Current understanding and controversy on brain access of GLP-1 and GLP-1 receptor agonists

Introduction

Incretins are defined as gut-produced hormones that facilitate postprandial insulin secretion from pancreatic islet β-cells. ^{[1, 2, 3, 4, 5, 6, 7, 8]} Such facilitation is tightly controlled in glucose concentration dependent manners, without causing hypoglycemia. The gut in mammalian species is known to produce two types of incretins. The first one is gastric inhibitory polypeptide (GIP) produced by gut endocrine K cells (also known as glucose-dependent insulinotropic polypeptide), mostly found in the duodenum. The second one is glucagon-like peptide-1 (GLP-1) produced by gut endocrine L cells, mostly found in the distal ileum and colon.^{[1,2,9, 10, 11, 12, 13]}

GIP, as the first incretin hormone, was identified by the Canadian scientist John Brown in University of British Columbia (UBC) and his colleagues in UBC and elsewhere back to later 1960 s and early 1970 s.^{[14, 15, 16, 17, 18, 19, 20]} For detailed information on the serendipity discovery of GIP and its functional studies, please see nice review articles and other publications elsewhere,^{[21, 22, 23]} including the “In Memoriam” composed by Timothy Kieffer (a formal PhD student of late professor John Brown) in 2017.^[24] We have yet seen the final success on developing GIP-based pharmaceutical agents for diabetes or other metabolic disorders. Mechanistic functional studies on GIP and its G-protein coupled receptor, GIPR. However, they have contributed to the development of dipeptidyl peptidase-4 (DPP-4) inhibitor Sitagliptin (commercially known as Januvia^®) as a diabetes drug by Merck & Co. in 2006.^[25,26] More recently, the dual GIP/GLP-1 receptor agonist Tirzepatide, has been shown to be more potent than a pure GLP-1 receptor agonist (GLP-1RA) in clinical trials on reducing body weight.^[27] Currently, Tirzepatide is sold under the brand name Mounjaro for type 2 diabetes (T2D) treatment and Zepbound for weight management and the treatment of obstructive sleep apnea.

In the late 1980 s, studies conducted by Habener and his colleagues at Harvard University ^{[28, 29, 30]} and by Holst and his colleagues at University of Copenhagen,^{[31, 32, 33]} demonstrated that GLP-1 is the 2^nd incretin hormone. Since then, efforts made by scientific researchers across multiple disciplines have led to the development of various GLP-1-based drugs or GLP-1RAs as therapeutic agents for T2D.

Active GLP-1 molecules include GLP-1^7–37 and GLP-1_{7–36 amide}.^{[28, 29, 30, 31, 32, 33]} They cannot serve as therapeutic agents per se, as their half-lives are only approximately 1.5–2.0 min in the circulation. The rapid degradation of the native GLP-1 molecules is mainly attributed to ubiquitously expressed endogenous enzymes known as DPP-4, which converts active GLP-1_7–37 and GLP-1_{7–36 amide} into inactive GLP-1_9–37 and GLP-1_{9–36 amide}. Another enzyme namely neutral endopeptidase 24.11 (NEP 24.11) can further cleave GLP-1_9–37 and GLP-1_{9–36 amide} into even smaller peptides.^{[34, 35, 36]} Paradoxically, certain smaller peptides, such as GLP-1_{28–36 amide} and GLP-1_{32–36 amide} were shown to possess certain metabolic beneficial effect.^{[34, 35, 36, 37, 38, 39]}

