Shared molecular signature in Alzheimer's disease and schizophrenia: A systematic review of the reelin signaling pathway

Shared molecular signature in Alzheimer’s disease and schizophrenia: A systematic review of the reelin signaling pathway Ana Isabel Valderrama-Mantilla c,1 , Celia Martín-Cuevas a,b,1,* , Ana G´ omez-Garrido a,b , Cristina Morente-Montilla a , Benedicto Crespo-Facorro a,b,c , Susana García-Cerro a,b a Instituto de Biomedicina de Sevilla (IBiS)/University Hospital Virgen del Rocío/CSIC/University of Sevilla, Manuel Siurot AV, Seville 41013, Spain b Spanish Network for Research in Mental Health (CIBERSAM, ISCIII), Monforte de Lemos AV, 3-5, Madrid 28029, Spain c Department of Psychiatry, School of Medicine, University of Seville, Manuel Siurot AV, Seville 41013, Spain ARTICLE INFO Keywords: Reelin RELN APOER2 DAB1 Schizophrenia Alzheimer’s disease ABSTRACT The Reelin signaling pathway, particularly the RELN-APOER2-DAB1 complex, has emerged as a key contributor to the neuropathology of Alzheimer’s disease (AD) and Schizophrenia (SZ). Despite being distinct clinical conditions, these disorders exhibit similar patterns of cognitive decline, including early disruptions in synaptic function and memory impairments. Notably, individuals with SZ have a 2–4 fold increased risk of developing AD or other dementias, highlighting potential shared molecular mechanisms, and positioning Reelin as a pivotal link between them. This systematic review explores the role of Reelin and its signaling components across these disorders. In AD, Reelin disruption correlates with hallmark features such as Tau hyperphosphorylation, amyloid-beta accumulation, and cognitive deficits. In SZ, alterations in Reelin signaling, including epigenetic modifications affecting RELN expression, are linked to disruptions in neuronal development and synaptic plasticity, particularly in the parietal and prefrontal cortices. Additionally, genomic studies reveal specific RELN variants and allelic imbalances that may influence disease severity and treatment response in SZ, suggesting RELN’s role as a potential biomarker for therapeutic outcomes. Region-specific Reelin alterations in both AD and SZ suggest differing impacts yet underscore a potential common molecular origin. Our findings highlight the Reelin pathway as a molecular convergence point, warranting further investigation as a therapeutic and diagnostic target for AD, SZ, and potentially other neuropsychiatric disorders. The interplay between genetic and epigenetic regulation of RELN may provide novel insights into neurodegeneration, with implications for personalized intervention strategies in AD and SZ. 1. Introduction Reelin, an extracellular matrix glycoprotein encoded by the RELN gene on chromosome 7q22, plays a vital role in neurodevelopment and the maintenance of adult central nervous system (CNS) functions (Botella-L´ opez et al., 2006;Jossin, 2020;Hattori and Kohno, 2021; Alexander et al., 2023). During early brain development, Reelin regulates critical processes such as neuronal migration and brain structure formation, ensuring proper organization of neural circuits (Jossin, 2020; Hattori and Kohno, 2021). In adulthood, Reelin continues to support synaptic plasticity, dendritic growth, spine formation, and synaptogenesis, all of which are essential for maintaining cognitive function and neural stability (Wasser and Herz, 2017; Yamakage et al., 2019; Jossin, 2020; S´ anchez-Hidalgo et al., 2022; Alexander et al., 2023). These dynamic roles highlight Reelin’s dual importance in both neurodevelopmental processes and the lifelong maintenance of synaptic integrity. Reelin undergoes proteolytic processing (Fig. 1) and exerts its biological effects by binding to its primary receptors: Apolipoprotein E receptor 2 (APOER2) and the Very low-density lipoprotein receptor (VLDLR) (Hiesberger et al., 1999; Bosch et al., 2016; Ogino et al., 2017; Wasser and Herz, 2017; Alexander et al., 2023). This interaction triggers intracellular signaling pathways through the phosphorylation of Disabled homolog-1 (DAB1) by Src family tyrosine kinases, regulating * Correspondence to: Hospital Universitario Virgen del Rocío, Avda. Manuel Siurot s/n, Sevilla 41013, Spain. E-mail addresses: [email protected] (A.I. Valderrama-Mantilla), [email protected] (C. Martín-Cuevas), [email protected] (A. G´ omez-Garrido), [email protected] (C. Morente-Montilla), [email protected] (B. Crespo-Facorro), [email protected] (S. García-Cerro). 1 These authors have contributed equally to this work and share first authorship. Contents lists available at ScienceDirect Neuroscience and Biobehavioral Reviews journal homepage: www.elsevier.com/locate/neubiorev https://doi.org/10.1016/j.neubiorev.2025.106032 Received 23 December 2024; Received in revised form 21 January 2025; Accepted 26 January 2025 Neuroscience and Biobehavioral Reviews 169 (2025) 106032 Available online 31 January 2025 0149-7634/© 2025 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). neuronal integrity, synaptic development, and Tau protein phosphorylation (Rice et al., 1998; Hiesberger et al., 1999; Wasser and Herz, 2017; Hattori and Kohno, 2021). Dysregulation of this pathway is linked to tauopathies, including Alzheimer’s disease (AD), where abnormal Tau phosphorylation contributes to neurodegeneration (Bock et al., 2004; Jossin et al., 2007; Alexander et al., 2023, Yi et al., 2024). Additionally, Reelin also interacts with membrane integrins (Sekine et al., 2012) and amyloid precursor protein (APP) (Herz and Chen, 2006; McCorkindale et al., 2022; Alexander et al., 2023), influencing neuronal migration, and amyloid processing (Herz and Chen, 2006; McCorkindale et al., 2022). Disruptions in these mechanisms are directly linked to the formation of amyloid-beta plaques (Selkoe, 2000; Herz and Chen, 2006; Alexander et al., 2023), a hallmark of AD. Consequently, these findings position Reelin as a critical target for understanding and potentially treating neurodegenerative diseases such as AD (Selkoe, 2000; Herz and Chen, 2006; McCorkindale et al., 2022; Alexander et al., 2023; Lopera et al., 2023). Reelin’s dysfunction is also associated with neuropsychiatric disorders like Schizophrenia (SZ) and bipolar disorder (BD) (Impagnatiello et al., 1998; Guidotti et al., 2000a.; Selkoe, 2000; Herz and Chen, 2006; S´ anchez-Hidalgo