Apoer2-ICD-dependent regulation of hippocampal ribosome mRNA loading

Background ApoE4, the most significant genetic risk factor for late-onset Alzheimer’s disease (AD), sequesters a pro-synaptogenic Reelin receptor, Apoer2, in the endosomal compartment and prevents its normal recycling. In the adult brain, Reelin potentiates excitatory synapses and thereby protects against amyloid-β toxicity. Recently, a gain-of-function mutation in Reelin that is protective against early-onset AD has been described. Alternative splicing of the Apoer2 intracellular domain (Apoer2-ICD) regulates Apoer2 signaling. Splicing of juxtamembraneous exon 16 alters the g-secretase mediated release of the Apoer2-ICD as well as synapse number and LTP, and inclusion of exon 19 ameliorates behavioral deficits in an AD mouse model. The Apoer2-ICD has also been shown to alter transcription of synaptic genes. However, the role of Apoer2 splicing for transcriptional regulation and its role in AD pathogenesis is unknown. Methods To assess in vivo mRNA-primed ribosomes specifically in hippocampi transduced with Apoer2-ICD splice variants, we crossed wild-type, cKO, and Apoer2 cleavage-resistant mice to a Cre-inducible translating ribosome affinity purification (TRAP) model. This allowed us to perform RNA-Seq on ribosome-loaded mRNA harvested specifically from hippocampal cells transduced with Apoer2-ICDs. Results Across all conditions, we observed ~ 4,700 altered ribosome-associated transcripts, several of which comprise key synaptic components such as extracellular matrix and focal adhesions with concomitant perturbation of critical signaling cascades, energy metabolism, translation, and apoptosis. We further demonstrated the ability of the Apoer2-ICD to rescue many of these altered transcripts, underscoring the importance of Apoer2 splicing in synaptic homeostasis. A variety of these altered genes have been implicated in AD, demonstrating how dysregulated Apoer2 splicing may contribute to neurodegeneration. Conclusions Our findings demonstrate how alternative splicing of the APOE and Reelin receptor Apoer2 and release of the Apoer2-ICD regulates numerous ribosome-associated transcripts in mouse hippocampi in vivo. These transcripts comprise a wide range of functions, and alterations in these transcripts suggest a mechanistic basis for the synaptic deficits seen in Apoer2 mutant mice and AD patients. Our findings, together with the recently reported AD-protective effects of a Reelin gain-of-function mutation in the presence of an early-onset AD mutation in Presenilin-1, implicate the Reelin/Apoer2 pathway as a target for AD therapeutics.

in the cleavage-resistant Apoer2 Δ16 KI mice (Apoer2 Δ16Δ19 ) enhances the synaptogenic effect and further exacerbates the synaptic dysfunction compared to those with the proline-rich domain (Apoer2 Δ16+19 ). This suggests that inclusion of exon 19 can attenuate the loss of Apoer2 cleavage independent of transcriptional regulation which relies on cleavage of the Apoer2-ICD (Telese et al., 2015;Wasser et al., 2014).
To probe for the role of Apoer2-ICD splice variants, we sought to uncover whether the Apoer2-ICD is necessary and su cient to regulate ribosome loading of key neuronal transcripts. This required a multimodel approach with three Apoer2 mouse lines -each lacking the release of the Apoer2-ICD and crossed to the Rosa26-TRAP (translating-ribosome a nity puri cation) mouse, which harbors a Cre-inducible GFP-tagged ribosomal subunit (Heiman et al., 2014;Heiman et al., 2008. Two of these lines express a cleavage-resistant Apoer2, Apoer2 Δ16±19 while the third is a Cre-inducible conditional knockout Apoer2 line, Apoer2 cKO . Upon lentiviral delivery of Cre with or without the Apoer2-ICD[± 19], we assessed the role of the Apoer2-ICD in regulating neuronal ribosome-associated transcripts and the impact of alternative splicing on this regulation. Importantly, this method allowed us to isolate and sequence mRNA-primed ribosomes speci cally from hippocampal cells in vivo which expressed exogenous soluble Apoer2-ICD, therefore ensuring changes to ribosome-bound mRNA were speci cally due to re-introduction of the Apoer2-ICD. When comparing our sequencing datasets to AD genetic risk loci, the ribosome association of 34 AD risk transcripts are differentially regulated. These ndings are consistent with new genetic linkages with the Reelin signaling pathway in AD pathogenesis (Bracher-Smith et al., 2022, Lopera et al., 2023, pointing to a likely underlying mechanism of Apoer2 regulation of the Reelin signaling pathway in AD.

