Apoer2/Lrp8: the undercover cop of synaptic homeostasis

Apolipoprotein E receptor 2 (ApoER2) is a receptor for the protein ApoE, the most common genetic risk factor for late-onset Alzheimer’s disease (AD). It is also a key modulator of synaptic homeostasis, in part through its effect on the expression of neuronal genes including those implicated in AD and other neuropsychiatric disorders. In this perspective, we highlight several genes affected by ApoER2 and its alternatively spliced forms and how aberrant expression can be rescued by the reintroduction of the ApoER2 intracellular domain in the mouse hippocampus.

The apolipoprotein E receptor 2 (ApoER2 in humans, Apoer2 in mice) works in concert with the multimodal regulator Reelin to orchestrate neuronal migration during development. As we age, the ApoER2/Reelin pathway becomes a key homeostatic regulator of synaptogenesis and neuronal excitability. The ability of ApoER2 to regulate such processes is partially dependent on the alternative splicing of ApoER2 mRNA, specifically the inclusion of exons 16 and 19 (Wasser et al., 2023). Exon 16 encodes an O-linked sugar domain which is necessary for the sequential cleavage and release of the ApoER2 extracellular domain and intracellular domain (ICD). Exon 19, on the other hand, encodes a proline-rich cytosolic domain which is required for the Reelin-induced enhancement of LTP by binding PSD95. The Apoer2-ICD has also been demonstrated to alter transcriptional networks, both in its membrane-bound, uncleaved form as well as in its cleaved, intracellular form. The latter mechanism acts through direct interactions with cis-regulatory elements (Balmaceda et al., 2014). Recent work from our laboratory utilized targeted purification of polysomal mRNA sequencing to parse out the effect of Apoer2 and its alternative splicing on the mouse brain translatome (Wasser et al., 2023). To study this, we reintroduced the Apoer2-ICD via lentiviral injection both with and without exon 19 (Apoer2-ICD and Apoer2-ICD^∆19) into the hippocampi of mice lacking Apoer2 completely (Apoer2^KO), mice specifically lacking Apoer2’s exon 16 (Apoer2^∆16), or mice lacking both exons 16 and 19 (Apoer2^∆16∆19). This experimental design allowed us to determine whether soluble ApoER2-ICD is necessary and/or sufficient to regulate neuronal transcription and translation. This work revealed ~4700 ribosome-associated transcripts altered across all the various ApoER2 genotypes (Apoer2^KO, Apoer2^∆16, Apoer2^∆16∆19) or the lentiviral overexpression of Apoer2-ICD or Apoer2-ICD^∆19 compared to their relevant controls. In particular, 443 of these genes were implicated in synaptic localization and/or function. The extent of these translatomic alterations is too broad to discuss in a single perspective, but their physiological categorization can help understand the effect of ApoER2 on neuronal function.

On a molecular level, the 443 synaptic genes regulated by Apoer2 splicing and/or the Apoer2-ICD fall into six categories: focal adhesion, neuron development, synapse organization, extracellular matrix, brain-derived neurotrophic factor (BDNF) signaling, and ion channel activity (Figure 1). The remainder of this section will be used to highlight key molecular pathways in these categories that are significantly affected by Apoer2 or its alternative splicing.

Figure 1.

Loss of Apoer2 leads to altered translation of proteins involved in focal adhesion, BDNF signaling, nervous system development, synapse organization, extracellular matrix, and ion channels.

Reintroduction of the Apoer2-ICD rescues the translation of many of these proteins. Parts of the figure were drawn based on pictures from Servier Medical Art. Servier Medical Art by Servier is licensed under a Creative Commons Attribution 3.0 Unported License (https://creativecommons.org/licenses/by/3.0/). Apoer2: Apoprotein E receptor 2; BDNF: brain-derived neurotrophic factor; ICD: intracellular domain.

In regard to focal adhesion, we found complete loss of Apoer2 (Apoer2^KO) leads to downregulation of p21-activated kinase (PAK) 2, 3, and 5 in the mouse hippocampus. These kinases are key regulators of the morphology, motility, and fate of neurons (Rane and Minden, 2014). Pak1, Pak2, Pak3, and Pak5 translation is also downregulated in hippocampi expressing cleavage resistant ApoER2 (Apoer2^∆16/Apoer2^∆16∆19). This not only elaborates the role of the Apoer2/Reelin pathway in neurogenesis, but it also suggests this pathway plays a role in synapse homeostasis. Loss of Pak3 in the mouse hippocampus leads to decreased overall spine numbers (Boda et al., 2004). This is the same effect seen in mouse models of Apoer2 deficiency (Apoer2^KO). Conversely, previous work from our laboratory has demonstrated that specific loss of Apoer2 exons 16 and 19 leads to increased spine density (Wasser et al., 2014). Wasser et al. (2014) also demonstrated increased levels of Apoer2 in hippocampi expressing only Apoer2^∆16 or Apoer2^∆16∆19. It should also be noted that Pak3 translation is upregulated in hippocampi lacking Apoer2 after the addition of the Apoer2-ICD^∆19, suggesting the soluble, short form of the Apoer2-ICD is sufficient to increase Pak signaling.

