic mutants (APPSA and APPSE), and the intersections between these conditions, at both differentiation days. Bellow: Barplots representing the number of interactors in the different groups on each differentiation day. B. Comparison of the numbers of APP interactors (APP included) on each differentiation day for Total, APPSA and APPSE interactors. C. Protein class classification according to PANTHER (Protein ANalysis THrough Evolutionary Relationships) for ‘Total’, APPSA and APPSE groups. D3 and D7, days 3 and 7 of neuronal differentiation. Only classes with n=2 interactors at t least one of the differentiation days were included. A. B. Total APP interactors (Wt+SA+SE) APP-SE interactors APP-SA interactors D3 D7 72 C. WT 163 WT Protein Classes @ D3 & D7 Total APP interactors APPSA Interactors APPSE Interactors preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 12 To start attributing putative functions to the phosphomutant interactors, a functional classification was performed based on the “Protein Class” ontology from PANTHER (Protein ANalysis THrough Evolutionary Relationships) database (Figure 1C) . In the Total APP interactors group, the dominance of RNA metabolism proteins is notable at both days, with around 60 interactors related to this class in each time point. Other classes mainly represented across differentiation time are Translational (9% at D3; 11% at D7 ) and Cytoskeletal proteins (4.5% at D3; 5.9% at D7). At D3 only two other classes have ³3% of interactors, namely Defense/immunity (3.7%) and Protein- binding activity modulator (3%). Some classes increased from D3 to D7, particularly ones related to post -translational modifications (PTMs), such as Metabolite interconversion enzymes (from 0.7% to 9.9%) and Protein modifying enzymes (0.7% to 5%), followed by Chaperones (0.7% to 3%). APPSA interactors account for most of the total APP interactors, which explains why their protein class distribution is very similar to the Total APP interactors one, with RNA metabolism proteins being the most abundant, followed by Translational and Cytoskeletal proteins, at both differentiation time points. Apart from a more pronounced decrease in RNA metabolism with time of differentiation, the remaining distributions for APPSA interactors mirror the total APP profile, with a similar increase in PTM-associated proteins (including metabolism-related ones). The APPSE group, however, presented some particularities. Although RNA metabolism remained the dominant class, it only represented one-third of the classes (orange arrow) and this class only slightly decreased at D7 (from 36.7% to 31.1%). At D3, proteins related to Translation (12.7%; vs 7.1% for APPSA) and Defence/immunity (double % relatively to APPSA) were more abundant in this group (blue arrows) . Of note, the Defence/immunity class is mainly composed of ribosomal protein subunits (RPS) and one chaperone, strengthening a pS655-dependent enrichment of Translation- related functions. At D7, there was an APPSE-specific growth in the Gene-specific transcriptional regulator (from 1.3% to 4.1%) and Protein-binding activity modulator (from 2.5% to 4.1%) classes, a trend not observed in APPSA . Conversely, the Metabolite interconversion and Chaperones classes increased less in percentage than in APPSA. Biological Processes and Pathways of APP S655 phosphomutants protein interactors Subsequent functional enrichment analyses were conducted to explore the biological processes (BP) and pathways enriched in the APP protein interactors groups. Figure 2 shows the APPSA - and the APPSE -enriched BP terms , colour-categorized by differentiation day . Terms were classified as mainly enriched in D3 ( not found ) or D7 (yellow nodes) if ≥75% of their protein members were detected only on that day. Otherwise, terms were considered enriched in both days (green nodes). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 13 Figure 2. Biological processes (BP) gene ontology (GO) terms for the APPSA (A) and APPSE (B) total groups. BPs specifically enriched at both differentiating days (D3 & D7) are presented in green and those enriched at D7 are in yellow. No terms only enriched at D3 were found. Terms were considered enriched in only one day if 75% or more of its protein members were found only in that day. Only terms with p<0.05 upon Benjamini-Hochberg multiple test correction are represented. The size of the term nodes is based on term’s significance (higher, the lower its p -value). Terms with more than 50% overlap are fused and only the more significant term is presented in this network. regulation of androgen receptor signaling pathway generation of precursor metabolites and energy protein refolding protein localization to lysosome telomere maintenance via telomerase negative regulation of protein catabolic process CRD-mediated mRNA stabilization thioredoxin peroxidase activity regulation of protein import into nucleus cellular response to steroid hormone stimulus negative regulation of catalytic activity cellular response to hypoxia ribonucleoprotein complex biogenesis regulation of RNA splicing spliceosomal complex assembly mRNA catabolic process protein-RNA complex assembly regulation of RNA stability purine-containing compound biosynthetic process RNA localization RNA splicing negative regulation of cell growth protein folding chaperone maintenance of protein localization in organelle intrinsic apoptotic signaling pathway in response to oxidative stress positive regulation of viral genome replication T cell migration mRNA export from nucleus androgen receptor signaling pathway regulation of spindle assembly translation regulatory ncRNA-mediated gene silencing positive regulation of translation positive regulation of intracellular protein transport nucleotide metabolic process primary miRNA processing negative regulation of phosphate metabolic process intrinsic apoptotic signaling pathway alternative mRNA splicing, via spliceosome glycolytic process purine ribonucleotide metabolic process negative regulation of protein modification process nicotinamide nucleotide metabolic process cellular oxidant detoxification telomere maintenance purine nucleotide metabolic process nucleotide biosynthetic process plasminogen activation regulation of nucleocytoplasmic transport mRNA alternative polyadenylation purine ribonucleoside triphosphate metabolic process regulation of protein stability positive regulation of nucleocytoplasmic transport co-transcriptional mRNA 3'-end processing, cleavage and polyadenylation pathway purine ribonucleoside triphosphate biosynthetic process mRNA stabilization negative regulation of nuclear-transcribed mRNA catabolic process, deadenylation-dependent decay negative regulation of catabolic process positive regulation of viral process regulation of protein localization to nucleus nuclear-transcribed mRNA catabolic process homeostasis of number of cells positive regulation of protein localization to nucleus RNA processing regulation of telomere maintenance via telomerase carbohydrate derivative catabolic process positive regulation of protein polymerization positive regulation of non-canonical NF-kappaB signal transduction post-transcriptional regulation of gene expression RNA transport regulation of oxidative stress-induced intrinsic apoptotic signaling pathway regulation of intracellular transport regulation of mRNA processing regulation of mRNA metabolic process regulation of alternative mRNA splicing, via spliceosome cytoplasmic translation translational initiation negative regulation of gene expression positive regulation of intracellular transport RNA destabilization negative regulation of mRNA metabolic process spliceosomal snRNP assembly regulation of translation positive regulation of type I interferon production regulation of synapse organization regulation of protein modification by small protein conjugation or removal regulation of translational initiation protein stabilization protein folding type I interferon production regulation of DNA-binding transcription factor activity negative regulation of protein modification by small protein conjugation or removal interferon-beta production nuclear export negative regulation of response to endoplasmic reticulum stress mRNA processing nucleocytoplasmic transport nucleobase-containing compound transport positive regulation of interferon-beta production import into nucleus negative regulation of ubiquitin-dependent protein catabolic process RNA catabolic process negative regulation of RNA splicing spindle assembly positive regulation of cytoplasmic translation mRNA metabolic process regulation of intracellular protein transport negative regulation of apoptotic signaling pathway mRNA splicing, via spliceosome DNA endonuclease activity negative regulation of kinase activity negative regulation of protein metabolic process regulation of intrinsic apoptotic signaling pathway negative regulation of translation protein localization to nucleus protein heterotetramerization negative regulation of intrinsic apoptotic signaling pathway regulation of cell morphogenesis nucleobase-containing compound catabolic process regulation of mitotic spindle organization A. APP SA D3 & D7 Day 7Biological processes regulation of mRNA metabolic process CPSF7 U2-type prespliceosome assembly CPSF6 co-transcriptional mRNA 3'-end processing, cleavage and polyadenylation pathway NUDT21 mRNA alternative polyadenylation mRNA processing SSBP1 DDX46 protein tetramerization PRPF4 PHB2 PHF5A PRDX3 SF3B5 positive regulation of DNA-binding transcription factor activity SNRPB2 RNA destabilization SNRPA1 RBM27 SF3B6 regulation of RNA stability MFAP1 regulation of RNA splicing HTATSF1 RPS3 SF3B3 NUMA1 SNRNP200 positive regulation of microtubule polymerization U2AF2 protein-RNA complex assembly SF3A1 mRNA metabolic process mRNA splicing, via spliceosome RPS26 SNRPF RTCB SNRPD2 DDX1 SNRPB RNA splicing PRPF6 TAF15 spliceosomal snRNP assembly FXR1 WTAP FASTKD3 SFPQ CAPRIN1 HNRNPA1 HSPA1B RBMX PABPC1 DDX17 GIGYF2 alternative mRNA splicing, via spliceosome ELAVL4 negative regulation of protein modification by small