Experimental Models and Subject details
All mice were housed under a 12:12 light:dark cycle and fed a normal chow diet. All animals were euthanized by inhalation of iso urane followed by decapitation according to strict regulations set by the National Institutes of Health Guide for the Care and Use of Laboratory Animals and the UT Southwestern Animal Care and Use Committee. All mice were maintained on a wild-type SV129 and C56BL/6J mixed background. The Apoer2 Δ16±19 mouse lines have been previously described (Wasser et al., 2014). The conditional Apoer2 knockout was created by anking exons 1 and 2 of the Lrp8 gene with LoxP sites. TRAP mice expressing a GFP-tagged ribosomal subunit (L10a:GFP) after Cre-induced excision of an upstream oxed stop codon were purchased from Jackson Labs (Rosa26fsTRAP, Jax no: 022367).

Constructs
All sequences used are listed in Table S11. For the luciferase assay, 2.6-kb of the promoter regions of either human Reln or mouse Lrp8 were cloned into the multiple cloning site of the Gaussia luciferase and secreted alkaline phosphatase reporter cloning vector (GeneCopoeia; pEZX-GA01; Cat#ZX103). The Apoer2-ICD expression plasmids were modi ed from (Beffert et al., 2005) to include an N-terminal 3x-FLAG with or without a C-terminal VP16. For the TRAP constructs, the 3x-FLAG and Apoer2-ICD sequences were cloned into the pLVX-IRES-ZsGreen1 Vector (Clontech Cat#632187) and the ZsGreen sequence was replaced with the Cre-recombinase sequence.
Luciferase assay HEK-293T or SH-SY5Y cells were co-transfected with dual reporter construct containing a CMV-driven secreted alkaline phosphatase (SEAP) and a Reelin promoter-driven secreted Gaussia luciferase (GLuc) reporter construct (GeneCopoeia, cat. no:pEZX-GA01) and either GFP or an Apoer2-ICD (Apoer2-ICD:VP16 construct (± exon 19, Apoer2-ICD[+ 19/Δ19] ± VP16). After 24 hours, the media was collected from transfected cells and the luciferase and control SEAP intensity was quanti ed. Luciferase signal was normalized to the SEAP signal for each well, then all values were normalized to media from cells transfected with an empty reporter construct and GFP. (3-5 wells per condition, 2 parallel measurements, at least 2 independent experiments). Lentiviral production HEK-293T cells at 70-80% con uency in 10-cm plates were co-transfected with 3ug lentiviral construct, packaging, and envelope plasmids (ratio -4:3:1) using Fugene6 transfection reagent following the manufacturers' instructions. Brie y, recombinant lentiviruses were produced by co-transfecting the cells with packaging (psPAX2; 2.25 µg) and the envelope vectors (pMD2.G; 0.75 µg) along with each lentiviral transfer vector (3 µg). Eight hours later, the cells were washed and fed with fresh culture medium containing 1mM sodium butyrate. The supernatant containing lentivirus particles was collected 48h after transfection followed by ltering through a 0.45µm lter and then concentrated to < 100ul with a 10/30 kDa Amicon ultra ltration lter. The concentrated lentivirus was layered over 10% sucrose (50mM Tris HCl pH 7.4, 100mM NaCl, 0.5mM EDTA, 10% sucrose) at a 4:1 ratio, then centrifuged at 14000 x g for 3 hours at 4°C. The subsequent pellet was resuspended in 1/100th of the original lentiviral media collected (Jiang et al., 2015).

Translating ribosome a nity puri cation (TRAP)
A nity matrix preparation. The matrix was prepared as described in (Heiman et al., 2014). For one hippocampus, the ratio of the components was 300ug Streptavidin MyOne T1 Dynabeads: 20µg biotinylated protein L: 50µg each of GFP antibodies 19C8 and 19F7 (100 µg total antibody). The appropriate amount of resuspended Dynabeads were washed once with 1X PBS (1:10 dilution of Phosphate-Buffered Saline (10X) from Invitrogen™ AM9625). Beads were then rotated end-over-end with biotinylated protein L in 1× PBS for 1 hour at room temperature and washed ve times with 3% BSA (in 1X PBS). The protein L-conjugated beads were then rotated end-over-end for 1 hour with the GFP antibodies in a low-salt buffer (in RNase-free water: HEPES (pH 7.3), 20mM; KCl, 150mM; MgCl 2 , 10mM; NP-40, 1%; add immediately before use: DTT, 0.5mM; cyclohexamide, 100µg/ml). Antibody-conjugated beads were not vortexed after this step. After washing three times with the low-salt buffer, the a nity matrix was aliquoted into tubes for individual hippocampal immunoprecipitations. Note: Before all bead washing steps, tubes were kept against the magnet for at least one minute before removing solutions to prevent loss of beads.
