Neurexins, primarily transcribed from three genes (in animals: Nrxn1, Nrxn2, Nrxn3; in humans: NRXN1, NRXN2, NRXN3) (Khoja et al., 2023), are pivotal presynaptic adhesion proteins within the nervous system. These proteins play critical roles in synapse formation and function, manifesting in two primary forms: the longer α-neurexins and the shorter β-neurexins (Reissner et al., 2013; Zhang et al., 2022; Trotter et al., 2023). α-neurexins, characterized by six large extracellular laminin/neurexin/sex hormone-binding (LNS) globulin domains and three interspersed epidermal growth factor (EGF)-like regions; along with β-neurexins, featuring only the sixth LNS domain and absent EGF-like regions, both play a pivotal role in synaptic function (Cuttler et al., 2021). These proteins, expressed at both excitatory and inhibitory synapses (Reissner et al., 2013), contribute to neurotransmission, synaptic plasticity, and neuronal development (Trotter et al., 2023). Their structures, particularly in the neurexin 1 protein, have been elucidated in various species, though human-specific structural data remains elusive (Cuttler et al., 2021).

The intricacies of neurexin functionality are underscored by their binding capability to various postsynaptic ligands, including neuroligins, cerebellins, and leucine-rich repeat transmembrane proteins (LRRTMs) (Francks, 2011; Asede et al., 2020). Such interactions underscore the multifaceted nature of neurexins in synaptic operations and their potential association with neurodevelopmental and neuropsychiatric disorders when dysfunctional (Kasem et al., 2018). Perturbations in neurexin signaling have been implicated in various neurodevelopmental and neuropsychiatric disorders, including autism spectrum disorders (ASD), schizophrenia (SCZ), bipolar disorder (BD), and attention deficit hyperactivity disorder (ADHD) (Kasem et al., 2018; Cuttler et al., 2021).

It should be noted that deletions of NRXN 1–3 and these neurodevelopmental and neuropsychiatric disorders are highly interconnected and not simple one-to-one relationships (Kasem et al., 2018). While not all individuals with NRXN deletions inevitably develop neurodevelopmental or neuropsychiatric disorders, a single NRXN deletion could potentially enhance the risk for a multitude of such disorders. Deletions of neurexin-1α, for instance, have been linked with a significantly elevated risk for SCZ, ASD, and intellectual disability (Reichelt et al., 2012). In contrast, a single disorder could also result from different deletions of neurexin genes. For example, dysfunctions across all three neurexin genes (i.e., NRXN 1–3) have been identified as contributors to a disrupted balance between excitatory and inhibitory neurotransmission in human ASD (Khoja et al., 2023). These findings further highlight the need to figure out the complex association between neurexins and these disorders.

Previous research has investigated the role of neurexins using both in vivo and in vitro approaches (Cuttler et al., 2021; Gomez et al., 2021). While genetic studies have greatly expanded our understanding of the involvement of neurexins in these disorders (Tromp et al., 2021), a significant gap remains in the functional analysis of these genetic aberrations. Understanding the precise molecular and cellular outcomes of neurexin dysfunction can pave the way for targeted therapeutic interventions. Herein lies the importance of in vitro human models, specifically induced pluripotent stem cells (iPSCs). By reprogramming somatic cells from patients into iPSCs, and subsequently differentiating these cells into neurons, researchers can create patient-specific neural models that recapitulate disease phenotypes (Jusop et al., 2023). This approach allows for the study of neurexin dysfunction in a cellular context that closely mirrors the patient’s genetic and epigenetic background (Lee et al., 2020). Meanwhile, somatic cells from healthy population could be used for comparison (Raab et al., 2014). On the other hand, animal models, particularly rodents, have traditionally served as invaluable tools for dissecting the physiological consequences of neurexin dysfunction (Bryda, 2013; Tromp et al., 2021). They offer insights into behavioral manifestations, neuronal circuitry alterations, and potential compensatory mechanisms that might arise due to the dysfunctional neurexin signaling. By employing both iPSCs and animal models, researchers could achieve a more comprehensive understanding that combines human-specific cellular insights with in vivo physiological observations (Rowe and Daley, 2019).

Yet, despite the existing plethora of studies on neurexin dysfunction’s relevance to neurodevelopmental and neuropsychiatric disorders, comprehensive systematic review encompassing both iPSC and animal model perspectives is lacking. This systematic review aims to bridge this knowledge gap by amalgamating existing research, shedding light on the molecular, cellular, physiological, and behavioral impacts of neurexin dysfunction, while highlighting both their commonalities and differences.

