Introduction to neuroplasticity

The brain comprises complex networks of neurons and synapses that work together to perform a wide range of functions, from regulating automatic processes to more advanced cognitive tasks. These circuits involve precisely controlled steps, ranging from neurogenesis to cell migration, differentiation, neuritogenesis, target identification, and synapse formation, elimination, and stabilization. Over time, these connections may be strengthened or weakened, leading to the formation of a highly efficient neural network that is essential for proper brain function [1,2,3].

Neuroplasticity can be defined as the nervous system’s ability to change its activity in response to intrinsic or extrinsic stimuli by reorganizing its structure, including alterations in neuronal networks involving changes in neuronal connectivity and function [4, 5]. Studies into neuroplasticity have demonstrated its involvement during development and in adulthood, in response to the environment, in learning and memory, and in recovery from brain lesions [6,7,8]. There are different forms of neuroplasticity in adults, acting by promoting neurogenesis, synaptogenesis, and synaptic and neurochemical plasticity [9].

The most appealing phenomenon of neuroplasticity appears to be adult neurogenesis, a process defined as the birth of new neurons in the adult brain. Neurogenesis comprises a series of steps that can be examined separately: proliferation, differentiation, migration, and cell survival [1, 2, 10, 11]. There is ongoing investigation into neurogenesis in various brain regions, including the striatum, cerebral cortex, amygdala, hypothalamus, hippocampus, and the lateral ventricles [12,13,14,15]. However, two well-researched areas where neurogenesis takes place are the subventricular zone (SVZ) along the lateral ventricle and the subgranular zone (SGZ) in the dentate gyrus (DG) of the hippocampus [1, 2, 16].

In the SVZ, a distinct subpopulation of multipotent cells characterized by their expression of glial fibrillary acidic protein (GFAP) and radial glial-like features, known as type B cells, are present. Following asymmetric division, these cells generate type C cells, also known as transit-amplifying cells, which undergo migration along the rostral migratory stream towards the olfactory bulb in mice, generating type A cells, or neuroblasts. Upon reaching their destination, these cells differentiate into local interneurons, specifically granule and periglomerular cells [5].

In the DG region of the hippocampus, neural stem cells (NSCs) are called radial glial-like cells (also known as R, RGL, or type 1 cells). In the adult hippocampus, most R cells are quiescent and share many markers expressed by NSCs in the embryo, including Nestin, GFAP, and sex-determining region Y-box 2 (Sox2). When activated, significant changes occur in their gene expression and cellular metabolism. R cells can either symmetrically duplicate, leading to the expansion of the R cell population, or they can divide asymmetrically. In the latter case, they give rise to a self-renewed R cell and a non-radial glia-like progenitor cell (also known as NR, type 2 cells, or IPCs: intermediate progenitor cells). Alternatively, R cells can generate two NR cells directly, neurons, or rarely astrocytes. IPCs are highly proliferative amplifying stem cells expressing the T-box brain protein 2 (Tbr2) and give rise to bipolar neuroblasts, which express doublecortin (DCX) and polysialylated-neuronal cell adhesion molecule (PSA-NCAM). Neuroblasts migrate tangentially along the SGZ before migrating short distances radially into the granule cell layer, turning into prospero homeobox protein 1 (Prox1)-positive neurons forming dendritic spines and synapses, and extending functionally connected axons before integrating into the hippocampal circuitry. These new integrated neurons express the neuronal protein NEUN and subsequently influence brain function [17, 18].

Synaptogenesis, another form of neuroplasticity, occurs continuously throughout an individual’s lifespan. It involves the maintenance of synapses, encompassing both axons and dendrites. The synapse formation is not limited to early development but also occurs in adult brains. This process plays a significant role in learning and memory [8, 19]. More than 90% of all excitatory synapses in the central nervous system (CNS) occur in structures called dendritic spines, which are small bumps that emerge from the surface of many neurons [20]. Dendritic spines are highly dynamic, and their stabilization and morphology are influenced by synaptic activity. This extrinsic regulation of spinal morphogenesis supports the development of an experience-dependent brain and the storage of information within the brain’s circuits [21, 22].

Furthermore, synaptic plasticity encompasses the alterations that occur at the synaptic level, specifically focusing on the mechanisms involving the trafficking and insertion of ionotropic glutamate receptors, N-methyl-D-aspartate receptor (NMDAR), and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPAR), into the synaptic membrane during long-term potentiation (LTP) and long-term depression (LTD). Additionally, synaptic plasticity involves neurochemical changes that influence neurotransmission by modulating the levels of transmitters, receptors, or transporters. These modifications ultimately impact neuronal signaling and contribute to the dynamic nature of synaptic function [23].

A comprehensive understanding of the regulatory mechanisms underlying neuroplasticity holds vital importance for future therapeutic interventions. Pathological alterations in synaptogenesis and synaptic plasticity are fundamental factors in numerous neurodegenerative disorders, including depression, Alzheimer’s disease, Parkinson’s disease, chronic pain, and addiction. Conversely, neuroplasticity also plays a role in brain recovery, with the potential to manipulate specific neuronal pathways and synapses, leading to significant implications for therapeutic and clinical interventions in neurodevelopmental and neurodegenerative pathologies. Consequently, the investigation of neuro-modulating molecules is of great value in the prevention and treatment of diseases. Certain molecules, such as purines, have been recognized for their regulatory role in influencing neurogenesis, synaptogenesis, and synaptic plasticity. The role of purinergic signaling and receptors during brain development has already been reviewed elsewhere [3, 24]. Additionally, the significance of purine nucleotide receptors in the adult brain has also been addressed in previous reviews [25, 26]. Here, we aim to elucidate extensively the role of purine nucleosides in neuroplasticity processes in the adult brain.

