Time vortex: the circadian–dopaminergic dialogue in Parkinson’s disease Authors: Meiru Zhou, Yuxiang Xu, Ye Liu, Jing Wang, Keyuan Hou, Zhaowu An, Jianshe Wei Published: 2026 ABSTRACT AbstractCircadian disruption is a prominent non-motor feature of Parkinson’s disease (PD) and is bidirectionally linked to PD pathology. It may interact with oxidative stress, inflammation, mitochondrial dysfunction, and proteostasis imbalance, though its causal role remains unclear. Circadian parameters may aid early diagnosis, monitoring, and time-oriented interventions. This review discusses how circadian abnormalities relate to prodromal and clinical PD, related mechanisms, and circadian-based therapies to improve quality of life.Subject terms: Parkinson's disease, Circadian rhythms and sleep Introduction PD is a progressive neurodegenerative disorder characterized primarily by motor dysfunction1. Its pathological hallmarks include progressive loss of dopaminergic neurons in the substantia nigra pars compacta, deposition of α-synuclein (α-syn), and formation of intracytoplasmic Lewy bodies2. Since James Parkinson first systematically described the disease in 1817, PD has become the second most common neurodegenerative disorder globally, following Alzheimer’s disease3. With the intensification of global population aging, the incidence of PD has shown a significant upward trend4. The clinical manifestations of PD exhibit marked heterogeneity, with classical motor symptoms including resting tremor, bradykinesia, muscle rigidity, and postural instability5. Recent research has increasingly focused on the potential diagnostic value of non-motor symptoms (e.g., hyposmia, rapid eye movement sleep behavior disorder, circadian rhythm disruption, autonomic dysfunction, and mood disorders) in early-stage detection5,6. Current consensus suggests that PD pathogenesis involves complex interactions among genetic susceptibility7, environmental factors (e.g., pesticide exposure)8, mitochondrial dysfunction9, α-synuclein aggregation-mediated proteostasis imbalance10, and neuroinflammatory responses11. Although levodopa and other medications can temporarily alleviate symptoms, they fail to halt disease progression and are associated with severe side effects such as wearing-off phenomenon and dyskinesia during long-term treatment12. This therapeutic impasse compels researchers to seek breakthroughs from novel biological perspectives. The circadian rhythm is an endogenous timekeeping system evolved by organisms to adapt to 24-h cyclic environmental changes caused by Earth’s rotation13. From unicellular cyanobacteria to higher mammals, this intrinsic biological oscillator coordinates metabolic, immune, and reproductive processes by anticipating periodic fluctuations in external cues like light and temperature, thereby conferring survival and competitive advantages14. At the molecular level, circadian regulation is primarily driven by transcription-translation feedback loops (TTFLs) of core clock genes: CLOCK/BMAL1 heterodimers activate the expression of period genes (Per, Cry), whose protein products undergo phosphorylation and translocate into the nucleus to inhibit CLOCK/BMAL1 activity, generating autonomous about 24-h oscillations15. This core mechanism exists not only in the suprachiasmatic nucleus (SCN) of the hypothalamus—the “central pacemaker”16—but also couples with peripheral tissue clocks (e.g., liver, adipose) through hormonal signals (e.g., corticosterone) and metabolites (e.g., NAD⁺), forming a hierarchical temporal regulatory network17. For clarity, the circadian clock system is hierarchically organized into the master clock and peripheral clocks18. In this review, circadian rhythm disruption refers to persistent dysfunction of the circadian timing system, manifested as an abnormal circadian phase (i.e., a misalignment relative to the external light–dark cycle), reduced amplitude (or weakened output strength) of circadian signals, increased fragmentation and/or reduced day-to-day stability of rhythms, or internal desynchrony between the suprachiasmatic nucleus (SCN)-driven central rhythm and peripheral oscillators19,20. From an operational perspective, real-world rest–activity rhythm (RAR) metrics derived from wearable actigraphy—relative amplitude (RA), interdaily stability (IS), and intradaily variability (IV)—are prioritized as the primary observable indicators21,22. Phase markers, including dim-light melatonin onset (DLMO) or melatonin rhythms, core body temperature rhythms, and, when available, peripheral clock-gene expression rhythms, are used as complementary measures23. The rapid socioeconomic development has been increasingly associated with circadian rhythm disruption in individuals, primarily attributable to frequent night shift schedules and irregular eating patterns24. Studies demonstrate that circadian misalignment contributes to various pathologies, including sleep disorders25, cardiovascular diseases, immune dysregulation, neurodegenerative disorders, and carcinogenesis. Circadian rhythm disruption has been recognized as a common pathological feature across multiple diseases26. Epidemiological studies have demonstrated that shift workers exhibit a 9% increased risk of developing type 2 diabetes27, a 3.3% elevated incidence of breast cancer, and the risk of cardiovascular death is similarly increased28. These findings highlight the critical role of circadian alignment in maintaining metabolic, oncological, and cardiovascular health. Mechanistic studies have revealed that deficiency of the core clock gene Bmal1 disrupts the rhythmicity of oxidative phosphorylation in cardiomyocytes, consequently elevating ROS levels29. Meanwhile, desynchronization of electrical activity in SCN neurons may compromise blood–brain barrier integrity, thereby facilitating the propagation of neuroinflammatory responses30. Notably, neuroinflammation plays a pivotal role in the pathogenesis of neurodegenerative diseases like PD. Concurrently, emerging evidence suggests that circadian disruption is closely associated with multiple clinical and pathological dimensions of PD31. On the one hand, PD patients frequently exhibit significant sleep disturbances and circadian rhythm disruptions32–34, which may exacerbate both motor and non-motor symptoms35, thereby influencing symptom burden and prognosis. On the other hand, experimental findings raise the possibility that circadian dysregulation may interact with PD-related pathophysiology through multiple mechanisms, including impaired synthesis and release of neurotransmitters, mitochondrial dysfunction, and dysregulation of the autophagy–lysosome pathway. However, whether circadian abnormalities primarily reflect early pathological involvement, downstream consequences of dopaminergic degeneration, or context-dependent modifiers of ongoing disease processes remains to be clarified. This review therefore aims to systematically synthesize current research advancements regarding the interplay between PD and circadian rhythms, with particular emphasis on the underlying biological mechanisms, evidence strength, and translational implications, thereby providing strategic directions for future investigations and clinical therapeutic development. Physiological mechanisms of circadian rhythms The circadian rhythm, as an endogenous timekeeping system, integrates environmental signals through the SCN—the central pacemaker—to coordinate spatiotemporal organization of physiological processes13,16. Located in the anterior hypothalamus, the bilaterally paired SCN