This systematic review was conducted in accordance with the 2020 guidelines and principles detailed in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement and checklist (Shan et al., 2022). The PICOS (Participants; Interventions; Comparisons; Outcomes; Study designs) are defined in Table 1.

We mainly performed a systematic search of PubMed and Google Scholar databases, targeting original peer-reviewed research articles published in the last 20 years from 10/05/2003 to 10/05/2023. The search terms included combinations of the following: [(Neurexin) OR (NRXN)) AND ((IPSC) OR (Induced Pluripotent Stem Cells) OR (Mouse) OR (Rat) OR (Animal)) AND ((Schizophrenia) OR (Autism) OR (Bipolar disorder) OR (Attention-deficit/hyperactivity disorder) OR (ADHD) OR (Neurodevelopmental disorders) OR (Neuropsychiatric Disorders)]. Only English language papers were included. Papers were initially screened by Title and Abstract Only, and excluded if deemed not relevant because the study was not on animal or iPSCs models, neurexin, or any neurodevelopmental/neuropsychiatric disorder. In a second round of screening, full-text papers were assessed for eligibility, and excluded if not relevant to our objectives. Additional search criterion was studies from the manually screening of the reference lists of extracted articles between 2003 and 2023, in order to identify additional studies that may fit the systematic review objective in the current article but were not identified by the PubMed and Google Scholar databases. In case of circular demonstration of research results combined with original empirical articles, only empirical studies were retained if a systematic or narrative review addressing the potential linkage between neurexin dysfunction and these disorders emerged (Shan et al., 2021).

During the second round of screening, we meticulously screened pertinent studies that utilized animal models or IPSCs as their primary methodologies. Emphasis was placed on retaining studies that elucidated synaptic/physiological activity and/or animal behavioral outcomes pertinent to neurodevelopmental and/or neuropsychiatric disorders as a result of neurexin dysfunction. Conversely, studies that lacked explicit mention of the potential association between neurexin dysfunction and these disorders were excluded.

For the studies employing animal models, the extracted data encompassed the targeted neurexin, type of animal used, generation strategy of the model, model viability, synaptic/physiological activities, principal behavioral findings, and the inferred neurodevelopmental or neuropsychiatric disease correlation. On the other hand, data retrieval from studies employing IPSCs comprised the targeted neurexin, donor details and sample count, IPSC model generation technique, genotypic profiles, cellular phenotypic outcomes, and the extrapolated disease correlation.

In addition, we utilized certain methods to evaluate bias risk in both in vitro and in vivo studies. The quality and risk for bias in the included in vitro (i.e., IPSC) studies were assessed by the adapted quasi-experimental studies appraisal tool by the Joanna Briggs Institute (Tufanaru et al., 2017). Each in vitro investigation was designated a risk status: low (where any detected bias would not substantively skew results), unclear (where potential bias introduces some result uncertainty), or high (indicating that bias could significantly compromise findings). For in vivo research, we employed the SYRCLE risk assessment tool, which refers to the Systematic Review Centre for Laboratory Animal Experimentation’s risk of bias tool, tailored for the included animal studies (Silveira et al., 2023). This instrument is structured around various bias categories: selection, performance, attrition, reporting, and other biases. The current systematic review involved a ten-question evaluation of the articles, with possible responses being “YES” (signifying minimal bias risk), “NO” (indicative of pronounced bias risk), and “UNCLEAR” (signifying indeterminate bias risk).

From our search on the PubMed and Google Scholar databases, we procured a total of 1754 records. Subsequent exploration through a ‘citation searching’ methodology yielded an additional 8 records. Firstly, 234 were identified as duplicates and thus removed. Following a preliminary assessment of titles and abstracts, we excluded 1,477 of the 1,528 entries. Subsequently, 51 articles progressed to an advanced screening phase, from which a further 24 were eliminated due to their irrelevance to our stipulated inclusion criteria or their categorization as systematic/narrative reviews upon thorough examination of their full text. Consequently, 27 articles formed the core of our review. Referring to Figure 1 for a detailed overview of our search and screening procedures. All these studies, without exception, probed the potential link between neurexin dysfunction and neurodevelopmental and/or neuropsychiatric disorders, utilizing either animal or IPSC models as their investigative platforms. A schematic representation of neurexin-neuroligin interaction is shown in Figure 2.