Purinergic signaling

Purines are a class of aromatic organic molecules essential to all cells. They form the structural basis for nucleic acid synthesis, function as second messengers, and are components of some coenzyme structures. Adenine nucleotides play a fundamental role in intracellular energy metabolism, while guanine nucleotides modulate G-protein activity.

Purines consist of nitrogenous bases derived from adenine and guanine. These include nucleotides with one or more phosphate groups, such as adenosine-5′-triphosphate (ATP), adenosine-5′-diphosphate (ADP), adenosine-5′-monophosphate (AMP), guanosine-5′-triphosphate (GTP), guanosine-5′-diphosphate (GDP), and guanosine-5′-monophosphate (GMP), as well as the nucleosides adenosine (ADO), inosine (INO), and guanosine (GUO). Furthermore, purines include the nitrogenous bases resulting from their catabolism, such as hypoxanthine, guanine, xanthine, and uric acid [27].

The purinergic transmission was first proposed by Burnstock in 1972 when ATP released extracellularly was attributed a role as a neurotransmitter. Ever since, ATP can be released alone or co-released with glutamate, γ-aminobutyric acid (GABA), dopamine, noradrenaline, 5-hydroxytryptamine, and acetylcholine in the CNS [28, 29]. Later, it was shown that GTP can also be co-released by exocytosis, including from ATP-containing synaptic vesicles, indicating that this nucleotide may have a role in neurotransmission [30].

Following their release, the ectonucleoside triphosphate diphosphohydrolase (E-NTPDases or apyrase/CD39) enzymes play a crucial role in cleaving triphosphorylated nucleotides (such as ATP and GTP) as well as diphosphorylated nucleotides (including ADP and GDP). Additionally, ectonucleotide pyrophosphatase/phosphodiesterases (E-NPPs) contribute to this process by hydrolyzing 5′-phosphodiester bonds in nucleotides, leading to the direct release of monophosphorylated nucleotides (AMP and GMP). Subsequently, these monophosphorylated nucleotides are further metabolized by ecto-alkaline phosphatase or ecto-5′-nucleotidase/CD73, resulting in the conversion of AMP into the nucleoside adenosine, and GMP into the nucleoside guanosine [31]. Adenosine may be metabolized through deamination by adenosine deaminase (ADA), generating another nucleoside, inosine (INO), or rephosphorylated to AMP by adenosine kinase.

The purine nucleosides are taken up by nucleoside transporters, and nucleoside and nucleotide levels can be regulated by the intracellular pathway. Adenosine can form INO, and then it is metabolized by the action of the enzyme purine nucleoside phosphorylase (PNP) producing hypoxanthine, which is converted to xanthine and uric acid by xanthine oxidase. GUO is converted to guanine by PNP, which is then deaminated by guanine deaminase producing xanthine and then uric acid. Additionally, the salvage purine pathway enzyme hypoxanthine–guanine phosphoribosyl transferase (HGPRT) can produce GMP or IMP from the condensation of guanine or hypoxanthine with 5′-phosphoribosyl, respectively [32].

Thus, besides intracellular nucleotides metabolism, the enzymatic activity of ectonucleotidases regulates nucleotides and nucleosides levels in the extracellular milieu and can activate a variety of purinergic receptors with different functions in the cell, which demonstrates the importance of these enzymes in purinergic modulation [33].

Adenine-derived purines have been found to exert diverse effects mediated through interactions with purinergic receptors, specifically P1 and P2 receptor types. P2 receptors are subdivided into P2X ligand-gated ion channel receptors responsive to ATP, comprising seven subtypes (P2X1-7); and P2Y are G-protein-coupled receptors (GPCR), comprising eight subtypes (P2Y1, P2Y2, P2Y4, P2Y6, P2Y11, P2Y12, P2Y13, P2Y14), responsive to ATP (P2Y11), ADP (P2Y1, P2Y12, and P2Y13), UTP (P2Y2 and P2Y4), UDP (P2Y6), and UDP-sugars (P2Y14) [34, 35].

The P1 receptors are metabotropic receptors (GPCR) that primarily respond to ADO as their main agonist. However, certain receptors within this family can also be modulated by the nucleosides INO and GUO [36,37,38,39].

The P1 family comprises four receptors: A1, which is expressed throughout the CNS; A2A, found in the striatum, nucleus accumbens, hippocampus, and cerebral cortex; A2B, with low expression in the brain; and A3, moderately expressed in the cerebellum and hippocampus [34, 40]. A1 receptor (A1R) and A3R are coupled to Gi/o protein and inhibit adenylyl cyclase, resulting in decreased cyclic AMP (cAMP) levels. On the other hand, A2AR and A2BR are coupled to Gs protein and activate adenylate cyclase, leading to increased cAMP levels [34].

The A1R plays a significant role in modulating excitatory transmission at both pre- and post-synaptic sites. On the other hand, adenosine A2A receptors (A2AR) are known to fine-tune synaptic plasticity, primarily by regulating the release of glutamate from presynaptic terminals [41], and are involved in regulating the uptake of both glutamate and GABA by astrocytes [42, 43]. Furthermore, studies have demonstrated the formation of oligomeric receptor complexes involving adenosine receptors. These complexes can involve homomeric receptor-receptor interactions between A1R-A1R or A2AR-A2AR, as well as heteromeric interactions between A1R and A2AR (besides interacting with GPCRs responsive to other ligands). The A1R-A2AR heteromer has been characterized in both heterologous expression systems and native tissue, and it has been shown to control synaptic glutamate release [44,45,46].