comprises approximately 20,000 neurons organized into ventrolateral (vlSCN) “core” and dorsomedial (dmSCN) “shell” subregions36. The vlSCN core is enriched with vasoactive intestinal peptide (VIP)-expressing neurons that directly receive blue light inputs from intrinsically photosensitive retinal ganglion cells (ipRGCs)37. These signals activate cAMP/PKA pathways via glutamate and pituitary adenylate cyclase-activating polypeptide (PACAP)38, dynamically modulating the phase of core clock genes. In contrast, the dmSCN shell predominantly contains arginine vasopressin (AVP)-expressing neurons39,40, which maintain rhythm robustness through stable molecular oscillations36,41. This “coupled oscillator” architecture ensures both environmental responsiveness and rhythmic stability42,43. The molecular machinery driving circadian rhythms relies on precisely regulated TTFLs44,45. During daylight hours, CLOCK (circadian locomotor output cycles kaput) and BMAL1 (aryl hydrocarbon receptor nuclear translocator-like protein 1) form heterodimers that bind E-box elements in core clock gene promoters, initiating transcription of Period (Per) and Cryptochrome (Cry) genes46. By evening, accumulated PER/CRY proteins form inhibitory complexes that translocate into the nucleus, blocking CLOCK/BMAL1 transcriptional activity to complete the negative feedback phase47–49. This regulatory process is accompanied by critical post-translational modifications: CK1α/δ-dependent phosphorylation promotes the ubiquitination and subsequent proteasomal degradation of PER proteins50, while FBXL3, as the substrate-recognition component of the SCF E3 ubiquitin ligase complex, modulates CRY stability through ubiquitination-mediated degradation51. Together, these coordinated PTMs enable daily molecular clock resetting, sustaining ~24-h oscillations. Beyond the core TTFL, auxiliary regulatory loops further stabilize circadian outputs. For instance, REV-ERBα/β and RORs compete for binding to ROR response elements (ROREs), rhythmically suppressing Bmal1 transcription, while DEC proteins prevent transcriptional overload by inhibiting Clock expression52. However, these ancillary pathways primarily modulate the amplitude and stability of oscillations driven by the core TTFL rather than constituting autonomous regulatory axes. This hierarchical architecture ensures that the TTFL remains the central pacemaking mechanism of circadian rhythms, whereas auxiliary circuits serve as stabilizing complements to preserve system-level resilience. The transcription-translation feedback loop (TTFL) maintains intrinsic rhythms in SCN neurons while simultaneously establishing a dynamic network with peripheral tissues through neural, endocrine, and metabolic signaling to coordinate systemic physiological functions temporally. Within the central nervous system, dopaminergic projections from the SCN regulate rhythmic tyrosine hydroxylase (TH) expression, optimizing motor function by ensuring peak dopamine synthesis and release during wakefulness52. The CLOCK protein remodels chromatin structure through its intrinsic histone acetyltransferase (HAT) activity, thereby enhancing the accessibility of synaptic plasticity-related genes including Bdnf53,54. Subsequently, BDNF activates TrkB receptors to facilitate hippocampal long-term potentiation (LTP) and dendritic spine structural remodeling, which collectively consolidate learning-associated memory formation55. This regulatory mechanism exhibits circadian rhythmicity, and its dynamic variations during the rest phase may optimize memory consolidation. Prefrontal cortical glutamate receptors exhibit circadian fluctuations that modulate neuronal excitability, influencing higher cognitive functions: resting-phase glutamate accumulation prevents neural network overactivation, while efficient sleep-phase clearance maintains next-day cognitive performance56,57. The SCN regulates adrenal gland activity via sympathetic innervation, driving peak corticosterone secretion during the early morning phase58. The corticosterone-glucocorticoid receptor (GR) complex subsequently modulates metabolism-related genes (e.g., Srebp1c) through epigenetic mechanisms, including chromatin remodeling59, thereby promoting hepatic lipid synthesis59,60. This regulatory process may facilitate the rapid mobilization of energetic substrates upon awakening to meet diurnal metabolic demands. Parasympathetic dominance in the afternoon enables acetylcholine and vasoactive intestinal peptide (VIP) to stimulate pancreatic β-cell insulin secretion, precisely matching muscle tissue glucose uptake requirements to circadian demands61,62. This neural-endocrine axis exemplifies predictive metabolic adaptation. Endocrine regulation involves light-mediated SCN suppression of pineal melatonin synthesis through the PVN-sympathetic pathway63, while darkness activates arylalkylamine N-acetyltransferase (AANAT)-dependent melatonin production64,65. Nocturnal melatonin peaks inhibit hypothalamic CRH release, reducing ACTH and cortisol to maintain HPA axis circadian balance—a bidirectional light-melatonin regulatory mechanism critical for sleep–wake cycles and stress responses66,67. The immune system demonstrates circadian regulation through enhanced NLRP3 inflammasome activity during late-night corticosterone troughs, optimizing spatiotemporal pathogen clearance. Studies demonstrate that morning vaccination enhances cytokine production in peripheral blood leukocytes upon restimulation68. This provides a molecular basis for the application of time therapy in immune intervention (Fig. 1). The circadian rhythm system of mammals: from central pacemaker to systemic physiological integration. Light information from ipRGCs entrains the SCN central pacemaker, organized into a ventrolateral “core” (VIP +) and dorsomedial “shell” (AVP +) that act as coupled oscillators. The core molecular clock is generated by the CLOCK/BMAL1–PER/CRY transcription–translation feedback loop, tuned by post-translational regulation and auxiliary REV-ERB/ROR/DEC circuits. SCN timing signals coordinate day–night physiology across systems, including dopaminergic function, synaptic plasticity/cognition, neuroendocrine–metabolic rhythms (corticosterone/insulin and melatonin–HPA coupling), and immune gating (e.g., NLRP3 activity), while synchronizing peripheral clocks. Circadian disruption is linked to metabolic, immune, and neurodegenerative pathologies. Created with BioGDP.com228. These multilayered circadian networks illustrate how molecular oscillators coordinate neuroendocrine, metabolic, and photic cues to maintain spatiotemporal precision in physiological processes. Disruption of these networks has been increasingly linked to metabolic syndrome, neurodegenerative diseases, and immune dysregulation. Pathogenesis of Parkinson’s disease PD, a multifaceted neurodegenerative disorder, is initially characterized by the selective degeneration of dopaminergic neurons in the substantia nigra pars compacta (SNc)1. As the disease advances, pronounced dopamine (DA) depletion within the nigrostriatal pathway disrupts basal ganglia circuitry, giving rise to hallmark motor symptoms such as bradykinesia, resting tremor, rigidity, and postural instability5. At the pathological level, the accumulation of misfolded α-synuclein into intraneuronal Lewy bodies represents a defining feature of PD10. According to Braak staging, α-synuclein pathology follows a stereotypical caudo-rostral progression, beginning in the dorsal motor nucleus of the vagus within the medulla oblongata and gradually