Circadian clock crosstalks with autism

Abstract Background The mechanism underlying autism spectrum disorder (ASD) remains incompletely understood, but researchers have identified over a thousand genes involved in complex interactions within the brain, nervous, and immune systems, particularly during the mechanism of brain development. Various contributory environmental effects including circadian rhythm have also been studied in ASD. Thus, capturing the global picture of the ASD‐clock network in combined form is critical. Methods We reconstructed the protein–protein interaction network of ASD and circadian rhythm to understand the connection between autism and the circadian clock. A graph theoretical study is undertaken to evaluate whether the network attributes are biologically realistic. The gene ontology enrichment analyses provide information about the most important biological processes. Results This study takes a fresh look at metabolic mechanisms and the identification of potential key proteins/pathways (ribosome biogenesis, oxidative stress, insulin/IGF pathway, Wnt pathway, and mTOR pathway), as well as the effects of specific conditions (such as maternal stress or disruption of circadian rhythm) on the development of ASD due to environmental factors. Conclusion Understanding the relationship between circadian rhythm and ASD provides insight into the involvement of these essential pathways in the pathogenesis/etiology of ASD, as well as potential early intervention options and chronotherapeutic strategies for treating or preventing the neurodevelopmental disorder.

disorder (ASD) were found to exhibit distinct melatonin secretion patterns, which could potentially influence sleep challenges and circadian rhythm disruptions.In another clinical research (Ballester-Navarro et al., 2021), it has been stated that both normal and abnormal sleepwake rhythms in individuals ranging in age from 18 to 41 years (autistic cases) might be associated with specific gene variants in the circadian clock (PER1) and melatonin pathway (N-acetyl serotonin Omethyltransferase [ASMT]) systems.A growing body of evidence indicates that the misalignment of the circadian system is associated with various chronic diseases including cardiovascular, metabolic, and neuropsychiatric disorders such as schizophrenia and ASD (Lorsung et al., 2021).
The neuropsychiatric disorder ASD is primarily caused by genetic factors (e.g., mutations in genes, chromosomal abnormalities, and copy number variations [CNVs]) and environmental factors as studied extensively.Several research studies have demonstrated a direct correlation between ASD and circadian rhythm disorders (Jin et al., 2018) (Abdul et al., 2022).Pregnant women who work night shifts may experience stress and disrupted circadian rhythms, and potentially leading to altered hormone levels, including melatonin and cortisol.These altered hormone levels are considered risk factors for ASD (Jin et al., 2018) (Taylor & Corbett, 2014), as indicated by studies such as those conducted by Nir et al. (1995) and Corbett et al. (2009).Environmental conditions including shift work and sleep disorders can impact maternal circadian rhythms (Wong et al., 2022), potentially affecting the development of fetal circadian rhythmicity (Sorensen et al., 2020).
Preterm birth is closely associated with ASD (Fitzgerald et al., 2020) (Allotey et al., 2018).Besides, preterm babies experience constant exposure to bright light and loud noises in neonatal intensive care units, which can disrupt their sleep patterns and impact their circadian development (Sorensen et al., 2020).The ultimate effect is the dysregulation of hormone levels including melatonin and glucocorticoids; this dysregulation has an impact on growth, development, and other health outcomes such as the increased risk of autism and impaired learning, vision, and neurobehavioral abnormalities (Allotey et al., 2018;Sorensen et al., 2020).ASD and attention deficit hyperactivity disorder have also been associated with changes in the functioning of various neurotransmitters like gamma-aminobutyric acid (GABA), norepinephrine, dopamine, and serotonin (Marotta et al., 2020;Eissa et al., 2018).These studies point out a causative link among the HPA axis, circadian rhythm disruption, and ASD.
For several behavioral complications of ASD patients, sleep and circadian rhythm disturbances have been given as explanatory factors (Yavuz-Kodat et al., 2020).Melatonin synthesis, essential for neurodevelopment, has been found to be low in autistic children (Kulman et al., 2000).Mutations in the ASMT enzyme of the melatonin synthesis pathway are reported in ASD patients pointing to a genetic link (Jonsson et al., 2010).Melatonin synthesis begins with tryptophan, followed by intermediates of 5-hydroxytryptophan, serotonin, and N-acetyl serotonin, respectively (Barrenetxe et al., 2004).Differences are seen in serotonin levels in the blood of children with autism.The blood levels of serotonin in children with advanced autism are higher than those of children with less advanced autism (Abdulamir et al., 2018).The disrup-tions in melatonin secretion and the serotonin system involved in the melatonin pathway are remarkable in ASDs, pointing out its association with circadian rhythm.The rhythmicity in the secretion of neurohormones, including cortisol, prolactin, and thyroid-stimulating hormone, also has a role in the pathology of autism (Nir et al., 1995;Taylor & Corbett, 2014).A dysfunctional HPA axis and aberrant cortisol levels are observed in autistic children (Lorsung et al., 2021).Mean cortisol levels are found to be higher among ASD patients with IQ < 55 (extremely low functioning ASD individuals), compared to those with IQ ≥ 70 (Putnam, 2015;Kidd et al., 2012).Activity-dependent changes in the efficacy of synaptic communication become particularly important in circadian rhythm disorders, which can increase the susceptibility to ASD during early neurodevelopment (Yenen & Çak, 2020) (Ebert & Greenberg, 2013).
This study focuses on the reconstruction of protein-protein interaction (PPI) networks of ASD and circadian rhythm to understand the connection between autism and the circadian clock.Up to this time, most published research has concentrated on the behavioral characteristics of people with ASD.Experimentally, animal models have been employed to explore the complex history of ASD, as it was impossible to create human neural cell cultures that could proliferate indefinitely (Pensado-López et al., 2020).Hence, it is of utmost significance to capture the comprehensive overview of these two networks in the combined form to understand the crosstalk between ASD and circadian rhythmicity as well as other metabolic pathways.
A graph theoretical analysis is performed to determine whether the network properties are biologically feasible.The gene ontology (GO) enrichment analysis provides detailed information about the essential cellular processes.This study presents a comprehensive circadian-ASD network indicating key components, that facilitate a deeper understanding of shared mechanisms underlying circadian rhythm disorder-related pathophysiology of ASD.