Tissue isolation and homogenization. Two months post-lentiviral injection, mice were anesthetized with iso urane, quickly decapitated and the whole brain removed and placed on an ice-cold metal sheet. The brain was halved and placed into an icy slurry of cyclohexamide-containing dissection buffer (in RNasefree water: HEPES (pH 7.3), 2.5mM; HBSS, 1X; glucose, 35mM; NaHCO 3 , 4mM; add immediately before use: cyclohexamide, 100ug/ml). Dissections were performed on the left hemisphere while keeping the other hemisphere in the slurry, the cerebellum and cortex were quickly dissected from the rst hemisphere and snap frozen in liquid nitrogen. The hippocampus was immediately homogenized in tissue-lysis buffer (1mL/25-50mg) (in RNase-free water: HEPES (pH 7.3), 20mM; KCl, 150mM; MgCl 2 , 10mM; add immediately before use: EDTA-free protease inhibitors, 1tab; DTT, 0.5mM; cyclohexamide, 100ug/ml; rRNasin, 10ul/ml; HEPES (pH 7.3), 20mM; SUPERase·In, 10ul/ml) with a 1-mL glass dounce homogenizer on ice followed by 10 strokes through a 23-gauge syringe. Lysates from the left hemisphere were incubated on ice while dissecting and homogenizing the right hippocampi. Lysates were centrifuged for 10 minutes at 2,000 x g at 4°C to remove nuclei (S2), and the supernatant was transferred to a fresh tube. A 1/9 volume of 10% NP-40 was added (1% nal), followed by gentle inversion to mix and a brief pulse spin to prevent lysate loss. A 1/9 volume of 300mM DHPC (30mM nal, prepared fresh each week in RNase-free water), followed by inversion to mix and incubated on ice for 5 minutes before centrifuging for 20 minutes at 20,000 x g at 4°C to remove mitochondria (S20). A small aliquot (~ 5%) was kept and stored at 4°C until after nal RNA puri cation. The remaining supernatant was transferred to a fresh tube for immunopuri cation.
Immunopuri cation. The appropriate amount of freshly washed and resuspended a nity matrix (See A nity matrix preparation section) was added to each lysate and incubated overnight (~ 18 hours) at 4°C with gentle end-over-end mixing. The next day, always keeping the tubes on ice, the beads were washed to reduce non-speci c binding prior to RNA puri cation as described in (Heiman et al., 2014). Tubes were pulse-centrifuged and beads were allowed to collect for at least one minute before each washing step. Brie y, tubes were quickly pulse-centrifuged then placed against the magnetic rack surrounded by ice. The GFP:L10-depleted lysate was removed and stored at -80°C. The beads bound to the RNA-GFP:L10 complexes were washed 4 times by resuspending beads by pipetting with 1mL of high-salt buffer (in RNase-free water: HEPES (pH 7.3), 20mM; KCl, 350mM; MgCl 2 , 10mM; NP-40, 1%; add immediately before use: DTT, 0.5mM; cyclohexamide, 100ug/ml). Of note, during each wash, beads were pipetted at least 3 more times after visible resuspension and bubbles were avoided. After removing the fourth wash, beads were removed from the magnet and warmed to room temperature before RNA puri cation.
RNA puri cation RNA was isolated from the GFP:L10-bound a nity matrix with the Absolutely RNA Nanoprep kit (Agilent, see Materials section) at room temperature. Brie y, 100µl of Nanoprep lysis buffer (with fresh ß-ME) was added to the beads. Tubes were vortexed and incubated for 10 minutes at room temperature. The tubes were then placed back on the magnet and the RNA-containing lysate was removed for puri cation according to the manufacturer's protocol with the following exception. Puri ed RNA was resuspended in 10uL of RNase-free water instead of the Elution buffer provided with the kit. IP RNA was aliquoted before freezing at -80°C. This elution buffer interfered with the accurate assessment of quality and approximate quantity of RNA by the 2100 Bioanalyzer with the Total RNA Pico chip (performed by the UTSW Genomics Sequencing Core). We obtained pico-to nanogram amounts of IP-RNA.