The remaining purine nucleosides, GUO and INO, are considered orphan ligands, i.e., specific receptors for these nucleosides have not yet been identified. However, studies have demonstrated that GUO exhibits a preferential binding site on rat brain membranes [47], besides interacting with a GPCR which binding seems not to be displaced by adenosine [48]. Although no specific receptor for GUO has been clearly identified or cloned thus far, several lines of evidence suggest a potential multi-target action, involving interaction sites on adenosine receptors, glutamate receptors, and potassium channels [49], a topic that warrants further evaluation.

There is evidence of GUO acting on the adenosinergic system by increasing the amount of adenosine available in the extracellular environment and binding with low affinity to adenosine receptors [27, 50,51,52]. GUO treatments exhibit a wide range of neuroprotective effects, many of which are abolished by selective adenosine receptor antagonists, including A1R and A2AR [49, 53,54,55,56].

Furthermore, it was recently shown that GUO may interact with A2AR when it is co-expressed with A1R, pointing to a putative interaction within the A1R-A2AR heteromer [56]. As for INO, it has been shown to display a higher affinity for A3R but can also bind to A1 and A2A receptors [57].

Both naturally occurring nucleosides are hydrophilic and do not readily cross bilayer membranes except by mediated or active transport. The currently known nucleoside transport activities are of two kinds: Na+-independent equilibrative nucleoside transporters (ENTs), mediating bidirectional transport according to the gradient, and Na+-dependent unidirectional concentrative nucleoside transporters (CNTs), which are capable of transporting nucleosides against an intracellular gradient of the nucleoside itself [58].

Purinergic signaling is essential in a variety of molecular mechanisms and signaling pathways that influence brain function. In this review, we will investigate the potential of purine nucleosides, specifically ADO, INO, and GUO, in modulating neuroplasticity in the adult brain.

The role of purinergic nucleosides in the neuroplasticity

Purinergic signaling plays an important role in mechanisms involved in neuroplasticity [59]. Nucleosides naturally occur in the extracellular space at low levels, but during challenging metabolic conditions such as injury, ischemia, and inflammation, their extracellular concentrations significantly increase, sometimes reaching millimolar levels. This increase is believed to act as a protective mechanism for neighboring nerve cells [50, 60].

Initially, studies exploring the protective or regenerative effects of nucleosides like guanosine (GUO) and inosine (INO) proposed that their primary role was intracellular, serving as a replenishment source for ATP levels in cells [61, 62]. However, subsequent research has demonstrated the neuroprotective effects of these nucleosides through the activation of adenosine receptors [63,64,65].

Additionally, research indicates that nucleosides, in combination with growth factors and other signaling substances, could synergistically enhance neuroplasticity and potentially exert combined control over adult neurogenesis by integrating intracellular signaling pathways [66, 67].

Effects of adenosine

ADO plays an important role as a neuromodulator within the CNS. Extensive research has revealed that ADO, as well as the manipulation of its receptors through pharmacological means, can effectively regulate neuroplasticity across various stages of neurogenesis, neuritogenesis, synaptogenesis, and synaptic plasticity (Table 1). In the following section, we will present studies that highlight the effects of adenosine on all these processes.

Table 1 Effects of adenosine on neuroplasticity
Full size table

The in vitro evaluation of neurogenesis in neurospheres derived from the SVZ of adult rodents showed that adenosine incubation decreased the number of transit-amplifying cells and immature-differentiating neurons. Moreover, specific activation of A1R using the N6-Cyclopentyladenosine (CPA) agonist reduced the population of immature-differentiating neurons. This suggests that A1R activation specifically inhibits the differentiation process from transit-amplifying cells to neuroblasts [68]. Additionally, A1R activation promotes the proliferation of NSCs through the activation of the mitogen-activated protein kinase kinase (MEK)/extracellular signal-regulated protein kinase (ERK) and protein kinase B (Akt) pathways [69].

In similar studies using neurospheres generated from the SVZ of adult rats and mice, ADO was found to inhibit neuronal differentiation at concentrations of 10, 50, or 100 µM [67, 68]. However, at a concentration of 1 µM ADO, conflicting results were observed. Benito-Muñoz and colleagues [68] reported inhibition of neuronal differentiation, while Mishra and colleagues [67] observed the proliferation of neuroprogenitor cells. This discrepancy could be attributed to the differences in the concentrations of trophic factors present in the culture medium, specifically 20 ng/ml epidermal growth factor (EGF) and 10 ng/ml fibroblast growth factor (FGF) in the former study, and 5 ng/ml EGF plus 2.5 ng/ml FGF-2 in the latter study. Therefore, it can be inferred that a low concentration of ADO combined with low concentrations of trophic factors may have a positive effect on neurogenesis, whereas a high concentration of ADO can inhibit this process.

In an in vivo protocol with adult rats, the intracerebroventricular administration of an A1R agonist (CPA, for 14 days) resulted in a decrease in the number of neuroblasts within the SVZ. This reduction was also evident in the olfactory bulb and was accompanied by the downregulation of genes associated with neurogenesis. Furthermore, the treatment with CPA stimulated astrogliogenesis, i.e., the generation of astrocytes [68].

Sleep deprivation is known to lead to an increase in intracellular levels of ADO, which in turn inhibits both adult neuronal cell proliferation and cognition. In adult rats subjected to 48 h of sleep deprivation, the administration of 8-cyclopentyl-1,3-dimethylxanthine (8-CPT), an adenosine A1R antagonist, helped to alleviate the decline in spatial reference memory. Additionally, 8-CPT treatment resulted in enhanced adult neuronal cell neurogenesis and proliferation, possibly due to the upregulation of brain-derived neurotrophic factor (BDNF) levels in the DG and CA1 regions of the hippocampus [70], which suggests a negative impact of A1R activation on neurogenesis.