protein conjugation or removal CIRBP negative regulation of mRNA metabolic process mRNA catabolic process Exclusive APP SE BP positive regulationof microtubulepolymerization positive regulation of DNA - binding transcription activity U2-type prespliciossome assembly protein tetramerization B. APP SE preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 14 The APPSA BP network was composed of 120 terms, 86% of which exclusive (not detected in APPSE). Prominent in this network were processes related to RNA metabolism, stability, transport and polyadenylation, together with clusters associated with translational initiation and nuclear- cytoplasmic transport/localization (all shared between D3 & D7; the nuclear-cytoplasmic transport/localization become more specialized at D7 , with added members ). Processes mainly enriched at D7 were related to PTM (mainly negative regulation of phosphorylation) , protein stability and nucleotide metaboli sm. Th ere were also smaller clusters of processes related to apoptotic signalling regulation , cellular response to steroid hormone, innate defence (all these shared at D3 & D7), plus protein refolding and telomere maintenance (enriched at D7). Relevant isolated BP terms of APPSA interactors included “Regulation of synapse formation ”, “Positive regulation of NF-kB signal transduction” (both days), plus "Protein localization to lysosome" and “regulation of cell morphogenesis” at D7. The network of enriched BP associated to the APPSE interactors was formed by 20 terms, most of them shared with APPSA (80%, non-delineated circles). Like APPSA, APPSE interactors were strongly associated with RNA -related processes on both days , particularly mRNA metaboli sm, splicing and stability, and protein–RNA complex assembly. Processes enriched at both days but exclusive of APPSE were “Protein tetramerization” (that shared proteins with an RNA processing cluster) and “U2-type prespliceossome assembly ” (sharing proteins with a spliceosome-related cluster). A role in RNA splicing was further strengthened with the D7-exclusive term “Regulation of RNA splicing”. Interestingly, two other terms were exclusive to phosphoS655 at D7 : “ Positive regulation of microtubule polymerization” and “Positive regulation of RNA -binding transcription factor activity”. Pathway enrichment analysis retrieved 47 pathways for APPSA interactors and 56 for APPSE interactors at D3, and 113 (APPSA) and 21 (APPSE) pathways for D7 interactors. Many of these were shared by both groups, with the top 20 shared pathways all related to RNA processing and translation, at both differentiation days (Supplementary Figure S2). Nevertheless, at D7 APPSE had a higher percentage of pathways related to translation ( APPSE D3: 35.7%, D7: 47.6% vs. APPSA D3: 31.9%, D7: 18.6%) and to RNA metabolism (APPSE D3: 32.1%, D7: 33.3% vs. APPSA D3: 34%, D7: 17.7%). In contrast, at D7 , APPSA showed greater representation of pathways related to transcription factors (APPSA D3: 4.3%, D7: 4.4% vs. APPSE D3: 3.6%, D7: 0.0%) and signalling (APPSA D3: 6.4%, D7: 12.4% vs. APPSE D3: 1.8%, D7: 0.0%). Exclusively enriched pathways were observed mainly at D3 for APPSE and at D7 for APPSA. At D3, APPSA had only three exclusive pathways (6.4%), two related to signalling and one to trafficking, whereas APPSE presented 12 exclusive pathways (21.4%). These included eight related to rRNA and translation, two linked to viral host response (potentially translation, given shared proteins with the previous terms), and two associated with neuronal differentiation: “Axon Guidance” and “Nervous system development”, associated to proteins PABPC1, RPS5, RPL38, RPS2, RPS18, RPL13, RPL30 and preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 15 RPS27 (Supplementary Figure S2, D3). At D7 of differentiation, a stage marked by a boost in APP interactors binding to APPSA, the pattern was reversed and APPSA presented a very high number of pathways, including 93 exclusive ones (82.3%). Top 20 APPSA pathways (prioritized by p-value) were all related to translation, including quality control of mRNA (Supplementary Figure S2, D7). APPSA exclusive pathways at D7 were generally related to translation (15), signalling events (14), RNA metabolism (13), regulation of expression (5), protein trafficking (5), PTM (4) and neuronal processes (4). APPSE binding partners form highly interconnected protein-protein interaction networks Protein-protein interactions (PPI) networks were constructed with all APPSE interactors identified at D3 or D7 ( Figure 3A left and right, respectively). Previously described APP interactors are outlined in red, includ ing 9 proteins at D3 and 10 at D7, while the remaining proteins represent novel APP interactors identified in this study (D3: 88.6%, D7: 86.1%). A closer look to the APPSE networks of Figure 3A reveals that six of the known APP binders – DDX1, H1-2, S100A8, SF3B6, and FUBP3 (52,53)– are exclusive or enriched APPSE binders, with the first two having FXR1 as a common binder (54–56). FXR1 also binds the D3+D7 exclusive SE interactor ATXN2, to which FUBP3 also binds (Figure 3A and B) (54). The D3 APPSE PPI network comprised 64 nodes connected by 340 edges, with a clustering coefficient of 0.42, indicative of a moderately modular organisation with a large interconnected component and a few peripheral groups. The D7 network ha d 60 nodes and 245 edges and , although the average number of neighbours slightly decreased from 10.6 to 8, the clustering coefficient remained the same (0.43), as well as the path length (2.2 for both days). Both networks connected most of the APPSE interactors identified in this work, revealing that many of these had already been reported to interact and likely participate in common functions. Only a small part of APPSE binders, 15 at D3 and 2 at D7, were excluded from the se PPI visualizations, given the absence of known interactors within the PPI. Automatic clustering of Figure 3A networks identified four protein clusters at D3 (Figure 3B, 1-4) and two clusters at D7 (Figure 3B, 5-6). Panther protein class analysis of each cluster indicated that D3’s cluster 1 is composed by RNA metabolism proteins (12 out of 14) , cluster 2 consisted of only translation-related proteins, cluster 3 of both RNA metabolism and translation proteins (4 and 2 out of 9, respectively), while the smaller cluster 4 has one cytoskeletal protein (DSP) and a cell adhesion molecule (DSG1). At D7, cluster 6 is mostly composed of RNA metabolism proteins (8 out of 10) and a scaffold/adaptor protein, and cluster 2 is also mainly composed by RNA metabolism proteins (6 out of 7) and a gene specific transcriptional regulator . Nearly all these clusters included exclusive and enriched APPSE interactors, indicating that S655 phosphorylation modulates interactions within key APP functional complexes. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 16 Figure 3. APPSE protein -protein interaction (PPI) networks at D3 and D7 of neuronal differentiation. A. Full APPSE interactors PPI network. B. Clusters formed by MCODE using the full APPSE networks. Node sizes in A. and B. are based on the node degree of each interactor in the total APPSE network. To further test if S655 phosphorylation promotes the formation of specific protein -protein complexes, PPI networks with only APPSE exclusive and enriched interactors were subsequently created (Figure 4A). A high number of interactions was visible, with the D3 network presenting 19 interactors (out of 26) and 34 edges, and the D7 network presenting 14 interactors (out of 21) and 19 edges. Hence, at both differentiation days, and using only experimentally validated interactions, more than two thirds of the APPSE exclusive/enriched interactors bind together in a network that includes APP, independently of the common APPSA/APPSE binders (in green in Figure 3). SNRPF KHDRBS3RTCB RPS27 DDX1 RPS18 NUDT21 HTATSF1 CPSF6 CHERP FXR2 H1-2 SF3B3 RPL38 SF3B5 EWSR1 PHF5A NUMA1 RPS2 PPP1CB RPS5 PPIH HNRNPK PCF11 CPSF7 CDKN2AIP PABPC1 S100A8 ATXN2 HBB SNRPB DSG1 SNRPB2 RBM14 SF3A1 MFAP1 SF3B6 WTAP APP DDX46 RPL13 RECQL RPL30 RBM27 CIRBP GIGYF2 SNRPA CCDC124 SFPQ ELAVL4 DSP INA JUP DCD CAPRIN1 SSBP1 NUFIP2 NIPSNAP1 UBAP2L LSM12 FXR1 PUF60 SNRPD2 DHX15 SNRPA PUF60 PRPF6 SNRNP200 HNRNPA1 RBMX SNRPD2 DDX17 TAF15 DDX46 PCBP1 SF1 U2AF2 PRPF4 DHX15 SNRPA1 A. B. D3 D7 SF1 U2AF2 PRPF4 DHX15 SNRPA1 SNRPA PUF60 PRPF6 SNRNP200 HNRNPA1 RBMX SNRPD2 DDX17 TAF15 DDX46 PCBP1 ATXN2 SNRPB DSG1 SNRPB2 SF3A1 SF3B6 DDX46 RPL13 RPL30 SNRPA SFPQ DSP JUPNUFIP2 UBAP2L LSM12 FXR1 PUF60 SNRPD2 DHX15 SNRPF RPS27 RPS18 NUDT21 CPSF6 CHERP FXR2 SF3B3 RPL38 SF3B5 PHF5A RPS2 RPS5 5 ATXN2 SNRPB DSG1 SNRPB2 SF3A1 SF3B6 DDX46 RPL13 RPL30 SNRPA SFPQ DSP JUPNUFIP2 UBAP2L LSM12 FXR1 PUF60 SNRPD2 DHX15 SNRPF RPS27 RPS18 NUDT21 CPSF6 CHERP FXR2 SF3B3 RPL38 SF3B5 PHF5A RPS2 RPS5 ATXN2 SNRPB DSG1 SNRPB2 SF3A1 SF3B6 DDX46 RPL13 RPL30 SNRPA SFPQ DSP JUPNUFIP2 UBAP2L LSM12 FXR1 PUF60 SNRPD2 DHX15 SNRPF RPS27 RPS18 NUDT21 CPSF6 CHERP FXR2 SF3B3 RPL38 SF3B5 PHF5A RPS2 RPS5 ATXN2 SNRPB DSG1 SNRPB2 SF3A1 SF3B6 DDX46 RPL13 RPL30 SNRPA SFPQ DSP JUPNUFIP2 UBAP2L LSM12 FXR1 PUF60 SNRPD2 DHX15 SNRPF RPS27 RPS18 NUDT21 CPSF6 CHERP FXR2 SF3B3 RPL38 SF3B5 PHF5A RPS2 RPS5 3 4 Phosphoexclusive interactors Phosphoenrichedinteractors Non-exclusive/non-enrichedSE groupinteractors PreviouslydescribedAPP interactors 621 FUBP3 MTHFD1 DDX46 EIF4H PCBP1 NIPSNAP1 EEF2 FABP5 GANAB PA2G4 PRKCSH TLN1 HIST1H1C CPS1 RPS27 LUC7L2 NCOA5 SF1 FXR2 U2AF2 NUMA1 PRPF4 DHX15 PPIH PPP1CB SNRPA1 TUBA1B SNRPA ANXA1 PSMA1 PHB2 PUF60 CPSF7 ACTB POLDIP3 PRPF6 CDKN2AIP SNRNP200 CD44 HNRNPA1 PRDX3RBMX MFAP1 SNRPD2 G3BP2 APP ARID1B DDX17 WTAP RPS26 ATXN2 RPS3 ATAD3A WBP11 GIGYF2 HTATSF1 SYMPK PTMA SSBP1 TAF15 RBM27 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 17 Figure 4. APPSE exclusive and enriched interactors at D3 and D7 of neuronal differentiation. A. APPSE exclusive (only detected on APP SE) and enriched (ratio APPSE / APPSA ≥ 2) interactors PPI network; node size in is static. B. Relative abundances of exclusive and enriched APPSE interactors in each group (APPWt, SE, SA), at day 