Synapses serve as the fundamental components facilitating neuronal communication and memory retention. The process of synaptic pruning, which involves the elimination of superfluous synaptic contacts, plays a critical role in the CNS, ensuring the proper development of neural circuits during the course of neural maturation and sustaining synaptic stability in adulthood. The establishment of accurate synaptic connections is of utmost importance for the optimal functioning of the brain. In the brain, there exists an activity-dependent mechanism for synaptic pruning: neurons selectively stabilize active synapses while getting rid of inactive ones, ensuring optimal synaptic connections throughout brain development [27]. Dysregulation of synaptic pruning is believed to lead to the onset and progression of various psychiatric disorders, such as depression and neurodegenerative diseases including Alzheimer's disease (AD) [28]. Additionally, abnormal synaptic pruning is also considered a mechanism contributing to CNS sensitization and the onset of chronic pain [29].

Microglial cells, as a subset of resident mononuclear phagocytes in the CNS, serve as the principal immune cells. They are extensively distributed throughout the brain and spinal cord, contributing to the maintenance of normal brain functions through the release of inflammatory cytokines, phagocytosis of apoptotic cells, synaptic pruning, regulation of synaptic plasticity, and formation of neural networks [30]. In the brain, astrocytes represent the predominant neuroglial cell type and play a vital role in directing synapse formation, plasticity, and restructuring to regulate neuronal functions [31]. Studies have shown that during brain development, the synaptic pruning directed by microglial and astrocytic cells is crucial in establishing the right neural connections. These cells remove weaker synapses to form the appropriate neural pathways, however, in pathological states, glial cells can engulf an excessive number of synapses, resulting in abnormal synaptic loss [32]. The phagocytic and chemotactic functions of microglial cells play a crucial role in synaptic pruning. Studies have shown that mice lacking the microglial phagocytic triggering receptor expressed on myeloid cells-2 (TREM2) [33] or the chemotactic factor C-X3-C Motif Chemokine Receptor 1 (CX3CR1) demonstrate an excessive presence of synapses in their brains [34]. Additionally, astrocytes significantly contribute to the process of synaptic pruning. Research findings have demonstrated that astrocytes utilize multiple EGF like domains 10 (MEGF10) and tyrosine-protein kinase Mer (MERTK) during the development of the retinal newborn system to facilitate synaptic elimination. However, when MEGF10 or MERTK is knocked out in astrocytes of mice, there is a notable decrease in their capacity to engulf surplus synapses by approximately 30% or 50%. This suggests that astrocytes actively contribute to the promotion of synaptic pruning by participating in the elimination process [35]. Additionally, in the adult hippocampal CA1 region, astrocytes play a more prominent role than microglia in eliminating both excitatory and inhibitory synapses through MEGF10-mediated continuous engulfment of excitatory synapses [36]. Hence, through phagocytic molecular mechanisms, various types of glial cells selectively remove synapses, adjusting their numbers.

In the brain development process, the complement system acts as an intermediary for glial cell phagocytosis and plays a role in synaptic pruning [37]. The central mechanism for synaptic pruning in the developing mammalian brain is now being attributed to the complement-dependent synaptic pruning by glial cells [28]. In 2007, a study initially demonstrated that synaptic pruning during development is mediated by complement C3 [38]. Further investigations have since indicated that the process of complement-mediated synaptic pruning involves the engulfment of weaker synapses labeled with C3 complement by microglial complement receptor 3 (CR3). The classical complement cascade pathway, facilitated by microglia, plays a crucial role in the non-autonomous mechanism of glial cell-mediated synaptic pruning during developmental synapse pruning [37]. Furthermore, the SRPX2 protein, which is expressed in neurons, functions as a complement inhibitor and has the ability to bind directly to C1q, thereby inhibiting its activity and regulating the process of complement-dependent synapse elimination. And mice with a knockout of SPRX2 [39], as well as mice lacking the microglial phagocytosis inhibitory factor CD47 [40], both exhibited a reduction in synaptic density. These findings provide compelling evidence that complement-dependent synaptic pruning, mediated by microglial cells, constitutes a fundamental mechanism in brain development. Moreover, the modulation of synaptic pruning is directly influenced by astrocytes and microglia through complement crosstalk. This is due to the presence of complement C3 in astrocytes and its cleavage fragment C3a receptor C3aR in microglia, which enhances the release of C3 and activates the microglia, thereby exacerbating abnormal synaptic pruning [41]. Additionally, in mice models of AD, astrocytes eliminate synapses in a C1q-dependent manner, resulting in pathological synaptic loss. Furthermore, astrocyte phagocytosis can compensate for the dysfunction of microglial phagocytosis [32].