Most studies suggest that A1R activation primarily affects neuronal differentiation, rather than influencing the proliferation of stem cells or neuroprogenitor cells. On the other hand, the impact of A2AR on the various stages of neurogenesis continues to be debated. Regarding the other adenosine receptors, specifically A2BR and A3R, they have garnered minimal focus in neurogenesis research and, as a result, are not well understood.

Activation of A2AR with CGS21680 in DG neurospheres obtained from early postnatal (P1-3) rats enhances the generation of newborn neurons without influencing overall cell proliferation. Furthermore, when combined with BDNF, A2AR activation amplifies the capacity of DG-derived cells to expand the proliferation pool of type 1 (radial glial-like cells) and type 2 (neuroprogenitor cells) [16]. Therefore, A2AR activation seems to promote neurogenesis.

Several studies using damage-inducing protocols have been used to study the effects of A2AR modulation on the neurogenic process. In mice with noise-induced hearing loss, a noticeable reduction in neurogenesis was observed in the DG. The A2AR agonist (CGS21680) effectively mitigated the detrimental impact on neurogenesis [71]. However, in a mouse model of alcohol addiction, the administration of an A2AR antagonist (ZM241385) resulted in decreased phosphorylation of ERK1/2 in the dorsal hippocampus. Additionally, this intervention increased the proliferation of neuroblasts within the dorsal hippocampus and improved memory function [72]. In an animal model of chemotherapy-induced cognitive impairment, the administration of the platinum-based chemotherapy cisplatin resulted in significant upregulation of A2AR and its downstream effectors, cAMP, and CREB in the hippocampus of adult mice. The blockade of A2AR using the antagonist (KW6002) effectively prevented cisplatin-induced deficits in neural progenitor proliferation and dendritic morphogenesis of adult-born neurons. This intervention also led to improvement in memory function and anxiety-like behavior without impacting brain tumor growth or the antitumor activity of cisplatin [73].

In humans, studies involving both cognitively normal and impaired individuals, gene-based association analysis revealed a significant association between the A2AR gene, ADORA2A, and hippocampal volume. Specifically, the minor allele of rs9608282 in ADORA2A was found to be associated with larger hippocampal volumes and better memory performance [74].

Beyond mammals, the purinergic system is involved in the proliferation of other organisms. For instance, in zebrafish, following hind fin amputation, adenosine has been shown to promote the proliferation of epidermal cells. The recruitment of progenitor cells in this process and additionally, the growth of sensory axons, are effects mediated through A2BR, highlighting its involvement [75].

These findings of ADO effect on neurogenesis involve a complex interplay of adenosine receptors, particularly A1R and A2AR. The activation of A1R appears to predominantly inhibit neuronal differentiation, as evidenced by a decrease in the number of neuroblasts and the promotion of astrogliogenesis. This effect is further corroborated by the negative impact of A1R activation on neurogenesis during conditions like sleep deprivation. On the other hand, A2AR activation seems to have a more supportive role in neurogenesis, enhancing the generation of new neurons and facilitating recovery in various damage-induced models. The role of A2AR in memory function and hippocampal volume in humans suggests its potential significance in cognitive health. However, there is a body of studies in which contrasting effects are observed with different adenosine concentrations and in varied experimental conditions, showing the need for further research to delineate more precise mechanisms and conditions under which adenosine receptors influence neurogenesis in both physiological and pathological states.

Neuroprogenitor cells play a critical role in the development of the nervous system. However, their transformation into mature neurons and their integration into the existing circuitry are essential for proper functioning. This integration hinges on the complex process of neurite formation, which serves as the structural foundation for neuronal connectivity and communication, to extend and establish intricate networks. Eventually, these networks give rise to the axon and dendritic arborization. Research on cell lines and neurospheres derived from the SVZ and DG has demonstrated that both P2 and P1 receptors can directly regulate neuritogenesis and neurite outgrowth, either independently or by modulating the actions of growth factors.

In the pheochromocytoma-derived PC12 cell line transfected with A1R and in primary cultures of cortical and hippocampal neurons, activation of A1R with CPA significantly suppressed nerve growth factor (NGF)-induced neurite outgrowth and promoted stress fiber formation [76]. However, SH-SY5Y neuroblastoma cells subjected to in vitro treatment with A1R (R-PIA) or A2AR agonist (CGS21680) exhibited notable morphological changes characterized by the emergence of neuritic processes. These findings show that activation of A1R or A2AR can induce neurite growth processes [77].

Impaired neuritogenesis, caused by p53 blockade, was effectively rescued by activating A2AR with CGS21680 in PC12 cells. This rescue mechanism was found to depend on the signaling of translin-associated protein X (TRAX) [78], a pathway known to be involved in synaptic plasticity and DNA repair [79]. Primary hippocampal cultured neurons from the DG of early postnatal (P1-3) rats showed increased dendritic branching following the application of an A2AR agonist (CGS21680). This observation highlights the role of A2AR in precisely modulating dendritic arborization [80]. Additionally, the stimulation of A2AR enhances NGF-induced neurite outgrowth in PC12 cells [81].

On the other hand, a study by Nakashima et al. (2018) revealed that the stimulation of A3R with the agonist 1-[2-chloro-6-[[(3-iodophenyl)methyl]amino]-9H-purin-p-yl]-1-deoxy-N-methyl-beta-D-ribofuranuronamide (2-Cl-IB-MECA), rather than A1R (CHA) or A2AR and A2BR (CGS21680 and BAY60-6583, respectively), promoted neurite outgrowth in rat retinal-derived ganglion cells. This effect was attributed to the increased phosphorylation of Akt, emphasizing the role of the Akt signaling pathway in A3R-mediated neurite outgrowth [82].