et al., 2022; Warren et al., 2024). Post-mortem studies have revealed reduced Reelin expression in key brain regions, including the prefrontal cortex, hippocampus, and caudate nucleus, in individuals with SZ (Impagnatiello et al., 1998; Guidotti et al., 2000b; Fatemi et al., 2005; Eastwood and Harrison, 2006; Imai et al., 2017; Arioka et al., 2018; Marzan et al., 2021; Alexander et al., 2023). This hypofunction is thought to disrupt GABAergic signalling and impair NMDA receptor function, contributing to an imbalance in excitatory-inhibitory (E/I) signalling, a hallmark of SC6Z pathology (Guerrin et al., 2021; S´ anchez-Hidalgo et al., 2022; Gillespie et al., 2024). Clinically, SZ and AD share overlapping features despite being distinct disorders, with the mechanisms linking these conditions remaining largely unexplored (White and Cummings, 1996). Both are characterized by significant cognitive impairments, structural and functional abnormalities in brain regions such as the hippocampus, and progressive declines in neural function (DeCarolis and Eisch, 2010). Notably, individuals with SZ are at a 2–4 fold increased risk of developing AD or other dementias later in life (Kochunov et al., 2021), underscoring a potential relationship between these two conditions. Adding to this clinical overlap, psychotic symptoms such as hallucinations and delusions, traditionally associated with SZ, are also reported in 40 %–60 % of AD patients. Emerging genetic evidence further supports this connection, with recent studies identifying 65 genes common to both SZ and AD, including the RELN gene (Guo et al., 2024). Together, this shared molecular framework, combined with the noted clinical convergence, strengthen the hypothesis that similar neuropathological pathways may underlie both disorders. In this context, the concept of "accelerated aging" has been proposed to account for the clinical, biological, and functional decline observed in SZ, echoing Emil Kraepelin’s original description of the disorder as "dementia praecox" (Kirkpatrick et al., 2008). This phenomenon aligns with the cognitive and structural brain changes shared by SZ and AD, suggesting a convergence of neurodevelopmental and neurodegenerative processes. Given Reelin’s central role in both neurodevelopment and synaptic function, and its involvement in neuropsychiatric and neurodegenerative disorders, this protein could serve as a unifying factor in understanding the neuropathology of SZ and AD. Consequently, this systematic review aims to explore the potential link between SZ and AD through a comprehensive analysis of Reelin expression in human samples, including postmortem studies and genetic models, emphasizing the role of the RELN-APOER2-DAB1 signaling pathway and genetic variants of the RELN gene in the pathogenesis of both clinical conditions (Fig. 2). 2. Method A systematic review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement (Moher et al., 2009) and a protocol was registered in PROSPERO (CRD42024596895). 2.1. Search strategy Medical Subject Headings (MeSH) terms were used to select the most relevant keywords for the search. The following search strategy was employed in PubMed and Scopus databases: ((Reelin) OR (RELN) OR (DAB1)) AND ((Alzheimer) OR (Schizophrenia) OR (neuropsychiatric disorders)). The search was conducted in December 2023. Articles included in the analysis focused on the involvement of Reelin and other components of its signaling pathway, such as APOER2,DAB1 and ADAMTS2 and 3, in the neuropathogenesis of AD and SZ in humans (postmortem tissue, cerebrospinal fluid (CSF), peripheral blood, saliva). The articles were evaluated by two independent reviewers. We also made a cross-reference search of included relevant studies and previous reviews and contacted study authors and experts for data clarification. Fig. 1. Reelin structure and proteolytic processing. The Reelin protein consists of three distinct regions: N-terminal, central, and C-terminal domains, separated by cleavage sites (Jossin et al., 2004, 2007; Jossin, 2020; Hattori and Kohno, 2021) that are tightly regulated by proteolytic enzymes such as metalloproteinases, including a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) (Jossin et al., 2004; Jossin, 2020; Hattori and Kohno, 2021). Created with Biorender.com. A.I. Valderrama-Mantilla et al. Neuroscience and Biobehavioral Reviews 169 (2025) 106032 2 2.2. Eligibility Inclusion criteria were selected to systematize the search in the databases so that only articles of interest on the role of Reelin in neuropsychiatric disorders in humans would be obtained. These inclusion criteria were studies on (i) the role of Reelin, its gene RELN, or its intermediary DAB1 in AD and SZ, (ii) studies involving human subjects, (iii) clinical trials, (iv) studies with case-control comparisons and (v) meta-analyses, with the goal of extracting additional relevant articles from their references. Exclusion criteria were (i) reviews and systematic reviews, (ii) psychiatric disorders other than AD and SZ, excluding content related to other disorders involving Reelin such as depression, epilepsy, autism spectrum disorder (ASD), or BD, (iii) studies using animal samples or based on animal models, (iv) studies that do not include Reelin, its gene RELN, or its intermediary DAB1 as a primary focus in AD or SZ and (v) articles published before 2010. 2.3. Study selection Two independent reviewers (AIV-M and CM-C) screened the titles and abstracts to identify studies that met the inclusion criteria outlined above using Rayyan (Ouzzani et al., 2016) software. The same researchers then reviewed the eligible full texts. The final list of articles was agreed to by consensus. Disagreements on eligibility were resolved by discussions with two additional reviewers (BC-F and SG-C). 