RNA ampli cation/ RNA-sequencing RNA ampli cation RNA ampli cation was performed as described in (Morris et al., 2011) with the exception that we used Superscript IV instead of Superscript III. aRNA concentration was estimated with the Agilent PicoChip before submitting aRNA for library preparation and RNA-sequencing. For RNAsequencing, we needed a minimum of 50ng of RNA, as we wanted to sequence samples separately, we performed a pilot experiment where we ampli ed 100pg pooled IP RNA from a subset of our IP RNA. This pilot experiment con rmed that we could successfully sequence ampli ed IP RNA. We then ampli ed 100pg of our highest quality IP RNA from each genotype and then assessed the quality and rough concentration before submitting 48 ampli ed RNA samples for library preparation and RNA-sequencing (6-9 mice/genotype, with at least one male or female represented per condition). Libraries were prepared with the Illumina TruSeq Stranded mRNA kit and run on NextSeq sequencing SE-75 (all across 3 ow cells) by the UTSW Genomics Sequencing Core.
"Rescued" transcripts were de ned as having a smaller log 2 FC and an adjusted p-value < 0.05. These adjusted p-values take into account the baseline effect of genotype compared to wildtype and the effect of the ICD in the genotype, thus increasing the power of the experiment. This is calculated by squaring the sum of each p-value and dividing this by the number of p-values summed (Colombo et al., 2017). Filtering by the adjusted genotype-rescue p-values is the most relevant measure of ICD effect. Heatmaps were created with Morpheus (https://software.broadinstitute.org/morpheus/) and hierarchical clustering was performed with Euclidian distance using complete linkage. Supervenns were created with the Compare

Data analysis
Statistical analyses were performed with Graphpad Prism 8 software using one-way and two-way ANOVA with Tukey's post hoc multiple comparisons test for exact multiplicity adjusted p-values between groups. All data sets were checked for normality with the D'Agostino & Pearson omnibus or KS normality test. If data was non-normal, the non-parametric Kruskal-Wallis test was performed with Dunn's post hoc multiple comparisons test.

Results
Apoer2-ICD regulation of ribosome-associated transcripts in vivo To determine how the Apoer2-ICD and its splicing regulate synapses, we crossed Rosa26-TRAP mice, which express a Cre-inducible GFP-tagged ribosomal subunit (L10a:GFP), to Apoer2 mutant mice which lack all or part of the Apoer2-ICD: a conditional Apoer2 knockout (Apoer2 cKO ) and two cleavage-resistant ]-IRES-Cre) to the hippocampi of these mice via lentiviral injection to demonstrate the su ciency of the Apoer2-ICD to alter the expression of key synaptic transcripts ( Fig. 1B-C). We next performed next generation RNA-Seq speci cally on hippocampal cells infected with Cre-expressing lentivirus, allowing us to precisely identify transcripts affected by the Apoer2-ICD within the complex physiology of the living brain.
We utilized this rich RNA-Seq dataset to answer key questions regarding the function of Apoer2 splicing on hippocampal mRNA primed ribosomes. First, we compared the basal effect of the loss of nuclear Apoer2-ICD transcriptional regulation by comparing the ribosome-associated transcripts from the hippocampi of Apoer2 KI/cKO to the wild-type Apoer2 injected with lentivirus expressing only Cre. We then identi ed whether the Apoer2-ICD variants could rescue these effects. Afterwards, we looked for the effects imparted by the expression of either Apoer2-ICD variants independent of genotype. To ensure the difference between Apoer2-ICD with and without exon 19 was not due to the inability of Apoer2-ICD[Δ19] to act as a transcriptional regulator, we tested the binding of our Apoer2-ICD constructs to the Reln (a known Apoer2-ICD-regulated gene) promotor using a Gaussia luciferase reporter construct driven by 2.5 kb of the Reln promoter (see methods). As the cleavage-resistant Apoer2 mice have elevated Apoer2 transcription and translation, we also evaluated whether these Apoer2-ICD splice variants bind the promoter of Apoer2 (Lrp8 promoter) (Figure S1A-C). We found no difference between the effects of Apoer2-ICD[+ 19] and Apoer2-ICD[Δ19] on luciferase expression controlled by Reln and Lrp8 promotors ( Figure S1D-G). This suggests that the Apoer2-ICD can regulate the transcription of at least a subset of its target genes independent of exon 19.