In an effort to gauge the risk of bias in both in vitro and in vivo studies, we have encapsulated our quality assessment findings in Figure 3 (pertaining to IPSC studies) and Figure 4 (relevant to animal studies). A noteworthy observation is that a predominant fraction of the in vitro studies showcased minimal bias risk. However, an anomaly was discerned for the criterion addressing the frequency of outcome measurements both pre and post-intervention/exposure. Herein, a palpable bias risk was evident, attributed largely to these studies’ inclination towards contrasting different cell lines (notably control vs. NRXN-deleted) rather than employing iterative measurements on identical cells prior to and subsequent to specific interventions. As for in vivo studies, we identified salient deficiencies in certain quality components. For instance, concerning the criterion, “Was the allocation sequence generated and executed appropriately?” a mere pair of studies (Esclassan et al., 2015; Boxer et al., 2021) demonstrated a minimal bias risk owing to their explicit randomization acknowledgment. Conversely, a sizable cohort exhibited pronounced bias risks due to ambiguous animal selection during result evaluation and/or a lack of randomized housing during experimentation. Furthermore, universal bias risks were detected in the realm of incomplete outcome data reporting. In addition, the absence of accessible study protocols introduced ambiguity in reporting bias, making it challenging to ascertain if the pre-determined primary and secondary outcomes were duly reported. Lastly, when assessed against the criterion, “Was the study apparently free of other problems that could result in high risk of bias?,” all studies uniformly exhibited a minimal bias risk.

The categorization of the 27 key studies according to their research models showed that 21 utilized animal models, while 6 employed IPSC models. An analysis of the neurexin targets within these studies revealed a distinct emphasis on nrxn1/NRXN1, especially Nrxn1α/NRXN1α. Notably, a substantial majority (more than 75%, n = 21) of these studies suggested a link to ASD as a potential associated condition, with other neurodevelopmental and neuropsychiatric disorders like schizophrenia, ADHD, and intellectual disability also being mentioned.

Within the 21 studies focusing on animal models, rodents, specifically rats and mice, were predominantly used in 19 of them, while Zebrafish and Drosophila melanogaster were each featured in one study. The research heavily favored Nrxn1α, explored in 14 studies, with less frequent examination of Nrxn2α, Nrxn3α, and Nrxnβ. Most studies adopted knockout (KO) strategies in mice, targeting specific Nrxn genes such as Nrxn1α, Nrxnβ, and Nrxn3. Variability in genetic backgrounds of these models (e.g., pure C57BL/6 J, mixed backgrounds) was noted, potentially contributing to phenotypic differences observed. For instance, while Etherton et al. (2009) and Armstrong et al. (2020) both researched Nrxn1α KO mice, their differing genetic backgrounds led to distinct behavioral outcomes, including increased aggression and impaired prepulse inhibition, respectively. Concerning model viability, the majority of studies successfully developed viable models. The earliest Nrxn loss of function mouse models were introduced by Missler et al. (2003) who generated single, double, and triple Nrxn1α/2α/3α KO mice. Tromp et al. (2021) highlighted the importance of maintaining at least two intact α-Nrxns for survival, based on variable survival rates in double and triple KO mice. In a more recent study, Cheung et al. (2023) discovered that triple Nrxn knockout in 5-HT neurons led to a significant reduction in serotonin neurons during early postnatal stages, indicating the viability of these models but with specific neural deficits. The studies consistently reported synaptic and physiological changes due to Nrxn gene disruptions. For example, Alabi et al. (2020) observed neural signaling disruptions in the striatum, and Born et al. (2015) noted a decrease in spontaneous transmitter release at excitatory synapses. Behavioral findings in these animal models displayed both similarities and variances. Commonly reported were deficits in social interactions and increased anxiety-like behavior, aligning with traits in ASD and Schizophrenia, as shown in studies like Dachtler et al. (2014) and Born et al. (2015). However, discrepancies were also apparent; Grayton et al. (2013) reported reduced locomotor activity and increased aggression, while Etherton et al. (2009) observed normal social and anxiety-like behaviors and locomotor activity. These variations suggest that the behavioral impacts of Nrxn disruptions can significantly differ due to the genetic background of the mice. The primary diseases correlated with these disruptions are predominantly ASD and schizophrenia, reflecting the strong link between Nrxn gene disruptions and these disorders. Nonetheless, the observed range of behavioral phenotypes, from social deficits to cognitive impairments, implies that the influence of these genes extends beyond these disorders, potentially affecting a wider spectrum of neurodevelopmental conditions.