Complement C1q serves as a pivotal mediator connecting microglia and synapses, thereby initiating the process of synaptic phagocytosis [42]. In mice and adult individuals, complement C1q is predominantly expressed within the brain and microglia [43]. In the CNS, C1q has been demonstrated to be upregulated as an initial response to injury and functions as the initiating protein for the classical complement cascade, leading to the generation of MAC, anaphylatoxins, and opsonins [44]. C1q plays a role in guiding synaptic pruning through glial cells. Studies show that under high-resolution microscopy, C1q co-labels with pre-synaptic or post-synaptic proteins [38]. Proteomic studies found that synapses marked with C1q have elevated levels of apoptotic indicators casepase-3 and membrane protein-5 in comparison to synapses labeled negatively for C1q, indicating that synaptic pruning associated with C1q exhibit mechanisms akin to apoptosis [45]. Synaptic pruning by complement C1q are typically associated with complement C3, its split fragments C3a and C3b, and the presence of the complement receptor CR3 on microglia. C1q, predominantly secreted by microglia, binds to target synapses and facilitates the cleavage of complement C3 on the synaptic surface into C3a and C3b. C3b is recognized as the "eat-me" signal, while its breakdown product, iC3b, is recognized by microglia's CR3, leading to microglia-mediated synaptic pruning in the affected brain region [46]. In mouse models of AD, an increase in C1q levels has been observed to be associated with synaptic specificity. The inhibition of C1q or the removal of C3 or CR3 has been found to result in a decrease in the number of microglia involved in phagocytosis, synaptic loss, and improvements in learning and memory-related tasks [47]. In the context of the developmental visual system, the synaptic pruning is regulated by components of the classical complement cascade. This process involves the activation of C1q, the synaptic pruning through complement fragment C3b, and the interaction with complement receptor CR3 present on microglia, which directly engulfs synapses expressing C3b [10].

The C1q/C3-CR3 co-mediated microglial synaptic pruning of the complement cascade pathway is a crucial factor in the pathogenesis of chronic pain and depression. Wang et al. [48] utilized the chronic restraint stress (CRS) methodology to establish a mouse model of depression. In the amygdala of the depressive mouse model, there was a significant increase in neuroinflammation and the expression of complements C1q and C3. This led to the activation of microglial cells, resulting in a decrease in synaptic content in the amygdala, as evidenced by reduced levels of Syn and PSD95 proteins. The aforementioned alterations resulted in the manifestation of behaviors resembling depression in mice. Notably, the absence of the C1q gene impeded the synaptic loss and depressive behaviors induced by CRS in mice, thereby illuminating the significant involvement of C1q/C3-mediated microglial activation and synaptic pruning in the mechanisms underlying the onset of depression [48]. This was observed specifically in the hippocampal CA1 region, leading to ameliorated depressive symptoms in mice with Parkinson's disease. Moreover, within the context of the Parkinson's depression model in mice induced by reserpine, the administration of botulinum neurotoxin A (BonT/A) resulted in a notable decrease in the expression levels of complement C1q and C3 proteins in the hippocampus of the depressed mice. This reduction was accompanied by a downregulation of microglial CR3 mRNA levels. Following the treatment, the markers Iba1 and CD68 in hippocampal microglia exhibited a significant decrease, suggesting a decrease in microglial activation. This treatment resulted in the restoration of synaptic density, as evidenced by the increased co-localization of the excitatory presynaptic vesicle protein vesicular glutamate transporter-2 (VGLut2) with PSD-95, and the decreased fluorescence co-localization of VGlut2 with Iba-1 and CD68, indicating a reduction in microglial phagocytic activity. Furthermore, the expression of pro-inflammatory factors, tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β), in microglial cells decreased, indicating the inhibitory effect of BonT/A on the C1q-C3/CR3-mediated complement cascade signaling pathway in the hippocampal CA1 region and subsequent alleviation of depression in Parkinson's mice [49]. In mouse models of visceral chronic pain induced by chronic stress, there was a significant increase in the expression levels of complement C1q and integrin alpha M (ITGAM) mRNA (which encodes the CD11b subunit of CR3) within amygdala microglial cells. Additionally, a heightened co-localization of C1q with the microglial activation marker Iba1 was detected through enhanced immunofluorescence. The utilization of Iba-1 to label the postsynaptic protein PSD95 demonstrated a significant increase in synaptic modifications by microglial cells. However, the use of CR3 antagonists could mitigate this synaptic pruning and ameliorate visceral hypersensitivity and pain [50]. In essence, the complement cascade pathway C1q/C3-CR3 likely plays a role in the intertwined pathogenic mechanisms of chronic pain and depression. This occurs by mediating synaptic pruning primarily within the amygdala and hippocampal microglial cells (refer to Fig. 1 and Table 1 for a summarized overview).