3 (D3) and day 7 (D7) of differentiation. Relative abundances are only shown if the interactor is detected on the group (FC ≥1,5 relative to control). Average values of a protein’s abundance in the replicate of a group, minus the median abundance of that protein in the empty GFP vector, in same day (D3 or D7). Figure 4B shows that the abundance of these APPSE exclusive and enriched interactors varied across differentiation time points . At D3, ELAVL4 and SF3B6 presented the highe st relative abundances, whereas TUBA1B was the most abundant at D7. Among the APPSE exclusive interactors, 15 proteins (IGHV3-66, H1-2, SNRPB, RPS2, IGHV3-72, RPS27, SF1, DERPC, INA, SSBP1, PPP1CB, NCOA5, PRKCSH, FUBP3, SYMPK) were only co-immunoprecipitated with A. RTCB ATXN2 NIPSNAP1 H1-2 SNRPB DDX1 PPIH SSBP1 RPS27 RPS2 HGHV3-66 DDX46 FASTKD3 DERPCHGHV3-72 HBB SF3B5 S100A8 FXR1 APP PHF5A KHDRBS3 ELAVL4 SF3B6 FXR2 DCD RTCB DCD FXR2 SF3B6 ELAVL4 KHDRBS3 PHF5A APP FXR1 S100A8 SF3B5 HBB HGHV3-72 DERPC FASTKD3 DDX46 HGHV3-66 RPS2 RPS27 SSBP1 PPIH DDX1 SNRPB H1-2 NIPSNAP1 ATXN2 RBMX DHX15 GANAB PLEKHA6 PRKCSH PPP1CB SSBP1 ATXN2 FUBP3 SYMPK PRPF6 WTAP SF1 U2AF2 TUBA1B NCOA5 APP RBM27RPS27 DERPC INA RPS27 RBM27 APP NCOA5 TUBA1B U2AF2 SF1 WTAP PRPF6 SYMPK FUBP3 ATXN2 SSBP1 PPP1CB PRKCSH GANAB DHX15 RBMX DERPC INA PLEKHA6 B. APPSE Exclusive interactor APPSE Enriched interactor APPSE Exclusive interactor APPSE Enriched interactor D3 D7 Phosphoexclusive interactors Phosphoenrichedinteractors PreviouslydescribedAPP interactors D3 D7 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 18 APPSE and not with APPWt, suggesting that a sustained S655 phosphorylation may be required for their binding to APP . Three of these interactors (DERPC, RPS27, SSBP1), together with ATXN2, were detected as exclusive APPSE interactors at both differentiation days . While the neuritogenesis-associated ATXN2 was the only interactor whose relative abundance increased with differentiation time (from ~60 at D3 to ~90 at D7), the abundance of the ribosome biogenesis and mitochondrial stability-associated DERPC, RSP27 and SSBP1 proteins decreased from D3 to D7 (about half for RPS27; SSBP1 became nearly undetectable). S655 phosphorylation potential functions in neuronal differentiation The smaller APPSE interactome suggests a functional specification for this phosphorylation. Further, most of the exclusive and enriched APPSE interactors interact with one another (Figure 4A), suggesting that they may act synergically as part of functional modules relevant to differentiation. To explore this potential functional specification in neuronal differentiation , we performed literature mining to extract the known functions of the APPSE exclusive and enriched interactors, followed by manual functional clustering. Table 1 groups these interactors by cellular process and outlines the proposed roles of each functional cluster in neuronal differentiation. Spliceosomal and core pre-mRNA processing factors (cluster 2) remained largely present at both time points, as well as proteins annotated for mRNA stability, transport and ribosome/translation functions (cluster 3, with several RBPs and ribosomal proteins) . These are related to neuronal differentiation by driving neuron-specific alternative splicing and mRNA processing programs that generate isoforms required for neuronal fate specification, axon guidance and synapse formation. In parallel, RNA-binding proteins, mRNA transport factors and ribosomal components support the localization and activity -dependent local translation of transcripts in neurites, growth cones and presynaptic compartments, all essential for neuronal maturation (57,58). Although the functional clusters remained the same, there were some fluctuations between D3 and D7, with increases in the representation of signal transducers and nuclear co -regulators (cluster 1) that were accompanied by decreases in the number of protein s in cluster 4 (proteostasis and stress) and cluster 5 (energy, mitochondrial biogenesis and maintenance) . Additionally, cytoskeletal and extracellular-matrix/adhesion proteins were only present at D7, since at D3 these functions were only indirectly represented by ELAVL4 and FXR1, that promote translation of mRNAs whose protein products are related to cytoskeleton-remodelling and neuritogenesis. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 19 Table 1. Functional clustering of enriched or exclusive APPSE interactors at D3 and D7. The proteins were classified based on evidence in articles and Uniprot linking them to a direct role in the functional groups. *, Proteins with an indirect role in the associated functional group. Diff., differentiation. Ref. References. Cluster 6 information at D3 is in italic as the indicated functions are indirect, derived from the protein’s main functions (induction of translation of specific proteins). Functional Group Dif. day Proteins Cellular Event(s) Main Function Potential Role in Neuronal Differentiation Ref. 1. Signalling: ligands, receptors, phospho- regulation, transcriptional regulation D3 ATXN2, DCD, H1-2 Signal sensing and PTM-mediated modulation; Gene expression control, histone modification. Translate extracellular and intracellular cues into rapid chromatin and post-translational modifications that change transcription factor activity and mRNA handling. Regulate neuronal gene accessibility and epigenetic reprogramming; Chromatin reshaping to permit neuron- specific transcription; Influence receptor abundance and trafficking. (59– 63) D7 ATXN2, FUBP3, NCOA5, PPP1CB Modulation of Ser/Thr dephosphorylation, regulation of nuclear localization of key transcriptional regulators and chromatin remodelling. Integrate growth and survival signals with nuclear co-regulator activity to adjust phosphorylation states, modulate chromatin remodelling and transcription factor function, fine-tune promoter and enhancer output. Control cell-cycle exit and survival programs, stabilize cytoskeletal regulators through phospho- dependent interactions (via PPP1CB), and shape receptor abundance during differentiation, promoting neuronal differentiation signalling. (63– 67) 2. Pre-mRNA processing & spliceosome D3 DDX1, DDX46, PHF5A, PPIH, PRPF31, RTCB, SF3B5, SF3B6, SNRPB Pre-mRNA processing, spliceosome assembly. RNA splicing, snRNP stabilization, RNP remodelling. Essential for neuron-specific transcript generation and alternative splicing programs during fate commitment; regulate isoform diversity critical for early differentiation. (68– 75) D7 DHX15, PRPF6, RBM27, RBMX, SF1, SYMPK, U2AF2, WTAP Spliceosome assembly, splice site recognition, 3′- end/processing scaffold, ribonucleoprotein remodelling. Ensure precise intron removal and generation of neuron-specific mRNA isoforms by assembling and stabilizing snRNP complexes and coordinating splicing with 3′-end processing. Drive production of mature, neuron-specific isoforms and alternative splicing programs required for late differentiation and functional maturation. (76– 83) 3. RNA stability, transport (local translation) and translational machinery D3 ATXN2, ELAVL4, FXR1, FXR2, RPS2, RPS27 RNA transport and localization to neurites, regulated initiation/elongation of translation, ribosome assembly and direct rRNA/protein methylation. RNA-binding proteins that regulate mRNA localization, stability, and translation efficiency; Ribosomal structure support translation of synaptic and cytoskeletal proteins during early differentiation. Maintain localized protein synthesis programs that drive neuronal polarity, growth cone dynamics and early neurite extension. (84– 91) D7 ATXN2, FUBP3, RBM27, RBMX, RPS12, RPS27 mRNA stabilization, localization, part of ribonucleoprotein granules, regulated translation; Ribosome structure/function. RNA-binding proteins that control mRNA availability and local translation in neurites; coordinate translation of transcripts needed for synapse formation. Ribosomal proteins RPS12 and RPS27 contribute to ribosome integrity and efficient translation during maturation. Maintain transcript pools and enable localized protein synthesis essential for neurite extension and synaptic maturation. Support increased, regulated translation of synaptic and structural proteins. (87,9 0,92– 96) 4. Proteostasis, PTMs, stress & inflammatory/ immune response D3 DCD, FXR1, PPIH, S100A8, SSBP1 Protein folding, proteostasis, stress, defence and repair responses. Maintain proteome integrity via PTMs and peptidyl-prolyl isomerization; Mount stress/inflammatory responses (S100A8, GSDMA, DCD) that impact survival and remodelling. Support protein biogenesis/folding under differentiation stress and coordinate adaptive inflammatory/stress signalling that can modulate neurite stabilization and survival. (97– 101) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 20 Functional Group Dif. day Proteins Cellular Event(s) Main Function Potential Role in Neuronal Differentiation Ref. D7 GANAB, PRKCSH, SSBP1 N-glycan trimming and calnexin/calreticulin chaperone cycle; ER quality control for secretory and membrane proteins. Trim glucose residues from nascent N-linked glycans to permit binding/release cycles with calnexin/calreticulin, promoting correct folding or targeting misfolded glycoproteins for ER quality control. Contributes to ER homeostasis under proteotoxic stress. Support the secretory pathway during the high demand for synaptic and adhesion molecules and ensure that only properly folded glycoproteins reach the neuronal surface during maturation. (99,1 02,10 3) 5. Mitochondrial function & metabolic support D3 FASTKD3, NIPSNAP1 , SSBP1 mtDNA maintenance; mitochondrial RNA processing. mtDNA replication / mitochondrial RNA handling; mitophagy signalling, energy regulation; quality control. Provide metabolic support for neurite elongation and axonogenesis; mitochondrial RNA processing adapts energy output during lineage transition. (99,1 04,10 5) D7 SSBP1 mtDNA maintenance; mitochondrial RNA processing. mtDNA replication / mitochondrial RNA handling; mitophagy signalling, quality control. Supply of ATP and Ca²⁺ buffering for neurite growth and