spreading to the SNc and neocortical regions69. This spatial-temporal cascade correlates with clinical trajectories, whereby non-motor symptoms frequently precede overt motor deficits. Circuit-level analysis reveals that SNc dopaminergic neuron loss diminishes striatal DA input10, attenuating activity in the direct pathway while enhancing output from the indirect pathway. This imbalance results in excessive inhibition of thalamocortical projections by the globus pallidus internus (GPi) and substantia nigra pars reticulata (SNpr), thereby contributing to motor dysfunction70,71. Beyond motor symptoms, accumulating evidence underscores the clinical relevance of non-motor features in PD pathophysiology. These include sensory impairments, rapid eye movement (REM) sleep behavior disorder (RBD), circadian rhythm disruption, autonomic dysregulation, and mood disturbances6. Together, these manifestations represent clinically important components of the disease spectrum and frequently accompany widespread neurodegeneration. Among these, hyposmia is the most prevalent sensory deficit and often precedes motor symptom onset by years or even decades, making it a valuable prodromal indicator. RBD, in particular, is regarded as a robust predictive marker in the prodromal phase, with approximately 50% of PD patients exhibiting REM sleep without atonia (RSWA) manifested as dream-enactment behaviors. Longitudinal cohort studies reveal that 38% of individuals with RBD eventually develop PD over an average follow-up period of 12.7 years72,73, reinforcing its role as an early biomarker of α-synucleinopathy. Elucidating the pathophysiological bases of clinical symptoms in PD is essential for developing individualized treatment strategies. A major clinical challenge is that motor symptoms attributable to dopaminergic deficiency, circadian rhythm disturbance, and autonomic dysfunction frequently coexist and overlap, necessitating multidimensional differentiation74. Specifically, the classic manifestations of dopaminergic deficiency include bradykinesia, resting tremor, and rigidity. Symptom severity often tracks the levodopa dosing cycle, presenting as wearing-off phenomenon or “on–off” fluctuations75,76. By contrast, circadian rhythm disruption is more commonly expressed as sleep–wake abnormalities and diurnal variation, and this symptom cluster is closely related to the light–dark transition77. Recognition of autonomic dysfunction requires careful attention to posture-dependent changes and specific physiological functions. Symptoms ranging from lightheadedness on standing suggestive of orthostatic hypotension to refractory constipation and urinary dysfunction often respond poorly to dopaminergic medications and may even be inadvertently exacerbated during medication adjustments78,79. Accordingly, integrating symptom patterns with medication timing, circadian context, and postural triggers can help disentangle primary and secondary contributors within complex symptom networks and enable more targeted interventions. In this context, sleep–wake cycle disturbances are particularly prominent in patients with PD, with over 80% exhibiting varying degrees of circadian rhythm misalignment80. Clinical studies have reported a delayed melatonin secretion phase and reduced amplitude of core body temperature rhythms, findings that are compatible with altered circadian regulation and possible involvement of SCN-related timing pathways81,82. In patients with PD, circadian rhythm disorders are often associated with sleep structure disturbances, including rapid eye movement (REM) sleep behavior disorder (RBD), and are also related to autonomic nerve dysfunction83. For example, orthostatic hypotension correlates with impaired diurnal regulation of baroreflex sensitivity83, while α-synuclein deposition in the enteric nervous system disrupts colonic motility rhythms, resulting in refractory constipation in up to 80% of PD patients84. Moreover, circadian misalignment is strongly linked to mood disturbances, including depression and anxiety, as well as to cognitive decline85,86. Clinical evidence demonstrates a positive correlation between the degree of melatonin phase delay and the rate of cognitive deterioration in PD patients with comorbid RBD87. Collectively, these findings support circadian rhythm disruption as an important non-motor and prodromal dimension of PD, closely linked to sleep–wake, autonomic, mood, and cognitive manifestations. In summary, circadian rhythm disruption is increasingly recognized as a clinically relevant temporal dimension of PD and a potential prodromal biomarker. However, whether these alterations primarily reflect downstream consequences of neurodegeneration, contribute to symptom burden, or modify disease vulnerability in specific contexts remains to be further clarified. Deciphering the molecular and neural mechanisms underlying circadian dysfunction in PD may help refine stage-specific preventive and therapeutic strategies. Circadian disturbances from prodromal to clinical PD Circadian rhythm abnormalities have attracted increasing attention in the prodromal stage of PD, particularly because they can be tracked longitudinally and may emerge before overt motor onset80,88. Clinical studies indicate that abnormalities in the sleep–wake cycle can be detected several years before the onset of motor symptoms, supporting the prodromal relevance of behavioral-level rhythm disruption80,88. However, early detectability does not by itself establish an upstream causal role. As research on prodromal cohorts and disease mechanisms advances, these alterations are no longer viewed merely as non-motor comorbidities but are increasingly integrated into multidimensional models of disease progression88,89. Nevertheless, current epidemiological evidence remains insufficient to determine whether these changes primarily reflect early pathological involvement of circadian networks or whether they themselves contribute causally to disease acceleration80. Consistent with the behavioral phenotype, early PD has been associated with reduced expression of core clock genes (e.g., BMAL1) in peripheral tissues (e.g., blood cells), accompanied by attenuated diurnal rhythmicity, providing clinical clues that circadian dysregulation may be involved in the disease process. Mechanistic studies provide more direct evidence90–92. Experimental evidence indicates that neuron-specific deletion of Bmal1 increases vulnerability to oxidative stress93,94, disrupts mitochondrial homeostasis and the autophagy–lysosome pathway95,96, and exacerbates α-syn-related neurotoxicity in animal models94. These findings support the mechanistic plausibility that core clock genes function not only as “timers” but also as regulatory hubs that coordinate metabolism and stress homeostasis, and that their impairment may establish a vulnerable substrate for downstream pathological cascades. Within this vulnerable context, immune remodeling emerges as a key hub linking early molecular alterations to later neuronal injury97,98. Microglia possess intrinsic circadian regulatory capacity, and circadian programs—modulated by nuclear receptors such as REV-ERBα—can bias inflammatory signaling toward easier engagement of pathways such as the NLRP3 inflammasome97,99,100. When circadian disruption persists, inflammatory responses may shift from transient and reversible states to more sustained activation, further amplified by modules such as the NLRP3 inflammasome, thereby promoting neuroinflammation that is more difficult to resolve97,101. Notably, immune