2.4. Data extraction, synthesis, and quality assessment The following information was extracted in triplicate from each study: first author, year, type of study and sample, study target, study subjects, clinical variables included, results and main findings. Of the thirty-seven included studies, thirty-six were evaluated using the checklist developed by the Clarity group at McMaster University (Available online: http://www.distillersr.com/wp-content/uploads/20 21/03/Tool-to-Assess-Risk-of-Bias-in-Case-Control-Studies-DistillerSR. pdf). This checklist is designed to assess the risk of bias in case-control studies across five key domains: the quality of exposure assessment, confidence that cases had developed the outcome of interest, appropriate selection of cases, appropriate selection of controls, and appropriate matching of cases and controls based on significant prognostic variables or statistical adjustments. Each study was classified into one of four categories for each domain: low risk of bias, probably low risk of bias, probably high risk of bias, or definitely high risk of bias (Figure S1). The remaining study, a meta-analysis conducted by Hui Gao et al. (2015), utilized the Newcastle-Ottawa Scale (NOS) (Wells et al., 2009) for quality assessment. The NOS applies a “star”rating system to evaluate the methodological quality across three areas: selection, comparability, and exposure. The scores on the NOS range from zero (indicating low quality) to nine stars (indicating high quality). Studies that scored fewer than 5 stars were classified as “low quality,”while those with 5 or more stars were considered “high quality.”The Supplementary Material was used for additional data and authors were contacted to request additional information. 3. Results 3.1. Data analysis: studies included and excluded A total of 1855 articles were retrieved from the databases (PubMed and Scopus). Of these, 131 duplicate references found by crossreferencing, 3 meta-analyses and 431 reviews were excluded. From the remaining studies, 1261 more were excluded for the following reasons: other psychiatric disorders (not AD or SZ) (227 articles), animal models studies (428), studies not focused on Reelin (220) and published prior to 2010 (386). Additionally, 95 articles were retrieved from 3 meta-analyses, of which 16 were excluded as duplicate references, and 71 studies were excluded for the following reasons: other psychiatric disorders (not AD or SZ) (17 articles), animal models studies (0), studies not focused on Reelin (22) and published prior to 2010 (32). The remaining 37 studies were included in the final analysis (Fig. 3). 3.2. Risk of bias The methodological quality of the included studies was assessed using the checklist provided by the Clarity group at McMaster University (Figure S1). Of the thirty-six studies evaluated with this scale, four did not specify the diagnostic criteria used to determine the development of the outcome of interest (Maloku et al., 2010; Liu et al., 2010; Ovadia and Shifman, 2011; Aberg et al., 2014). Six studies reported that cases were selected from a defined catchment area where diagnostic procedures were likely to have changed over time (Kuang et al., 2011; B¨ onsch et al., 2012; Aberg et al., 2014; Feh´ er et al., 2015; Ho et al., 2020; Nie et al., 2021). Eight studies indicated that controls were selected from a different population than the cases (Liu et al., 2010; Kuang et al., 2011; B¨ onsch et al., 2012; Aberg et al., 2014; Feh´ er et al., 2015; Nabil Fikri et al., 2017; Xu et al., 2020; Nie et al., 2021). Twenty-five studies did not employ any method to control the casecontrol groups for demographic features or clinical characteristics (e.g., age and sex) (Wedenoja et al., 2010; Liu et al., 2010; Ovadia and Shifman, 2011; Antoniades et al., 2011; Kuang et al., 2011; Verbrugghe et al., 2012; B¨ onsch et al., 2012;Notter and Knuesel, 2013;Cuchillo-Ib´ a˜ nez et al., 2013;Aberg et al., 2014;Dong et al., 2015;Feh´ er et al., 2015;Cuchillo-Ib´ a˜ nez et al., 2016;Mata-Balaguer et al., 2018;Han et al., 2019;L´ opez-Font et al., 2019;Sozuguzel et al., 2019;Bai et al., 2019;Xu et al., 2020;Lid´ on et al., 2020;Nie et al., 2021;Ramsden et al., 2022;Bracher-Smith et al., 2022;Lopera et al., 2023;Ramsden et al., 2023). Sixteen studies were classified as having a low risk of bias, with all domains rated as “green”or with only one domain rated as “yellow” (Li et al., 2011; Kramer et al., 2011; Habl et al., 2012;Notter and Knuesel, 2013;Bufill et al., 2015;Hornig et al., 2015;Dong et al., 2015; Nabil Fikri et al., 2017;Luo et al., 2019;L´ opez-Font et al., 2019;Bai et al., 2019;Ho et al., 2020;Zhou et al., 2022;Ramsden et al., 2022; Fig. 2. Graphical abstract of this review. Created with Biorender.com. A.I. Valderrama-Mantilla et al. Neuroscience and Biobehavioral Reviews 169 (2025) 106032 3 L´ opez-Font et al., 2022;Ramsden et al., 2023). Twelve studies showed an unclear risk of bias, with one domain rated as “red”or two domains rated as “yellow”(Maloku et al., 2010; Wedenoja et al., 2010; Antoniades et al., 2011; Verbrugghe et al., 2012; Cuchillo-Ib´ a˜ nez et al., 2013; Cuchillo-Ib´ a˜ nez et al., 2016;Mata-Balaguer et al., 2018;Han et al., 2019;Sozuguzel et al., 2019;Lid´ on et al., 2020;Bracher-Smith et al., 2022;Lopera et al., 2023). Eight studies were assessed as having a high risk of bias, with two domains rated as “red”or three or more domains rated as “yellow”or “red”(Liu et al., 2010; Ovadia and Shifman, 2011; Kuang et al., 2011; B¨ onsch et al., 2012; Aberg et al., 2014; Feh´ er et al., 2015; Xu et al., 2020; Nie et al., 2021). Our risk of bias analysis categorized studies from categories B and C as having an “unclear”risk of bias (yellow category). The meta-analysis by Hui Gao et al. (2015) identified several biases using the NOS scale. In the selection domain, there was a risk of bias due to the lack of specified criteria for defining cases and controls and uncertainty regarding whether participants adequately represented the target population. For comparability, studies employed statistical adjustments for covariates such as age and sex, reducing the risk of bias by controlling for confounding factors. In terms of exposure and outcome assessment, standardized measurement methods were used, which decreased the risk of bias. However, the non-response rate among case and control groups was not evaluated. Consequently, these biases in the individual studies included in the meta-analysis could accumulate and potentially distort the overall conclusions, affecting the validity and applicability of the final results. 