synaptogenesis; ensure mitochondrial quality during high metabolic demand. (99,1 06) 6. Cytoskeleton & membrane remodelling (includes neurite outgrowth and maturation) D3 Indirectly: ELAVL4*, FXR1* Cytoskeletal remodeling, neurite extension. Coordinate local synthesis of cytoskeletal and adhesion proteins (e.g. GAP43, Tau, CaMKIIα, DSP, …) to drive neurite extension and synapse formation. Promote axon/dendrite growth, growth-cone guidance and maturation of synaptic structures. (101, 107– 109) D7 INA, PLEKHA6, TUBA1B Intermediate filament formation, microtubule polymerization, membrane- cytoskeleton scaffolding, cell adhesion. Structural remodelling, neuritogenesis, axon/dendrite stabilization, trafficking, and guidance for synaptogenesis. Directly enable neurite extension, spine formation and membrane trafficking. (63,1 10– 113) APPSE interactors “Q96QA5”, “P68871”, “A0A0C4DH42”, “A0A0B4J1Y9” and “P0CG12” were excluded from Table 1 since some of them are probable contaminants, and others miss detailed information or valid processes concerning neuronal development. PhosphoS655 positively modulates APP-induced neuritogenesis A graphical representation of the APPSE exclusive and enriched interactors’ potential location and functions on neuronal differentiation at D3 and D7, is presented in Figure 5A. We also performed immunocytochemistry analyses of the potential subcellular regions of APPSE binding to some of these interactors, selected based on their abundance and/or functional relevance (Figure 5B). Microphotographs of differentiating SH-SY5Y cells at D3 (Figure 5B, left panels) showed ATXN2 localized in puncta/spots throughout the cell. Its signal was generally faint, except for some regions near the plasma membrane (PM) along existing projections , and at sites from wh ich new projections could emerge, where it appears to co-localize with APPSE. FRX2, an enriched APPSE interactor and ATXN2-binder, was localized to the nucleus, cytoplasmic puncta and PM. FXR2 showed some co-localization with APPSE at all these locations, but particularly in perinuclear vesicles and Golgi, as well as in various puncta at the PM. Co-localization was much less evident with APPSA (Supplementary Figure S4). The APPSE-enriched interactor ELAVL4/HuD, highly abundant in neurons, displayed a granular cytoplasmic distribution and was highly enriched in the nucleus. ELAVL4 showed some co-localization with APPSE in various cellular puncta, but mainly at perinuclear vesicles/Golgi and PM (including several projections/pre-neurites). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 21 Figure 5. APPS exclusive and enriched interactors organized in functional clusters in neuronal differentiating cells. A. Schematic figure created using BioRender.com. The terms listed are: (1) Signal transduction (including ligand binding, receptor trafficking and phosphoregulation) and gene expression, (2) mRNA processing and spliceosome, (3) RNA stability, transport and tr anslation (including local translation and translational machinery), (4) Proteostasis, PTMs, stress & immune response, (5) Mitochondrial function & metabolic support and (6) Cytoskeleton & membrane remodeling (including regulation of neurite outgrowth and synaptogenesis). B. Immunocytochemistry analysis of the subcellular distribution of APPSE (GFP tagged) and various of its exclusive or enriched interactors (in magenta) in SH-SY5Y cells differentiated with retinoic acid for 3 (D3) or 7 (D7) days. Signal transduction and Gene expression 1 mRNA processing & Spliceosome 2 RNA stability, transport and translation 3 Proteostasis, PTMs, Stress & Immune response 4 Mitochondrial function & Metabolic support 5 Cytoskeleton & Membrane remodelling 6 D3 D7 B. SE APP-GFP ATXN2 Overlay SE APP-GFP ATXN2 Overlay SE APP-GFP ELAVL4 Overlay SE APP-GFP FXR2 Overlay SE APP-GFP INA Overlay SE APP-GFP FUBP3 Overlay C. D. E. A. H1-2, DCD, ATXN2 RTCB, PHF5A, SNRPB, PRPF31, SF3B5, SF3B6, PPIH, DDX46, DDX1 FXR1, FXR2, ELAVL4, ATXN2, RPS2, RPS27 1 2 3 PPIH, S100A8, SSBP1, DCD, FXR1 4 4 5 NIPSNAP1, FASTKD3, SSBP1 6 ELAVL4 *, FXR1* Nucleous Nucleolus Ribosomes Smooth ER Rough ER Golgi apparatus Cytoskeleton Mitochondria Cytoskeleton3 6 Axon 3 Cytoskeleton 6 Dendrites 1 4 4 RBMX, SF1, PRPF6, DHX15, RBM27, U2AF2, SYMPK, WTAP ATXN2, FUBP3, RBM27, RBMX, RPS12, RPS27 2 3 PRKCSH, GANAB, SSBP1 4 4 5 SSBP1 6 6 INA, TUBA1B, PLEKHA6 Nucleolus Nucleous Ribosomes Smooth ER Rough ER Golgi apparatus Cytoskeleton PPP1CB, NCOA5, ATXN2, FUBP3 Cytoskeleton1 1 3 3 4 4 6 Axon 4 Dendrites Mitochondria preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 22 Conversely, ELAVL4 co-localization with APPSA was minimal, and ELAVL4 appeared to be less visible at the nuclei and PM of APPSA-transfected cells (Supplementary Figure S4). At D7 (Figure 5B, right panels), the APPSE interactors analysed were ATXN2, FUBP3 (alias FBP3, also an ATXN2-binder), and the cytoskeleton-related INA. ATXN2 displayed a punctate distribution throughout the cell but was highly enriched near the PM, particularly in cell projections/pre-neurites and growth cones. ATXN2 co-localized with both APPSE and APPSA in all these sites (ATXN2 D7 panels in Figure 5B and Supplementary Figure S4), what suggests that ATXN2 does not bind to APPSA due to a lack of affinity, but may participate in some related protein complexes/functions. FUBP3 was predominantly located in the nucleus (nucleoplasm) and at/near the PM, particularly along projections/pre-neurites and growth cones. FUBP3-APPSE co-localizing spots were visible at the PM and projections, and to a lesser extent in the nuclei. A similar co-localization pattern was observed for FUBP3 and APPSA (Supplementary Figure S4). Finally, APPSE was also observed to co-localize with cytoskeleton-associated proteins like INA (intermediate filament protein with reported interactions with APP (114), despite not listed on BioGrid ), in cytoplasm and PM, particularly in membranar projections/pre -neurites. In APPSA -expressing cells, INA staining appeared as small puncta in the nuclei (highly enriched) and cytoplasm, with little to no co- localization with APPSA. Noticeably, the signal intensity of some of these proteins (e.g. ELAVL4 and INA), appeared to increase in cells with higher levels of APP SE expression, particularly in APPSE ones, what may be due to induced expression and/or reduced degradation , but warrants confirmation but other techniques. Further and related, there were cells where highly expressed APPSE seemed to trap its interactors at the Golgi and cytoplasm (Supplementary Figure 5A). Other important note was that at D7 of RA-induced differentiation, more than half of the APPSA -expressing cells exhibited abnormal morphology (e.g. rounded shape, altered nuclei structure), suggestive of an early apoptotic state. This phenotype was more evident for cells with higher t ransfection levels, while cells with low to moderate expression seemed less affected. Given the potential involvement of some of its exclusive and enriched interactors in neuritogenic processes, and some visual indication of a more elongated phenotype in APPSE transfected cells (relatively to APPS A-transfected ones) , we assessed the impact of S655 phosphorylation on neurite outgrowth and elongation . The number of protruding processes was scored, and their lengths measured, in APPSE, APPSA and APPWt -expressing cells (Figure 5C-E). APPSE expression favoured the elongation of processes, resulting in a significant increase in their mean length (Figure 5C) and a higher percentage of longer processes (35 -50 µm and >50 µm; Figure 5D), an indirect indicator of differentiated neurite -bearing cells (Figure 5E). Notably, overexpression of APPWt already enhanced some of these parameters, while APPSA remained similar to control cells (transfected with empty vector). Conversely, the total number of processes per cell were slightly higher for APPSA cells, although not reaching significance in these conditions. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 23

In this work, we aimed to characterize the S655 phosphorylation-dependent APP interactome under a neuronal differentiation context, using mass spectrometry to identify proteins interacting with different APP -GFP fusion proteins carrying S655 mutations. From the 234 potential APP interactors retrieved by our working methodology, 65 were already known APP interactors (27.7% of ID proteins), including APP itself, which was detected in all groups. Another known APP binder, the PP1 catalytic subunit PPP1CB, was only retrieved in the APPSE group on D7, in accordance with its reported function in dephosphorylating APP at this residue (115). These support that the experimental methodology was sufficiently robust. The results showed a smaller set of APPSE interactors compared to APPSA, indicating that S655 phosphorylation drives functional sub -specialization and/or enrichment in specific functional subsets, while likely preventing some interactions. Differential interactions should be derived from altered conformation of the C-terminus due to the presence of the negatively charged phosphate group within the YTSI sorting motif (116,117), what increases/reduces the affinity of binders, and/or by differential localization of the APPSA vs APPSE species. Our previous studies already showed enhanced trafficking of APPSE from the TGN to the PM and vice versa, while APPSA remained for longer at the Golgi and lysosomes (34,37). Since the APPSE species is more exported to the PM and vesicles (exocytic and early endosomes), it is expected to interact less with proteins abundant in the ER and Golgi apparatus, for example. Regarding the nature of APP interactors, functional enrichment analyses revealed that the ‘Total interactors’ and the APPSA groups were very similar, in line with expected temporary occurrence of p hosphoS655 APP. In all groups and timepoints, the main functions of APP interactors were related to RNA metabolism (including splicing) and translation (including ribosomal structure). At D7, the APPSA interactome was associated to signalling mechanisms (including apoptotic) and