Glucagon like peptide-1 (GLP-1) is a peptide called enteropancreatin secreted by small intestinal L cells. Physiologically, it can bind to GLP-1 receptor on pancreatic beta cells through endocrine, activate adenylate cyclase, increase intracellular cAMP levels, promote glycolysis and ATP production, and enhance glucose-stimulated insulin secretion. The direct regulation of GLP-1 on insulin expression is significant for type 2 diabetes (T2D) patients with simple glucose tolerance but normal insulin secretion (1). It can avoid adverse reactions and potential risk of pancreatic gland degeneration caused by direct insulin treatment (2). It is worth noting that GLP-1 can easily spread to the central system and exert neuroprotective and neuropathic actions in various neurodegenerative and cerebrovascular disorders (Alzheimer’s disease (AD), Parkinson’s disease (PD), stroke, etc.) through biological signal transduction (3–5). However, due to the short half-life of endogenous GLP-1 in the body (only about 2 minutes), it will be rapidly degraded by dipeptidyl peptidase-4 (DPP-4) and neutral endopeptidase. Therefore, GLP-1 derivatives with medium to long acting properties have become the main GLP-1RA for clinical applications, such as exendin-4, exenatide, liraglutide, lixisenatide, semaglutide, etc. (6–8). These peptide-based GLP-1RA have been widely studied globally for improving cognitive dysfunction in patients with neurodegenerative diseases (AD and PD) (ID: NCT03456687; NCT01255163; NCT03439943; NCT03659682; NCT01843075; NCT02953665) (9–12).

There are two main categories of GLP-1 receptor agonists, small molecule compounds and peptide-based GLP-1 derivatives as shown in Figure 1 , including exenatide, liraglutide, dulaglutide, etc. Many literatures proved that GLP-1 derivatives performed neuroprotective effects and inhibited neuronal apoptosis (13–15). It was reported that cerebral expression of GLP-1 receptor protein reduced following subarachnoid hemorrhage (SAH), but markedly increased by liraglutide (15). Liraglutide also alleviated SAH-induced early brain injury and inflammation. Exendin-4 not only reduced the infarct size of ischemic stroke, improved the neurologic deficit scores in ischemia/reperfusion-induced rats, but also reduced the potential rtPA-induced hemorrhagic risk through the Wnt/β-catenin signaling pathway (16).

There are also a type of non-peptide small molecules, which are mainly extracted from herbal medicine, such as Cornus officinalis, Gardenia jasminoides, Rehmannia glutinosa, Lamiophlomis rotata, and Strychnos nuzi (17–20). These small molecule compounds mainly belong to iridoid glycosides, including geniposide, Shanzhiside methylester (SM), 8-O-acetyl-SM, morroniside, catalpol, genipin methyl ether and so on (18, 21). These GLP-1RA could be used to treat diabetes and its complicated obesity, fatty liver, hypertension and various cardiovascular diseases (22). In addition, accumulating studies demonstrated that GLP-1RA performed neuroprotective effects in a variety of neurodegenerative diseases (mainly AD, PD and stroke) (23–25).

Even for patients without complications of diabetes, their cognitive functions were still significantly improved (26, 27). The notable therapeutic impact of peptide-based GLP-1 derivatives in clinical neurodegenerative disorders had fostered enthusiasm among researchers to develop small molecule GLP-1RA. It was reported that small molecule GLP-1RA could reduce chronic neuropathic pain and exert neuroprotective effects by improving synaptic plasticity and ameliorating neuronal death (28, 29). GLP-1RA inhibited inflammatory responses in chronic pain through the GLP-1R/PI3K/Akt/ERK-1/2/IL-10 signaling pathway and reduced the release of inflammatory factors such as IL-6, IL-1, tumor necrosis factor alpha (TNF-α) (30). In addition, Wang Y.X. also found that the analgesic effects of various small molecule GLP-1RA such as lemairamin mainly relied on their promotion of IL-10 secretion and subsequent IL-10R/β endorphin signaling pathways, which effectively alleviated bone cancer pain and neuropathic pain (31–33). In these studies related to analgesia, shanzhizhi methyl ester performed a dose-dependent and persistent (>4 hours) effects against allodynia (17). In rodent models of neurodegenerative diseases, a series of iridoid glycosides represented by geniposide and mononucleoside could effectively resist cellular oxidative damage and counteract neuronal apoptosis (19, 28). In summary, as a promising class of small molecule GLP-1RA, iridoid glycosides represent a potential therapeutic avenue for the management of neurodegenerative disorders.