Together, these data demonstrate the role of adenosine receptors in regulating neuritogenesis. The suppression of neurite outgrowth by A1R activation contrasts with the promotion of neuritic processes by A2AR and A3R. Neuronal development and plasticity need the involvement of A2AR in dendritic branching and the enhancement of NGF-induced neurite outgrowth. Additionally, Akt phosphorylation may be the pathway in neuronal differentiation stimulated by A3R.

Adenosine receptors play a role in synaptic transmission by being present in synapses, where they actively modulate the release of various neurotransmitters like serotonin, acetylcholine, GABA, and glutamate. The literature extensively describes how ADO plays a role in reducing the magnitude of basal synaptic transmission through A1R activation in various brain regions [83,84,85,86].

In the absence of ambient adenosine action, an impairment was observed in both short-term plasticity and LTP. In hippocampal slices of rats and mice, blocking A1R (DPCPX), enzymatically degrading extracellular adenosine, or genetically deleting A1R resulted in a noticeable increase in mossy fiber synaptic responses [85].

In a comparative study examining the role of A1R in hippocampal synaptic plasticity, two age groups were analyzed: young rats (5–6 weeks) and old rats (2 years). Using an A1R antagonist (DPCPX), results indicated a mild excitatory effect on synaptic transmission that induced LTD and promoted depotentiation in both animal groups. These findings demonstrated that endogenous adenosine limits LTD and depotentiation in both young adult and aged rats through A1R activation [87].

In adult mice hippocampal slices, the A1R antagonist (DPCPX) was found to decrease the amplitude of LTP in the ventral hippocampus (VH) and significantly increase the magnitude of LTD in both VH and the dorsal hippocampus (DH). The study also investigated the involvement of A2AR, and the antagonist SCH58261 did not alter the magnitude of LTD in the VH but did induce the manifestation of LTP in the DH. These data highlight the role of these receptors in modulating synaptic strength in either the VH, associated with stress, emotion, and affection, or the DH, which is related to cognitive functions [83, 88].

Silencing the A2AR gene in adult mouse astrocytes leads to an increase in astrocytic arborization complexity while also causing deficits in spatial memory and compromising hippocampal synaptic plasticity. The effects include a reduction of LTP magnitude and a shift of LTD towards LTP. These findings suggest distinct roles for astrocytic and neuronal A2AR in synaptic plasticity [89].

Despite few studies evaluating the A3R, in adult rat hippocampus, activation of adenosine A3R by Cl-IBMECA induced LTP and attenuated LTD [90].

The modulation of synaptic responses by adenosine exhibits complex dynamics in specific brain regions. In a study exploring the connection between the olfactory bulb output and layer 1a of the anterior piriform cortex, using slices from adult mice, ADO at different concentrations (30 or 100 µM) demonstrated a more potent attenuation effect on low-frequency inputs compared to high-frequency inputs. As a result, this led to a decrease in the likelihood of neurotransmitter release and a significant attenuation of short-term depression mechanisms [86]

In the hypothalamus, using slices from 21–28-day-old hypocretin-EGFP transgenic mice, it was observed that ADO inhibits LTP in hypocretin neurons. However, when an A1R antagonist (8-CPT) was used, it facilitated LTP induction. This suggests that endogenous adenosine, acting through A1R, is crucial in suppressing excitatory synaptic transmission and LTP induction in hypocretin neurons [84]. Additionally, ADO reduced the amplitude of GABAergic inhibitory post-synaptic currents (IPSCs), an effect blocked by the A1R antagonist (CPT) in these neurons [91].

The analysis of corticostriatal synapses in parasagittal brain slices of BAC transgenic mice revealed that the coactivation of FGF and A2AR (CGS21680) facilitates the induction of LTP in striatopalidal medium spiny neurons. Furthermore, this study also demonstrated the coactivation of FGF and A2AR-induced neuritogenesis in PC12 cells and increased the density of dendritic spines in primary cultured neurons [92].

A recent finding suggests that the absence of A2AR may lead to alterations in synaptic protein expression, pointing to potential disturbances in synaptic communication and neuronal function. The A2AR knockout (KO) mice exhibited significant impairment in working memory, spatial memory, and emotional learning, accompanied by a reduction in the production of newborn neurons. An abnormal increase in synaptophysin protein levels, observed both pre- and postsynaptically, was found in the hippocampus and prefrontal cortex (PFC) of these mice. In contrast, a decrease in SNAP25 protein levels, a marker for presynaptic proteins, was observed in these brain regions [93].

In conclusion, this body of research provides evidence of how adenosine receptors modulate synaptic transmission and plasticity in various brain regions. For A1R, there is evidence that it reduces basal synaptic transmission and affects plasticity, with its antagonism causing varied synaptic responses. Then, for A2AR, its role is complex: silencing it in astrocytes impacts memory and synaptic plasticity while activating it influences LTP and dendritic spine density. Notably, A2AR deficiency leads to altered synaptic protein expression and cognitive impairments. Finally, for A3R, which has been less studied, it appears to induce LTP and attenuate LTD in the rat hippocampus (Fig. 1).