protein localization terms, including the "Protein localization to lysosome" consistent with the expected prolonged localization of APPSA in this organelle (34). Overall, APPSE seems to be more associated to specific translation events, microtubule and neuronal terms. Nuclear functions related to RNA processing are likely associated with the AICD fragment rather than the full-length APP, since it is the main form trafficked to the nucleus. AICD has been implicated in nuclear transcriptional regulation through the formation of a complex with the adaptor protein Fe65 and the histone acetyltransferase Tip60 (118), although its precise contribution for transcriptional regulation remains debated (119). Colocalization studies have shown that APP CTFs, such as AICD, accumulate within intranuclear compartments enriched preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 24 in splicing factors, suggesting a potential involvement in alternative splicing regulation (120). Given that the human brain exhibits a uniquely high level of alternative splicing, changes in splicing decisions can significantly influence protein structure, mRNA localization, translational efficiency, and decay. In neurons, such regulatory networks a re essential for their development and for maintaining polarity and synaptic plasticity. APP may interact with diverse RNA-binding proteins (RBPs) or be incorporated into messenger ribonucleoprotein complexes to affect the post-transcriptional fate of multiple transcripts (119,121). The APPSE interactome was observed to be enriched in pre -mRNA processing and spliceosomal components (e.g. SF3B5/6, PRPF6, U2AF2, DHX15, PPIH), suggesting that the AICD fraction generated from APPSE may preferentially engage with these nuclear splicing hubs(124). Regarding the APPSE-specific or enriched partners , these were observed to form highly interconnected, dense PPI clusters , and many have known functions in neurite outgrowth. Some co-localized with APPSE not only at perinuclear vesicles, but also at the PM, growth cones and (pre)neurites. Additionally, APPSE expressing cells generally have pre -neurites and neurites more elongated. All this indicates that S655 phosphorylation likely induces the formation of specific APP-including “neurodifferentiation hub” (protein macrocomplexes with role in neuritic elongation) rather than only simply adding or removing single binders to APPSA-containing complexes/functions. A central APPSE -exclusive interactor at both time points is ATXN2, a multifunctional RBP involved in RNA metabolism, stress granule dynamics, endocytosis and calcium -related signalling (122). ATXN2 binds factors such as EGFR , known for its neuritogenic properties, ACTN1, DDX1, FXR1/FXR2, and FMRP, linking APPSE to growth -factor signalling, actin remodelling, endocytosis, and stress -responsive RNP condensates that regulate mRNA stability and translation in neurite s (8,54,86,123). Given ATXN2’s known roles in stabilizing specific mRNAs and organizing stress granules, its selective association with APPSE suggests that pS655 APP is preferentially recruited into RNP assemblies that coordinate receptor signalling (e.g., EGFR), actin dynamics and RNA handling during early neuronal differentiation. This is consistent with our imaging data, where ATXN2 puncta accumulate near membranes, growth cones and actin -rich structures, co-localizing with APPSE in both early projections and more elaborated neurites. A particularly striking feature of the APPSE -enriched interactome is the predominance of RBPs and translational regulators with well-established roles in neurite outgrowth and synaptic development. ELAVL4/HuD, FXR1, FXR2 and FUBP3, all appear as SE-exclusive or enriched interactors at different time points. ELAVL4/HuD stabilizes and promotes the translation of key neuronal mRNAs, including GAP -43, Tau and NRN1, which are required for growth cone preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 25 formation, axon extension and regeneration. Decreased HuD function impairs GAP-43 mRNA stability and leads to defective neurite outgrowth, while HuD upregulation is associated with enhanced neurite extension, recovery after injury , and decreased Aβ production (107,125). Concordantly, phosphorylation of APP at S655 has been linked to decreased Aβ production (37,126). Notably, within the D3 APPSE interactor network is EWSR1, a transcription repressor that interacts with ELAVL4/HuD. Disruption of EWSR1 function in knockout mouse and zebrafish models results in severe developmental phenotypes, underscoring the importance of this interactor for neuronal development and survival (127,128). The preferential association of ELAVL4 with APPSE at D3, and their co -localization at projections and PM suggest that pS655 APP may scaffold a HuD-containing complex that boosts local expression of growth cone and cytoskeleton regulators when neurites start to emerge. The Fragile X family proteins FXR1 and FXR2 are involved in translational control at neurites. FXR1 and FXR2 regulate mRNA translation efficiency, often through miRNA -dependent mechanisms, and share targets and structural features with FMRP. This protein, for instance, controls Map1b mRNA translation during Sema3A-mediated axon guidance and interacts with APP and CaMKII pathways via CYFIP1/eIF4E complexes (129). FXR1 displays a distinct expression profile, being more prominent in muscle and certain neuronal populations, and has been implicated in translation regulation, RNA stability and local mRNA control in response to signalling cues (108). FUBP3 binds 3 ′ UTR elements to modulate translation of specific transcripts, has been linked to neuronal functions (e.g. FGF9) and, more recently, to amyloid- β related signalling in neurons (92,130). At later differentiation stages (D7), more APPSE interactors are associated to cytoskeleton and potentially to membrane/ECM coupling, with INA, TUBA1B and PLEKHA6 appearing as exclusive/enriched partners. INA ( a neurofilament protein) and TUBA1B (αtubulin) are structural elements required for neurite elongation and vesicular transport, with alterations in their normal physiology potentially leading to neurodegeneration (110,131,132). In conclusion, phosphorylation at S655 changes APP binding partners over the course of neuronal differentiation, acting like a switch to increase or decrease APP affinity to specific protein complexes. These phosphorylation -driven changes in protein binding suggest that APP can take on different functional roles at different stages of neuron development, affecting specific RNA processing and protein production in the cell , with consequences for neuritogenesis, namely APPSE-driven neurite elongation. While these findings are preliminary and mainly hypothesis-generating, they provide testable directions for follow-up experiments to understand how S655 phosphorylation controls APP and cell behaviour. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 26 Acknowledgments This work was supported by the PT Foundation for Science and Technology (FCT, I.P.), and the European Regional Development Fund (FEDER), via programs Portugal2020 and 2030, Centro2020 and 2030 and COMPETE2030, by funding the Institute of Biomedicine (iBiME D; UIDB/04501/2020) and projects GoBack ( PTDC/CVT-CVT/32261/2017) and Reconnect (COMPETE2030-FEDER-00891600). The MS experiments were performed in the Biomolecular Mass Spectrometry and Proteomics group at Utrecht University (NL), under a grant of the PRIME- XS consortium ( PRIME-XS 292 ). Microphotographs were acquired in the LiM Platform for Advanced Optical Imaging of iBiMED -UA, member of the Portuguese Platform for Bioimaging (PPBI), a node of the EuBi European Bioimaging Network.

1. Andrew RJ, Kellett KAB, Thinakaran G, Hooper NM. A Greek Tragedy: The Growing Complexity of Alzheimer Amyloid Precursor Protein Proteolysis. Journal of Biological Chemistry. 2016;291(37):19235–44. doi:10.1074/jbc.R116.746032 2. Wang J, Fourriere L, Gleeson PA. Advances in the cell biology of the trafficking and processing of amyloid precursor protein: impact of familial Alzheimer’s disease mutations. Biochem J. 2024;481(19):1297–325. doi:10.1042/bcj20240056 PubMed PMID: 39302110. 3. Nithianandam V, Bukhari H, Leventhal MJ, Battaglia RA, Dong X, Fraenkel E, et al. Integrative analysis reveals a conserved role for the amyloid precursor protein in proteostasis during aging. Nat Commun. 2023;14(1):7034. doi:10.1038/s41467-023- 42822-1 4. Müller UC, Deller T, Korte M. Not just amyloid: physiological functions of the amyloid precursor protein family. Nat Rev Neurosci. 2017;18(5):281–98. doi:10.1038/nrn.2017.29 5. Dunot J, Ribera A, Pousinha PA, Marie H. Spatiotemporal insights of APP function. Curr Opin Neurobiol. 2023;82:102754. doi:https://doi.org/10.1016/j.conb.2023.102754 6. Young-Pearse TL, Bai J, Chang R, Zheng JB, LoTurco JJ, Selkoe DJ. A critical function for beta-amyloid precursor protein in neuronal migration revealed by in utero RNA interference. J Neurosci. 2007/12/28. 2007;27(52):14459–69. doi:10.1523/jneurosci.4701- 07.2007 PubMed PMID: 18160654. 7. Hoe HS, Lee KJ, Carney RS, Lee J, Markova A, Lee JY, et al. Interaction of reelin with amyloid precursor protein promotes neurite outgrowth. J Neurosci. 2009/06/12. 2009;29(23):7459–73. doi:10.1523/jneurosci.4872-08.2009 PubMed PMID: 19515914. 8. da Rocha JF, Bastos L, Domingues SC, Bento AR, Konietzko U, da Cruz ESOAB, et al. APP Binds to the EGFR Ligands HB-EGF and EGF, Acting Synergistically with EGF to Promote ERK Signaling and Neuritogenesis. Mol Neurobiol. 20201002nd ed. 2021;58(2):668–88. doi:10.1007/s12035-020-02139-2 PubMed PMID: 33009641. 9. Bamford RA, Zuko A, Eve M, Sprengers JJ, Post H, Taggenbrock R, et al. CNTN4 modulates neural elongation through interplay with APP. Open Biol. 20240515th ed. 2024;14(5):240018. doi:10.1098/rsob.240018 PubMed PMID: 38745463. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 27 10. Tyan SH, Shih AY, Walsh JJ, Maruyama H, Sarsoza F, Ku L, et al. Amyloid precursor protein (APP) regulates synaptic structure and function. Mol Cell Neurosci. 2012/08/14. 