Across all conditions, the ribosome association of ~ 4,700 transcripts were altered (Fig. 1C, D, Table S1-4). Approximately half of the altered transcripts were basal differences between the Apoer2 cKO/KI and the wild-type when injected with lentivirus expressing only Cre (Fig. 1C). In addition, the majority of ribosome-associated mRNAs were restored when Apoer2-ICD[± 19]-IRES-Cre was injected and a smaller portion were either not rescued or rescued by only one of the Apoer2-ICD splice variants (Fig. 1C). The other half of the overall altered transcripts were a result of genotype-independent effects of lentiviral expression of the Apoer2-ICD in the wild-type or Apoer2 cKO/KI mice (Fig. 1D). Here we nd that in the Apoer2 WT , the majority of transcripts are similarly regulated by both Apoer2-ICD splice variants, with a smaller portion regulated by only one of the Apoer2-ICD splice variants; however, across the Apoer2 cKO/KI mice the effects of the Apoer2-ICD variants are more diverse (Fig. 1D).
The overall dataset was highly enriched for synaptic compartments and neuronal processes/pathways as well as diseases of the brain by ToppFun enrichment analysis . This work ow linked ~ 443 of these synaptic genes to six core clusters: synapse organization, neuron development, focal adhesion, extracellular matrix (ECM), Fragile X Syndrome, BDNF signaling, and ion channel activity (Fig. 1F). In addition, we have also manually gathered all the transcripts annotated under these categories and expanded the network to include those not annotated in the SynGo database, resulting in a total of 1,301 transcripts. The residual 208 SynGO synaptic transcripts not included in the minimal synaptic network in genotypes in opposing directions, 11 transcripts were similarly regulated in the Apoer2 cKO and Apoer2 Δ16Δ19 with opposite effects of Apoer2 Δ16+19 (up, OPRK1, RCC2; down, ROCK1, SLC1A1, ARID1B, BRWD1, DSP, KDM4B, ND6, RALGAPA2, ZNF644), 4 were common between Apoer2 Δ16+19 and Apoer2 Δ16Δ19 with opposite effect in Apoer2 cKO (ACTA2, SETD2, NCOA2), and 2 were common between Apoer2 cKO and Apoer2 Δ16+19 with opposite effect in Apoer2 Δ16Δ19 (up, RASAL2; down, COPG2) (Fig. 2B). This demonstrates the ability of exon 19 splicing to affect synaptic homeostasis in the absence of its normal g-secretase mediated cleavage.
As all Apoer2 KI/cKO mice lack the release of the Apoer2-ICD, we expected to nd more than just 4.2% of the differentially regulated transcripts in common between them when compared to Apoer2 WT . To nd similar regulatory modules within these > 2000 transcripts only, we reduced our criteria within these genes to |log 2 FC| > 0.5 ( Fig. 2C-D). With this reduced criterion, we identi ed 237 up-and 680 down-regulated transcripts in all three Apoer2 KI/cKO (605 total, Table S5). Eighty-one of these are annotated in GSEA  Table S5). Focal adhesion transcripts were both up-and down-regulated. The other top enrichments for up-regulated transcripts were response to growth factor, neuron projection development, and matrisome. The top enrichments for down-regulation were post-synaptic transcripts, Signaling by Rho GTPases, and mRNA metabolic process (Fig. 2E, Table S5).
There were 17 up-and 108 down-regulated synaptic ribosome-associated transcripts with similar differential regulation in the Apoer2 KI/cKO lines compared to wild-type controls, suggesting potential key transcripts that are regulated by the Apoer2-ICD. Eighty-ve of these transcripts were in our minimal synaptic network from Only 416 transcripts were not shared between at least two of the genotypes. Of the 416 differentially changed in only one genotype, Apoer2 cKO has 37 up-and 33 down-regulated (up-regulation in neuron death: APOE, CASP3, MAPK8, TFAP2B), Apoer2 Δ16+19 has 120 up-and 51 down-regulated, and Apoer2 Δ16Δ19 has 97 up-and 78 down-regulated. The differential regulation of these synaptic ribosomeassociated transcripts could provide insight into the differential phenotypes observed in these mouse lines.