Fig. 1
figure 1

Schematic summary of adenosine (ADO) effects on adult brain neuroplasticity. ADO modulates the proliferation of Neural stem cells (NSCs) and Intermediate progenitor cells (IPCs); promotes neurogenesis, synaptogenesis and modulates synaptic plasticity. The effects of ADO on neurogenesis and synaptic plasticity rely on ADO levels and involve a complex interplay of adenosine receptors, particularly A1R and A2AR. Different studies obtained conflicting results depending on the protocols applied, which resulted in similar effects with A1R or A2AR agonists and antagonists. Detailed effects and references are outlined in Table 1. The figure was created using BioRender.com

Full size image

Effects of inosine

There are few studies evaluating the impact of inosine on neurogenesis, neuritogenesis, and synaptic plasticity, and most of them focus on assessing its neuroprotective properties (Table 2). These studies have yielded evidence regarding the diverse effects of INO in different physiological contexts. INO has been shown to possess anti-inflammatory properties [94], anti-tumoral effects [95], and the ability to modulate immune responses [64, 96]. It has also demonstrated beneficial effects in cardiovascular diseases [97], and respiratory diseases, with some of these effects being mediated through adenosinergic receptors [39, 60, 98]. Additionally, INO has exhibited neuroprotective properties in various contexts [39, 61, 62, 64, 99], including neurodegenerative diseases [100, 101], and neurological damage [102]. However, its role in neuroplasticity is still not fully understood and requires further investigation.

Table 2 Effects of inosine on neuroplasticity
Full size table

In aged female rats, the administration of INO significantly improved learning and memory. This treatment also lowered lipid peroxidation and nitrite levels while raising the levels of reduced glutathione and superoxide dismutase. INO treatment also reduced the number of pyknotic neurons in the hippocampal CA1 region. [94].

In a model of Huntington’s disease (HD), INO demonstrated promising effects by mitigating neurotoxicity induced by 3-nitropropionic acid (3-NP). It enhanced motor function and decreased oxidative stress, inflammation, and neuronal death in the striatum. These beneficial effects occurred through the activation of the A2AR, which in turn triggered the BDNF/TrkB/ERK/CREB signaling pathway [37].

Following corticospinal tract injury in male rats, INO treatment stimulated uninjured pyramidal cells in the sensorimotor cortex to extend the process to denervated areas, indicating its potential for promoting axonal sprouting and neural circuit reorganization [103].

There is a therapeutic potential of INO in spinal cord injury recovery, as observed in adult rats with spinal cord injuries, where daily oral administration of INO (for 7 or 10 days) improved locomotion, micturition, and neuronal survival [104]. In a focal brain injury model, administering INO intracerebroventricularly for 28 days increased the number of crossed axons. This treatment nearly fully restored paw function and was associated with corticospinal axons emerging from the uninjured area and crossing into the territory of the injured pathway [105].

Additionally, INO promotes neural repair and functional recovery after stroke. In adult rats with a unilateral stroke targeting the forelimb motor region, continuous INO delivery altered gene expression in corticospinal neurons of the undamaged hemisphere. It stimulated axon collateral formation and the development of synaptic-like structures on the denervated side of the spinal cord, thereby restoring impaired forelimb function [106].

In cultured cerebellar neurons exposed to chemical hypoxia induced by rotenone, INO was more effective than ADO and GUO in enhancing the number of neurites. This effect occurred via the MAPK signaling pathway, showing INO’s potential as a neurite outgrowth modulator under hypoxic conditions [63].

In hippocampal slices from aged mice and adult rats, INO reduced field excitatory post-synaptic potentials, an effect that was blocked by the adenosine A1 receptor antagonist DPCPX. Furthermore, INO diminished the magnitude of long-term potentiation (LTP), a result that was prevented by DPCPX, the adenosine A2A receptor antagonist SCH58261, adenosine deaminase (ADA), and by inhibiting the equilibrative nucleoside transporter (ENT). These observations suggest that the impact of INO on hippocampal synaptic transmission and plasticity is not due to the direct activation of adenosine receptors, but due to an indirect modification of the tonic activation of adenosine receptors through a nucleoside transporter-mediated modulation of the extracellular levels of adenosine [107]. Although its mechanism of action is not clearly defined, INO demonstrates protective behavioral impacts, including neuritogenesis, axonal sprouting, a decrease in excitatory potential, and ultimately, neuroprotection (Fig. 2).

Fig. 2
figure 2

Illustration of inosine (INO)-induced neuroplasticity. INO exhibits neuroprotective and neurotrophic effects. Studies indicate INO’s role in promoting neuritogenesis, enhancing neuronal survival, and increasing axonal sprouting. Detailed effects of INO and corresponding references are provided in Table 2. This figure was created using BioRender.com

Full size image

Effects of guanosine

Guanosine, an endogenous guanine-based nucleoside, has gained significant interest for its neuroprotective and neurotrophic properties. These properties include promoting neurogenesis, enhancing neuronal survival, facilitating neurite outgrowth, and improving synaptic plasticity, as detailed in Table 3.

Table 3 Effects of guanosine on neuroplasticity
Full size table

In vitro studies using primary cultures have revealed the mitogenic effects of GUO on rat astrocytes. These effects are mediated through intricate mechanisms involving extracellular ADO signaling and soluble factors derived from microglia [108].

In cultures of cerebellar granule neurons, GUO has been shown to exert proliferative effects, mediated by A2AR, NMDAR, and non-NMDAR receptors [109]. The proliferative effects of GUO appear to be mediated by several signaling pathways, including ERK, calcium-calmodulin-dependent kinase-II (CaMKII), protein kinase C (PKC), phosphatidylinositol-3-kinase (PI3-K), and protein kinase A (PKA) [109, 110].