2012;51(1–2):43–52. doi:10.1016/j.mcn.2012.07.009 PubMed PMID: 22884903. 11. Rajão-Saraiva J, Dunot J, Ribera A, Temido-Ferreira M, Coelho JE, König S, et al. Age- dependent NMDA receptor function is regulated by the amyloid precursor protein. Aging Cell. 20230126th ed. 2023;22(3):e13778. doi:10.1111/acel.13778 PubMed PMID: 36704841. 12. Ma F, Akolkar H, Xu J, Liu Y, Popova D, Xie J, et al. The Amyloid Precursor Protein Modulates the Position and Length of the Axon Initial Segment. The Journal of Neuroscience. 2023;43(10):1830–44. doi:10.1523/jneurosci.0172-22.2023 13. Klein S, Goldman A, Lee H, Ghahremani S, Bhakta V, Center UCG, et al. Truncating mutations in APP cause a distinct neurological phenotype. Ann Neurol. 2016;80(3):456– 60. doi:https://doi.org/10.1002/ana.24727 14. Puzzo D, Privitera L, Leznik E, Fà M, Staniszewski A, Palmeri A, et al. Picomolar amyloid- beta positively modulates synaptic plasticity and memory in hippocampus. J Neurosci. 2008;28(53):14537–45. doi:10.1523/jneurosci.2692-08.2008 PubMed PMID: 19118188. 15. Weyer SW, Zagrebelsky M, Herrmann U, Hick M, Ganss L, Gobbert J, et al. Comparative analysis of single and combined APP/APLP knockouts reveals reduced spine density in APP-KO mice that is prevented by APPsα expression. Acta Neuropathol Commun. 2014;2(1):36. doi:10.1186/2051-5960-2-36 16. Hick M, Herrmann U, Weyer SW, Mallm JP, Tschäpe JA, Borgers M, et al. Acute function of secreted amyloid precursor protein fragment APPsα in synaptic plasticity. Acta Neuropathol. 2015;129(1):21–37. doi:10.1007/s00401-014-1368-x 17. Klevanski M, Herrmann U, Weyer SW, Fol R, Cartier N, Wolfer DP, et al. The APP Intracellular Domain Is Required for Normal Synaptic Morphology, Synaptic Plasticity, and Hippocampus-Dependent Behavior. J Neurosci. 2015/12/15. 2015;35(49):16018–33. doi:10.1523/jneurosci.2009-15.2015 PubMed PMID: 26658856. 18. Pousinha PA, Mouska X, Raymond EF, Gwizdek C, Dhib G, Poupon G, et al. Physiological and pathophysiological control of synaptic GluN2B-NMDA receptors by the C-terminal domain of amyloid precursor protein. Slutsky I, editor. Elife. 2017;6:e25659. doi:10.7554/eLife.25659 19. Bergmans BA, Shariati SA, Habets RL, Verstreken P, Schoonjans L, Muller U, et al. Neurons generated from APP/APLP1/APLP2 triple knockout embryonic stem cells behave normally in vitro and in vivo: lack of evidence for a cell autonomous role of the amyloid precursor protein in neuronal differentiation. Stem Cells. 2010/01/06. 2010;28(3):399–406. doi:10.1002/stem.296 PubMed PMID: 20049903. 20. Aydin D, Weyer SW, Muller UC. Functions of the APP gene family in the nervous system: insights from mouse models. Exp Brain Res. 2011/09/21. 2012;217(3–4):423–34. doi:10.1007/s00221-011-2861-2 PubMed PMID: 21931985. 21. Rem PD, Sereikaite V, Fernández-Fernández D, Reinartz S, Ulrich D, Fritzius T, et al. Soluble amyloid-β precursor peptide does not regulate GABAB receptor activity. Chao M V, Huguenard JR, Tyagarajan SK, Mulle C, editors. Elife. 2023;12:e82082. doi:10.7554/eLife.82082 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 28 22. Rice HC, de Malmazet D, Schreurs A, Frere S, Van Molle I, Volkov AN, et al. Secreted amyloid-β precursor protein functions as a GABABR1a ligand to modulate synaptic transmission. Science (1979). 2019;363(6423):eaao4827. doi:doi:10.1126/science.aao4827 23. Vieira S, Rebelo S, Esselmann H, Wiltfang J, Lah J, Lane R, et al. Retrieval of the Alzheimer’s amyloid precursor protein from the endosome to the TGN is S655 phosphorylation state-dependent and retromer-mediated. Mol Neurodegener. 2010;5(1):40. PubMed PMID: doi:10.1186/1750-1326-5-40. 24. Triaca V, Sposato V, Bolasco G, Ciotti MT, Pelicci P, Bruni AC, et al. NGF controls APP cleavage by downregulating APP phosphorylation at Thr668: relevance for Alzheimer’s disease. Aging Cell. 20160413th ed. 2016;15(4):661–72. doi:10.1111/acel.12473 PubMed PMID: 27076121. 25. Branca C, Shaw DM, Belfiore R, Gokhale V, Shaw AY, Foley C, et al. Dyrk1 inhibition improves Alzheimer’s disease-like pathology. Aging Cell. 2017;16(5):1146–54. doi:https://doi.org/10.1111/acel.12648 26. Tamayev R, Zhou D, D’Adamio L. The interactome of the amyloid beta precursor protein family members is shaped by phosphorylation of their intracellular domains. Mol Neurodegener. 2009/07/16. 2009;4:28. doi:10.1186/1750-1326-4-28 PubMed PMID: 19602287. 27. Zhang T, Chen D, Lee TH. Phosphorylation Signaling in APP Processing in Alzheimer’s Disease. Int J Mol Sci. 20191227th ed. 2019;21(1). doi:10.3390/ijms21010209 PubMed PMID: 31892243. 28. Gandy S, Czernik AJ, Greengard P. Phosphorylation of Alzheimer disease amyloid precursor peptide by protein kinase C and Ca2+/calmodulin-dependent protein kinase II. Proc Natl Acad Sci U S A. 1988/08/01. 1988;85(16):6218–21. PubMed PMID: 3137567. 29. Lee MS, Kao SC, Lemere CA, Xia W, Tseng HC, Zhou Y, et al. APP processing is regulated by cytoplasmic phosphorylation. J Cell Biol. 2003;163(1):83–95. doi:10.1083/jcb.200301115 30. Suzuki T, Nairn AC, Gandy SE, Greengard P. Phosphorylation of Alzheimer amyloid precursor protein by protein kinase C. Neuroscience. 1992/06/01. 1992;48(4):755–61. PubMed PMID: 1630623. 31. Hu YB, Ren RJ, Zhang YF, Huang Y, Cui HL, Ma C, et al. Rho-associated coiled-coil kinase 1 activation mediates amyloid precursor protein site-specific Ser655 phosphorylation and triggers amyloid pathology. Aging Cell. 20190709th ed. 2019;18(5):e13001. doi:10.1111/acel.13001 PubMed PMID: 31287605. 32. Ramelot TA, Nicholson LK. Phosphorylation-induced structural changes in the amyloid precursor protein cytoplasmic tail detected by NMR. J Mol Biol. 2001/03/29. 2001;307(3):871–84. doi:10.1006/jmbi.2001.4535 PubMed PMID: 11273707. 33. Tam JH, Cobb MR, Seah C, Pasternak SH. Tyrosine Binding Protein Sites Regulate the Intracellular Trafficking and Processing of Amyloid Precursor Protein through a Novel Lysosome-Directed Pathway. PLoS One. 20161024th ed. 2016;11(10):e0161445. doi:10.1371/journal.pone.0161445 PubMed PMID: 27776132. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 29 34. Vieira S, Rebelo S, Domingues S, da Cruz e Silva E, da Cruz e Silva O. S655 phosphorylation enhances APP secretory traffic. Mol Cell Biochem [Internet]. 2009;328(1):145–54. Available from: http://dx.doi.org/10.1007/s11010-009-0084-7 35. da Rocha JF, da Cruz e Silva OA, Vieira SI. Analysis of the amyloid precursor protein role in neuritogenesis reveals a biphasic SH-SY5Y neuronal cell differentiation model. J Neurochem. 2015/04/24. 2015;134(2):288–301. doi:10.1111/jnc.13133 PubMed PMID: 25903790. 36. Da Rocha JF, Da Cruz E Silva OAB, Vieira SI. Analysis of the amyloid precursor protein role in neuritogenesis reveals a biphasic SH-SY5Y neuronal cell differentiation model. J Neurochem. 2015 Jul 1;134(2):288–301. doi:10.1111/JNC.13133 PubMed PMID: 25903790. 37. Vieira SI, Rebelo S, Esselmann H, Wiltfang J, Lah J, Lane R, et al. Retrieval of the Alzheimer’s amyloid precursor protein from the endosome to the TGN is S655 phosphorylation state-dependent and retromer-mediated [Internet]. 2010. doi:10.1186/1750-1326-5-40 38. Orsburn BC. Proteome Discoverer—A Community Enhanced Data Processing Suite for Protein Informatics. Proteomes 2021, Vol 9, Page 15. 2021 Mar 23;9(1):15. doi:10.3390/PROTEOMES9010015 PubMed PMID: 33806881. 39. Tyanova S, Temu T, Cox J. The MaxQuant computational platform for mass spectrometry-based shotgun proteomics. Nature Protocols 2016 11:12. 2016 Oct 27;11(12):2301–19. doi:10.1038/nprot.2016.136 PubMed PMID: 27809316. 40. Cai T, Yang F. Strategies for Characterization of Low-Abundant Intact or Truncated Low- Molecular-Weight Proteins From Human Plasma. Enzymes. 2017;42:105. doi:10.1016/BS.ENZ.2017.08.004 PubMed PMID: 29054267. 41. Fijalkowski I, Willems P, Jonckheere V, Simoens L, Damme P Van. Hidden in plain sight: challenges in proteomics detection of small ORF-encoded polypeptides. microLife. 2022;3:uqac005. doi:10.1093/femsml/uqac005 PubMed PMID: 37223358. 42. Bindea G, Mlecnik B, Hackl H, Charoentong P, Tosolini M, Kirilovsky A, et al. ClueGO: A Cytoscape plug-in to decipher functionally grouped gene ontology and pathway annotation networks. Bioinformatics. 2009;25(8):1091–3. doi:10.1093/BIOINFORMATICS/BTP101, PubMed PMID: 19237447. 43. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: A software Environment for integrated models of biomolecular interaction networks. Genome Res. 2003 Nov;13(11):2498–504. doi:10.1101/GR.1239303, PubMed PMID: 14597658. 44. Aleksander SA, Balhoff J, Carbon S, Cherry JM, Drabkin HJ, Ebert D, et al. The Gene Ontology knowledgebase in 2023. Genetics. 2023 May 1;224(1). doi:10.1093/GENETICS/IYAD031 PubMed PMID: 36866529. 45. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, et al. Gene ontology: Tool for the unification of biology. Nat Genet. 2000 May;25(1):25–9. doi:10.1038/75556 PubMed PMID: 10802651. 46. Kolberg L, Raudvere U, Kuzmin I, Adler P, Vilo J, Peterson H. G:Profiler-interoperable web service for functional enrichment analysis and gene identifier mapping (2023 update). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 30 Nucleic Acids Res. 2023 Jul 5;51(W1):W207–12. doi:10.1093/NAR/GKAD347 PubMed PMID: 37144459. 47. Uhlén M, Fagerberg L, Hallström BM, Lindskog C, Oksvold P, Mardinoglu A, et al. Proteomics. Tissue-based map of the human proteome. Science. 2015 Jan 23;347(6220). doi:10.1126/SCIENCE.1260419 PubMed PMID: 25613900. 48. Stark C, Breitkreutz BJ, Reguly T, Boucher L, Breitkreutz A, Tyers M. BioGRID: a general repository for interaction datasets. Nucleic Acids Res. 2006;34(Database issue). doi:10.1093/NAR/GKJ109 PubMed PMID: 16381927. 49. Utriainen M, Morris JH. clusterMaker2: a major update to clusterMaker, a multi-algorithm clustering app for Cytoscape. BMC Bioinformatics. 2023 Dec 1;24(1):1–28. doi:10.1186/S12859-023-05225-Z/FIGURES/10 PubMed PMID: 37020209. 50. Bateman A, Martin MJ, Orchard S, Magrane M, Adesina A, Ahmad S, et al. UniProt: the Universal Protein Knowledgebase in 2025. Nucleic Acids Res. 2025 Jan 6;53(D1):D609– 17. doi:10.1093/NAR/GKAE1010 PubMed PMID: 39552041. 