Complement component C3, which plays a crucial role in the classical, alternative, and lectin pathways of complement activation, occupies a central position in the activation process, rendering it a pivotal focus for therapeutic interventions [51]. C3a, an activation fragment produced upon C3 activation via the three complement pathways, can be released by astrocytes and bind to the G-protein coupled transmembrane receptor C3aR on microglia, thereby exerting an influence on the functionality of these microglial cells. Consequently, it is plausible that the C3/C3a-C3aR pathway facilitates interactions between astrocytes and microglia [52]. Multiple studies have demonstrated that the activation of C3/C3a-C3aR signaling plays a significant role in the development and advancement of chronic pain and depression in mice exposed to chronic unpredictable mild stress (CUMS) or lipopolysaccharide (LPS)-induced depression. In an attempt to investigate this phenomenon, Li et al. [53] conducted an experiment wherein human umbilical cord mesenchymal stem cells (hUC-MSCs), known for their immunomodulatory and anti-inflammatory properties, were administered to mice subjected to the CUMS model. The findings of this study revealed that the administration of either hUC-MSCs or C3aR antagonists effectively ameliorated depression and anxiety-like behaviors induced by CUMS, and reduced the M1 polarization of microglia and pro-inflammatory cytokine levels (TNF-α and IL-10) in hippocampal neurons, and the activation of the complement pathway was due to the activation of C3a in astrocytes, which further activated the C3aR receptor on microglia, promoting its M1 polarization process [53]. Zhang et al. [25] conducted a study on mice with depression induced by CUMS or LPS. They found that the C3/C3aR pathway showed an early upregulation in the PFC, suggesting its potential as an early indicator of depression onset. Additionally, blocking the C3/C3aR signal not only alleviated depressive-like behaviors in mice but also suppressed the translocation of microglial signal transducer and activator of transcription 3 (STAT3) and the expression of inflammatory cytokines in the PFC. This blockade facilitated synaptic phagocytosis, highlighting its therapeutic potential in depression treatment. Furthermore, the administration of an IL-1R antagonist resulted in a noteworthy reduction in the NF-κB pathway and C3 levels within the PFC of mice subjected to CUMS. This blockade also impeded the activation of the IL-1R/pNF-κB/C3 pathway in neurotoxic A1 astrocytes [25]. Gynostemma pentaphyllum, a commonly employed herbal medicine, has been scientifically validated to possess potent anti-inflammatory properties through its saponin extract. Cheng et al. [54] conducted a study that demonstrated the ability of Gypenoside XVII, an active component of Gynostemma pentaphyllum, to inhibit the C3/C3aR/STAT3/cytokine signal pathway in the PFC of mice. This inhibition resulted in a decrease in microglia activation and a reduction in excessive synaptic pruning in the PFC of mice subjected to CUMS, leading to antidepressant effects. These findings suggest that the activation of abnormal synaptic pruning in depression, mediated by astrocytes-microglia, is facilitated through the C3/C3a-C3aR complement signaling pathway. Moreover, astrocytes-microglia are also implicated in the development of chronic pain through the C3/C3a-C3aR complement pathway. Previous studies have demonstrated abnormal activation of complement C3 in the dorsal horn of the spinal cord in rats with CCI, which contributes to the initiation of NP and the development of pain hypersensitivity [55]. In a subsequent investigation conducted by Mou et al. [56], it was observed that C3/C3aR levels were elevated in the dorsal horn of rats in the CCI model. Furthermore, the administration of C3 or C3aR antagonists via intrathecal route resulted in a reduction in M1 polarization in microglia and alleviated symptoms of NP. Zhu et al. [57] conducted a study in which they utilized thoracotomy and intraperitoneal injection of LPS to induce the expression of neurotoxic A1 astrocytes in rats. Through intrathecal injection, they antagonized the C3aR in microglia and observed that the downregulation of C3aR inhibited the activation of LPS-induced neurotoxic A1 astrocytes and M1 polarization of microglia. Additionally, this downregulation increased the presence of A2 astrocytes and resulted in a reduction in both the mechanical pain threshold and the incidence of chronic pain in mice. In light of these findings, it is evident that the C3/C3a-C3aR complement pathway is central to the molecular crosstalk and synaptic pruning between astrocytes and microglia. This interaction likely plays a vital role in the combined mechanisms underpinning chronic pain and depression (Fig. 2 and Table 1).