Many literatures indicated that GLP-1RA were also considered as the latent intervention strategy against ischemic stroke to improve the pathological symptoms and prognosis. The principal neurological function regulated by GLP-1RA could be summarized as follows ( Figure 2 ): (1) GLP-1RA liraglutide and exendin-4 effectively inhibited neuronal apoptosis (15), and ultimately improved early brain injury by suppressing NLRP3 inflammasome and inflammatory reaction in microglia (34–36); (2) Exendin-4 meliorated vascular permeability by inhibiting blood-brain barrier breakdown and leakage (37), and reduced the hemorrhagic risk of ischemic insults via the Wnt/beta-catenin signaling pathway (16). Besides, exendin-4 also dilated cerebral arterioles, raised cerebral blood flow and finally reversed the cerebral ischemia (38). (3) GLP-1RA (e.g. morroniside, exendin-4) promoted M2 polarization in microglia by upregulating nuclear factor erythroid 2-related factor 2 (28), and further improving synapse growth and neuronal regeneration via the cAMP/PKA pathway (39), and finally reduced cognitive and memory impairment induced by stroke (40). GLP-1RA appeared to exert neuroprotective effects primarily by regulating microglia, which were the most significant immune cells in the central nervous system (41).

As a new type of hypoglycemic drug, GLP-RAs are currently mainly used in clinical practice to evaluate the prognosis of various complications in T2D patients, such as stroke, kidney injury, fatty liver, and cardiovascular disease (42, 43). Compared with other new hypoglycemic drugs such as SGLT2 and DPP-4, GLP-1RA have a unique preventive effect in reducing the risk of stroke as shown in Table 1 (43, 45). With the ageing of society, stroke has emerged as the leading cause of mortality and major cause of disability worldwide, greatly increasing the burden of human society (46). According to the degree of severity, stroke-related disability led to premature death with disability-adjusted life-years (DALYs) (47). There was a notable discrepancy in the proportion of deaths resulting from hemorrhagic stroke among different racial and ethnic groups. The proportion of deaths due to hemorrhagic stroke among middle-aged and older people (≥35 years) in the 1990s was 38% for Asians, 32% for Hispanics, 24% for blacks, and 18% for whites (47, 48). The high rates of intracerebral hemorrhage (ICH) in Asian populations have been attributed to the high prevalence of hypertension, smoking and metabolic syndrome (49). Using Caucasians as a reference, the adjusted relative risk (RR[95%CI]) of ICH in Asians was 1.6[1.1-2.3] (P=0.01) (50, 51). The mortality rate increased by 32% from 1990 to 2019 (52). In 2019, there were 2.2 million deaths and 45.9 million DALYs after stroke insults in China (53). By 2050, the burden of stroke in individuals (aged ≥65 years) will increase significantly: 104.7% in incidence, 218.5% in prevalence, 100.0% in mortality, and 58.9% in DALYs (54).