Exendin-4 (Ex-4) is a polypeptide with 39 amino acid residues, isolated from the venom of the Gila monster (Heloderma suspectum).^[40] Ex-4 shares 53% of its amino acid residue identity with mammalian GLP-1, while its half-life is much longer (∼10 min) due to its resistance against the enzymatic cleavage mediated by DPP-4. In 1992, Eng’s team demonstrated the incretin feature of Ex-4 in both rat and guinea pig models.^[41,42] In 1993, Goke et al. (in Eng’s team) further demonstrated that Ex-4 is a potent agonist of GLP-1R.^[43] Truncated Ex-4 known as Exendin (9–39) serves as a powerful antagonist of GLP-1R, which has been broadly utilized in functional studies of GLP-1/GLP-1R signaling cascade in the brain and elsewhere. GLP-1R is also a G-protein-coupled receptor. Both protein kinase A (PKA) and exchange protein directly activated by cAMP (Epac) have been shown to mediate functions of GLP-1/GLP-1R signaling activation in pancreatic islets and elsewhere.^{[7,44, 45, 46]}

A synthetic version of Ex-4, namely Exenatide, developed by Amylin Pharmaceuticals and Eli Lilly and Co., was the first GLP-1-based therapeutic agent for T2D. It was approved by the Food and Drug Administration (FDA) in 2005, with the brand name Byetta^®. Over the past two decades, nearly a dozen chemically modified versions of human GLP-1 have been developed as therapeutic agents, including Liraglutide (Victoza^®) and Semaglutide (Ozempic^®), the products of Novo Nordisk Inc.

Extensive investigations over the past three decades have shown that GLP-1 and GLP-1RAs exert their metabolic and other beneficial functions far beyond serving as an incretin hormone. In other word, functions of GLP-1 do not entirely dependent on insulin. This explains why GLP-1RAs are effective in T2D patients with insulin resistance.^[9] Indeed, GLP- 1 is a pluripotent hormone, targeting various organs including the brain, gut, heart, liver, lung, and adipose tissues. For extra-pancreatic functions of GLP-1 and its based diabetes drugs, please see nice review and other publications elsewhere.^{[2,4,6,47, 48, 49, 50, 51, 52, 53, 54, 55]}

It is also worth mentioning that GLP-1 based research has been drawing extensive global attention. Over the past five years, top scientists in this field, including Joel F. Habener, Jens J. Holst, Daniel J. Drucker, Svetlana Mojsov, and Lotte B. Knudsen, have been recognized by a battery of prestigious international awards, commented recently by our team and by others.^{[5,56, 57, 58, 59, 60]} The top awards include Warren Alpert Foundation Prize, Canada Gairdner International Award, Wolf Prize in Medicine, Tang Prize in Medicine, Lasker in Clinical Research, and most recently, the Breakthrough Prize in Life Sciences Award.

GLP-1 based drugs also promote weight loss

Obesity and overweight may lead to the development of various metabolic diseases including T2D, heart and cardiovascular disorders, stroke, metabolic associated fatty liver disease (MAFLD), infectious diseases, and others.^{[61, 62, 63, 64, 65]} Based on World Health Organization (WHO), worldwide adult obesity has more than doubled since 1990, while adolescent obesity has quadrupled. In 2022, 43% of adults in the world were overweight and 16% of our world population were living with obesity. We have also seen even astonishing elevation on children and adolescents overweight and obesity. These changes have not only increased the economic burdens globally but also worsened the overall quality of our lives. Hence, brain function of GLP-1 and its based drugs has been drawing significant global attention.

Brain function of GLP-1 was initially demonstrated by Bloom’s team in Imperial College London back to 1996 in a rat model with the brain 3^rd ventricle native GLP-1 injection,^[66,67] followed by intensive investigations in North America, Asian countries, European nations, and elsewhere. For further updated knowledge on the anorexic function and neuroprotective effects of GLP-1 and GLP-1RAs in animal models and in clinical studies, please see excellent review articles elsewhere.^{[51,68, 69, 70, 71, 72]}