imbalance and circadian disruption form a bidirectionally coupled system: circadian disturbance can dysregulate inflammatory responsiveness, whereas persistent neuroinflammation can in turn destabilize circadian rhythms via effects on the hypothalamic SCN and its downstream outputs, further eroding rhythm stability101,102. A similar bidirectional coupling is observed between the circadian system and α-syn pathology. On the one hand, circadian disruption can coincide with reduced efficiency of protein-clearance processes, facilitating abnormal α-syn accumulation and intercellular spread103–105. On the other hand, α-syn deposition can involve sleep–wake regulatory circuits (e.g., the locus coeruleus and nucleus accumbens), indicating that circadian abnormalities may shift from a downstream readout of pathology to a factor that can further amplify selected disease-related processes106–108. If these processes persist, they collectively increase the risk of dysfunction in the nigrostriatal pathway and injury to dopaminergic neurons, ultimately associating with the emergence of motor symptoms. From a time-series perspective, before motor onset, dopaminergic terminals and axons in the striatum may already show degeneration and reduced synaptic transmission, whereas neuronal cell bodies in the SNc can remain relatively preserved109,110. This pattern—functional impairment preceding overt structural loss—provides a neurobiological basis for prominent non-motor symptoms during the prodromal stage and supports the existence of a window for early intervention. As PD progresses into the clinical stage, pathological burden can exceed compensatory capacity, irreversible neuronal loss accelerates, and motor symptoms emerge and progressively worsen111,112. In summary, behavioral rhythm alterations years before diagnosis provide traceable early signals80, and peripheral molecular clock abnormalities extend these observations to the biological level113. Mechanistic evidence further suggests that core clock genes participate in stress regulation and protein-processing homeostasis, thereby supporting a vulnerability-modifying framework114. In parallel, microglial immune regulation and α-syn pathology each form mutually reinforcing feedback loops with the circadian system. Together, these observations support the prodromal relevance of circadian abnormalities, while the causal position of these changes across the transition to clinical PD remains to be fully resolved101,115. Interactions between circadian clock genes and genetics of PD Circadian clock genes have been increasingly linked to PD-related molecular and clinical phenotypes116. Non-motor manifestations of PD—particularly sleep–wake rhythm disturbances and circadian alterations in endocrine function—precede the onset of motor symptoms33,117. Historically, these rhythm abnormalities were largely viewed as secondary consequences of dopaminergic neurodegeneration occurring after disease onset. However, recent studies have raised a key question: under specific conditions, perturbation of the circadian timing system may interact with PD susceptibility pathways25,116,118. These findings support mechanistic plausibility for circadian–genetic crosstalk, but do not establish a uniform upstream causal role for circadian disruption in PD initiation. In population genetics, studies have reported associations between variants in core circadian clock genes (e.g., CLOCK, ARNTL/BMAL1, PER, CRY) and PD-related phenotypes or specific clinical outcomes118,119. In a Chinese whole-exome sequencing cohort of patients with early-onset or familial PD (n = 1997), investigators stratified rare coding variants in PER1 and subsequently performed within-cohort genotype–phenotype association analyses119. The results indicated that carriers of deleterious PER1 variants were more likely to develop motor complications such as dyskinesia. The same sequencing study further reported that functional loss-of-function variants in CRY1 were more prevalent in the early-onset or familial PD population119. Collectively, these findings suggest that circadian rhythm-related genes are not merely background components of the timing system, rather, under specific genetic contexts, they may represent modifiers within the broader PD risk landscape120,121. Individuals harboring specific rhythm-gene variants may have a less stable endogenous timing system from early life, and this inherent vulnerability could increase sensitivity to aging- or environment-related stressors such as shift work and light disturbance122,123. At the molecular and cellular levels, the interplay between circadian genes and genetic susceptibility to PD can be viewed as bidirectional regulation. On one hand, circadian genes influence diurnal variation in oxidative stress, mitochondrial homeostasis, and neuroinflammatory activity, thereby creating a fluctuating survival microenvironment for neurons expressing specific PD risk genes124–126. Previous studies have shown that the BMAL1/CLOCK heterodimer can regulate rhythmic activity at the SNCA promoter and modulate the rhythmic expression of numerous downstream targets via non-classical E-box-like elements, leading to diurnal differences in oxidative stress responses, mitochondrial function, inflammation-related processes, and autophagic flux94,127,128. On the other hand, in individuals carrying PD risk genes (e.g., SNCA, LRRK2, Parkin), mitochondrial function and proteostasis are often already compromised105,129–131. Under these conditions, perturbation of core clock-gene expression by aging or environmental desynchronization may further narrow the capacity of cells to cope with α-syn aggregation or mitochondrial damage, particularly during circadian phases when antioxidant defenses are relatively weak94,121,124,132. It is worth noting that either environmental intervention that disrupts the rhythm or knockout of the Bmal1 gene can exacerbate the vulnerability of dopaminergic neurons to neurotoxins94,133. It is worth noting that the evidence from the disruption of the core clock mechanism genes is not entirely equivalent to the evidence from behavioral or environmental circadian rhythm disorders. Taken together, these findings support a vulnerability-modifying role for an intact rhythm system in neuronal survival. More importantly, this interaction can become self-reinforcing. The hallmark pathological feature of PD—abnormal α-syn aggregation—can itself interact with the core molecular clock134,135. In various PD animal models, the accumulation of α-syn has been observed to interfere with the transcription–translation feedback loop of clock genes, for example, by affecting the stability of core clock proteins or their ability to bind DNA elements, thereby weakening circadian amplitude35,114,135. Thus, rhythm-system vulnerability conferred by genetic background or PD-related risk render neurons more susceptible to damage across the day–night cycle94,114. Moreover, early accumulation of pathological proteins can further impair normal molecular-clock function, leading to substantial disruption of rhythm-regulated protective physiological processes (e.g., mitophagy and rhythmic release of inflammatory factors)97,136,137. This cycle can amplify local pathological changes into systemic circadian disruption and ultimately aggravate the core neurodegenerative process. Thus, an individual’s genetic risk is not a fixed number but is constantly increased or decreased in the daily and nightly physiological rhythm fluctuations. This perspective also provides