Homo sapiens protein-protein introduction data
Homo sapiens PPI data are obtained from the BioGRID database (Stark et al., 2006).Within these data, only experimental system types, specifically physical interactions, are used, and genetic interactions are excluded.As the confidence score is connected to the reliability of PPI from physical data, interactions with confidence scores below 0.85 are removed.In cases where a different scoring system produces data with scores exceeding 1, the average of interaction scores within the same publication is calculated.Interaction data with scores higher than 1.5 times the calculated average are then selected.As a result, a set of 102,991 PPI data is collected as reliable for further analysis.A Python code is developed to facilitate filtering process.Additionally, a dataset of physical PPI (experimental) with confidence scores higher than 0.85, obtained from the String database, is appended (106,669 PPIs).The total number of unique PPIs is 209,658.

Construction of circadian PPI networks
The "circadian" term is used to search the Gene Ontology Consortium web page (Carbon et al., 2009)  Moreover, after careful literature research, an additional refined dataset consisting of 1435 human circadian gene products is identified (Li et al., 2016) and processed as explained above, resulting in 1417 PPIs and referred to as "CPPIN2."

Construction of autism PPI networks
Through a literature review, a total of 106 susceptibility genes associated with autism in H. sapiens are identified (Iakoucheva et al., 2019).
Additionally, a list of 3590 GO ID numbers linked to these susceptibility genes is compiled.The network created from 106 gene products is referred to as "Autism PPI Network 1 (APPIN1

Construction of circadian and autism PPI joint network
In the autism and circadian joint network, the proteins from APPIN5 and CPPIN2 are combined, resulting in the construction of a network consisting of 1844 proteins and 2394 interactions.The more compre-hensive APPIN5 is selected, encompassing circadian proteins within its clusters.CPPIN2 is chosen due to its more specific circadian network compared to CPPIN1.This combined network is named combined circadian-autism PPI (CAPPIN) and has undergone a detailed study.

Topological analysis and clusters and GO enrichment analysis
Graph theoretical analysis is used to conduct topological analysis of the reconstructed PPI networks.Properties such as node degree, hubs (highly connected nodes), and shortest path lengths among indirectly connected nodes, network diameter, and mean path length are determined using the Network Analyzer plugin (ver.2.7 and 4.4.6)(Doncheva et al., 2012) within Cytoscape (ver.3.6.1 and 3.9.0)(Shannon et al., 2003).The list of binary interacting proteins is given as the input, and the collected output includes different topological network parameters such as network clustering coefficient, network diameter, network radius, characteristic path length, average number of neighbors, network density, and node degree.MCODE plug-in (v1.5.1 and v2.0.0) is utilized to identify clusters (highly interconnected regions) within the network (Bader & Hogue, 2003).The node score cutoff is set to 0.2 for cluster expansion, the Kcore value is set to 2, and the maximum depth from seed is set to 100.
The GO Tool developed by Princeton University (Boyle et al., 2004) is used to analyze the GO terms through enrichment analysis.The significance level is set to 0.05, indicating that only terms enriched with a p-value of .05 or lower are taken into consideration.

Network decomposition analysis (NDA)
To comprehend the information flow (topology) within the circadianautism PPIN, linear paths connecting the center protein involved in circadian rhythm to the transcription factor in the nucleus related to ASD are determined using the NetSearch algorithm (Steffen et al., 2002).NetSearch draws all possible linear paths of a specified length through the interaction map, starting at any membrane protein and ending at any DNA-binding protein.Hence, the protein-protein interaction graph (comprising directed interactions) is used as the input for the NetSearch algorithm.Additionally, the other inputs (the circadian rhythm-related center protein) and the output (an autism-related transcription factor representing a cellular response) need also be defined for the linear path calculations.This linear path analysis is conducted to obtain insights into network crosstalk, and the molecules' involvement in the CAPPIN pathways.

Reconstruction and topological analysis of PPI networks of circadian and autism systems
The circadian (CPPIN), autism (APPIN), and CAPPIN networks are created by following the steps outlined in Section 2 (Figure 1).A graph theoretical analysis is conducted to calculate various topological properties, including the clustering coefficient, node connectivity, the shortest path lengths among indirectly connected nodes, network diameter, centrality measures, and characteristic path length.The clustering coefficient indicates the degree to which a network exhibits a clustering tendency (Masuda et al., 2018).The clustering coefficients for CPPIN2, APPIN5, and CAPPIN are 0.139, 0.225, and 0.151, respectively (as presented in Tables 1 and 2), and these values are consistent with the literature-reviewed values of normal human cells.The network clustering coefficients for bone cells are estimated as 0.217 and 0.261 for normal and cancer PPIN, respectively (Rahman et al., 2013;Sahoo et al., 2016).In another study from the literature (Dahiya et al., 2019), the clustering coefficient of the human brain-specific PPIN is 0.108.The diameter of the network represents the distance between the two farthest points (Junker & Schreiber, 2008).Compared to the same dataset of bone cells, where the bone PPIN diameter is 7 (Rahman et al., 2013;Sahoo et al., 2016), the diameter of circadian, autism, and combined PPIN diameters are slightly higher, that is, 18, 18, and 14, respectively.Closeness centrality indicates the distance of a protein in the network to all other proteins (Junker & Schreiber, 2008).
The change in the closeness centrality with the number of neighboring proteins demonstrates a linear relationship.When comparing these values with the literature, the characteristic path length and network diameter closely resemble those reported for various human-related networks (Table 3).Based on these network properties, the CPPINs, APPINs, and CAPPIN are considered small-world networks, similar to many other complex biological networks (Latora & Marchiori, 2001;Watts & Strogatz, 1998).
The properties of the nine main circadian genes (PER1, PER2, PER3, CLOCK, ARNTL [BMAL1], CRY1, CRY2, CSNK1E, and NPAS2) in CPPIN2 are tabulated in Table 4. Closeness centrality demonstrates the distance of a protein to other proteins (Junker & Schreiber, 2008) which indicates the effective spread of information through the network.Among these main nine circadian proteins, CSNK1E stands out