Many systematic reviews with network meta-analysis summarized the risk of noval hypoglycemic drugs on diabetic complications, including cardiovascular events, nonfatal stroke, renal composite outcomes and fatty liver (42–45). By using random-effects model, the network meta-analysis indicated that the risk profile of DPP-4i was same as placebo in all the outcomes (44). Distinctly, both SGLT-2i and GLP-1RA significantly decreased the risk of cardiovascular events (odds ratios (OR)[95%CI], GLP−1RA: 0.87[0.82-0.93]; SGLT−2i: 0.88[0.82-0.95]), and renal composite outcome (GLP−1RA: 0.86[0.78-0.94); SGLT−2i:0.59[0.52-0.67]) when compared to placebo ( Table 1 ). In addition, only administration of GLP-1RA resulted into lower nonfatal strokes (OR, 0.88[0.77-0.99]) than those receiving placebo. P-rank scores confirmed that GLP-1RA reduced the risk of nonfatal stroke by 80.6%.

The similar viewpoints were also supported by other critical reviews and clinical trials (55–58), and inferred that GLP-1RA could be considered in patients at a high risk of, or with established cardiovascular diseases to improve the prognosis of cardiovascular events (especially nonfatal stroke), whilst SGLT2i could be recommended for patients with heart failure or chronic kidney diseases.

Although multiple cytokines (IL-4, IL-10, etc.) together inhibited neuroinflammation in neurological disorders, their functions were significantly different. M2 microglia could be divided into several subtypes such as IL-4-induced M2a and IL-10-induced M2c. The synergistic induction of IL-4 and IL-10 could enhance the expression of M2a-related genes, produce a large amount of CCL24 (eotaxin-2) and promote eosinophil migration (104). In clinical practice focusing on prognostic indicators for ischemic stroke patients found that there was a close negative correlation between IL-10 and NIHSS scores (P=0.0006), but not between IL-4 and NIHSS scores (P=0.088) (105). Comparing the prognosis of ischemic stroke patients between survivors and non survivors, the IL-10 level in the survivors group was significantly higher than that in the non survivors group (P=0.006) (106). Although there is much preclinical evidence as mentioned above, few clinical trials was reported to evaluate the efficacy of GLP-1RAs in the treatment of ischemic stroke or diabetes complicated by ischemic stroke. A letter reported the clinical manifestations of the GLP-1RA exenatide in acute stroke patients for the first time, including treatment regimens and adverse reactions (107). The results showed that 11 patients experienced mild nausea (n=6) and vomiting (n=5), with no serious adverse reactions or deaths. Given the high incidence of vomiting, the author suggested prophylactic administration of antiemetics. The combination of exenatide and antiemetic agents had good safety and tolerability, with no adverse effects on neurological function and prognosis. In addition, we also found that the clinical rationale and protocol design of the TEXAIS (Treatment With Exenatide in Acute Ischemic Stroke) trial (Trial registration: ClinicalTrials.gov/ANZCTR NTA1127 and ACTRN12614001189617) (108, 109). The latest results from the 2023 TEXAIS trial indicated that the use of exenatide (5 µg, twice daily injections for 5 days) did not significantly reduce neurological damage within 7 days in patients with acute ischemic stroke (trial registration numbers: ACTRN12617000409370 and NCT03287076). However, in the exenatide treatment group, the primary outcome rate was 61.2% (n=170), compared to 56.7% in the standard treatment group (n=171), with an adjusted ratio of 1.22 [95% CI, 0.79-1.88] (P=0.38)., which meant that the proportion of subjects who achieved the desired effect in exenatide treatment group was higher (110). GLP-1RAs may be an optimal option for ischemic stroke treatment.