3.3. Role of the Reelin signaling pathway in AD Literature included in this systematic review centered on the role of the Reelin signaling pathway in AD is reported in Table 1. Postmortem studies have indicated that early-stage accumulation of phosphorylated Tau (pTau) in AD is closely linked to the disruption of the APOER2-DAB1 pathway, which is highly associated with Reelin signaling, occurring across various neuroanatomical regions, particularly in the entorhinalhippocampal areas (Ramsden et al., 2023). This disruption is believed to accelerate neurodegeneration and disease progression, exacerbating memory deficits and promoting Tau hyperphosphorylation, a key factor in AD pathology (Ramsden et al., 2022, 2023). Similarly, neuronal populations vulnerable to AD show high APOER2 protein and gene expression, which is thought to contribute to the formation of neuritic plaques (Ramsden et al., 2022, 2023,Table 1). Additionally, lipid peroxidation of the APOER2-DAB1 pathway has also been associated with AD progression (Ramsden et al., 2022, 2023; Yi et al., 2024). An epistatic interaction between APOE and the RELN-DAB1 pathway has also been identified, highlighting the importance of this pathway in AD neuropathogenesis (Bracher-Smith et al., 2022). New significant loci across the AD genome, including those mapping to DAB1, further support the connection between the RELN-DAB1 pathway and AD. Specifically, single nucleotide polymorphisms (SNPs) RELN-rs528528 and RELN-rs607755 were associated with AD risk in men but not women, suggesting a sex-specific genetic predisposition to AD (Feh´ er et al., 2015). Increased DAB1 mRNA expression in the human dorsolateral prefrontal cortex has also been linked to protection against AD (Gao et al., 2015), while DAB1-exon 22 C/G (rs362691) has been significantly associated with AD neuropathogenesis (Antoniades et al., 2011). These findings emphasize the complexity of genetic factors involved in AD and the critical role of the RELN-DAB1 pathway in its pathology (Antoniades et al., 2011). Likewise, associations between various genotypes in the Reelin signaling pathway and AD risk have also been identified, with risk SNPs such as RELN-rs2299356 and protective SNPs like RELN-rs528528, PLK2-RS15009, PLK2-rs702723, and CAMK2A (rs3756577 and rs3822606), underscoring the pathway’s significance in the disease (Bufill et al., 2015). Furthermore, patients with sporadic AD exhibited lower protein levels of APOER2 in CSF and a higher Reelin/APOER2 ratio (L´ opez-Font Fig. 3. PRISMA flow diagram of the reviewing process –systematic selection for inclusion of exclusion. Created with Biorender.com. A.I. Valderrama-Mantilla et al. Neuroscience and Biobehavioral Reviews 169 (2025) 106032 4 Table 1 Main findings of studies on the role of Reelin signaling pathway in Alzheimer’s Disease. First author and year Type of study and sample Study target Study subjects Clinical variables included Results Main findings Lopera et al. (2023) Genetic study. Study of the second case in the world with proven extreme resistance to autosomal dominant AD. Male with resilience to AD. Male remains free of AD symptoms until the age of 67, despite carrying a mutation (PSEN1-E280A) of high risk for AD, a very high load of amyloid plaque and entorhinal Tau tangles. ●Resistance to the development of AD despite the risk mutation is linked to a variant of RELN called COLBOS. ●COLBOS binds to VLDLR and APOER2 receptors. ●It is involved in a gain of function, with an increased ability to activate DAB1 and reduce human Tau phosphorylation. The RELN COLBOS variant that confers resistance to the onset of early AD symptoms in the high-risk mutation patient suggests a role for RELN signaling pathways in resistance to dementia. Ramsden et al. (2023) Postmortem tissue and peripheral blood in vivo. APOER2 expression and accumulation of APOER2-DAB1 pathways components. Samples taken from amygdala, hippocampus and medial temporal gyrus. Study of the components of the APOER2-DAB1 and RELN-DAB1 pathway in the neuropathogenesis of AD. 64 rapid autopsy cases. Clinic-pathological spectrum of AD. Diagnosis according to NIA-Reagan criteria. Includes postmortem interval, Braak stages, Thal stages. Cognitive assessment criteria: MiniMental Status Exam MMSE, 0.30. pTau is one of many components of the APOER2-DAB1 pathway that accumulate in multiple neuroanatomical (entorhinal-hippocampal) locations in the early stages of AD development. It supports the concept that disruption of APOER2DAB1 enables AD neurodegeneration and disease progression. ●APOER2-DAB1 signaling suppresses Tau phosphorylation. ●Populations of vulnerable neurons strongly express APOER2. ●Multiple components of the APOER2-DAB1 pathway accumulate leading to neuritic plaques. They correlate with disease progression. ●Intermediates of the RELN-DAB1 pathway accumulate in neurons that express APOER2 extracellularly, forming plaques. Bracher-Smith et al. (2022) Genetic study. Study of ApoE4 homozygous individuals. Samples from the UK Biobank. To identify if ApoE4 carriers have other genetic predisposing factors in the progression of AD. 5390 individuals. 288 cases and 5102 controls. ●Subjects older than 65 years homozygous for ApoE4. ●Patients with sex chromosome aneuploidy are excluded. ●The control group is obtained from subjects who are negative for AD but positive for other dementias. The DAB1-RELN pathway is closely related to ApoE expression and therefore directly involved in the neuropathogenesis of AD. ●No significant SNP associations are found at the ApoE locus with AD. ●New significant loci are identified throughout the AD genome by mapping DAB1. ●The DAB1-RELN pathway is associated with AD. ●There is an epistatic interaction between ApoE and the DAB1RELN pathway. L´ opez-Font et al. (2022) CSF. To assess whether proteolytic fragments of Reelin are affected in CSF in AD. 43 cases versus 14 controls. Differentiation according to sex, age (79 years) and mutated ApoE status. ●Decreased Reelin length of 420 kDa. ●310 kDa N-terminal increase in AD patients due to Ct site excision. ●500 kDa aberrant Reelin species in AD. Appearance of aberrant Reelin in AD subjects suggests the involvement of Reelin in AD. Ramsden et al. (2022) Postmortem tissue. In vitro biochemical experiments. Involvement of the Reelin-APOER2-DAB1 pathway in anatomical vulnerability for sporadic AD. 26 patients versus controls. Cognitively healthy, with mild cognitive impairment or sporadic AD. Differences in sex, age and postmortem interval. Cognitive assessment according to MMSE. ●APOER2 is expressed in areas of the entorhinalhippocampal pathway involved in memory. ●Reelin-APOER2-DAB1 and other components of the cascade are expressed in neuritic plaques in AD cases. ●Markers of the ReelinAPOER2-DAB1 pathway are associated with disease progression and cognitive deficits The findings provide information to link lipid peroxidation of APOER2 and components of the Reelin-APOER2-DAB1 pathway to ApoE receptors in sporadic AD. Lid´ on et al. (2020) Postmortem frontal cortex tissue and CSF samples from Hospital Clinic of Barcelona and G¨ ottingen, Germany. To assess the potential role of CSF Reelin expression as a diagnostic and prognostic method. 