Studies evaluating the neuroprotective effect of GUO have shown that it increases glutamate uptake, reducing its levels in the extracellular milieu [36, 54]. In addition, a study on ligand interaction with ionotropic glutamate receptors revealed that GUO decreases glutamate binding to NMDAR and AMPAR, probably limiting their activation [111]. Then, given GUO’s direct and/or indirect modulation of glutamatergic system receptors, such as AMPAR and NMDAR, it is plausible that this molecule contributes to neuroplasticity, especially in glutamatergic synapses.

GUO treatment enhances the proliferation and differentiation of neural stem cells derived from the hippocampal DG and the SVZ through the activation of the cAMP-CREB pathway and increased expression of BDNF [112, 113]. The effect of GUO on increasing the number of neurospheres from adult mice was reduced by removing ADO from the culture medium [112]. This GUO effect might also be due to the increased extracellular ADO levels since low ADO levels are involved in increased proliferation. However, additional studies are necessary to confirm such a hypothesis.

The impact of GUO on neuronal differentiation has also been investigated using the SH-SY5Y neuroblastoma cell line. GUO induces cell-cycle arrest in these cells, making them more responsive to differentiation processes favoring the dopaminergic/adrenergic phenotype [114].

In vivo studies on adult mice have shown that chronic GUO treatment promotes hippocampal neurogenesis. In one study involving adult female Swiss mice, oral administration of GUO (5 mg/kg, once daily for 21 days) led to a significant increase in the number of NeuroD-positive immature neurons in both the hippocampal SGZ and the ventral SGZ. This increase in neurogenesis was accompanied by antidepressant-like effects [115]. Similarly, in adult male C57Bl/6 mice, intraperitoneal GUO administration (8 mg/kg, once daily for 26 days) enhanced the proliferation of neuroprogenitor cells (Bromo-deoxyuridine-positive, BrdU + cells) and the generation of immature neurons (BrdU + /doublecortin-DCX + cells) in the SGZ. Consistent with the previous study, GUO treatment also exhibited antidepressant-like effects in these mice [112].

The investigation of GUO effects on healthy mice reveals promising therapeutic potential for conditions like stroke, Parkinson’s disease, and spinal cord injuries, demonstrating that GUO can effectively promote neuroplasticity and regeneration.

Following a photothrombosis-induced stroke model in mice, GUO treatment enhanced neurogenesis and angiogenesis. This effect was accompanied by increased levels of BDNF and vascular endothelial growth factor (VEGF), suggesting their involvement in the long-term functional recovery observed after stroke [116].

In a Parkinson’s disease model induced by the neurotoxin 1-methyl-4-phenylpyridinium (MPP +) in rats, GUO treatment (8 mg/kg, i.p., once a day for 8 weeks) demonstrated beneficial effects. It reduced apoptosis, promoted cellular proliferation in the substantia nigra pars compacta (SNc) and subventricular zone, and increased the number of dopaminergic neurons in the SNc. GUO treatment also ameliorated disease symptoms and increased the population of FGF-2-expressing cells [117].

Furthermore, in the peripheral nervous system, administration of GUO has been shown to promote spinal cord remyelination by enhancing the proliferation and differentiation of endogenous adult oligodendroglia progenitor cells [118].

GUO has been shown to play a significant role in neuritogenesis and synaptogenesis, modulating synaptic connectivity and facilitating the formation of functional neuronal circuits. In studies involving undifferentiated PC12 cells, GUO synergistically enhances NGF-dependent neurite outgrowth by activating the constitutive isozyme Heme-Oxigenase (HO)-2 and inducing the expression of HO-1, enzymes responsible for carbon monoxide (CO) synthesis. This activation stimulates soluble guanylate cyclase and increases intracellular cGMP, leading to enhanced neurite outgrowth [119].

In addition, GUO has shown a positive effect on neuritogenesis in primary cultures of rat hippocampal neurons and human SH-SY5Y neuroblastoma cells, as demonstrated by increased neurite number and branching [114, 120, 121]. These effects may be attributed to the synthesis and release of neurotrophic factors such as NGF, FGF-2, and transforming growth factor β (TGFβ), which are activated by GUO treatment [31, 121,122,123]. However, neuritogenesis was not observed in the primary culture of cerebellar granule neurons within 24 h of GUO treatment, or neurons exposed to conditioned medium of GUO-treated astrocytes [124].

GUO treatment has emerged as a positive factor for synaptogenesis, as studies have shown that it promotes synapse development and functional connectivity. In vivo studies have demonstrated that GUO treatment increases both the number and size of synaptic buttons, and it enhances the projection of axonal branches and the development of synaptic buttons in retinal ganglion cells [125].

In rats subjected to traumatic brain injury (TBI), various detrimental effects, including mood disturbances (anxiety-like behavior) and memory dysfunction, were observed. Additionally, there was a decrease in the expression of synaptic proteins (synaptophysin) and proteins involved in neuroplasticity (BDNF and CREB), neuronal loss, and increased astrogliosis and microgliosis in the hippocampus. Chronic GUO treatment at a dose of 7.5 mg/kg, administered intraperitoneally and initiated 1 h after TBI, prevented all these long-term behavioral, neurochemical, and morphological alterations induced by TBI. These effects were dependent on adenosine A1R rather than A2AR [55].

In models of major depressive disorder, hippocampal atrophy and a decrease in synaptic contacts have been observed. GUO has demonstrated an antidepressant-like effect in rodents undergoing the tail suspension test. This effect was associated with the activation of AMPA receptors, voltage-dependent calcium channels (VDCC), and the BDNF/TrkB signaling pathway. GUO treatment also led to a rapid increase in the phosphorylation of P70S6K (Thr389) and the content of synapsin I, a marker of presynaptic protein, in the hippocampus. However, no significant changes in dendritic spine density and remodeling were observed in the ventral DG region [126].