Apoer2 activity is also linked to overall neuronal growth and survival through the mTOR system: loss of Apoer2 leads to decreased translation of Tsc1, the gene that encodes harmartin, a key inhibitor of mTOR activation (Rehbein et al., 2021). Conversely, the addition of the Apoer2-ICD with or without exon 19 rescues this effect. This suggests Apoer2 acts, in part, to increase Tsc1 levels, leading to inhibition of mTOR-dependent protein synthesis and anabolic/catabolic homeostasis. In terms of neuronal growth, loss of Apoer2 cleavage decreases BDNF translation in the mouse hippocampi, an effect that is also rescued by the addition of the Apoer2-ICD. Both the effect of the Apoer2-ICD on Tsc1 and BDNF suggests Apoer2 cleavage is in part responsible for protein synthesis homeostasis through Tsc1 regulation and for activating anabolic processes through BDNF.

Regarding Apoer2-dependent effects on ion channel function, one of the most down-regulated transcripts in our analysis of Apoer2^KO brains (but not brains expressing either Apoer2^∆16 or Apoer2^∆16∆19) was Slc1a1 which encodes excitatory amino acid transporter 3 (Eeat3). Eeat3 is one of the key glutamate transporters in the central nervous system (CNS) and is responsible for buffering the excitatory effects of glutamate within the synaptic cleft. Translation of this transporter is rescued by reintroduction of the Apoer2-ICD with or without exon 19. Interestingly, this is only the case when the Apoer2-ICD is reintroduced into hippocampi lacking Apoer2 and not when the Apoer2-ICD is reintroduced into hippocampi expressing either Apoer2^∆16 or Apoer2^∆16∆19. Impaired cleavage of Apoer2 has no effect on Eeat3, suggesting Apoer2 functions at the level of the membrane (as opposed to cleavage-dependent diffusion of the Apoer2-ICD) to affect neuronal glutamate buffering capacity. Interestingly, mutations in Slc1a1 have been linked to psychiatric disorders such as obsessive-compulsive disorder (Arnold et al., 2006). How Apoer2 regulates Slc1a1 translation in a cleavage-independent manner is unknown, but it most likely employs Apoer2/Reelin signaling, further underscoring the importance of this pathway in overall synaptic homeostasis.

While several transcripts are differentially translated in only one mutant Apoer2 genotype, one class of focal adhesion proteins, the laminins, are upregulated in hippocampi from mice lacking Apoer2 (Apoer2^KO) as well as in mice expressing either cleavage resistant form of Apoer2 (Apoer2^∆16/Apoer2^∆16∆19). Specifically, the translation of laminin 4 and laminin 5 are both upregulated in mice expressing Apoer2^∆16 and mice lacking Apoer2 completely (Apoer2^KO) while only laminin 4 is upregulated in mice expressing Apoer2^∆16∆19. These abnormalities are corrected by the addition of the Apoer2-ICD with or without the inclusion of exon 19. Laminin 4 and laminin 5 have been shown to stabilize synapses within the neuromuscular junction and the CNS while degradation of laminin is also required for maintenance of LTP (Omar et al., 2017). This suggests poor laminin homeostasis secondary to loss of Apoer2-ICD release may result in synaptic dysregulation like what is seen in hippocampi expressing cleavage-resistant Apoer2 (Apoer2^∆16/Apoer2^∆16∆19) or hippocampi lacking Apoer2 completely (Apoer2^KO). Further, the reintroduction of the Apoer2-ICD alone appears sufficient to maintain laminin-dependent synaptic function.

The molecular consequences of both loss of Apoer2 altogether (Apoer2^KO) and Apoer2 cleavage (Apoer2^∆16/Apoer2^∆16∆19) demonstrate two distinct pathways that affect the neuronal translatome: in one pathway, intact Apoer2 is necessary to act through Reelin/Apoer2 signaling, while in the other pathway, Reelin/Apoer2 signaling is less important and release of the Apoer2-ICD is necessary as a “second messenger”. The multifaceted nature of this receptor yearns for further investigation, and our previous work defines a sturdy foundation on which to build future studies of how Apoer2 maintains synaptic homeostasis.