51. Smith CL, Eppig JT. The mammalian phenotype ontology: Enabling robust annotation and comparative analysis. Wiley Interdiscip Rev Syst Biol Med. 2009 Nov;1(3):390–9. doi:10.1002/wsbm.44 PubMed PMID: 20052305. 52. Duce JA, Smith DP, Blake RE, Crouch PJ, Li QX, Masters CL, et al. Linker Histone H1 Binds to Disease Associated Amyloid-like Fibrils. J Mol Biol. 2006 Aug 18;361(3):493– 505. doi:10.1016/j.jmb.2006.06.038 PubMed PMID: 16854430. 53. Oláh J, Vincze O, Virók D, Simon D, Bozsó Z, Tokési N, et al. Interactions of pathological hallmark proteins: tubulin polymerization promoting protein/p25, beta-amyloid, and alpha- synuclein. J Biol Chem. 2011 Oct 30;286(39):34088–100. doi:10.1074/JBC.M111.243907 PubMed PMID: 21832049. 54. Youn JY, Dunham WH, Hong SJ, Knight JDR, Bashkurov M, Chen GI, et al. High-Density Proximity Mapping Reveals the Subcellular Organization of mRNA-Associated Granules and Bodies. Mol Cell. 2018 Feb 1;69(3):517-532.e11. doi:10.1016/j.molcel.2017.12.020 PubMed PMID: 29395067. 55. Wheat A, Yu C, Wang X, Burke AM, Chemmama IE, Kaake RM, et al. Protein interaction landscapes revealed by advanced in vivo cross-linking-mass spectrometry. Proc Natl Acad Sci U S A. 2021 Aug 10;118(32). doi:10.1073/pnas.2023360118 PubMed PMID: 34349018. 56. Fasci D, Ingen H Van, Scheltema RA, Heck AJR. Histone interaction landscapes visualized by crosslinking mass spectrometry in intact cell nuclei. Molecular and Cellular Proteomics. 2018 Oct 1;17(10):2018–33. doi:10.1074/mcp.RA118.000924 PubMed PMID: 30021884. 57. Furlanis E, Scheiffele P. Regulation of neuronal differentiation, function, and plasticity by alternative splicing. Annu Rev Cell Dev Biol. 2018 Oct 6;34:451. doi:10.1146/annurev- cellbio-100617-062826 PubMed PMID: 30028642. 58. Sasaki Y. Local Translation in Growth Cones and Presynapses, Two Axonal Compartments for Local Neuronal Functions. Biomolecules. 2020 May 1;10(5):668. doi:10.3390/biom10050668 PubMed PMID: 32344905. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 31 59. Kim JM, Kim K, Punj V, Liang G, Ulmer TS, Lu W, et al. Linker histone H1.2 establishes chromatin compaction and gene silencing through recognition of H3K27me3. Sci Rep. 2015 Nov 19;5. doi:10.1038/SREP16714 PubMed PMID: 26581166. 60. Zhang Y, Reinberg D. Transcription regulation by histone methylation: interplay between different covalent modifications of the core histone tails. Genes Dev. 2001 Sep 15;15(18):2343–60. doi:10.1101/GAD.927301 PubMed PMID: 11562345. 61. Nachmani D, Gutschner T, Reches A, Diederichs S, Mandelboim O. RNA-binding proteins regulate the expression of the immune activating ligand MICB. Nature Communications 2014 5:1. 2014 Jun 13;5(1):4186-. doi:10.1038/ncomms5186 PubMed PMID: 24924487. 62. Lager TW, Conner C, Keating CR, Warshaw JN, Panopoulos AD. Cell surface GRP78 and Dermcidin cooperate to regulate breast cancer cell migration through Wnt signaling. Oncogene. 2021 Jun 10;40(23):4050–9. doi:10.1038/S41388-021-01821-6 PubMed PMID: 33981001. 63. Nonis D, Schmidt MHH, van de Loo S, Eich F, Dikic I, Nowock J, et al. Ataxin-2 associates with the endocytosis complex and affects EGF receptor trafficking. Cell Signal. 2008 Oct;20(10):1725–39. doi:10.1016/j.cellsig.2008.05.018 PubMed PMID: 18602463. 64. Foley K, McKee C, Nairn AC, Xia H. Regulation of Synaptic Transmission and Plasticity by Protein Phosphatase 1. J Neurosci. 2021 Apr 7;41(14):3040–50. doi:10.1523/JNEUROSCI.2026-20.2021 PubMed PMID: 33827970. 65. Koptan DMT, Rasheed Bahgat DM, Abdelrasool AA, Allam RSHM, Elgengehy FT, Abdel Baki NM, et al. Analysis of Nuclear Receptor Coactivator 5 (NCOA5) Messenger RNA Expression and rs2903908 Single Nucleotide Polymorphism of NCOA5 in an Egyptian Cohort with Behçet’s Disease: A Single-Center Case-control Study. Ocul Immunol Inflamm. 2022 Aug 18;30(6):1436–46. doi:10.1080/09273948.2021.1889610;JOURNAL:JOURNAL:IOII20;WGROUP:STRING:P UBLICATION PubMed PMID: 34255592. 66. CHEN Y, CUI S, GUO Y, CHEN D. Co-regulator NCOA5 and cancer. Biocell. 2022 Mar 16;46(7):1643–50. doi:10.32604/BIOCELL.2022.019142 67. Davis-Smyth T, Duncan RC, Zheng T, Michelotti G, Levens D. The far upstream element- binding proteins comprise an ancient family of single-strand DNA-binding transactivators. J Biol Chem. 1996;271(49):31679–87. doi:10.1074/JBC.271.49.31679 PubMed PMID: 8940189. 68. Trotman JB, Braceros KCA, Cherney RE, Murvin MM, Calabrese JM. The control of polycomb repressive complexes by long noncoding RNAs. Wiley Interdiscip Rev RNA. 2021 Nov 1;12(6):e1657. doi:10.1002/WRNA.1657 PubMed PMID: 33861025. 69. Popow J, Englert M, Weitzer S, Schleiffer A, Mierzwa B, Mechtler K, et al. HSPC117 is the essential subunit of a human tRNA splicing ligase complex. Science. 2011 Feb 11;331(6018):760–4. doi:10.1126/SCIENCE.1197847 PubMed PMID: 21311021. 70. Alam SS, Kumar S, Beauchamp MC, Bareke E, Boucher A, Nzirorera N, et al. Snrpb is required in murine neural crest cells for proper splicing and craniofacial morphogenesis. DMM Disease Models and Mechanisms. 2022;15(6). doi:10.1242/DMM.049544 PubMed PMID: 35593225. 71. Teng T, Tsai JH, Puyang X, Seiler M, Peng S, Prajapati S, et al. Splicing modulators act at the branch point adenosine binding pocket defined by the PHF5A-SF3b complex. Nat preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 32 Commun. 2017 May 25;8(1):15522-. doi:10.1038/NCOMMS15522;TECHMETA PubMed PMID: 28541300. 72. Makarova O V., Makarov EM, Liu S, Vornlocher HP, Lührmann R. Protein 61K, encoded by a gene (PRPF31) linked to autosomal dominant retinitis pigmentosa, is required for U4/U6*U5 tri-snRNP formation and pre-mRNA splicing. EMBO J. 2002 Mar 1;21(5):1148– 57. doi:10.1093/EMBOJ/21.5.1148 PubMed PMID: 11867543. 73. Sun C. The SF3b complex: splicing and beyond. Cell Mol Life Sci. 2020 Sep 1;77(18):3583. doi:10.1007/S00018-020-03493-Z PubMed PMID: 32140746. 74. Rajiv C, Jackson SRE, Cocklin S, Eisenmesser EZ, Davis TL. The spliceosomal proteins PPIH and PRPF4 exhibit bi-partite binding. Biochem J. 2017 Nov 1;474(21):3689. doi:10.1042/BCJ20170366 PubMed PMID: 28935721. 75. Yang F, Bian T, Zhan X, Chen Z, Xing Z, Larsen NA, et al. Mechanisms of the RNA helicases DDX42 and DDX46 in human U2 snRNP assembly. Nat Commun. 2023 Dec 1;14(1). doi:10.1038/S41467-023-36489-X PubMed PMID: 36797247. 76. Silla T, Schmid M, Dou Y, Garland W, Milek M, Imami K, et al. The human ZC3H3 and RBM26/27 proteins are critical for PAXT-mediated nuclear RNA decay. Nucleic Acids Res. 2020 Mar 18;48(5):2518–30. doi:10.1093/NAR/GKZ1238 PubMed PMID: 31950173. 77. Elliott DJ, Dalgliesh C, Hysenaj G, Ehrmann I. RBMX family proteins connect the fields of nuclear RNA processing, disease and sex chromosome biology. International Journal of Biochemistry and Cell Biology. 2019 Mar 1;108:1–6. doi:10.1016/j.biocel.2018.12.014 PubMed PMID: 30593955. 78. Selenko P, Gregorovic G, Sprangers R, Stier G, Rhani Z, Krämer A, et al. Structural basis for the molecular recognition between human splicing factors U2AF65 and SF1/mBBP. Mol Cell. 2003 Apr 1;11(4):965–76. doi:10.1016/S1097-2765(03)00115-1 PubMed PMID: 12718882. 79. Tanackovic G, Ransijn A, Ayuso C, Harper S, Berson EL, Rivolta C. A missense mutation in PRPF6 causes impairment of pre-mRNA splicing and autosomal-dominant retinitis pigmentosa. Am J Hum Genet. 2011 May 13;88(5):643–9. doi:10.1016/j.ajhg.2011.04.008 PubMed PMID: 21549338. 80. Feng Q, Krick K, Chu J, Burge CB. Splicing quality control mediated by DHX15 and its G- patch activator SUGP1. Cell Rep. 2023 Oct 31;42(10):113223. doi:10.1016/J.CELREP.2023.113223 PubMed PMID: 37805921. 81. Wang X, You B, Yin F, Chen C, He H, Liu F, et al. A presumed missense variant in the U2AF2 gene causes exon skipping in neurodevelopmental diseases. J Hum Genet. 2023 Jun 1;68(6):375–82. doi:10.1038/S10038-023-01128-2 PubMed PMID: 36747105. 82. Kolev NG, Steitz JA. Symplekin and multiple other polyadenylation factors participate in 3’-end maturation of histone mRNAs. Genes Dev. 2005 Nov 1;19(21):2583–92. doi:10.1101/GAD.1371105 PubMed PMID: 16230528. 83. Huang Q, Mo J, Liao Z, Chen X, Zhang B. The RNA m6A writer WTAP in diseases: structure, roles, and mechanisms. Cell Death Dis. 2022 Oct 1;13(10). doi:10.1038/S41419-022-05268-9 PubMed PMID: 36207306. 84. Khan FA, Fang N, Zhang W, Ji S. The multifaceted role of Fragile X-Related Protein 1 (FXR1) in cellular processes: an updated review on cancer and clinical applications. Cell preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 33 Death & Disease 2024 15:1. 2024 Jan 18;15(1):72-. doi:10.1038/s41419-023-06413-8 PubMed PMID: 38238286. 85. Silvestri B, Mochi M, Garone MG, Rosa A. Emerging Roles for the RNA-Binding Protein HuD (ELAVL4) in Nervous System Diseases. Int J Mol Sci. 2022 Dec 1;23(23):14606. doi:10.3390/IJMS232314606 PubMed PMID: 36498933. 86. Cha IJ, Lee D, Park SS, Chung CG, Kim SY, Jo MG, et al. Ataxin-2 Dysregulation Triggers a Compensatory Fragile X Mental Retardation Protein Decrease in Drosophila C4da Neurons. Mol Cells. 2020;43(10):870. doi:10.14348/MOLCELLS.2020.0158 PubMed PMID: 33115979. 87. Li L, Wang M, Huang L, Zheng X, Wang L, Miao H. Ataxin-2: a powerful RNA-binding protein. Discover Oncology. 2024 Dec 1;15(1):298. doi:10.1007/S12672-024-01158-Y PubMed PMID: 39039334. 88. Hsu SH, Hung WC. Protein arginine methyltransferase 3: A crucial regulator in metabolic reprogramming and gene expression in cancers. Cancer Lett. 2023 Feb 1;554:216008. doi:10.1016/J.CANLET.2022.216008 PubMed PMID: 36400311. 89. Dong J, Hinnebusch AG. uS5/Rps2 residues at the 40S ribosome entry channel enhance initiation at suboptimal start codons in vivo. Genetics. 2021 Jan 1;220(1):iyab176. doi:10.1093/GENETICS/IYAB176 PubMed PMID: 34791232. 90. Montellese C, van den Heuvel J, Ashiono C, Dörner K, Melnik A, Jonas S, et al. USP16 counteracts mono-ubiquitination of RPS27a and promotes maturation of the 40S ribosomal subunit. Elife. 2020 Mar 1;9. doi:10.7554/ELIFE.54435 PubMed PMID: 32129764. 91. Fernández E, Li KW, Rajan N, Rubeis S De, Fiers M, Smit AB, et al. FXR2P Exerts a Positive Translational Control and Is Required for the Activity-Dependent Increase of PSD95 Expression. The Journal of Neuroscience. 2015 Jun 24;35(25):9402. doi:10.1523/JNEUROSCI.4800-14.2015 PubMed PMID: 26109663. 