Network's hub protein analysis
The node connectivity identifies highly connected proteins that participate in significant numbers of interactions and hold critical roles in the organization of the cellular protein interaction network.The top 10 highly connected proteins, referred to as hub proteins, are listed in Table 5A and B.
In circadian network CPPN1, a crucial hub protein is the ubiquitin C or polyubiquitin-C protein, known for its stress-protective function (Aken et al., 2016;Bianchi et al., 2018).Circadian rhythm translation-transcription feedback loops are mediated by ubiquitination, and ubiquitination intermediates play a role in maintaining the stability of the primary circadian proteins within the circadian network (Stojkovic et al., 2014) Takata et al., 2018).
CTNNB1 protein and CREB-binding protein (CREBBP) exist in all autism networks, except for the APPIN2.DYNC1H1 and STX1A proteins are found in three of the five networks and ANK2, SIN3A, DLG2, The interactome network CPPIN2 is subdivided into clusters of proteins with extensive and robust interactions.Cluster 1 from CPPIN2, which has been extracted from Cytoscape, containing essential circadian genes.
NRGXN1, HRAS, SNAP29, CTBP1, and TCF4 are found in two of the five autism networks.Changes in the synapse function and gene regulation may contribute to the development of ASD.ANK2 is a high-risk ASD gene product, and ANK2 mutations are often associated with ASD (average IQ) (Yang et al., 2019).The ANK2 gene is involved in cell motility, activation, and proliferation processes (Crawley et al., 2016;Iwakawa et al., 2015).Syntaxin-1A (STX1A), a specific protein of the nervous system, plays a role in the docking of synaptic vesicles and the presynaptic membrane; it affects serotonin (5-HT) release in the brain and GABAergic neurotransmission, possibly contributing to attention-deficit/hyperactivity disorder (Wang et al., 2019).The common ASD-related proteins SIN3A, CTNNB1, and CTBP1 interact with other shared proteins: MECP2 (associated with SIN3A), CHD8-PTK7-TCF7L2 (associated with CTNNB1), and EHMT1-FOXP1 (associated with CTBP1) (Crawley et al., 2016).Different publications have indicated that mutations in the CTNNB1 gene are associated with both cancer and autism (Dong et al., 2016).The point that draws attention from these highly connected proteins is that ribosomal proteins form most of the hub proteins in the APPIN networks.The CREBBP protein takes part in the transcriptional coactivation of numerous different transcription factors.Rubinstein-Taybi syndrome (RTS), classified within the ASD family, arises from a mutation in the gene coding for CREBBP (Stef et al., 2007)).Similar to ASD, the primary symptoms of Rubinstein-Taybi disorder encompass the gradual development of cognitive and motor skills (Galéra et al., 2009).The hub protein CREBBP (Table 5A) also featured in several clusters (associated with the transcription by RNA polymerase II processes) of GO enrichment analysis of 49 common proteins of CPPIN2 and APPIN5.The relationship between autism and ribosomal genes/proteins requires further investigation.Another protein listed in Table 5A, DYNC1H1, is linked to the regulation of cellular processes.Mutations in this DYNC1H1 gene may potentially contribute to neurological disorders (Hoang et al., 2017), resulting in neuromuscular and sensory deficits (Schlager et al., 2014).TCF7L2 is linked to the Wnt-signaling pathway and is associated with schizophrenia (Alkelai et al., 2012).With a prominent expression in the brain, TCF4 actively participates in neurodevelopment.It collaborates with class II bHLH transcription factors Math1, HASH1, and neuroD2 (Navarrete et al., 2013).The TCF4 gene is also associated with memory function and plays a regulatory role in neurodevelopmental pathways (Forrest et al., 2018).Haploinsufficiency in TBL1XR1 leads to ID with dysmorphism, though it does not consistently result in autistic behavior (Pons et al., 2015).TBL1XR1 contributes to the regulation of the Wnt/β-catenin signaling pathway, influencing various stages of brain development and processes linked to cell fate determination and stem cell renewal (Noelanders & Vleminckx, 2017).APC protein interacts with beta-catenin and plays a critical role in multiple cellular processes.The deletion of the APC gene was observed in individuals with autism (Barber et al., 1994).NRXN1, associated with neurodevelopmental disorders, also plays a role in ASD due to CNVs in NRXN1 gene (Kim et al., 2008).

F I G U R E 3
The interactome network combined circadian-autism protein-protein interaction network (CAPPIN) is subdivided into clusters of proteins with extensive and robust interactions.Clusters: (a) 1, (b) 10, (c) 12, (d) 30 extracted from Cytoscape.