Neuroinflammation refers to the immune response occurring in the CNS that is initiated by microglia and astrocytes. This response is characterized by the activation of resident glial cells, including microglia and astrocytes, the release of cytokines and chemokines, and the activation and infiltration of leukocytes [58]. Several studies have suggested that the cerebral cortex and its subdivisions, such as the PFC, anterior cingulate cortex, amygdala, hippocampus, and median raphe nucleus, play a crucial role in the co-occurrence of pain and depression. The coexistence of chronic pain and depression is associated with central inflammation, specifically involving TNF-α, L-1β, and IL-6, as well as variations in peripheral cortisol levels [59]. Additionally, persistent neuroinflammation plays a vital role in the initiation and perpetuation of concurrent chronic pain and depression [60]. Consequently, targeted interventions aimed at reducing neuroinflammation have the potential to enhance the management of comorbid depression and chronic pain [61].

In the CNS, the activation of complement plays a crucial role in inducing inflammatory responses within immune effector cells, thereby protecting neurons from potential hazards and toxins [62]. The classical complement pathway, which encompasses essential components such as C1q, as well as the resultant activation byproducts C3a and C5a and their corresponding receptors, serves as the foundation for the complement-driven inflammatory and noninflammatory processes within the brain [63]. Additionally, the activation of glial cells serves as a significant mediator of complement-induced inflammation in the CNS. In response to inflammatory challenges, neurons within the CNS produce various components, including C1r, C1s, C4, C2, and C3, which contribute to the formation of the C1 complex (consisting of C1q, C1r2, and C1s2). Subsequently, C3 is cleaved into C3b/iC3b, which activates microglial cells through the CR3 receptor, leading to synaptic alterations. Following C3 cleavage, C5a is generated along with C3b and C3a, which further promote inflammation-associated microglial and astrocytic cells to increase neurotoxicity, ultimately resulting in the impairment of neuronal function and eventual neuronal death [64]. The findings of this study indicate that in the LPS-induced mouse model of neuroinflammation, there is an upregulation of complement C1q levels in the mouse hippocampus. This upregulation is accompanied by an increase in synaptic phagocytosis by microglial cells, resulting in a higher loss of synapses and subsequent cognitive impairment in the mice. However, it was observed that neutralizing C1q signaling can prevent these changes, suggesting that activated microglial cells and the complement cascade C1q signaling may play a role in the synaptic loss and cognitive impairment observed in the LPS-induced neuroinflammation mouse model [65]. In another study involving neurons and glial cells stimulated by LPS, the levels of C3 were also significantly increased, which can enhance LPS-induced neuroinflammation and neurodegeneration through the Mac1/NOX2 pathway [66]. To conclude, the activation of glial cells via the complement cascade is crucial in the pathogenesis of central neuroinflammation.

The interaction between C1q and microglia, as a key activator of the classical complement pathway, has been demonstrated to effectively suppress neuroinflammation by inhibiting the secretion of pro-inflammatory cytokines [67]. This inhibitory effect of C1q on neuroinflammation is also implicated in the development of depression and chronic pain. In a study conducted by Madeshiya et al. [68], a mouse model of depression was established using the learned helplessness paradigm. Their findings revealed that the knockout of C1q intensified the learned helplessness behaviors induced by electric foot shocks in mice. Furthermore, there was a significant increase in the mRNA levels of pro-inflammatory cytokines TNF-α, IL-6, and IL-1β in the mouse PFC, along with an increase in M1-polarized microglia, suggesting that the absence of C1q worsened depressive-like behaviors in mice and is associated with induced neuroinflammation and M1 polarization of microglia. Moreover, studies that employed in vivo, nucleus-specific calcium signaling perturbation strategies, supplemented with gene mapping, bioinformatics, and functional analyses, identified that C1q promotes activity-dependent spinal remodeling and persistent pain hypersensitization in a manner related to inflammatory hypersensitivity reactions [69].