92. Gau BH, Chen TM, Shih YHJ, Sun HS. FUBP3 interacts with FGF9 3′ microsatellite and positively regulates FGF9 translation. Nucleic Acids Res. 2011 May;39(9):3582. doi:10.1093/NAR/GKQ1295 PubMed PMID: 21252297. 93. Cai T, Cinkornpumin JK, Yu Z, Villarreal OD, Pastor WA, Richard S. Deletion of RBMX RGG/RG motif in Shashi-XLID syndrome leads to aberrant p53 activation and neuronal differentiation defects. Cell Rep. 2021 Jul 13;36(2). doi:10.1016/J.CELREP.2021.109337 PubMed PMID: 34260915. 94. Chowdhury TA, Luy DA, Scapellato G, Farache D, Lee ASY, Quinn CC. Ortholog of autism candidate gene RBM27 regulates mitoribosomal assembly factor MALS-1 to protect against mitochondrial dysfunction and axon degeneration during neurodevelopment. PLoS Biol. 2024 Oct 1;22(10):e3002876. doi:10.1371/JOURNAL.PBIO.3002876 PubMed PMID: 39480871. 95. Demirci H, Wang L, Murphy IV F V., Murphy EL, Carr JF, Blanchard SC, et al. The central role of protein S12 in organizing the structure of the decoding site of the ribosome. RNA. 2013 Dec;19(12):1791. doi:10.1261/RNA.040030.113 PubMed PMID: 24152548. 96. Remacha M, Jimenez-Diaz A, Santos C, Briones E, Zambrano R, Rodriguez Gabriel MA, et al. Proteins P1, P2, and P0, components of the eukaryotic ribosome stalk. New preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 34 structural and functional aspects. Biochem Cell Biol. 1995;73(11–12):959–68. doi:10.1139/O95-103 PubMed PMID: 8722011. 97. Gaither LA, Borawski J, Anderson LJ, Balabanis KA, Devay P, Joberty G, et al. Multiple cyclophilins involved in different cellular pathways mediate HCV replication. Virology. 2010 Feb 5;397(1):43–55. doi:10.1016/j.virol.2009.10.043 PubMed PMID: 19932913. 98. Schittek B. The multiple facets of dermcidin in cell survival and host defense. J Innate Immun. 2012 Jun;4(4):349–60. doi:10.1159/000336844 PubMed PMID: 22455996. 99. Tan K, Fujimoto M, Takii R, Takaki E, Hayashida N, Nakai A. Mitochondrial SSBP1 protects cells from proteotoxic stresses by potentiating stress-induced HSF1 transcriptional activity. Nat Commun. 2015 Mar 12;6:6580. doi:10.1038/NCOMMS7580 PubMed PMID: 25762445. 100. Tian Q, Li Z, Yan Z, Jiang S, Zhao X, Wang L, et al. Inflammatory role of S100A8/A9 in the central nervous system non-neoplastic diseases. Brain Res Bull. 2024 Nov 1;218. doi:10.1016/j.brainresbull.2024.111100 PubMed PMID: 39396712. 101. Shimizu H, Hohjoh H. FMRP, FXR1 protein and Dlg4 mRNA, which are associated with fragile X syndrome, are involved in the ubiquitin-proteasome system. Sci Rep. 2023 Dec 1;13(1). doi:10.1038/S41598-023-29152-4 PubMed PMID: 36732356. 102. Porath B, Gainullin VG, Cornec-Le Gall E, Dillinger EK, Heyer CM, Hopp K, et al. Mutations in GANAB, Encoding the Glucosidase IIα Subunit, Cause Autosomal-Dominant Polycystic Kidney and Liver Disease. Am J Hum Genet. 2016 Jun 2;98(6):1193. doi:10.1016/J.AJHG.2016.05.004 PubMed PMID: 27259053. 103. Kozlov G, Gehring K. Calnexin cycle – structural features of the ER chaperone system. FEBS J. 2020 Oct 1;287(20):4322. doi:10.1111/FEBS.15330 PubMed PMID: 32285592. 104. Princely Abudu Y, Pankiv S, Mathai BJ, Håkon Lystad A, Bindesbøll C, Brenne HB, et al. NIPSNAP1 and NIPSNAP2 Act as “Eat Me” Signals for Mitophagy. Dev Cell. 2019 May 20;49(4):509-525.e12. doi:10.1016/j.devcel.2019.03.013 PubMed PMID: 30982665. 105. Boehm E, Zornoza M, Jourdain AA, Magdalena AD, García-Consuegra I, Merino RT, et al. Role of FAST kinase domains 3 (FASTKD3) in post-transcriptional regulation of mitochondrial gene expression. Journal of Biological Chemistry. 2016 Dec 9;291(50):25877–87. doi:10.1074/jbc.M116.730291 PubMed PMID: 27789713. 106. Piro-Mégy C, Sarzi E, Tarrés-Solé A, Péquignot M, Hensen F, Quilès M, et al. Dominant mutations in mtDNA maintenance gene SSBP1 cause optic atrophy and foveopathy. J Clin Invest. 2020 Jan 2;130(1):143–56. doi:10.1172/JCI128513 PubMed PMID: 31550237. 107. Bronicki LM, Jasmin BJ. Emerging complexity of the HuD/ELAVl4 gene; implications for neuronal development, function, and dysfunction [Internet]. 2013. doi:10.1261/rna 108. Khan FA, Fang N, Zhang W, Ji S. The multifaceted role of Fragile X-Related Protein 1 (FXR1) in cellular processes: an updated review on cancer and clinical applications. Cell Death Dis. 2024 Jan 1;15(1). doi:10.1038/S41419-023-06413-8 PubMed PMID: 38238286. 109. Whitman SA, Cover C, Yu L, Nelson DL, Zarnescu DC, Gregorio CC. Desmoplakin and talin2 are novel mRNA targets of fragile X-related protein-1 in cardiac muscle. Circ Res. 2011 Jul 22;109(3):262–71. doi:10.1161/CIRCRESAHA.111.244244 PubMed PMID: 21659647. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 35 110. Yuan A, Rao M V., Sasaki T, Chen Y, Kumar A, Veeranna, et al. α-Internexin Is Structurally and Functionally Associated with the Neurofilament Triplet Proteins in the Mature CNS. The Journal of Neuroscience. 2006 Sep 27;26(39):10006. doi:10.1523/JNEUROSCI.2580-06.2006 PubMed PMID: 17005864. 111. Aiken J, Buscaglia G, Bates EA, Moore JK. The α-Tubulin gene TUBA1A in Brain Development: A Key Ingredient in the Neuronal Isotype Blend. J Dev Biol. 2017 Sep 1;5(3):8. doi:10.3390/JDB5030008 PubMed PMID: 29057214. 112. Sluysmans S, Méan I, Xiao T, Boukhatemi A, Ferreira F, Jond L, et al. PLEKHA5, PLEKHA6, and PLEKHA7 bind to PDZD11 to target the Menkes ATPase ATP7A to the cell periphery and regulate copper homeostasis. Mol Biol Cell. 2021 Nov 1;32(21). doi:10.1091/MBC.E21-07-0355 PubMed PMID: 34613798. 113. Carri NG, Rubin K, Gullberg D, Ebendal T. Neuritogenesis on collagen substrates. Involvement of integrin-like matrix receptors in retinal fibre outgrowth on collagen. Int J Dev Neurosci. 1992;10(5):393–405. doi:10.1016/0736-5748(92)90029-Y PubMed PMID: 1492591. 114. Norstrom EM, Zhang C, Tanzi R, Sisodia SS. Identification of NEEP21 as a β-Amyloid Precursor Protein-Interacting Protein In Vivo That Modulates Amyloidogenic Processing In Vitro. The Journal of Neuroscience. 2010 Nov 17;30(46):15677. doi:10.1523/JNEUROSCI.4464-10.2010 PubMed PMID: 21084623. 115. Da Cruz Silva EEF, Da Cruz Silva EOA, Zaia CT, Greengard P. Inhibition of Protein Phosphatase 1 Stimulates Secretion of Alzheimer Amyloid Precursor Protein. Molecular Medicine 1995 1:5. 1995 Jul 1;1(5):535–41. doi:10.1007/BF03401590 PubMed PMID: 8529119. 116. Ramelot TA, Gentile LN, Nicholson LK. Transient structure of the amyloid precursor protein cytoplasmic tail indicates preordering of structure for binding to cytosolic factors. Biochemistry. 2000 Mar 14;39(10):2714–25. doi:10.1021/BI992580M PubMed PMID: 10704223. 117. Ramelot TA, Nicholson LK. Phosphorylation-induced structural changes in the amyloid precursor protein cytoplasmic tail detected by NMR. J Mol Biol. 2001 Mar 30;307(3):871– 84. doi:10.1006/jmbi.2001.4535 PubMed PMID: 11273707. 118. Cao X, Südhof TC. A transcriptionally [correction of transcriptively] active complex of APP with Fe65 and histone acetyltransferase Tip60. Science. 2001 Jul 6;293(5527):115–20. doi:10.1126/SCIENCE.1058783 PubMed PMID: 11441186. 119. Lee YJ, Ch’ng TH. RIP at the Synapse and the Role of Intracellular Domains in Neurons. Neuromolecular Med. 2020 Mar 1;22(1). doi:10.1007/S12017-019-08556-4 PubMed PMID: 31346933. 120. Muresan Z, Muresan V. A phosphorylated, carboxy-terminal fragment of β-amyloid precursor protein localizes to the splicing factor compartment. Hum Mol Genet. 2004 Mar 1;13(5):475–88. doi:10.1093/HMG/DDH054 PubMed PMID: 14722157. 121. Characterizing G-quadruplex Mediated Regulation of Amyloid Precursor Protein Expression [Internet]. 2016 Oct. doi:10.17918/ETD-7245 122. Costa RG, Conceição A, Matos CA, Nóbrega C. The polyglutamine protein ATXN2: from its molecular functions to its involvement in disease. Cell Death Dis. 2024 Jun 1;15(6):1– 14. doi:10.1038/S41419-024-06812-5;SUBJMETA PubMed PMID: 38877004. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint 36 123. Eich F, Nonis DF, Sen NE, Nowock J, Auburger G. Ataxin-2 Binds Alpha-Actinin-1 [Internet]. 2019 Nov 30. doi:10.20944/PREPRINTS201911.0399.V1 124. Konietzko U. AICD Nuclear Signaling and Its Possible Contribution to Alzheimers Disease. Curr Alzheimer Res. 2012 Feb 17;9(2):200–16. doi:10.2174/156720512799361673 PubMed PMID: 21605035. 125. Amadio M, Pascale A, Wang J, Ho L, Quattrone A, Gandy S, et al. nELAV proteins alteration in Alzheimer’s disease brain: a novel putative target for amyloid-beta reverberating on AbetaPP processing. J Alzheimers Dis. 2009;16(2):409–19. doi:10.3233/JAD-2009-0967 PubMed PMID: 19221430. 126. Tam JHK, Rebecca Cobb M, Seah C, Pasternak SH. Tyrosine binding protein sites regulate the intracellular trafficking and processing of amyloid precursor protein through a novel lysosome-directed pathway. PLoS One. 2016 Oct 1;11(10). doi:10.1371/JOURNAL.PONE.0161445, PubMed PMID: 27776132. 127. Perrone-Bizzozero N, Bird CW. Role of HuD in nervous system function and pathology. Frontiers in Bioscience - Scholar. 2013 Jan 1;5 S(2):554–63. doi:10.2741/S389 PubMed PMID: 23277068. 128. Azuma M, Embree LJ, Sabaawy H, Hickstein DD. Ewing sarcoma protein ewsr1 maintains mitotic integrity and proneural cell survival in the zebrafish embryo. PLoS One. 2007 Oct 3;2(10). doi:10.1371/JOURNAL.PONE.0000979 PubMed PMID: 17912356. 129. Hörnberg H, Holt C. RNA-binding proteins and translational regulation in axons and growth cones. Front Neurosci. 2013 May 23;7(7 MAY):46243. doi:10.3389/fnins.2013.00081 130. Yao J, Li Y, Liu X, Liang W, Li Y, Wu L, et al. FUBP3 mediates the amyloid-β-induced neuronal NLRP3 expression. Neural Regen Res. 2025 Jul;20(7):2068–83. doi:10.4103/NRR.NRR-D-23-01799 PubMed PMID: 39254567. 131. Gartz Hanson M, Aiken J, Sietsema D V., Sept D, Bates EA, Niswander L, et al. Novel α- tubulin mutation disrupts neural development and tubulin proteostasis. Dev Biol. 2016 Jan 15;409(2):406–19. doi:10.1016/J.YDBIO.2015.11.022 PubMed PMID: 26658218. 132. Lucas L, Tsoi PS, Quan MD, Choi KJ, Ferreon JC, Ferreon ACM. Tubulin transforms Tau and α-synuclein condensates from pathological to physiological. Nat Commun. 2026 Mar 3. doi:10.1038/s41467-026-69618-3 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.26.714600doi: bioRxiv preprint