new ideas for the early warning and intervention of PD. Circadian rhythms and central glial inflammatory gating in PD During PD pathogenesis, neuroinflammation is widely regarded as an important amplification pathway rather than a passive accompanying phenomenon. When circadian rhythms are disrupted, microglia are more likely to shift from a relatively protective state toward a pro-inflammatory phenotype, thereby enhancing inflammatory responses and promoting activation of the NLRP3 inflammasome97. At the molecular level, activation of the NLRP3 inflammasome and the subsequent caspase-1-mediated maturation and release of IL-1β and IL-18 have been repeatedly documented in patient samples and in various animal models97,138,139. Following inflammasome activation, ASC specks can be released and propagate between cells. Under stimulation by α-syn preformed fibrils (PFFs), these particles are associated with enhanced inflammasome signaling, accompanied by clustering of reactive microglia, an increased pathological burden of α-syn, and aggravated dopaminergic neuron injury, ultimately leading to deterioration of motor function139–141. During this inflammatory amplification process, the circadian rhythm system provides a defined regulatory entry point, and its main effect is manifested in the influence of microglia-intrinsic clock genes on inflammation-related transcriptional programs. Among these genes, REV-ERBα (NR1D1) serves as a key node in the clock output pathway and can inhibit NF-κB-dependent transcription and suppress NLRP3 transcription97,142. In the substantia nigra of MPTP-induced PD mice, the diurnal fluctuations of REV-ERBα were significantly weakened or even lost, and the diurnal fluctuations in microglial morphology and inflammatory markers were no longer apparent97. Studies have shown that, in the BV2 microglial cell line, the REV-ERBα agonist SR9009 can effectively reduce the levels of phosphorylated NF-κB p65 (p-p65), NLRP3, ASC, cleaved caspase-1, and IL-1β/IL-18—key readouts related to inflammasome activation—and inhibit the inflammatory activation induced by MPP⁺ or α-synuclein PFFs97,143,144. It should be noted that inflammatory signals can, in turn, perturb the molecular clock system. In PD, this relationship forms a reciprocal loop. Across multiple experimental systems, pro-inflammatory cytokines such as TNF-α have been shown to weaken CLOCK–BMAL1-driven transcription at E-box-dependent promoters, thereby suppressing the expression of core clock genes145–147. In the context of central nervous system inflammation, the phase (peak timing) of microglial Bmal1 expression shifts, accompanied by the rhythm disorder of its downstream target genes147. This suggests that, when inflammatory stimulation persists, circadian regulation in microglia tends to weaken, making the inflammatory state more likely to persist over the long term and less likely to return to its original pattern of circadian oscillation. However, the direction of this target’s effect is closely dependent on the model context. In the early stages of the APP-KI Alzheimer’s disease model, the CLOCK/BMAL1 negative feedback loop in microglia is already impaired. Under these conditions, treatment with the REV-ERBα agonist SR9009 increases the expression of inflammatory cytokines and exacerbates cognitive impairment147,148. This observation suggests that, even for the same molecular target, immunomodulatory outcomes can differ across distinct proteinopathy backgrounds or under different circadian states. Circadian rhythm and mitochondrial redox vulnerability in PD Disruption of circadian rhythms may exacerbate the accumulation of oxidative stress through multiple mechanisms and thereby interact with PD-related vulnerability. The selective vulnerability of dopaminergic neurons in the SNc is closely related to their autonomous pacemaker electrical activity149. During pacemaker activity, these neurons rely on L-type Ca²⁺ channels (mainly CaV1.3) to sustain subthreshold Ca²⁺ influx, thereby increasing intracellular Ca²⁺ load and enhancing coupling with mitochondrial metabolism, ultimately leading to increased basal mitochondrial oxidative load150,151. ROS not only damages cell membrane integrity through lipid peroxidation and interferes with neurotransmitter release and Ca²⁺ homeostasis, but also directly oxidizes dopamine molecules, generating neurotoxic quinone compounds152,153. Therefore, when energy production efficiency decreases and oxidative stress increases simultaneously, the risk of structural damage is further exacerbated154. The circadian system’s influence on this vulnerability is more evident in the regulation of metabolic homeostasis by cell-intrinsic clock genes, rather than merely in overt disruptions of behavioral rhythms. In conditional genetic models, TH⁺ neuron loss in the SNc was observed after Bmal1 (Arntl) knockout either in neuronal lineages or selectively in TH⁺ dopaminergic neurons. This loss was accompanied by transcriptional changes in oxidative phosphorylation (OXPHOS)-related gene programs and PD-associated pathways94. Notably, when a T20 light–dark cycle (10 h:10 h) induced pronounced behavioral rhythm disruption, the number of SNc TH⁺ neurons did not decrease94. This “genetic perturbation effective, environmental desynchronization ineffective” control demonstrates that altering the external light–dark cycle to disrupt behavioral rhythms does not necessarily lead to SNc neuronal loss. Conversely, direct intervention in core clock genes is more likely to cause imbalance in mitochondrial gene expression programs, accompanied by neuronal damage. This regulatory relationship manifests as a bidirectional interaction in the context of PD pathology. The cellular redox state and energy-sensing pathways can influence BMAL1/CLOCK activity and the post-translational modification of clock proteins via signaling axes such as NAD⁺–AMPK–SIRT1, thereby allowing mitochondrial metabolic state and oxidative stress levels to feed back onto the circadian oscillator155,156. Moreover, ROS and inflammatory signaling are mutually amplifying, ROS can activate pathways such as NF-κB and MAPK to promote the production and release of pro-inflammatory factors, whereas an inflammatory microenvironment can enhance NADPH oxidase activity, forming a self-reinforcing positive feedback loop157–159. In α-syn-related models, the expression of core clock genes, including BMAL1, PER2, and CLOCK, is downregulated, accompanied by altered BMAL1 acetylation, further supporting this view89,159. Overall, under the combined influence of mitochondrial metabolic stress and α-syn pathology, core cellular clock components are frequently disrupted, and this disruption may further impair the cell’s capacity to maintain homeostasis under repeated stimulation. Therefore, evaluation of intervention efficacy at this level should not rely solely on improvements in behavioral or circadian phenotypes as the primary criteria. It is also essential to assess whether mitochondrial dysfunction and oxidative damage have been substantially alleviated. Taking time-restricted feeding (TRF) in the MPTP mouse model as an example, under a casein-based diet, TRF downregulated monoamine oxidase B (MAO-B) and certain dopamine metabolism-related indicators, while improving some measures of motor performance. However, it failed to restore striatal DA levels or prevent neuronal loss160. Conversely, under a soy protein–based diet, the same TRF protocol was accompanied by recovery of DA levels, improved survival of TH⁺ neurons, and reduced