Together, studies discussed here demonstrate a positive impact of GUO on cell proliferation, neurogenesis, and neuritogenesis. While the exact mechanisms underlying many of these effects are still unclear, they suggest a potential interplay between purinergic and glutamatergic systems (Fig. 3).

Fig. 3
figure 3

Schematic depiction of guanosine (GUO)-induced neuroplasticity. GUO promotes the proliferation of neural stem cells (NSCs), astrocytes, oligodendroglia progenitor cells (OPCs), and intermediate progenitor cells. GUO induces neurogenesis, enhances neuronal survival, promotes neurite outgrowth, and enhances synaptic plasticity. Details regarding effects and references can be found in Table 3. This figure was generated using BioRender.com

Full size image

Clinical perspective

Neurodegenerative and neuropsychiatric diseases, as well as injuries resulting from trauma or stroke, pose significant challenges as current treatments often fail to achieve full recovery. Thus, identifying new therapeutic targets is important to enhance the quality of life for affected individuals. In this regard, the purinergic system has emerged as a promising avenue, with numerous studies highlighting its crucial role in modulating neuroplasticity across various contexts. Nucleosides, including ADO, INO, and GUO, have shown the ability to influence neuroplasticity at different stages of neurogenesis and synaptic modulation.

Purine nucleosides show a promising therapeutical strategy in recent clinical trials, serving as either potential standalone treatments or adjuncts in various treatment approaches.

A comprehensive search was carried out by entering “adenosine” under the “intervention/treatment” category, which yielded a total of 745 studies as results (https://clinicaltrials.gov/search?intr=adenosine, accessed on 08/17/2023). ADO is being used in several disorders, such as heart disease (NCT00881686, NCT04392362, NCT02681913), in COVID-19 (NCT04588441) among others. In the peripheral nervous system, studies are using ADO in chronic neuropathic pain (NCT00349921), and the use of BVT.115959 (A2AR Agonist) in diabetic patients with neuropathic pain (NCT00452777).

ADO has a short half-life in circulation; however, its local extracellular levels can be increased through the inhibition of ENT1, an adenosine reuptake transporter. Recent studies have explored the use of ticagrelor, a P2Y12R antagonist approved for preventing thrombotic events in patients with acute coronary syndrome (ACS) or myocardial infarction (MI), to achieve this purpose. Ticagrelor binds to ENT1, leading to elevated local adenosine levels. This increase in adenosine has been associated with potential benefits such as anti-inflammatory effects and enhanced coronary vasodilation. Moreover, there is promise that this mechanism may improve cognitive function and reduce the risk of dementia and Parkinson’s disease in individuals with vascular high-risk factors [127].

In the research referring to inosine (https://clinicaltrials.gov/search?intr=inosine, accessed on 08/16/2023), 45 studies appear, showing this nucleoside being tested alone or in association with other drugs, such as isoprisone (Inosine pranobex—inosine acedoben dimepranol), mainly tested in immunocompromised patients (NCT00002060), or Cytoflavin.® (a combination drug comprising inosine, nicotinamide, riboflavin, and succinic acid), used in head trauma (NCT05935787, PMID:20,517,223), stroke (NCT05935787, NCT05297851), and for the prevention of cognitive decline following major surgery among elderly patients (NCT03849664). Cytoflavin is designed to enhance cerebral blood flow, activate metabolic processes, and improve neuroplasticity in the central nervous system [128]

Recent studies with the administration of INO were found to be safe and well-tolerated. However, these studies primarily focused on increasing urate levels as a natural antioxidant rather than exploring inosine’s potential as a neuromodulator. Future studies employing different strategies and appropriate statistical power to assess efficacy may provide valuable insights into the possible neuroprotective role of inosine [129, 130]. In neurodegenerative diseases, this nucleoside has been tested in amyotrophic lateral sclerosis [131], multiple sclerosis [132], and Parkinson’s disease [129]. In this last study, which reached phase 3, participants with PD received oral INO or placebo. However, the results did not support the use of inosine as a treatment for early PD [129].

Upon investigating guanosine (https://clinicaltrials.gov/search?intr=guanosine, accessed on 08/16/2023), the search yielded a total of 13 studies, but none of the studies use the GUO nucleoside alone or associated with other drugs. However, given its safety profile and its demonstrated protective and trophic effects, guanosine should be considered for evaluation, particularly in the aftermath of brain-damaging situations, such as ischemia. In such scenarios, guanosine could potentially play a dual role initially preventing excitotoxicity and inflammation, and subsequently aiding in neural reorganization and plasticity [65].

In the context of purine metabolism, allopurinol, a drug widely used to control high levels of uric acid due to xanthine oxidase activity inhibition, was also tested as a possible strategy in refractory epilepsies [133]. Additionally, it was evaluated as adjunctive therapy for conditions like poorly responsive schizophrenia (NCT00864825), refractory aggressive behavior, and mania (NCT01092221). By inhibiting purine catabolism, it could elevate nucleoside levels, thus contributing to the observed beneficial effects [134,135,136,137,138,139,140,141].

Overall, this review has provided compelling evidence for the roles of purine nucleosides in modulating critical processes such as neurogenesis, neuritogenesis, synaptogenesis, and synaptic plasticity in various brain regions. Moreover, purine nucleosides have shown to be safe in clinical trials, suggesting their potential as a treatment for brain injuries. Together, these findings highlight the diverse functions of these nucleosides in shaping neural networks and their significant involvement in the remodeling response of the adult central nervous system following brain insults.