246 postmortem samples, of which: ●AD N =55. Samples are randomly distributed between Germany and Spain. Healthy subjects, AD and mild cognitive ●Increase of RELN mRNA in frontal cortex in advanced stages of AD. Not observed in stages of mild cognitive impairment. CSF Reelin mRNA levels cannot be considered a diagnostic method, but are a prodromal marker of AD. (continued on next page) A.I. Valderrama-Mantilla et al. Neuroscience and Biobehavioral Reviews 169 (2025) 106032 5 Table 1 (continued) First author and year Type of study and sample Study target Study subjects Clinical variables included Results Main findings ●Mild cognitive impairment N=40. ●Controls N=41. impairment are included. ●Decreased levels of Reelin in CSF in dementia patients compared to controls. Han et al. (2019) RNA sequencing of postmortem hippocampal tissue. To identify the association of three exon and SNP skipping events of RELN and NOS1 genes in AD, as well as their involvement in β-amyloid accumulation. 24 AD cases versus 50 cognitively healthy elderly controls. Initial age, sex and years of education. Classification according to Braak scale. ●Decreased expression of two exons of RELN and one of NOS1 in AD. ●The RELN SNP rs362771 is related to cortical β-amyloid levels. ●This SNP is a regulatory splicing element. Exon skipping events and SNPs that impact the splicing process in the human hippocampus play a role in the pathogenesis of AD by participating in the accumulation of β-amyloid. L´ opez-Font et al. (2019) CSF sample. Evaluation of the efficiency of Reelin signaling in AD, focusing on the relationship between Reelin and APOER2 in CSF. ●Sporadic AD N=14. ●Autosomal Dominant Alzheimer’s Disease (ADAD) N =7. ●Controls N=7. Patients with sporadic AD and ADAD. Differentiation by age. ●Patients with sporadic AD express lower levels of APOER2 in CSF and a higher Reelin/APOER2 ratio. ●ADAD carriers of PSEN1 mutations had higher APOER2 levels and lower Reelin/APOER2 levels. ●APOER2 levels in CSF may be a suitable measure to assess altered Reelin signaling in AD. ●Differences in Reelin expression between sporadic AD and ADAD. Mata-Balaguer et al. (2018) Postmortem tissue from the Institute of Neuropathology of Barcelona. To explore the effect of β-amyloid (Aβ42) on DNA methylation of the RELN promoter at the Cterminal end. 30 cases: 12 women and 18 men. Patients aged 75 ±10 years. They are classified according to Braak stage. ●Aβ42 decreases DNA methyltransferase levels. ●Methylation of the RELN promoter does not change with Aβ42 administration. ●APOER2 levels are found to be lower after Aβ42. ●Aβ42-induced Reelin signaling is associated with increased Reelin expression in AD. ●APOER2 levels decrease while Reelin levels increase. ●APOER2 exerts a modulatory role on Reelin expression. Cuchillo-Ib´ a˜ nez et al. (2016) Postmortem tissue. Frontal cortex. Study of the interaction between Reelin and β-amyloid in the brain. Patients in advanced stages of AD according to the Braak classification. Difference by age and sex. ●Reelin shows interaction with β-amyloid, especially in late stages of AD. ●Increased Reelin, with reduced DAB1 and APOER2 internalization. ●Reelin expression is increased in AD, but interaction with Aβ hinders its function. ●APOER2 fragments in CSF could be a measure to assess the efficiency of Reelin signaling, which is decreased in AD. Gao et al. (2015) Genetic study. Exploration of the genetic implications of AD and its biological pathways. 17 genes are studied in 3 independent AD datasets. Caucasian population. Of the genes studied, the expression of DAB1 mRNA in human dorsolateral prefrontal cortex stood out. ●High expression of DAB1 mRNA in the prefrontal cortex is associated with protection against the development of AD. Bufill et al. (2015) Peripheral blood sample. Involvement of the Reelin pathway in the risk of developing AD or mild cognitive impairment (MCI). ●AD N =121. ●MCI N =94. ●Controls N=198. 40 SNP variants in 8 genes related to the RELN pathway are genotyped in a Spanish population. ●Association between RELN (rs528528 and rs2299356), PLK2 (rs15009 and rs702723) and CAMK2A (rs3756577 and rs3822606) genotypes and AD is found. ●RELN-rs2299356 association with AD risk. ●Protective effect of RELN-rs528528, PLK2rs15009 and PLK2rs702723 genotype. ●CAMK2A-rs756577 genotype associated with reduced risk of AD. ●There is a link between certain genotypes of the Reelin signaling pathway, MCI and AD. ●Protective and risk genotypes within the RELN signaling pathway exist for AD. Feh´ er et al. (2015) Genetic study. Association between genetic variants of RELN and the risk of developing AD. 432 cases with 308 controls from a Hungarian population. Patients with lateonset AD of 70 ±5 years are included. SNPs rs528528 and rs607755 are linked to AD in men. Not in women. ●rs528528 and rs607755 are associated with AD risk, specifically in males. Notter and Knuesel (2013) Postmortem human hippocampal tissue and CSF from the lateral ventricles. Investigation in humans of the already reported age-related reduction of Reelin in mice and its 8 AD cases versus 8 controls. Individuals of the same age and clinical stage from the Netherlands Brain Bank, ●Presence of Reelin fragments containing Nand C-termini in the amyloid bodies (CAm), ●The presence of Reelin in CAm may be linked to disturbances in neuronal transport in (continued on next page) A.I. Valderrama-Mantilla et al. Neuroscience and Biobehavioral Reviews 169 (2025) 106032 6 et al., 2022). In contrast, carriers of presenilin 1 (PSEN1) mutations associated with autosomal dominant AD (ADAD) displayed higher APOER2 protein levels and a lower Reelin/APOER2 ratio (L´ opez-Font et al., 2019). These findings suggest that CSF APOER2 protein levels may reflect alterations