Module detection and analysis of PPI networks
In biological systems, scale-free networks are composed of clustered regions called modules (Bader & Hogue, 2003).CPPIN2 is divided into modules using the "Cytoscape" Network Analyzer MCODE plugin.Twenty clusters, characterized by highly associated domains, are obtained and visualized in Cytoscape (Table 6A).The cluster with the significantly higher score, Cluster 1, comprises 24 proteins and 218 interactions (edges) between these proteins (Figure 2).This cluster has nine primary circadian clock gene products and ribosomal proteins related to the circadian system.
To conduct a comprehensive analysis of APPIN5, it is also segmented into subclusters using the "Cytoscape" Network Analyzer MCODE plug-in.A total of 16 clusters, characterized by highly associated domains, are extracted and visually represented using Cytoscape.
The significant cluster, Cluster 1, comprises four proteins and a higher score in MCODE results.It includes the primary circadian clock gene products, ARNTL, ARNTL2, PER1, and NPAS2 (Table 7A).The identification of a cluster associated with circadian clock proteins within the pure autism network holds considerable significance.However, due to the small size of Cluster 1, it is not directly feasible to deduce a clear or significant connection between circadian clock proteins and ASD.Nevertheless, several studies (Ballester-Navarro et al., 2021;Vallée et al., 2020;Yang et al., 2016) have provided insights indicating the interrelation between the circadian clock and autism.The thalamo-cortico-amygdala pathway, essential for complex emotional memory (Garcia et al., 2000), has been reported as dysfunctional in autism (Nicholas et al., 2007), these studies on neuronal PAS domain protein 2 (NPAS2) reveal that NPAS2-deficient mice subjected to behavioral tests exhibit deficits in the long-term memory.NPAS2 plays a regulatory role in acquiring specific types of memory, linking NPAS2 to autistic disorder.Moreover, a study (Nicholas et al., 2007) on the hypothesis that clock genes are implicated in autistic disorder revealed noteworthy associations in the PER1 and NPAS2 genes.7A).Neurexin/neuroligin interactions in the neuronal synaptic cleft are crucial for the proper functioning of the synaptic network and neurotransmission (Südhof, 2008).A specific mRNA or translation process stimulates the local translation of significant ASDrelated proteins in the synaptic domain.Translation of the mRNAs for NLGN1, NLGN2, and NLGN3 is negatively regulated by Fragile X mental retardation protein (FMRP), loss of this translational repressor, FMRP, leads to Fragile X syndrome, associated with ASD (Joo & Benavides, 2021).Loss-of-function mutations in the NRXN1, NLGN3, and NLGN4 genes in ASD (Joo & Benavides, 2021;Tabuch et al., 2007) and disruption of SH3 and SHANK (multiple ankyrin repeat domains) lead to abnormal behaviors in ASD (Joo & Benavides, 2021;Monteiro & Feng, 2017).Some mutations in genes known to modulate synaptic plasticity are associated with rare cases of autism (Klauck, 2006).
Therefore, the contribution of these synaptic plasticity-related genetic factors to the development of abnormal central nervous system (CNS) in ASD requires further research.
To gain a deeper understanding of the combined network of circadian and autism-related proteins, CAPPIN is divided into clusters using the "Cytoscape" Network Analyzer MCODE plug-in.A total of 30 unique clusters are identified and visualized using Cytoscape (Table 8A).The highest scoring Cluster 1 (comprising 14 proteins and 145 edges) is included in the highest scoring cluster in CPPIN2.
Within this combined network, seven additional clusters containing six or more proteins, are identified (Figure 3).Analyzing these clusters revealed a significant correlation between autism and circadian proteins involved in DNA and RNA transcription processes.The identification of Cluster 10, which is directly connected to the Wnt signaling pathway (Figure 3b), points out the importance of this pathway.The Wnt signaling pathway plays a crucial role in tumorigenesis (Zhan et al., 2017).Furthermore, numerous studies (Bae & Hong, 2018;Kwan et al., 2016;Zhang et al., 2014) have demonstrated the association between the Wnt signaling pathway and autism.The intensified transcription within the Wnt pathway is strongly correlated with the development of ASD (Mbadiwe & Millis, 2013).The binding of β-catenin to the LEC/TCF influences the transcription of genes within the Wnt pathway (Mbadiwe & Millis, 2013).Upon examining Cluster 10, the CTNBB1 protein, which functions as the hub protein of the CAPPIN, plays crucial roles in the canonical Wnt signaling pathway and is intricately linked to autism (Kwan et al., 2016

Go enrichment analysis
The significant GO terms of the proteins within all clusters obtained through MCODE are identified using the GO Tool provided by Princeton University (Boyle et al., 2004).These enriched GO terms of the  circadian proteins in CPPIN2-Cluster 1 are presented along with their corresponding p-values (Table 6A ).One of the molecular functions identified is blue light photoreceptor, which refers to a series of molecular signals triggered upon sensing of blue light by a photoreceptor molecule within a wavelength between a range of 400 and 470 nm (Tosini et al., 2016;Weber et al., 2019).The GO category of "blue light photoreceptor activity" exhibits an extremely low frequency within the genome, with only 2 out of 19,751 genes associated with it.
Thus, the function of the blue light photoreceptor holds significance in comprehending the impact of light on the circadian clock, particularly given the correlation of the CRY1 and CRY2 genes with specific molecular functions.Additionally, other GO terms with higher cluster frequencies include transcription regulatory region sequence-specific metabolic pathways.This term describes proteins that selectively and non-covalently interact with specific sequences of DNA and RNA in regulatory regions, controlling transcription of sections of DNA and RNA molecules.The most significant finding is that the GO terms with high corrected p-values are associated with circadian rhythm, glucocorticoid receptor modulation, ribosome biogenesis, and circadian control of gene expression (Table 6B).
The significant GO terms of the proteins in Clusters 2 and 16 of APPIN5 are tabulated with p-values in Table 7B and C The enriched GO terms for CAPPIN Cluster 1 primarily pertain to cytoplasmic translation or translation, glucocorticoid receptor signaling pathway and its regulation, ribonucleoprotein complex biogenesis, cellular response to endogenous stimulus, histone acetylation, and peptidyl-lysine acetylation processes (Table 8B).MAPK3 (present in  9).
The significant role of histone acetylation in the context of ASD has recently been elucidated and warrants further in-depth investigation.
Alterations in histone acetylation have the potential to induce modifications in gene transcription, playing a crucial role in the development of ASD.Notably, the mRNA levels of ASD-related risk genes like NLGN1, SHANK2, SHANK3, and CNTNAP2 are influenced by the prenatal suppression of histone deacetylase family, which are linked to neurodevelopmental disorders (Tseng et al., 2022) et al., 2022).