Liraglutide (Saxenda^®) is the first GLP-1RA therapeutic agent being approved by FDA for adults and adolescents (ages ≥ 12) with obesity or overweight and comorbidities. In a 56-week, randomized, placebo-controlled trial (SCALE trial), 3.0 mg of daily Liraglutide shown mean body weight loss of 8.0%.^[73] In 2021, the second GLP-1RA, Semaglutide (Wegovy^®) became approved by FDA for the treatment of obesity or overweight. Superior efficacy of Semaglutide (Wegovy^®) was demonstrated in STEP 1 trial, whereas mean body weight lost nearly achieved 15% over 68-week trial by giving 2.4 mg once per week.^[74] It’s interesting to note that Semaglutide (Ozempic^®) has been approved for T2D and cardiovascular risk reduction but not yet for weight loss application. The dual GIP/GLP-1 receptor agonist Tirzepatide (Zepbound^®) was approved in 2023 and is the most effective GLP-1RA to date for body weight reduction in individuals with obesity or overweight.^[75] When compared with Semaglutide, Tirzepatide treatment results 5% or greater weight loss with similar extent of gastrointestinal adverse events.^[76] Several clinical trials on additional GLP-1-based agents including triple agonists (Retatrutide),^[77] oral formation (Orforglipron),^[78] and combination therapy (CagriSema) are in the late phase of trials.

The strong body weight lowering effect of above GLP-1RAs, as well as the dual GIP/GLP-1 receptor agonist Tirzepatide, expanded the application of incretin-based therapeutics into the novel avenues including the anti-obesity. Even with the profound body weight lowering effects seen in preclinical studies and clinical trials, it is still puzzling to many researchers and clinicians on the mechanism of action of those GLP-1-based therapeutics. There is no doubt that target sites for GLP-1 or GLP-1RAs to exert their anti-obesity function are located with the brain. It, however, remains to be determined exactly how gut produced GLP-1 or peripherally administrated GLP-1RAs enter the brain to exert their anorectic and other beneficial functions. Alternatively, can gut produced native GLP-1 or peripherally administrated GLP-1RAs as therapeutic agents exert their brain function without physically entering the brain, but via other means. An alternative mean is via the vagal afferent neuron mediated organ-organ communication. Here we will discuss some key literatures on addressing the above questions during the past two decades, including the highlight of controversies in this active field. We have then composed a summary, along with presenting our perspective view.

Barriers between the brain and the peripheral circulation system

There are three barriers between the brain and the peripheral circulation system, known as the blood brain barrier (BBB), the blood-cerebrospinal fluid (CSF) barrier defined here as BCB, and the brain circumventricular organs (CVOs) (Figure 1).^[79] We have presented a brief introduction on each of them in below.

Figure 1

Structures of blood-brain barrier and blood cerebrospinal fluid, and locations of the seven circumventricular organs in the brain. BBB: blood-brain barrier; BCB: blood cerebrospinal fluid (CSF) barrier (BCB); CVOs: circumventricular organs; ISF: interstitial fluid.

In mammals, BBB is a semipermeable border of endothelial cells that control the transfer of chemicals or solutes between the peripheral circulatory system and the central nervous system. Such a barrier effectively protects the brain from damages that can be induced by harmful substances in the peripheral circulation. In normal physiological conditions, BBB passage of circulating molecules can be achieved via either the paracellular route or via the transcellular route. The passive paracellular route only allows the passage of molecules that are smaller than 500 Da. Native GLP-1 and Liraglutide are, however, 3298.7 Da and 3751.2 Da, respectively. The long-term effective GLP-1RA Semaglutide is 4113.6 Da. The transcellular route refers to either transporter- or receptor-medicated access that across the BBB. A few early studies have suggested that peripheral leptin administration may across BBB via such transcellular mechanism,^{[80, 81, 82]} raising the possibility that GLP-1 may also enter the brain utilizing a similar mechanism.