The cleavage products of the complement system, namely C3a and C5a, have been observed in various systems as mediators that elicit neuroinflammation. By means of their seven-transmembrane coupled receptors, C3aR and C5aR as well as C3a and C5a have the ability to stimulate glial cells to generate pro-inflammatory cytokines and participate in a range of diseases [70]. In general, microglial cells express C3aR and C5aR on their surface. Activation of these receptors can augment the chemotaxis of both immune and glial cells, leading to the production and release of inflammatory factors in a Ca2+-dependent manner [71]. The binding of C5a to C5aR has the potential to enhance the secretion of pro-inflammatory cytokines and intensify the activation of pro-inflammatory microglial cells via the p38 mitogen-activated protein kinase and extracellular signal-regulated kinase pathways [72].

The involvement of the C3a-C3aR and C5a-C5aR axis in chronic pain and depression can be attributed to their ability to induce central neuroinflammation. Prior research has demonstrated that C3a possesses the ability to enhance the calcium current in dorsal root spinal neurons and the intracellular calcium flow induced by capsaicin. Consequently, this mechanism serves to activate and sensitize pain receptors during the pain perception process [23]. Furthermore, investigations have indicated that the neuropeptide TLQP-21 Trifluoroacetate can elicit hyperalgesia and NP by activating the C3aR1 receptors located on the microglial cells of the dorsal spinal cord [73]. The findings from research conducted on the CUMS model demonstrate that the augmentation of the C3a-C3aR signaling pathway has the potential to induce M1 polarization of microglial cells, leading to heightened levels of pro-inflammatory cytokines, such as TNF-α and IL-1β, within the hippocampus. Consequently, this can result in the manifestation of anxiety and depressive-like behaviors in mice [53]. Additionally, another investigation proposes that within the same CUMS model, the inhibition of C3aR effectively diminishes the expression of NOD-like receptor protein 3 inflammasomes and the subsequent inflammatory response in the hippocampus. As a result, this intervention exhibits an improvement in depressive-like behaviors observed in mice [74]. Crider et al. [20] propose that in mice with a chronic stress model, monocytes expressing C3a-C3aR migrate into the PFC, leading to increased levels of the inflammatory cytokine IL-1β in the PFC. However, when C3aR is knocked out, the recruitment of PFC monocytes is significantly reduced, resulting in decreased IL-1β levels in the PFC and improved depressive-like behaviors. These findings suggest that C3a-C3aR may play a role in the development of chronic pain and depression by promoting neuroinflammation.

The activation of C5a-C5aR has been extensively investigated in the context of chronic pain. C5a, when bound to C5aR, exhibits potent inflammatory properties by inducing the production of chemotactic factors and pro-inflammatory effects. This subsequently leads to the upregulation of inflammatory cytokines such as IL-6, TNF-α, and PGE-2, while simultaneously reducing the secretion of anti-inflammatory factors, resulting in the manifestation of hyperalgesia [75]. Moreover, C5a has the ability to mediate NP at sites of nerve injury, within dorsal root ganglia (DRGs), and in the spinal cord. Specifically, at injury sites, C5a recruits macrophages and T cells through C5aR activation, thereby contributing to the development of neural pain. Additionally, C5a in DRGs may directly sensitize primary sensory neurons. In the spinal cord, C5aR is primarily expressed in neural glial cells, and these neural glial cells can be stimulated by C5a, leading to their activation [26]. Recent clinical studies have indicated a significant correlation between levels of C5a and C5 in both plasma and cerebrospinal fluid, and the manifestation of depression [11,76,77]. Consequently, it can be postulated that an elevation in plasma complement C5a concentration may serve as an indicator of a persistent inflammatory process, potentially resulting in compromised integrity of the blood-brain barrier (BBB). Subsequently, the infiltration of the peripheral complement system into the CNS may occur, thereby triggering the activation of central immune-inflammatory agents and exacerbating symptoms associated with depression [78]. Therefore, the complement C5a in plasma may penetrate brain tissues a few hours after BBB leakage, inducing neuroinflammation, which affects the CNS [79]. To conclude, the complement pathways C3a-C3aR and C5a-C5aR may participate in the comorbidity mechanisms of chronic pain and depression through inducing central neuroinflammation (Fig. 3 and Table 2 for summary).