Network decomposition analysis (NDA)
Network  proteins with path lengths of 6, 7, and 8, the AXIN1 protein emerges as the most interactive protein with CSNK1E.The characteristic path length of CAPPIN was determined to be 5.896.Therefore, the pathways of 6 steps are visualized in Figure 4, whereas the most probable outcomes of pathways with steps of 5-6-7-8 (path lengths) are depicted in Figure 5.
Bottleneck proteins, recognized as central connectors that are crucial to numerous shortest paths in an interaction network, control most information flow (signal transduction) from ARNTL to the TCF4 protein.ARNTL and CSNK1E function as the hub/bottleneck proteins in the network.Thus, mutations in these proteins may potentially disrupt the circadian system leading to disorders, including ASD.A recent study on AXIN1 indicates a correlation with autistic traits (Smedler et al., 2021).The role of CTNNB1 gene in the regulation of cognitive and autistic-like behaviors has been reported in several publications (Dong et al., 2016).TLE3 is a target of FOXP2, a transcription factor located on chromosome 7q31 related to autistic disorder (Benítez-Burraco et al., 2018).
TA B L E 8 B Important gene ontology (GO) terms of the proteins in Cluster 1 of combined circadian-autism protein-protein interaction network (CAPPIN) obtained using the Gene Ontology Tool developed by Princeton University (Boyle, et al.,2004).(hub/bottleneck proteins), and TCF4.These pathways provide valuable insights into the complex protein-protein interactions that underlie the connection between circadian rhythms and autism spectrum disorder (ASD).

Circadian clock crosstalks with many pathways
The human body possesses a circadian rhythm that regulates numerous physiological activities across various organizational levels, maintaining a 24-h transcription-translation feedback loop.There are nine core gene products associated with circadian rhythms: Period 1 (PER1), Period 2 (PER2), Period 3 (PER3), Casein Kinase Iε, CLOCK, ARNTL (BMAL1), TIM, CRY1, and CRY2 (Lévi, 2002).These gene products also regulate cell proliferation and could play a critical role in the cell cycle.
Furthermore, studies on micro-peripherical tissues suggest that the biological clock is activated in response to DNA damage.Particularly, PER2 is implicated in the DNA-damage response process in rodents (Fu & Lee, 2003).
In the regulation of circadian rhythms, the BMAL1 (ARNTL) gene that interacts with the CLOCK or NPAS2 genes gives rise to the formation of BMAL1:NPAS2 or BMAL1:CLOCK complexes within the cytoplasm.These resultant complexes play a role in activating various other circadian genes within the nucleus, working in conjunction with the E-box: PER1/2, CRY1/2, and DEC1/2.PER1/2-CRY1/2 and DEC1/2 complexes function to inhibit the activity of heterodimeric transcription factors BMAL1:NPAS2 or BMAL1:CLOCK, and this inhibition varies based on the tissue.Consequently, these heterodimers participate in initiating their transcription in a negative loop (Li, 2019).
An important point to note is the influence of the P53 gene by PER1 and PER2.Specifically, PER2 directly interacts with P53, preventing its ubiquitination and thereby leading to P53 stabilization and modulation of its nuclear import (Gotoh et al., 2015(Gotoh et al., , 2014(Gotoh et al., , 2016)).Notably, the absence of P53 can result in altered PER2 expression within the SCN (Miki et al., 2013).Both P53 and its downstream effector P21 (activated by P53) play pivotal roles in DNA damage repair pathways and the inhibition of tumor formation (Li, 2019).Examining the main circadian pathway (as depicted in Figure 6), it is evident that impaired circadian genes or gene products fail to activate P53, which, in reality, functions to suppress tumor formation.
BMAL1:CLOCK heterodimer whose activation is modulated by CREBBP functions as a transcriptional activator within the circadian rhythm and plays a pivotal role in regulating the expression of F I G U R E 6 Mammalian protein signaling pathway with positive and negative feedback loops.This figure illustrates circadian genes and their feedback loops in the regulation of mammalian circadian rhythms, with positive and negative feedback mechanisms.It also highlights the pivotal role of the P53 tumor suppressor gene.circadian genes (Lee et al., 2010).However, the MYC protein, recognized as an oncogene, disrupts the formation of the BMAL1-CLOCK dimers by binding to the E-box.Furthermore, MYC takes part in activating another negative regulator REV-ERB, leading to suppressing BMAL1 expression and consequently impacting the circadian clock (Altman et al., 2015).A well-functioning circadian rhythm is indispensable for maintaining good health.