In the hiPSC studies analyzed, hiPSC lines originated from diverse sources such as skin biopsies and peripheral blood mononuclear cells, collected from both healthy subjects and patients with disorders including ASD and schizophrenia. While the generation of these iPSC models consistently utilized cellular reprogramming techniques, there was a slight variation in the genetic focus across different studies. The primary attention was on NRXN1α, with investigations into the impacts of NRXN1α+/− deletions, exemplified by Avazzadeh et al. (2019, 2021), and bi-allelic NRXN1-alpha deletions, as reported by Lam et al. (2019). These studies primarily explored intragenic deletions affecting NRXN1, uncovering a spectrum of cellular phenotypes indicative of impaired neuronal function, such as the altered dynamics of calcium and sodium and modifications in neurotransmitter release, as detailed by Avazzadeh et al. (2019, 2021) and Pak et al. (2021). The cellular phenotypes identified exhibited both commonalities and differences; for instance, Avazzadeh et al. (2021) and Pak et al. (2021) observed changes in ion dynamics and neurotransmitter release, whereas Lam et al. (2019) noted a shift in cells towards a radial glia-like identity. Predominantly, the genetic and cellular alterations were linked to ASD and schizophrenia. Nevertheless, the observed range of symptoms and associated conditions, such as seizures, intellectual disability, and developmental delays, indicate that the influence of these genetic modifications might extend well beyond these disorders. This broader impact is exemplified by Avazzadeh et al. (2021), who associated NRXN1α+/− deletions with ASD and related symptoms, and Flaherty et al. (2019), who connected NRXN1 deletions to schizophrenia and bipolar disorder with psychosis.

Our review systematically examines Nrxn knockout/knock-in (KO/KI) animal models and human-induced pluripotent stem cell (hiPSC) studies on neurexins, highlighting their complex roles in various neurodevelopmental and neuropsychiatric conditions. Utilizing diverse methods, from CRISPR/Cas9 genome editing to hiPSC-derived neuronal cultures, these studies showcase the significant effects of neurexin variations, notably in disorders like ASD and schizophrenia (Avazzadeh et al., 2019; Flaherty et al., 2019). Neurexin dysfunction is also linked to other conditions such as ADHD, intellectual disability, seizures, and developmental delays. Tables 2, 3 provide in-depth information on the specifics of each hiPSC study and Nrxn KO/KI animal investigation, including target neurexins, the animals used, model generation strategies, model viability, synaptic or physiological activities, key findings regarding animal behaviors, and related diseases.

Recognizing the genetic link between neurexin variants and an increased risk for neurodevelopmental and neuropsychiatric disorders, our study seeked to understand the neuronal functions of these genes and their potential role in initiating, progressing, and exacerbating mental disorders. We have expanded upon previous reviews by generating animal models of Nrxns and integrating relevant data from hiPSC studies to better evaluate the mental conditions associated with these genes (Tromp et al., 2021).

This systematic review comprehensively explores the connections between neurexin dysfunction and a range of neurodevelopmental and neuropsychiatric disorders, employing a dual approach of animal models and hiPSC studies. The review highlights the significant implications of neurexin dysregulation, particularly in ASD and schizophrenia, while also acknowledging the broad spectrum of other neurodevelopmental disorders influenced by neurexin anomalies. The findings emphasize the need for more standardized methodologies in future research, particularly in the development of animal and hiPSC models. Additionally, the review underscores the importance of broadening our investigative focus beyond the primarily studied NRXN1 to include other less explored neurexin variants. This approach is crucial for deepening our understanding of these complex disorders and advancing towards personalized therapeutic interventions. The potential of hiPSC models as powerful tools in this research domain is particularly noted, promising to bridge existing gaps and propel translational research forward.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

DS: Conceptualization, Methodology, Resources, Validation, Writing – original draft, Writing – review & editing. YS: Conceptualization, Methodology, Writing – original draft, Writing – review & editing. YZ: Conceptualization, Methodology, Data curation, Writing – original draft, Writing – review & editing. CH: Investigation, Methodology, Validation, Visualization, Writing – review & editing. WX: Formal analysis, Methodology, Validation, Visualization, Writing – review & editing. ZL: Investigation, Validation, Visualization, Writing – review & editing. FG: Conceptualization, Investigation, Visualization, Writing – review & editing. QO: Formal analysis, Investigation, Visualization, Writing – review & editing. ZiD: Data curation, Investigation, Methodology, Writing – review & editing. ZhD: Conceptualization, Investigation, Methodology, Resources, Validation, Writing – review & editing.