inflammatory responses160. This “same intervention, different backgrounds, different outcomes” pattern indicates that improvements in motor behavior and neuroprotection are not always synchronized, necessitating evaluation using both structural and pathological endpoints (Fig. 2). Reciprocal coupling between the intrinsic clock and mitochondrial redox stress in SNc dopaminergic neurons. SNc dopaminergic neurons depend on CaV1.3 L-type Ca²⁺ channels for autonomous pacemaking, increasing mitochondrial coupling and basal ROS. ROS promotes lipid peroxidation, disrupts Ca²⁺ homeostasis, oxidizes dopamine to toxic quinones, and—when GSH buffering fails—drives structural damage. The cell-intrinsic clock (BMAL1/CLOCK) regulates OXPHOS programs, while NAD⁺–AMPK–SIRT1 and redox signaling feedback to modify clock activity, forming a bidirectional loop that is further perturbed in α-syn models (clock gene downregulation and altered BMAL1 acetylation). Time-restricted feeding (TRF) in MPTP mice highlights context dependence: food access was restricted to an 8-h active-phase window (ZT13–ZT21; 1 h after lights off to 3 h before lights on) under a 12-h light/dark cycle, with MPTP administered on day 21 and outcomes assessed on day 28160. Motor improvement may occur without DA recovery/neuron protection on a casein diet, whereas soy protein can align TRF with DA restoration, TH⁺ survival, and reduced inflammation—supporting evaluation with both behavioral and structural endpoints. Created with BioGDP.com228. Circadian rhythm and the α-syn–protein homeostasis axis in PD Abnormal α-syn aggregation and cell-to-cell spread, together with a reduction in protein clearance, constitute a central hub of protein homeostasis imbalance in PD161. When the production of misfolded proteins persistently exceeds the system’s degradative capacity, α-syn can progressively form soluble oligomers and insoluble fibrillar aggregates162. Some of these aggregates exhibit seed-like properties that promote conformational conversion of additional α-syn and are associated with local pathological accumulation and spread to adjacent regions163,164. In the degradation pathway, dysfunction of the autophagy–lysosome pathway is strongly associated with α-syn aggregation165. Wild-type α-syn (i.e., lacking familial SNCA mutations) can be delivered to lysosomes for degradation via chaperone-mediated autophagy (CMA), whereas variants such as A53T and A30P are impaired at the step of lysosomal membrane receptor-mediated translocation166,167. This not only reduces their own clearance but also hinders the lysosomal entry of other CMA substrates, thereby increasing the likelihood that the local system reaches “clearance saturation”. The ubiquitin–proteasome system (UPS) also contributes to the stability of α-syn-related proteins and to global cellular protein turnover, and it is coupled to the turnover of clock proteins168,169. However, with respect to more directly measurable outcomes—aggregate burden and propagation—current evidence is more focused on the endocytosis–lysosome axis170–172. Accordingly, the discussion on rhythm and α-syn pathology is more suitable to focus on the uptake of α-syn by glial cells, the lysosomal processing ability, and degradation efficiency. In astrocytes, the effects of BMAL1 on inflammatory state and protein-handling capacity show marked context dependence. At stages of low aggregate burden, when homeostasis is largely preserved, astrocyte-specific Bmal1 deletion can induce spontaneous activation and increased expression of pro-inflammatory mediators127,128, accompanied by downregulation of the glutathione S-transferase (GST)-associated defense program, thereby weakening neuronal support and increasing vulnerability in vitro128. By contrast, under pathological conditions characterized by a high aggregate burden, the same manipulation can upregulate Bag3 and enhance BAG3-dependent selective macroautophagy as well as endocytic–lysosomal clearance, improving the uptake and degradation of α-syn (and tau) and thereby reducing pathological burden and limiting α-syn spread128. Accordingly, the direction of Bmal1-dependent regulation in disease progression is contingent on pathological stage and microenvironment. Feedback regulation also has a clear molecular substrate, α-syn aggregation and the ensuing pro-inflammatory signaling can persistently suppress BMAL1 and other circadian clock nodes in glial and myeloid cells via the NF-κB–miR-155 axis173, thereby blunting diurnal variation in immune responses and protein handling and promoting reciprocal reinforcement between protein accumulation and inflammation174. Importantly, causal inferences drawn from genetic deletion of core clock genes are not interchangeable with those drawn from environmentally induced circadian disruption. Neuron-specific or TH⁺ dopaminergic neuron-specific Bmal1 deletion (e.g., TH-Cre) can reduce SNc TH⁺ neuron numbers even in the absence of PFF injection94. However, in the α-syn PFF model, this does not necessarily further exacerbate TH⁺ loss following injection. Likewise, circadian disruption induced by photoperiod manipulation does not reliably reproduce this neuronal-loss phenotype94. These observations indicate that improvements in circadian measures do not necessarily translate into an altered neurodegenerative trajectory. Both should therefore be evaluated in parallel using endpoints such as SNc TH⁺ neuron counts, striatal TH immunoreactivity, and α-syn pathological burden. From an interventional perspective, metabolic regulation and feeding-time scheduling can influence pathological burden and clearance capacity in certain α-syn models175,176. In an rAAV–α-syn overexpression model, intermittent fasting initiated after pathology had formed was associated with improved motor performance, accompanied by reduced injury to dopaminergic neurons and axon terminals, partial restoration of striatal dopamine levels, enhanced autophagy, increased clearance of phosphorylated α-syn (p-α-syn), reduced insoluble species, and downregulation of inflammation-related transcriptional programs176. Accordingly, within the same experimental system, concurrent quantification of α-syn seed amplification activity (α-syn SAA), p-α-syn, insoluble-species burden, lysosomal flux markers, and behavioral endpoints may help distinguish “symptom improvement” from changes in “pathological burden” and delineate the translatable boundaries of rhythm- and metabolism-based interventions across models and pathological stages. Circadian rhythm and the blood–brain barrier gating mechanism in PD Disruption of circadian rhythms can influence PD-related pathological processes not only within the central nervous system but also by altering peripheral immune homeostasis, potentially allowing peripheral inflammatory signals to more readily cross the blood–brain barrier and may facilitate peripheral-to-central inflammatory signaling under specific conditions177–179. In the context of PD pathology, the effects of circadian disruption may not necessarily present as a generalized elevation in overall inflammatory levels; instead, they are more likely to be reflected in pronounced time-of-day differences in how peripheral inflammatory signals access the central nervous system180,181. When the phase, amplitude, or stability of circadian rhythms is disrupted, neurovascular unit (NVU) regulation of barrier-crossing events may no longer follow its normal circadian pattern, making peripheral inflammation more likely to compromise the blood–brain barrier at specific times