Dysfunction of metabolic pathways due to mutations in the clock-controlled genes
Autism is influenced by the dysfunction of significant pathways within the CNS.NPAS2, also known as a member of PAS protein 4, is a transcription factor protein that plays a crucial role in sleep homeostasis and maintaining circadian behaviors under normal light/dark conditions.NPAS2 protein emerges prominently in clusters of excessively associated proteins (MCODE clusters), as well as in the result of GO enrichment analysis.Impaired memory and sleep disturbances (symptomatic of ASD) have been investigated using mouse models.Npas2 (−/−) knockout mice have shown disruptions in complex emotional memory (Garcia et al., 2000), indicating the critical role of non-REM sleep homeostasis and a reduction in total sleep time (Franken et al., 2006).Furthermore, a specific protein variant, NPAS2 471 Leu/Ser, has been associated with seasonal affective disorder and diurnal preference in humans (Johansson et al., 2003).In the case of impaired transcription/translation feedback loop, the binding of CLOCK:BMAL1 or NPAS2:BMAL1 is inhibited, resulting in the downregulation of both the period and cryptochrome genes' transcription (Lorsung et al., 2021;Ye et al., 2014).
Mutations in the clock-controlled genes are frequently observed in individuals with ASD.Although individuals with ASD have been studied for genes strongly associated with circadian rhythm, statistical analyses have revealed variations among these ASD individuals (Yang et al., 2016).The polymorphisms in Npas2 such as cytosine/thymine SNP in intron 3 (NPAS2_X3_C_T) (Nicholas et al., 2007), mutations in PER1 including cytosine → guanine SNP (Per1_rs885747) and a cytosine/adenine SNP (Per1_rs6416892) (Nicholas et al., 2007), as well as mutations in PER2 (proline/alanine substitution at amino acid 1228), and mutations in PER3 (an arginine/glutamine substitution at amino acid 366) have been identified in individuals with ASD.These variations have a detrimental impact on gene function.These clinical findings underscore the association of circadian rhythm genes/gene products PER2 and PER3 in the pathogenesis of ASD.Additionally, the mutations in NR1D1 have also been linked to abnormal ASD brain development (Lorsung et al., 2021).RORα/β is known to activate the transcription of BMAL1 and NFIL3 (Preitner et al., 2002;Ueda et al., 2005).However, RORα cannot function properly in the case of impaired NR1D1.
Another pathway linked to ASD is the canonical WNT/β-catenin pathway which becomes upregulated under the dysregulation of circadian rhythm (sleep disorder) (Vallée et al., 2020).This pathway initiates the metabolic reprogramming of cellular energy metabolism in individuals with developmental cognitive disorders (Vallée et al., 2020).

4.2.1
Oxidative stress and its link to ASD and circadian clock Oxidative stress is a mechanism through which prenatal and postnatal problems contribute to the development of ASD.In more than 100 publications investigating any relationship between oxidative stress and ASD/Rett syndrome, an association has been consistently reported, and these studies have explored abnormalities in microglial activation, maternal antibodies to fetal brain tissue, genetic mutations affecting the immune system, and cytokine abnormalities, all linking oxidative stress with ASD.The effects of oxidative stress on ASD provide insights into the influence of parental age on ASD.Older men's spermatozoa are more vulnerable to oxidative stress, potentially leading to DNA fragmentation, a factor associated with neuropsychiatric disorders (Mandic-Maravic et al., 2019).
Due to a weakened antioxidant defense mechanism, children with ASD seem to be more susceptible to oxidative stress, leading to elevated lipid peroxidation.This observation suggests that early antioxidant treatment could potentially enhance prognosis by mitigating oxidative stress before it inflicts further irreversible brain damage (Bjørklund et al., 2020).
From blood pressure and sleep/wake cycles to cellular signaling pathways that play crucial roles in health and diseases, oxygen and circadian rhythmicity are essential for maintaining homeostasis in various physiological processes.The ability of the human body or cells to regulate internal systems, such as redox levels and circadian rhythms, could be compromised when subjected to significant stress.Impairment in redox regulation and circadian rhythms can lead to various adverse consequences at both the cellular and organismal levels, including the development of heart disease, neurodegenerative conditions, and cancer (Fanjul-Moles &López-Riquelme, 2016; Lananna & Musiek, 2020;Wilking et al., 2013).The impacts of oxidative stress and dysregulated circadian rhythms have been extensively studied, and the molecular mechanisms that connect these two processes are well established and observed in the reconstructed networks of the present research.

The connection between the circadian clock and autism within mTor signaling pathway
The mTOR signaling pathway plays an important role in controlling significant functions within neurons, encompassing neurodevelopment and intracellular metabolism (Liu & Sabatini, 2020;Wullschleger et al., 2006).Recent studies (Cao, 2018;Cao et al., 2011) have demonstrated the influence of the circadian clock on mTOR activities.These activities exhibit robust diurnal oscillations across various systems.Furthermore, mTOR signaling impacts the duration, synchronization, and entrainment of central and peripheral circadian clocks (Singla, Mishra, Cao, 2022;Singla, Mishra, Lin, et al. 2022).The study, which focused on the association between autism and circadian genes in mice, demonstrated that mice lacking the Bmal1 gene exhibited differences in the expression of genes associated with autism and ataxia and also displayed dysregulated pathways, including overactive mTORC1 signaling (Liu et al., 2022).Moreover, mTOR creates a complex signaling network that combines external and internal signals to influence mRNA translation (Liu & Sabatini, 2020;Wullschleger et al., 2006).
The study on autism (Neves-Pereira et al., 2009) demonstrates that the eukaryotic initiation factor 4E (EIF4E) gene, which plays a role in translation control, is implicated in the disorder.Specifically, specific mutations in the EIF4E promoter region have been linked to autism.In the context of APPIN5, EIF4E serves as a hub protein within the autism network, while also being closely interconnected with the mTOR signaling pathway and the circadian clock.One of the downstream processes of this pathway is the activation of the EIF4E, a protein responsible for binding to the mRNA cap structure (Romagnoli et al., 2021).The results of this study point to a need on a thorough evaluation of the mTOR pathway and a review of ribosomal proteins.