It was noted that no clinical trials had yet been completed on AD, PD, or other neurodegenerative disorders. Thus clinical data on GLP-1 agonists in brain-related conditions remain limited. However, in a double-blinded trial recruiting 54 participants with type 2 diabetes were randomized to liraglutide (1.8 mg/day) or placebo for 26 weeks (111). The mRNA expressions of tumor necrosis factor-α (p = 0.004) and interleukin-1β in peripheral blood mononuclear cells were downregulated (p = 0.046) in the liraglutide-treated group. The clinical evidence of GLP-1RAs in anti-inflammatory effect was also proved in semaglutide, exenatide and others (112–115). Chronic neuroinflammation is considered an important factor in cognitive and memory impairment in neurodegenerative diseases. It was reported that exendin-4 prohibited amyloid-β-induced microglial activation, thereby limiting neuroinflammation, reducing the levels of TNF-α, C1q and IL-1α; and improving recognition (116). Similar activities of other GLP-1 mimetics were also demonstrated in a rodent PD model (117). Another study also indicated that exendin-4 improved memory impairment by dampening the AMPK/NF-κB pathway, reducing levels of IL-1β and TNF-α, and increasing synaptic protein levels (40). Besides, liraglutide could also improve rat memory and cognition function by reduced neuronal apoptosis, tau phosphorylation, and β-site APP cleaving enzyme 1 levels (118). For a long time, scientists believe that chronic neuroinflammation is crucial in cognitive impairment, and that GLP-1RAs protect synaptic and learning functions from neuroinflammation (119, 120).

A genome-wide association study examining common genetic features of AD had identified neuroinflammatory pathways (mainly associated with TNF) to be a key feature of AD, and a driver of risk for these chronic neurodegenerative impairment (121). Plenty of nonsteroidal anti-inflammatory drugs (NSAIDs) were recruited in many clinical trials for AD treatment (122). NSAIDs primarily reduce inflammation by inhibiting cyclooxygenase enzymes (COX-1 and COX-2), which are responsible for the production of prostaglandins, which are involved in the inflammatory response, including the promotion of pain, fever, and tissue damage. Therefore, the anti-inflammatory effect of NSAIDs is mainly applied to peripheral tissues and for short-term inflammation with the limited CNS penetration. Besides, both GLP-1RAs and SGLT2 inhibitors showed the obvious inhibition on neuroinflammation (123, 124). SGLT2 inhibitors could block the reabsorption of glucose in the kidneys, leading to increased glucose excretion and improved glycemic control in type 2 diabetes. The new preclinical evidence also suggested that they might have neuroprotective properties through reducing oxidative stress, modulating inflammatory pathways, and enhancing mitochondrial function (125, 126). However, their effects in humans with brain disorders are still under investigation.

This review was mainly summarized on the basis of research data, and therefore had limitations in many aspects: (1) Although we already knew that GLP-1RA inhibited the activation of microglia and promoted their polarization towards the M2 type, the regulatory mechanism of their induction of polarization in naive CD4⁺ T cells was still unknown, in particular their impact on the proportion of Tregs subtypes in T cells at the sites of infarct lesion. (2) The interaction between microglia and T cells may be bidirectional, as microglia express various molecules involved in antigen presentation and T cell regulation, such as MHC II, CD11c, Dectin-1, etc (127). Do GLP-1RA have a regulatory effect on microglial antigen presentation and microglial-T cell balance? (3) Clinical evidence for GLP-1RA in stroke is still lacking. The latest TEXAIS results in 2023 were negative with a certain trend, which might be limited by the small sample size, short treatment duration and lack of medium to long-term therapeutic evaluation. Current research on GLP-1RA in ischemic stroke should be expanded to include large-scale, long-term follow-up clinical trials to further assess its efficacy and safety. Additionally, evaluating the long-term use of GLP-1RA, particularly across different populations, could provide valuable insights into its therapeutic potential and risk profile, which will contribute to developing more effective treatment strategies and improving the prognosis and quality of life for patients with ischemic stroke and diabetes (128, 129).

Finally, this study elucidated the immunosuppressive and neuroregenerative functions of GLP-1RA based on the "Tregs-microglia-neuron/NPC" axis theory. Given the advantages of small-molecule GLP-1RA, such as rapid diffusion, good BBB permeability and multiple modes of administration, small molecule compounds will be one of the important development trends of GLP-1RA. In the future, we look forward to the clinical research evidence of small-molecule GLP-1RA intervening in ischemic stroke or T2D complicated by ischemic stroke.