and to elicit a stronger inflammatory response within the brain180,182,183. Studies have shown that circadian disruption can bias peripheral macrophages toward a more pro-inflammatory state, and that they can impair tight-junction integrity by secreting matrix metalloproteinases (MMPs), increasing barrier vulnerability and thereby promoting peripheral–central immune crosstalk184–187. Consistent with this, in a PD cohort, dynamic contrast-enhanced MRI detected impaired blood–brain barrier (BBB) integrity, with more pronounced increases in parameters such as Ktrans in pathology-relevant regions, including the substantia nigra179,188,189. These findings suggest that BBB dysfunction may contribute to the disease phenotype rather than being merely an incidental accompaniment. Recent work further indicates that the BBB is not a static barrier but a dynamic interface that varies with circadian rhythms15. A study that focused on BBB efflux via ABCB1/P-gp as a key readout showed that, in mice, BBB clearance of efflux substrates fluctuates over a 24-h period. Moreover, when the intrinsic endothelial clock was disrupted (e.g., endothelial-specific Bmal1 deletion or Cry1/2 deletion), this diurnal variation disappeared, indicating that endothelial circadian mechanisms contribute to regulating BBB transport function15,180,182. At the same time, diurnal changes in neuronal network activity show an inverse relationship with P-gp-mediated efflux, and endothelial circadian gene deletion weakens this relationship, suggesting that BBB diurnal dynamics are shaped not only by peripheral rhythms but may also be inversely modulated by the central network state181,190. Under inflammatory stimulation, these diurnal differences are often more readily observed. For example, with LPS challenge, the same stimulus can markedly change the extent of BBB disruption and the ensuing inflammatory response in the brain simply by shifting the time of administration. When given in the evening or at night, more pronounced tracer extravasation and stronger inflammation-related transcriptional changes are more likely to occur180,191. This temporal difference is attenuated or even abolished under constant-darkness conditions, and microglial exhaustion can block the “nighttime exaggeration” of the BBB extravasation phenotype. Together, these findings indicate that BBB susceptibility to inflammatory insults—and the subsequent repair process—are time-dependent. When this framework is considered in the pathological context of PD, the question shifts from “Is there a vulnerable time window?” to “Why is that window more prone to being breached?” The chronic inflammatory milieu, background oxidative stress, and α-syn-related proteotoxicity in PD create a persistent stress environment for the NVU179,192. On the one hand, α-syn-related pathology may make peripheral inflammatory signals more likely to translate into measurable increases in BBB permeability. For example, in the presence of α-syn pathology, LPS-induced increases in BBB permeability are amplified, suggesting that a higher pathological load raises the likelihood of BBB functional impairment under inflammatory challenge188,193,194. In parallel, changes in transporters that are closely related to BBB gating are heterogeneous. Some studies in PD cohorts have reported reduced P-gp function, whereas in an early, treatment-naïve PD cohort, no obvious P-gp impairment was observed181,195,196. This inconsistency suggests that alterations in the transporter axis may be strongly related to disease stage, clinical phenotype, or coexisting pathological features, making it difficult to draw straightforward inferences. However, circadian variation in BBB structural proteins is not consistently reproducible across models, time windows, or detection methods. By contrast, functional readouts—such as solute exchange and clearance capacity—and diurnal differences elicited under inflammatory stimulation are more readily observed across studies182,190. These observations suggest that the key question now shifts from whether the BBB exhibits diurnal variation to why such variation weakens or disappears in the context of PD. Possible explanations include impairment of endothelial clock function, reduced coordination between neuronal activity and endothelial function, or dysregulated signaling at the microglia–vascular perivascular interface182. Moreover, whether these changes can be captured by longitudinal population imaging (i.e., repeated scans across multiple time points) together with wearable-derived rhythm metrics—and ultimately used to identify potential windows for intervention—remains an important direction for future investigation (Fig. 3). Circadian gating of the blood–brain barrier and its breakdown in Parkinson’s disease. Schematic model of BBB/NVU dynamics under (left) healthy conditions with an intact circadian rhythm versus (right) Parkinson’s disease (PD) with circadian disruption. In the healthy state, endothelial tight junctions preserve barrier integrity and rhythmic ABCB1/P-glycoprotein (P-gp)-mediated efflux supports time-of-day-dependent clearance of circulating substrates, limiting toxin accumulation and restraining peripheral-to-central inflammatory signaling; macrophages remain relatively quiescent and microglia are predominantly surveillant. In PD with circadian disruption, peripheral macrophages are biased toward a pro-inflammatory phenotype and secrete matrix metalloproteinases (MMPs) that weaken tight-junction “stitches,” increasing BBB vulnerability and facilitating entry of peripheral inflammatory mediators. In the presence of PD-related stressors (including α-synuclein proteotoxicity), this heightened barrier leakiness promotes microglial activation and neuroinflammatory amplification, reinforcing a pathogenic loop between peripheral inflammation, BBB dysfunction, and central immune responses. Created with BioGDP.com228. Intervention measures and outlook Although the causes and clinical course of PD vary widely, current treatments still focus primarily on symptomatic management, and therapeutic responses differ substantially among individuals197. As the disease progresses, non-motor symptoms—including motor fluctuations, sleep disturbances, and autonomic dysfunction—continue to accumulate, thereby narrowing the patient population and clinical contexts in which long-term sustained benefits can be achieved118,198. Increasing evidence indicates that circadian rhythm disruption is not merely a concomitant feature of PD, but is tightly intertwined with sleep–wake regulation, emotional and cognitive processes, and autonomic nervous system function. In individuals with idiopathic REM sleep behavior disorder (iRBD)—a prototypical prodromal population—rhythm abnormalities can now be systematically captured and are associated with clinical phenotypes and imaging markers of neurodegeneration, providing a practical entry point for earlier identification and implementable intervention199–202. At the clinical application level, the advantage of circadian rhythm indicators is that they can be readily obtained in everyday settings and facilitate long-term follow-up203. Using wrist-worn accelerometers, 24-hour rest–activity rhythm (RAR) monitoring combined with a nonparametric analytic framework can reliably derive key indices—such as relative amplitude (RA), diurnal stability (IS), and rhythm fragmentation (IV)—from multi-day free-living data, thereby shifting rhythm disruption from a conceptual description to a quantifiable measure199,203,204. A cohort study