Ribosome biogenesis and its relation to ASD and circadian clock
The GO enrichment analyses highlight the significance of ribosome biogenesis in relation to ASD that could potentially contribute to its causative factors and related neurodevelopmental disorders.In autism, the expression of ribosomal protein-encoding genes is notably elevated (upregulated), suggesting an intensified ribosome biogenesis process and consequently heightened protein synthesis (Lombardo, 2021).The increased copy number of ribosomal genes within the dendrites alters the level of dendritic translation, thereby representing a risk factor for neuropsychiatric disorders (Porokhovnik & Lyapunova, 2019).As a result of disturbed "translational homeostasis," individuals with abundant ribosomes (due to moderate or high copy numbers of ribosomal genes) may develop ASD.In a study by Tran et al. (2020), new variants have been identified in six candidate genes (CHM, ENPP1, IGF1, LAS1L, SYP, and TBX22) enriched in an ASD cohort.These genes are associated with various neurological conditions and play roles in neuronal biological processes, such as synaptic processes, regulation of transport, and ribosome maturation.Therefore, the genetic predispositions within these genes are reported to be causative factors for ASD (Tran et al., 2020).Another ribosomal gene with missense mutations is RPL10 gene, which exhibits high expression in mouse hippocampusa critical site of action for learning, memory, and social and affective functions.Consequently, functionally altered RPL10 protein may contribute to cognitive malfunction, which is also observed in autism (Klauck, 2006).On the other hand, another study suggests that RPL10 might not have a significant impact on ASD susceptibility (Gong et al., 2009).
The circadian clock regulates the transcription of translation initiation factors and orchestrates the clock-dependent rhythmic activation of signaling pathways involved in their regulation.This intricate process governs the temporal translation of a subset of mRNAs essential for ribosome biogenesis.Consequently, the circadian clock plays a pivotal role in synchronizing the transcription and translation processes crucial for ribosome biogenesis (Benegiamo et al., 2016;Jouffe et al., 2013).Recent research has examined the downregulated genes for RNA-binding proteins in the transcriptome of induced pluripotent stem cell neurons (iNeurons) to model the RTS (Larizza et al., 2022).Intriguingly, many individuals with RTS display traits similar to those observed in individuals with autism.This study has identified several downregulated gene products in RTS, including ribosomal subunits and nucleolar proteins such as NOP58 and fibrillarin.These proteins form complexes with snoRNAs and play a key role in driving posttranscriptional changes required for rRNA maturation.Fibrillarin is essential for the epigenetic regulation of ribosomal genes, and NOP58-associated snoRNA levels are modulated by NOP58 interactor BMAL1, a transcriptional regulator of the circadian rhythm.Despite these findings, the precise involvement of ribonucleoprotein particles in the production of rhythmic gene expression remains a mystery.Informatics aimed at integrating these findings with circadian transcriptome datasets are required to elucidate the posttranscriptional processes underlying circadian control, whereas the target RNAs for many of the RNA-binding proteins are being discovered.

CONCLUSION
Autism is a complex disorder involving over 1000 genes.Alongside genetic factors, various environmental influences contributing to ASD have been identified, including maternal nutritional factors and obesity, maternal stress, maternal immunological response/dysfunction, earlylife immune insults, parental age, heavy metals, and air pollution.In the present study, a network-based approach is conducted to identify significant pathways and genes/gene products within both individuals (pure autistic and pure circadian) and combined networks of circadian rhythm and ASD.Numerous key hub gene products are highlighted, showing significant enrichment in areas such as ribosomal proteins, protein synthesis, regulatory pathways, and the Wnt/β-catenin signaling pathway.These findings merit further investigation through experimental validation studies.Several fundamental processes that pertain to synaptic function are also revealed to be regulated by specific genes associated with ASD.These processes encompass ribosome maturation and mRNA regulation.Thus, changes in synaptic function and gene regulation may serve as causative factors for ASD.
Given that the translation of significant proteins associated with ASD occurs within the synaptic region, maintaining translational homeostasis could potentially prevent or mitigate this complex and multifactorial disorder.
The enriched GO terms for combined network CAPPIN are primarily related to cytoplasmic translation or translation, glucocorticoid receptor signaling pathway and its regulation, ribonucleoprotein complex biogenesis, cellular response to endogenous stimulus, and processes such as histone acetylation and peptidyl-lysine acetylation.
Histone deacetylases have recently been implicated in ASD and other neurodevelopmental disorders characterized by similar clinical and diagnostic comorbidities (e.g., epilepsy, anxiety, and intellectual impairment).The application of epigenetic drugs, such as histone deacetylase inhibitors (Larizza et al., 2022)