Apoer2-ICD regulation in the Apoer2 WT Across all Apoer2 WT conditions overexpressing either Apoer2-ICD, there were 1,034 differentially regulated transcripts (Fig. 3, S3B, Table S1). Compared to the Apoer2 WT injected with lentivirus expressing only Cre, the inclusion of one of the Apoer2-ICDs up-and down-regulated 718 and 314 transcripts, respectively. ] is LRP3, whose expression is increased by Apoer2 and is reduced in the frontal cortex of postmortem AD brains (Cuchillo-Ibanez et al., 2021). The top functional enrichment categories for these up-regulated transcripts were post-synapse, cell junction organization, and cellular component morphogenesis ( Figure S4A, Table S6).
Of the 314 down-regulated transcripts, 203 are reduced by both Apoer2-ICDs, 6 by Apoer2-ICD[+ 19], and 105 by Apoer2-ICD[Δ19]. The top functional enrichment categories for these down-regulated transcripts were cell surface receptor signaling pathways involved in cell-cell signaling, nervous system development, and cellular component morphogenesis ( Figure S4A, Table S6 (Table S2). The top functional enrichment categories for these up-regulated transcripts were extracellular matrix/focal adhesion and response to growth factor ( Figure S4B, Table S7).
Within our minimal synaptic network from Fig. 1F, there are 164 transcripts regulated overall in the Apoer2 cKO conditions (Fig. 4A-B) (Table S3). The top functional enrichment categories for these up-regulated transcripts were cell junction organization, head development, and signaling by receptor tyrosine kinases ( Figure S4C, Table S8).
Of  (Table S3). The top functional enrichment categories for these down-regulated transcripts were synaptic signaling, cell cycle, and post-synapse ( Figure S4C, Table S8).   (Table  S4). The top functional enrichment categories for these up-regulated transcripts were post-synapse, cellular component morphogenesis, and axon ( Figure S4D, Table S9).  (Table S4). The top functional enrichment categories for these transcripts were post-synapse, Rho GTPases signaling, and cellular component morphogenesis ( Figure S4D, Table S9).
Within our minimal synaptic network from Fig. 1F, there are 241 SynGO transcripts regulated overall in the Apoer2 Δ16Δ19 conditions (Fig. 6). One hundred and sixty-three transcripts are either up-or down-regulated transcripts are differentially translated in one or more experimental conditions compared to Apoer2 WT (Fig. 7A). Over 50% of these play a critical role in APP/Aβ metabolism, including APP itself, as well as three gamma-secretase subunits (PSEN1, PSEN2 and APH1B). Ten of these are involved in Aβ toxicity or clearance (ABCA7, ACE, BIN1, CD2AP, CR1, CTSB, CTSH, GRN, PTK2B and SORL1), and nine play a role in tau toxicity. Other functional roles of these AD risk genes include regulation of immune response, endocytosis, cytoskeleton, neuron projection, and synaptic function. In Apoer2 cKO and cleavage-de cient lines, only CD2AP is signi cantly regulated in the same direction (down). If we consider transcripts that are signi cantly altered in at least one Apoer2 cKO or cleavage-de cient line, we nd four with similar down-regulation (log2FC <= -0. As transcription does not always re ect ribosome-associated transcripts, we sought to uncover how the Apoer2-ICD regulates the synaptic translatome using a multi-model approach with three genetic Apoer2 mouse lines -each lacking the Apoer2-ICD. Here we demonstrate a large network of differentially regulated ribosome-associated transcripts when the Apoer2-ICD is either not present or not released. This network comprises 15% of all the annotated mouse genes. Of these, half were signi cantly altered by just the loss of the Apoer2-ICD release and the majority were rescued by reintroducing either Apoer2-ICD with or without the alternatively spliced exon 19. Without the Apoer2-ICD, approximately 30% of the differentially regulated ribosome-associated transcripts were similarly altered across all Apoer2 cKO and cleavage-de cient mouse lines -mice with very different phenotypes -suggesting a common Apoer2-ICD regulatory module. From cell adhesion to synaptic scaffolds and the core machinery of chemical neurotransmission, the Apoer2 translatome spans almost all aspects of the synapse and intersects with many brain disorders (AD, schizophrenia, bipolar disorder, autism, epilepsy, and depression) unveiling a novel synaptic role for Apoer2.