in Reelin signaling in AD, highlighting differences in protein expression between sporadic AD and ADAD (L´ opez-Font et al., 2022). Regarding alterations specifically related to Reelin, variants in its gen RELN, such as SNP rs362771, have been linked to amyloid-beta accumulation (Han et al., 2019). Genetic studies have also revealed that the RELN-COLBOS variant, which binds to VLDLR and APOER2 receptors, is associated with resistance to early AD symptoms in individuals with high-risk mutations (Lopera et al., 2023). This COLBOS variant exhibits a gain-of-function effect, enhancing DAB1 activation and reducing human Tau phosphorylation (Lopera et al., 2023). These findings underscore the crucial role of the Reelin signaling pathway in dementia resistance, particularly in the context of genetic risk factors for AD. Notably, upregulation of Reelin has also been observed as a compensatory response to stress related to Tau or amyloid-beta, even before the onset of dementia symptoms, suggesting that Reelin may serve as a premorbid marker in AD (Cuchillo-Ib´ a˜ nez et al., 2013; Han et al., 2019). In the same vein, increased presence of Nand C-terminus-containing Reelin fragments has been detected in corpora amylacea, age-related deposits found abundantly in the molecular layer of the hippocampus in AD patients (Notter and Knuesel, 2013). Altogether, these findings suggest that alterations in Reelin may precede the appearance of dementia symptoms in AD (Kramer et al., 2011). Altered RELN mRNA levels were also noted in CSF from AD patients (Cuchillo-Ib´ a˜ nez et al., 2016;Lid´ on et al., 2020). Finally, L´ opez-Font et al. (2022) showed a marked reduction in the full-length 420 kDa form of Reelin in AD compared to controls, alongside a notable increase in the N-terminal 310 kDa form, attributed to cleavage at the C-terminal site. Additionally, an aberrant 500 kDa form of Reelin was also identified in AD patients (L´ opez-Font et al., 2022). Regarding epigenetic mechanisms, in vitro administration of amyloid-beta 42 led to decreased levels of DNA methyltransferase (DNMT) (Mata-Balaguer et al., 2018). Interestingly, while RELN promoter methylation remained unaffected in AD brain, APOER2 protein and mRNA levels decreased, and Reelin levels increased, supporting the view that APOER2 could exert a modulatory role on reelin expression (Mata-Balaguer et al., 2018). Additionally, Mata-Balaguer et al. propose that measuring soluble N-terminal ApoER2 fragments in CSF and Table 1 (continued) First author and year Type of study and sample Study target Study subjects Clinical variables included Results Main findings accumulation in neuritic varicosities along the olfactory limbic tracts. Amsterdam are included. spherical age-related deposits. ●Increased density of these deposits in the hippocampal molecular layer in AD. the ageing process, resulting in the accumulation of metabolites in neuritic varicosities. ●Aging changes in Reelin levels could be involved in CAm formation by affecting cytoskeleton dynamics. ●The presence of CAm is an indicator of the degenerative state of neuritic compartments. Cuchillo-Ib´ a˜ nez et al. (2013) Postmortem tissue from human frontal cortex. To determine whether β-amyloid-induced alteration of Reelin will lead to a signaling error in the Reelin pathway, contributing to the pathogenesis of AD. 5 cases with AD of 66 ±7 years versus 5 controls of 73 ±2 years. UIPA Tissue Bank, Madrid. Sporadic cases of AD in Braak stages V-VI. Controls without dementia or brain pathology. 73 ±2 years. Postmortem interval 6 hours. ● β-amyloid-induced Reelin variants show a reduced ability to negatively regulate tau phosphorylation via GSK3B and DAB1. ●If Reelin signaling is defective, there is a positive regulation of the 14–3–3 protein that drives tau phosphorylation through modulation of GSK3B. ● β-amyloid-induced Reelin species result in a reduced ability to bind APOER2. ●Reelin in AD fails to form physiologically active dimers. ● β-amyloid induces the expression of abnormal Reelin species. ●It is the altered expression of Reelin by β-amyloid that results in a disruption of the Reelin signaling pathway. ●Association of deregulation of β-amyloid and tau phosphorylation through Reelin. ●Reelin directly involved in the pathogenesis of AD. Kramer et al. (2011) Postmortem limbic and neocortical region tissue. Genetic differences in older adults with and without dementia, according to neurofibrillary tangle (NFT) burden. 299 nondemented with autopsy: ●185 with high NFT burden. ●114 with low NFT burden. ●>65 years old. ●Deceased with autopsy. ●Diagnosis of nondementia. ●>1 clinical assessment within 1 year prior to death. ●DNA available. ●Caucasian. ●Diagnosis of AD according to DSM-IIIR. ●Reelin signaling pathways are involved in the phosphorylation of tau, the main component of NFTs. ●Reelin is involved in the phosphorylation of tau directly or through β-amyloid. ●RELN variants are related to the pathogenesis of cognitive health. ●Positive regulation of Reelin could represent a compensatory response to tauor β-amyloidlinked stress associated with AD, even before the manifestation of dementia. ●Reelin premorbid marker. A.I. Valderrama-Mantilla et al. Neuroscience and Biobehavioral Reviews 169 (2025) 106032 7 intracellular C-terminal ApoER2 fragments in frontal cortex extracts may serve as a reliable indicator of Reelin signaling impairment in patients with AD. 3.4. Role of the Reelin signaling pathway in SZ Studies included in this systematic review examining the role of the Reelin signaling pathway in SZ are summarized in Table 2. Research has highlighted the critical involvement of the transcription factor early growth response 3 (EGR3) in regulating RELN signaling by binding to its promoter region (Nie et al., 2021). Overexpression of EGR3 leads to neurite overgrowth, which can be reversed through RELN deletion. Notably, patients with SZ exhibit decreased levels of EGR3 and RELN mRNA in peripheral blood, while postmortem brain tissue shows overexpression of both factors (Nie et al., 2021). Taken together, this evidence highlights a significant correlation and strongly suggests a critical role for EGR3 in the pathogenesis of SZ (Nie et al., 2021). Furthermore, a deficit in Reelin expression has been shown to impact GABAergic Purkinje neuron development during cerebellar maturation (Maloku et al., 2010). Reduced Reelin expression was found in specific SZ brain regions, particularly in the left prefrontal cortex, suggesting a potential role in neurological development and deficiencies in synaptic plasticity associated with the disorder (Habl et al., 2012). Alterations in peripheral Reelin expression were also noted, with SZ patients exhibiting higher Reelin protein concentrations compared to controls (Hornig et al., 2015). Genetic studies have also identified positive associations between Reelin and SZ, providing valuable insights into the genotype-phenotype correlations associated with this disorder. A notable 4-RELN-SNP haplotype comprising rs362814, rs39339, rs540058, and rs661575 (Bai et al., 2019), alongside RELN-rs362719 (Kuang et al., 2011), demonstrated a strong association with SZ, whereas another haplotype, TCTC, appeared to confer a protective effect (Luo et al., 2019). These genetic associations have also been linked to the severity and specific characteristics of SZ symptoms (Luo et al., 2019). However, some authors suggest that the entire RELN gene may not be directly implicated in SZ pathogenesis, as evidenced by the lack of significant associations with SNPs rs155333, rs6465938, and rs2535764 (Xu et al., 2020). Interestingly, RELN is also believed to modulate responses to antipsychotic treatment, due to the association of the SNPs rs362814, rs362626, and rs2237628 with treatment response. This suggests that specific alterations in various RELN SNPs may serve as predictive biomarkers for treatment efficacy (Xu et al., 2020). Additionally, sex-specific associations between RELN SNPs, such as rs7341475, and SZ underscore differences in disease development and pathogenesis between men and women (Liu et al., 2010; Li et al., 2011; Ovadia and Shifman, 2011; Sozuguzel et al., 2019), highlight the complexity of the contributing factors to the disorder (Ovadia and Shifman, 2011). Other genes involved in synaptic plasticity and neurological development linked to Reelin signaling, including APOE, APOER2, VLDLR, and DAB1, have also been implicated in memory performance and cognitive function in SZ patients (Verbrugghe et al., 2012). Finally, genetic evidence supports RELN as a modifier gene in SZ pathogenesis, with consistent associations between certain alleles and symptom severity, particularly regarding cognitive impairment and both positive and negative symptoms (Wedenoja et al., 2010). From an epigenetic perspective, Aberg et al. (2014) and Zhou et al. (2022) showed that the CpG island methylation rate in the RELN gene promoter region is elevated in control subjects compared to SZ patients, with no significant differences identified between SZ subtypes I and II (Aberg et al., 2014; Zhou et al., 2022). Notably, changes in RELN promoter methylation induced by cognitive rehabilitation therapy were associated with improvements in frontotemporal functional connectivity in SZ patients, as well as enhanced cognitive performance, suggesting that epigenetic modifications in RELN may promote cognitive plasticity and improve recognized disease markers (Ho et al., 2020). Additionally, higher RELN promoter methylation in peripheral blood was observed in men relative to women, indicating once again a potential role of RELN methylation in SZ pathogenesis, particularly among males (Nabil Fikri et al., 2017). Moreover, a significant reduction in global DNA methylation, including the methylation of RELN gene, was observed in SZ subjects, especially in men, compared to healthy twins, supporting the hypothesis of an epigenetic model of SZ, wherein DNA methylation plays a crucial role in gene expression regulation and disease progression (B¨ onsch et al., 2012). Lastly, increased binding of the enzyme DNMT1, which catalyzes the transfer of methyl groups to specific CpG sites in DNA, to GABAergic and glutamatergic RELN promoters was noted in SZ patients, although this binding did not correlate with increased promoter methylation (Dong et al., 2015). This DNMT1 binding was primarily localized to the cortex and not the cerebellum, indicating a neuron-specific mechanism that is likely independent of DNA methylation activity (Dong et al., 2015). 4. Discussion The concept of "dementia praecox" in SZ is crucial for understanding the temporal aspects of the disorder. Emil Kraepelin categorized SZ within a group of clinical syndromes leading to "processes of deterioration," which he termed "dementia praecox" (Kirkpatrick et al., 2008; Angst, 2002). The author divided endogenous psychoses into two main groups: manic-depressive psychoses, which he considered curable, and dementia praecox, deemed incurable. Kraepelin further refined this classification by distinguishing between two variants of dementia praecox: progressive, with continuous deterioration, and intermittent episodes without irreversible deficits (Angst, 2002). This historical perspective aligns with the idea of SZ having a neurodegenerative component, similar to AD (Caspi et al., 2024). In this context, the Reelin signaling pathway, particularly the RELN-APOER2-DAB1 complex, has emerged as a critical factor in the neuropathology of both AD and SZ, offering further insight into the neurodegenerative aspects of these disorders (Antoniades et al., 2011; Verbrugghe et al., 2012; Bracher-Smith et al., 2022; Ramsden et al., 2023). In this systematic review, we highlight that early pTau accumulation in AD reflects disease progression (Cuchillo-Ib´ a˜ nez et al., 2016;Andrade-Guerrero et al., 2023;Zhang et al., 2024) and is linked to disruptions in the APOER2-DAB1 signaling pathway (Cuchillo-Ib´ a˜ nez et al., 2016; Lid´ on et al., 2020;Ramsden et al., 2023) as well as decreased Reelin levels in CSF (Notter and Knuesel, 2013;L´ opez-Font et al., 2019;Lid´ on et al., 2020;L´ opez-Font et al., 2022). Neuritic plaques formation has also been associated with these pathways (Ramsden et al., 2022, 2023). Additionally, amyloid-beta-induced variants of Reelin impair Tau phosphorylation, suggesting that amyloid-beta may accelerate AD progression by disrupting the RELN pathway (Ramsden et al., 2022, 2023). These findings highlight the role of the RELN-APOER2-DAB1 signaling pathway in AD progression. In parallel, the Reelin signalling pathway is also involved in the neuropathogenesis of SZ through a complex network of molecular interactions (Maloku et al., 2010; Habl et al., 2012; Hornig et al., 2015; Nie et al., 2021). Altered Reelin expression, particularly reductions in the short isoform N-R2 (Fig. 1), has been identified in brain areas associated with SZ (Ovadia and Shifman, 2011). As here showed, genetic and biochemical analyses suggest disruptions in the RELN-DAB1 signalling pathway contribute to SZ pathogenesis, with rare variants of DAB1 being implicated (Impagnatiello et al., 1998; Guidotti et al., 2000a; Luo et al., 2019; Xu et al., 2020) (Nawa et al., 2020). Interestingly, DAB1 cKO mice exhibited behavioral abnormalities reminiscent of symptoms observed in patients with SZ (Imai et al., 2017), supporting a link between Reelin levels, synaptic plasticity, and neurological development in SZ. Additionally, associations between specific RELN SNPs and SZ have been identified (Liu et al., 2010; Kuang et al., 2011; Ovadia and Shifman, 2011; Luo et al., 2019; Sozuguzel et al., 2019; Xu et al., 2020), with some variants linked to social withdrawal and impulsivity A.I. Valderrama-Mantilla et al. Neuroscience and Biobehavioral Reviews 169 (2025) 106032 8 Table 2 Main findings of studies on the role of Reelin signaling pathway in Schizophrenia. Zhou et al. (2022) Peripheral blood. Association between methylation of the CpG island in the RELN promoter region and the types of positive (Type I) and negative (Type II) SZ. 200 cases versus 200 controls. ●Diagnosis of SZ according to DSM-IV. ●Between 18 and 65 years old. ●First episode of SZ. ●No treatment 2 weeks before admission. ●Exclusion of psychiatric comorbidities. ●Positive and negative symptoms. ●Chinese population. ●Methylation rate of the CpG island in the RELN promoter region was higher in the control group. ●There is no difference between type I and type II. ●Methylation of the CpG island in the promoter region of RELN is associated with SZ, but not with its clinical subtype. Nie et al. (2021) Postmortem tissue. Sample of prefrontal cortex (Brodmann Area 9) and parietal cortex. Search for new transcription factors (TFs) of the Reelin pathway involved in SZ. ●Postmortem tissue: 5 cases versus 5 controls. ●Peripheral blood: 40 cases versus 40 controls from a population in northwest China. ●Diagnosis of SZ according to DSM-IV. Controls without psychiatric history. It includes sex, comorbidities, age, postmortem interval and alcoholism. ●In peripheral blood: patients with medication, drug use, DM, HTA and tumors are excluded. ●EGR3 is a TF that regulates Reelin signaling by binding to its promoter region. ●Overexpression of EGR3 induces neurite outgrowth, reversible by deletion of RELN. ●The amount of EGR3 and RELN in peripheral blood in SZ subjects were decreased and correlated with each other. ●EGR3 and RELN are overexpressed in postmortem SZ tissue. ●EGR3 is a TF of the Reelin pathway and regulates neurite proliferation. ●EGR3 via the Reelin signaling pathway plays a role in the molecular mechanisms of SZ. Ho et al. (2020) Saliva and peripheral blood samples. Epigenetic and behavioral involvement in molecular changes of disease evolution after cognitive rehabilitation (CR). 35 SZ hospitalized patients versus 15 healthy controls. Patients admitted to cognitive rehabilitation with similar pretest IQ conditions. Diagnosis of SZ according to DSM-IV. Age less than 40 years. Hospitalization of less than 8 days. No family history of SZ. ●After CR, changes in the promoter of Reelin CpG. ●Improved intrafrontal and frontotemporal functional connectivity. ●Increased cognitive performance. ●Changes in Reelin methylation result in increased global efficacy and improved cognitive performance. ●CR-driven epigenetic modifications on RELN promote cognitive plasticity and improve known markers of disease. Xu et al. (2020) A total of 15 RELN SNPs are genotyped. Evaluation of the influence of RELN on the response to antipsychotic medication. 260 unrelated hospitalized patients with SZ. Gender and age of onset (18’65 years) are included. Han Chinese population. DSM-IV diagnosis. The SNPs studied rs155333, rs6465938 and rs2535764 are not significant. ●RELN affects the outcome of antipsychotic treatment. ●RELN SNPs are predictive biomarkers of response to antipsychotics. Luo et al. (2019) Genetic analysis. 30 SNPs single nucleotid polymorphisms of RELN. Contribution of RELN to the predisposition and severity of SZ symptoms. 102 cases of unrelated SZ patients versus 169 healthy controls. Samples from the Third People’s Hospital in Zhongsgan, randomly recruited from volunteers. Southern Han Chinese. Patients with positive and negative symptoms, paranoid or undifferentiated SZ, who had a fist onset or recurrence after stopping treatment at least 1 month ago. Diagnosis according to DSM-IV. Difference between sex, age, drugs, family history of SZ. It shows positive associations between RELN and SZ. Provides information on genotypephenotype correlation for SZ. It suggests that RELN is involved in both the development and the severity and characteristics of SZ symptoms. ●Haplotype 4-SNP consisting of rs362814, rs39339, rs540058 and rs661575 highly associated with SZ. ●Haplotype TCTC is a protective factor for SZ. ●Haplotype 4-SNP is associated with social avoidance, impulsive and dangerous behavior in subjects with SZ. ●Haplotype composed of rs2229864, rs2535764 and rs262355 is related to hostility. ●rs727708 is related to both SZ risk and severity. Sozuguzel et al. (2019) Genetic analysis. Isolation of genomic DNA. Involvement of RELN rs7341475 variant in SZ. 105 cases of SZ patients versus 137 healthy controls. Turkish population. Age between 18 and 73 years. Sex differentiation. Diagnosis of SZ according to DSM-V. There was only association between RELN rs7341475 and SZ in the female gender. Samples from Kocaeli Hospital. ●rs7341475 association with SZ in women. ●No rs7341475 association in the general population. Bai et al. (2019) Genetic study. Four SNPs (rs1062831, rs3808039, rs362746 and rs736707) were genotyped in the RELN gene. Association of RELN and SZ in samples classified by sex. 1536 participants. Chinese population. Differentiation by sex. ●rs362746 is associated with SZ. ●No sex-specific role is found. ●RELN is a gene involved in the pathogenesis of SZ. (continued on next page) A.I. Valderrama-Mantilla et al. Neuroscience and Biobehavioral Reviews 169 (2025) 106032 9