F
I G U R E 1 Protein-protein interaction networks and Venn diagram: (a) APPIN5: This significant network focuses on interactions and highlights hub proteins and their connections related to autism spectrum disorder (ASD); (b) CPPIN2: This network emphasizes circadian clock-related interactions; (c) combined circadian-autism protein-protein interaction network (CAPPIN): The CAPPIN visualizes comprehensive protein-protein interactions, combining APPIN5 and CPPIN2, showcasing the significant interactions between autism and the circadian clock; (d) Venn diagram of interactions: This diagram provides an overview of the overlaps and unique interaction numbers between these protein interaction networks.
. The enriched GO terms for Cluster 2 are mainly related to circadian rhythm and its regulation, as well as the redox state.Cluster 16 has enriched GO terms associated with the regulation of transcription by RNA polymerase II, regulation of RNA biosynthetic process, RNA metabolic process, regulation of nitrogen compound metabolic process, canonical Wnt signaling pathway involved in regulation, and more.
APPIN and CAPPIN) andRPS6KA4 (present in CPPIN2 and CAPPIN)    are involved in the positive regulation of peptidyl-lysine acetylation and have been found to be influenced by antidepressants and related molecular pathway analogs(Drago et al., 2011).The enriched GO terms for CAPPIN Cluster 10 are primarily related to Wnt signaling pathway, regulation of macromolecule biosynthetic process, regulation of RNA biosynthetic process, and midbrain dopaminergic neuron differentiation.Indeed, there are 49 shared proteins present in CPPIN2 and APPIN5.The enriched GO terms associated with these proteins primarily revolve around regulation processes, including the regulation of transcription by RNA polymerase II, regulation of RNA metabolic process, regulation of transcription, DNA-templated, regulation of transport, regulation of localization, regulation of ion transport, histone modification and acetylation, peptidyl-lysine modification and acetylation, as well as multicellular organism development (Table decomposition analysis (NDA) involves breaking down a network into fundamental pathways that capture the time-invariant topological structure.NDA serves the purpose of defining the overall capabilities of a cell.To conduct a detailed analysis of information flow, the CPPIN is dissected into linear pathways connecting transcription factors and various proteins within the PPI network.Linear pathways originating from the central circadian gene product (ARNTL [BMAL1]) to the transcriptional factor gene (TCF4) are examined using the NetSearch Algorithm.TCF4 functions as a transcription factor that regulates a wide array of developmental processes.Deficiencies in TCF4 have been linked to various human disorders, including Pitt-Hopkins syndrome, an ID condition, and ASD.The total count of pathways from ARNTL to TCF4 is estimated to be 1350.The participation of proteins in these linear paths indicates their significance in signal transduction, as any signaling network is composed of linear paths.Therefore, the participation percentages of each protein in linear paths of the CAPPIN are calculated to gain insights into the roles of these proteins in transmitting information flow from ARNTL to TCF4.The proteins most frequently encountered in these linear paths are identified by MATLAB.Among these , RPL14, ARNTL, RPS25, CRY2, RPL10, RPL21, RPL3, CRY1, RPS5, , EIF5B, RPL14, ARNTL, RPS25, CRY2, RPL10, RPL21, CSNK1D, RPL3, CRY1, RPS5, RPS23 F I G U R E 4 Six-protein length pathways of combined circadian-autism protein-protein interaction network (CAPPIN) from ARNTL to TCF4 (CPPIN: ARNTL and CSNK1E (hub/bottleneck proteins), AXIN1, CTNNB1, TLE3, TLE4; APPIN: CTNNB1, TCF4).This figure illustrates the 6-steps pathways within the CAPPIN that connect ARNTL to TCF4.The central role of ARNTL and CSNK1E as hub/bottleneck proteins is highlighted, along with the significant involvement of AXIN1, CTNNB1, TLE3, and TLE4 in these pathways., SIN3A, SMARCC2, GFAP, CACNA1C, LDB1, PPP2R5D, NDEL1, CHD8, TBL1XR1, STXBP1, FOXP2, GABRB3, FOXP1, GABRB2, CTNNB1, MKX, ARNTL, NRCAM, ANK2, SCN2A, NPAS2, ARID1B, NCOA1, , SIN3A, SMARCC2, GFAP, CACNA1C, LDB1, PPP2R5D, NDEL1, CHD8, TBL1XR1, STXBP1, FOXP2, GABRB3, FOXP1, GABRB2, CTNNB1, MKX, ARNTL, NRCAM, ANK2, SCN2A, NPAS2, ARID1B, NCOA1, TCF7L2, CREBBPF I G U R E 5The most likely pathways of combined circadian-autism protein-protein interaction network (CAPPIN)(5-6-7-8 steps)  from ARNTL to TCF4.This figure illustrates the most likely pathways within the CAPPIN, depicting a series of 5-6-7-8 steps that connect ARNTL to TCF4.CAPPIN represents a combined network involving critical proteins such as ARNTL, CSNK1E in the treatment of neurodevelopmental disorders arising from RNA-binding proteins could potentially restore the activity of lysine (K) acetyltransferases (KAT) in various neurodegenerative conditions, including autism.The present study introduces a novel perspective on metabolic mechanisms and unveils potential key proteins/pathways such as ribosome biogenesis, redox, and oxidative stress.Consequently, it highlights that it could mitigate the development of ASD triggered by environmental factors like maternal stress, or advanced parental age (which is linked to spermatozoids of older men and oxidative stress).Ultimately, comprehending the connection between circadian rhythm and ASD provides valuable insights into the role of these pathways in the etiology and pathogenesis of ASD.Initiating antioxidant treatment early on could potentially enhance the trajectory of the medical condition by mitigating oxidative stress before it leads to irreversible brain damage.Furthermore, emerging treatments are under investi-gation, targeting epigenetic marks like DNA methylation and histone acetylation, with the hope of addressing or preventing autism.
to create a dataset comprising GO-terms related to the circadian function.All three types of GO terms (biological process, molecular function, and cellular component) are thoroughly examined to identify the terms that demonstrate sig-

TA B L E 1
The result of analyzing autism protein-protein interaction (PPI) networks.
Main nine circadian genes properties in the protein-protein interaction (PPI) circadian network.The highest degree proteins in the autism protein-protein interaction (PPI) networks.
(De Rubeis et al., 2014;9)gree proteins in circadian and combined circadian-autism protein-protein interaction (PPI) networks.tocauseX-linkedintellectualdisability (ID)(De Oliveira et al., 2019), suggesting a potential link to ASD.The central component of the circadian clock, CSNK1E (Casein Kinase 1, Epsilon), determines the length of thcircadian period.It holds a high degree (34 and 37) within the circadian CPPIN and the combined network CAPPIN, respectively, and is the most interacting protein among the nine primary circadian pro-teins.Notably, CSNK1E is determined as a gene significantly enriched in damaging de novo mutations in ASD cases(De Rubeis et al., 2014; MCODE clustering results for CPPIN2. Two single nucleotide polymorphisms (SNPs) within each gene reached TA B L E 6 A Important gene ontology (GO) term functions of proteins in Cluster 1 in APPIN5 including circadian main genes.
. Furthermore, prenatal exposure to environmental factors has the potential to induce modifications in epigenetic markers within utero through various mechanisms, including chemical modifications like DNA methylation or protein modifications such as histone acetylation/deacetylation.These changes have been implicated in contributing to TA B L E 7 B TA B L E 8 A MCODE clustering results for combined circadian-autism protein-protein interaction network (CAPPIN).