Part of this module is enriched in integrins, collagens, laminins, and a variety of cadherin-and cytoskeleton-binding proteins (Fig. 2), which are key players in focal adhesion. Reelin's protein abundance is reduced in neurons, overall ribosome-associated transcript numbers are also reduced. This includes the regulation of the critical regulator of synaptic homeostasis, translation, and cytoskeletal dynamics -activity-regulated cytoskeletal protein (ARC). Reelin signaling leads to the release of the Apoer2-ICD, and in this study, we observed down-regulation of ARC in the translatomes of all ApoER2-ICD cKO/KI genotypes, further implicating the Reelin/Apoer2 pathway in ARC signaling.
Likewise, Reelin, in conjunction with integrins, enhances ARC translation in an mTOR-dependent manner (Dong et al., 2003). We observed decreased ribosome associated transcripts of several translation regulators across all genotypes; however, whether these effects are a product of overall reduced Reelin function or a key component of how the lack of the Apoer2-ICD impacts protein translation/mRNA transcription could not be deduced.
Translatome differences between Apoer2 cKO and cleavagede cient mouse lines: Clues to phenotypic differences Understanding the subtle differences between the Apoer2 cKO and cleavage-de cient mouse models could Altered ribosome-associated transcripts provide insight into synaptic dysfunction in Apoer2 KI models While transcripts associated with overarching pathways (BDNF signaling, ion channel activity, focal adhesion, ECM, synapse organization, nervous system development, and Fragile X Syndrome) have already been discussed, it is key to understand how several of these transcript changes may alter synaptic transmission and plasticity. For this discussion, we will only include transcripts that were signi cantly dysregulated in at least one genotype and rescued with both Apoer2-ICDs, thus demonstrating su ciency of the Apoer2-ICD to regulate these ribosome-associated transcripts. In Apoer2 Δ16+19 cells, expression of Prkcb is decreased. Pkcrb has been shown to increase the abundance of the readily releasable pool, increasing baseline synaptic transmission (Chu et al. 2014). This correlates with Another synaptic hub protein that is implicated in AD and schizophrenia, β-catenin (CTNNB1), is a core participant in Wnt-signaling, and the number of ribosome-associated CTNNB1 transcripts is downregulated in the Apoer2 cKO (Fig. 4). The Wnt-signaling pathway, much like Reelin signaling, regulates brain , an effect mediated through down-regulation of TNIK (Traf2-and Nck-interacting kinase) (Yuan et al., 2021). TNIK is up-regulated by either Apoer2-ICD (Fig. 3), which provides another mechanism by which reduced Reelin signaling could impart risk for psychosis. These experiments reveal a landscape of critical synaptic proteins altered in the absence of the Apoer2-ICD. Further exploration is required to understand how these changes affect the synaptic environment; however, our study, together with the protective nature of a Reelin gain-of-function mutation (Lopera et al., 2023) demonstrates the importance of Reelin/Apoer2 signaling for AD pathogenesis.
Limitations of the study A major limitation of this study is the inability to quantify the proteomic consequences of the Apoer2-ICD.
Because we are reintroducing the Apoer2-ICD through viral infection, only a subset of cells (~ 10%) will express our construct. Therefore, the signal-to-noise ratio of bulk RNA-Seq or proteomics would be too low to ascertain Apoer2-ICD-dependent changes in vivo. For this reason, we utilized our Cre-dependent TRAP protocol to isolate mRNA strictly from lentivirus-infected -and thus ICD expressing -cells. While this approach cannot reveal differential protein expression in the infected cells, the speci city and sensitivity of the technique allowed us to detect changes of ribosome-primed transcripts secondary to the presence or absence of Apoer2-ICD.
Further, while our approach does not speci cally differentiate between hippocampal cell-types, it does re ect the physiological context of the hippocampus where Apoer2 is expressed in all cells. By complementing our data analysis with single cell sequencing databases, we were nonetheless able to ascribe the majority of the mRNA expression changes to neurons (Karlsson et al., 2021) ( Figure S5), demonstrating the effect of Apoer2-ICD expression primarily on hippocampal neurons. Further information and requests for resources, reagents, and data should be directed to and will be ful lled by the lead contact: Dr. Catherine Wasser.

Competing Interests
The authors declare no competing interests.
Joachim Herz is a cofounder of Reelin Therapeutics.