Separate from the discussion on Apoer2’s effect on a molecular level, Apoer2 and its ICD also regulate the translation of many transcripts associated with prevalent neuropsychiatric disorders such as autism spectrum disorder (ASD) and schizophrenia, as well as transcripts associated with neurodegenerative diseases such as AD. One of the most classic concepts of synaptic dyshomeostasis in neuropsychiatric disorders is the idea of excitation-inhibition imbalance (Sohal and Rubenstein, 2019). Apoer2 has been shown to affect synaptic homeostasis, but the loss of either full-length Apoer2 (Apoer2^KO) or Apoer2 cleavage (Apoer2^∆16/Apoer2^∆16∆19) significantly decreases the translation of Gabra3 which encodes a subunit of the GABA receptor, a major inhibitory neurotransmitter receptor in the CNS. Gabra3 mutations have been linked to both schizophrenia and ASD, and decreased translation of this transcript leads to decreased inhibition at GABAergic synapses, an effect seen in various psychiatric disorders. This, coupled with the increased glutamate receptor expression in mouse hippocampi expressing cleavage resistant Apoer2 (Apoer2^∆16/Apoer2^∆16∆19) further suggests a shifted excitation-inhibition imbalance caused by Apoer2 disruption (Wasser et al., 2014). Translation of Nedd4, a transcript associated with ASD and schizophrenia, was also significantly decreased in hippocampi from all mutant Apoer2 genotypes (Apoer2^KO/Apoer2^∆16/Apoer2^∆16∆19). Nedd4 is a ubiquitin E3-ligase responsible for the ubiquitination of GluA1, leading to the internalization of AMPA receptors (the major excitatory neurotransmitter receptors in the CNS) (Louros and Osterweil, 2016). Decreased internalization of AMPA receptors further impairs the excitation-inhibition imbalance at Apoer2-deficient synapses, leading to the conclusion that Apoer2 functions as a neuromodulator, not just in terms of canonical plasticity, but also in non-canonical plasticity by regulating the translation of key excitatory and inhibitory receptor subunits (e.g., Gabra3) or the proteins that regulate these subunits (e.g., Nedd4).

Finally, Apoer2 has a direct effect on several transcripts associated with AD. This is not surprising since Apoer2 is a receptor for ApoE. The ApoE4 isoform is the greatest genetic risk factor of late-onset AD. Moreover, Apoer2 alterative splicing is altered in AD brains, and the Apoer2 ligand Reelin has been shown to be protective in AD (Wasser et al., 2023). Our laboratory has further demonstrated that the AD-risk-related ApoE4 traps Apoer2 in the endosomal compartment where it is prevented from reinsertion into the synaptic membrane and thus participation in synaptic plasticity (Pohlkamp et al., 2021). How ApoER2 and its alternative splicing imparts its effect on AD has remained relatively under-studied prior to our translatomic evidence demonstrating several genes associated with AD pathogenesis are regulated by ApoER2 or its cleavage. In mouse hippocampi lacking Apoer2 (Apoer2^KO) and mouse hippocampi expressing cleavage-resistant Apoer2 (Apoer2^∆16/Apoer2^∆16∆19), there is a decrease in the translation of Cd2ap. This transcript codes for a key regulator of cytoskeletal remodeling which confers on neurons the ability to structurally adapt and facilitate synaptic plasticity (Tao et al., 2019). This transcript is also a known gene mutated in a population of Han Chinese patients with AD (Gao et al., 2022). Reintroduction of the Apoer2-ICD into mice lacking Apoer2 rescues the Cd2ap expression, suggesting AD may be accelerated, in part, by down-regulation of CD2AP cytoskeletal synaptic regulation and poor synaptic homeostasis.

Overall, we have hypothesized that the prime cause of ApoE4-dependent AD is stalling of the endocytosed ApoE4-Apoer2 complex in the early endosome. This leads to impaired Apoer2 activity through 2 mechanisms: reduction of the Reelin-Apoer2 signaling pathway and impaired cleavage of the Apoer2 receptor. Regardless of the mechanisms, it is clear that Apoer2 plays a pivotal role in synaptic regulation, both by affecting the translation of transcripts associated with synaptic homeostasis as well as transcripts associated with ASD, schizophrenia, and AD. From a therapeutic perspective, correction of Apoer2-related pathologies depends on proper maintenance of Apoer2 at the plasma membrane as well as proper cleavage of the Apoer2-ICD. Our study has demonstrated that exogenous administration of the Apoer2-ICD is sufficient to rescue several of the effects of Apoer2 dysregulation (Wasser et al., 2023). However, by ensuring Apoer2 remains at the plasma membrane to act as both a Reelin receptor and as a source of soluble Apoer2-ICDs would ameliorate the aberrant mRNA translation signature seen in Apoer2 mutants. In our previous work, we have reported on the inhibition of endosomal recycling of Apoer2 to the plasma membrane by ApoE4 and on the use of a sodium-proton exchanger inhibitor for restoring the impaired recycling kinetics (Pohlkamp et al., 2021). The mechanism by which ApoE4 impairs ApoE recycling is beyond the scope of this perspective, but it further emphasizes the critical role of ApoER2 and its cleavage on synaptic homeostasis, and how maintenance of surface ApoER2 is a tractable and therapeutically rational target for ApoER2-related disorders such as AD.

This work was supported by NIH grants NS093382, NS108115, AG053391, HL063762 (to JH). JH was further supported by Blue Field Project to Cure FTD, BrightFocus Foundation (A20135245 and A2016396S), Harrington Discovery Institute, the Alzheimer’s Association, and a Circle of Friends Pilot Synergy Award. JH is a cofounder of Reelin Therapeutics, Inc.