in PD showed that, compared with controls, patients commonly showed lower RA and IS and higher IV, these RAR indices were statistically associated with motor symptom severity and autonomic symptom burden, supporting their feasibility as practical, implementable indicators for longitudinal follow-up202. At the pre-onset stage, changes in the RAR pattern can also be detected in individuals with iRBD199,205. More importantly, recent multimodal naturalistic monitoring studies have linked greater RAR fragmentation with reduced striatal dopamine transporter uptake on DAT-SPECT and other risk cues, thereby providing a more robust basis for using RAR as a stratification variable in high-risk populations206,207. Phase assessment that complements wearable monitoring can be achieved by measuring the melatonin secretion rhythm. Studies have shown that melatonin rhythm amplitude is reduced in patients with PD and is associated with daytime sleepiness208. However, this indicator is relatively sensitive to disease stage and medication status. Dopaminergic therapy can alter melatonin secretion levels and may affect the correspondence between melatonin phase and actual sleep onset time208–210. Therefore, although phase-assessment methods such as dim light melatonin onset (DLMO) are relatively mature and are better suited to mechanistic studies or small-sample intervention trials aimed at clarifying mechanisms210, their sampling costs are high, making them difficult to implement in routine outpatient screening at present211. A more practical approach is to reserve them for validating and stratifying specific subgroups and for evaluating intervention mechanisms. At the intervention level, melatonin and its receptor agonists can be used both to regulate circadian rhythms (as chronobiotics) and to improve sleep212. However, multiple clinical trials have shown that melatonin does not improve PD motor symptoms, regardless of whether immediate-release or sustained-release formulations are used; UPDRS-III scores show no significant change after treatment213,214. Its clinical advantages lie in a relatively clear dosing regimen and the easier uniformity of prescriptions. Nevertheless, the magnitude of improvement appears to be more sensitive to dose, formulation, timing of administration, and participants’ baseline circadian status215,216. Existing studies generally suggest good safety and improvements in some sleep measures, but variability in dosing protocols and overall study quality still leave the conclusions uncertain217,218. Future studies should use consistent circadian metrics to distinguish “sleep improvement” from “circadian adjustment” and evaluate these outcomes separately. Regarding RBD management, current guidelines provide conditional recommendations for clonazepam and immediate-release melatonin34,219. In patients with PD and in the prodromal stage, when these treatments are used to reduce the risk of nocturnal behavioral events and follow-up incorporates indicators such as RAR fragmentation or phase shifts34,220, they can help determine whether “symptom improvement” is accompanied by measurable changes in circadian metrics. At the same time, lifestyle and behavioral interventions (regular sleep schedules, exercise, and light exposure) can serve as a standardized foundational approach because they are highly accessible and low risk221,222. By reducing interference from external behavioral factors and improving comparability across studies, this approach helps avoid misattributing baseline differences to direct effects of circadian factors. Overall, the circadian system is closely coupled with PD in a bidirectional manner. When applying this knowledge to clinical practice, the priority is not to generalize circadian disruption as the primary driver, but to establish actionable, implementable plans. This includes identifying real-world rhythm variables that can be measured reliably (primarily RAR indices, supplemented by research-level phase measures such as DLMO)223,224, defining controllable intervention steps (e.g., parameterized prescription protocols for light therapy and the timing of circadian-modifying medications)225, and, from a disease-stage perspective, determining whether circadian disruption at each stage is more likely to represent a modifiable factor or an indicator of disease state226,227. Building on this framework, higher-quality randomized controlled trials, together with appropriate stratification strategies, should be used to verify the reproducible benefits of circadian interventions on clinical outcomes such as sleep, daytime functioning, and motor fluctuations. In addition, it should be examined whether these benefits are accompanied—within specific subgroups—by corresponding changes in indicators of disease progression, thereby providing a more actionable, iterative, and time-specific intervention approach for the comprehensive management of PD (Fig. 4). Circadian-informed care framework for Parkinson’s disease across disease stages. This schematic diagram has established a treatment framework based on circadian rhythms for different stages of Parkinson’s disease. Connecting the wearable rest–activity rhythm (RAR) indicators in the real world (RA/IS/IV) with risk stratification (iRBD, DAT-SPECT), complementary phase assessment (melatonin/DLMO), and targeted intervention measures (melatonin/agonist, phototherapy, lifestyle). Outcomes are tracked in parallel as symptoms vs rhythm endpoints, highlighting that sleep improvement ≠ circadian alignment. Created with BioGDP.com228. Conclusion This review indicates that circadian rhythm disruption is bidirectionally associated with multiple pathological and clinical features of PD. Circadian abnormalities may serve as potential biomarkers for early diagnosis and disease monitoring, and may provide a reference framework for phased assessment and intervention based on circadian rhythms. Future studies should further elucidate the complex mechanisms linking circadian rhythms and Parkinson’s disease, clarify whether these abnormalities reflect early pathological changes, subsequent disease progression, or feedback amplification at different stages, and evaluate the potential effects of circadian-targeted interventions on circadian parameters, symptom burden, and related pathological processes. Overall, these insights provide a scientific foundation for circadian-based strategies in the prevention and treatment of PD, with the potential to optimize patient management and improve their quality of life. Acknowledgements This work was supported partly by the National Natural Science Foundation of China (81271410, 32161143021), Henan Provincial Science and Technology Key Projects Program (252301420027, 252102311160), Henan Provincial Medical Science and Technology Key Projects Program (SBGJ202402075). Author contributions M.Z. and Y.X. conceptualized the title, prepared the initial draft, and designed the figures. Y.L., J.W., K.H. and Z.A. helped in drafting the manuscript, revised the manuscript, and prepared the final draft. Y.X. and J.W. helped in preparing the final draft, critically revised the manuscript, and supervised the project. All authors have read and approved the manuscript. Data availability No datasets were generated or analysed during the current study. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Yuxiang Xu, Email: xuyx@henu.edu.cn. Jianshe Wei, Email: jswei@henu.edu.cn. References Associated Data Data Availability Statement No datasets were generated or analysed during the current study.