Another way for circulating substance to enter the brain is via BCB. By crossing BCB, the circulating substance can enter CSF. We have learned that the choroid plexus (CP) of BCB displays fundamentally different properties from that of BBB.^[83] With relatively high permeable capillaries, CP provides the central nerves system (CNS) with a high turnover rate of fluid that contain micronutrients, peptides, and hormones for the neuronal networks. It has also been reported that peripherally produced leptin can bind to leptin receptors that are expressed in ependymal cells, crossing the ependymal barrier.^[84] If GLP-1 or GLP-1RAs can enter the brain via BCB, one would anticipate that GLP-1RA level in CSF would increase following a long-term GLP-1RA treatment.

Additional sites for circulating molecules to enter the brain are through various CVOs, bordering the 3^rd and 4^th ventricles of the brain.^[85] Those structures in the brain are characterized by highly permeable capillaries, unlike those in the rest of the brain area where BBB is present at the capillary level. Although the term “CVO” was proposed in 1958 by Dr. Helmut O. Hofer, the penetration of blood-borne dyes into small specific CVO regions was demonstrated in the early 20^th century. The permeable CVOs enabling the rapid neurohumoral exchange include the subfornical organ (SFO), the area postrema (AP), the vascular organ of lamina terminalis (VOLT), the median eminence (ME), neurohypophysis (NH), the pineal gland (PG), and subcommissural organ (SCO). Figure 1 illustrates the structures of BBB, BCB and the locations of the seven CVOs, along with their abbreviations. For detailed functions of these CVOs, please see review articles elsewhere.^[86,87]

Can native gut produced GLP-1 and peripherally administrated GLP-1RAs access the brain via BBB

Back to 2002, Kastin and colleagues have assessed the influx of radioisotope labelled [Ser⁸] GLP-1 into the brain following its peripheral administration [i.e., intravenous (iv) injection).^[88] [Ser⁸] GLP-1 was demonstrated to possess similar biological effects when compared with the native GLP-1 but has greater stability.^[88] The authors reported that they have detected a rapid influx with the multiple-time regression analysis when ¹²⁵I-[Ser⁸] GLP-1 was peripherally administrated.^[88] Furthermore, there was no self inhibition by excess doses of the unlabeled [Ser⁸] GLP-1 either via iv injection or via in situ brain perfusion. In addition, the authors reported that they observed no inhibition of influx by the GLP-1R antagonist Exendin (9–39). Based on the lack of inhibition of the influx of radiolabelled GLP-1 with unlabelled GLP-1R antagonist, the authors made their suggestion that the rapid entry of those radiolabelled GLP-1 into the brain does not require the participation of GLP-1R.^[88] Similar investigations was then performed with the GLP-1RA Exenatide by the same research group.^[89] Brain access of relatively larger sized GLP-1RAs (human origin GLP-1 with structural modifications) such as Liraglutide has also been assessed recently, utilizing the similar radioisotope labeling approaches.^{[90, 91, 92, 93, 94]}

Although Kastin and colleagues have suggested that GLP-1 and Ex-4 can rapidly access the brain parenchyma in a receptor-independent manner,^[88,89] their suggestion is against the classical dogma that BBB only allows its paracellular access of molecules smaller than 500 Da, regardless of receptor presence.

In 2020, Fu and colleagues presented their investigation, showing a set of totally different observations along with their different conclusions.^[95] They have iv injected fluorescence (FAM) labeled-GLP-1 or Ex-4 into the rats, followed by the detection of their appearance as well as PKA activity in various brain regions. They have also demonstrated that rat brain microvascular endothelial cells can rapidly uptake fluorescence-labeled GLP-1, the event can be blunted with the utilization of the GLP-1R antagonist Exendin (9–39). Hence, they concluded that GLP-1 crosses the BBB through active trans-endothelial transport, which requires the GLP-1R. This observation is somehow like brain access of peripheral leptin, via a receptor mediated transcellular mechanism, as we have mentioned above.^{[80, 81, 82]} This conclusion, however, is not supported by recent studies, indicating that hypothalamic endothelial cells do not express GLP-1R,^{[96, 97, 98]} including one review article discussed by Buller and Blouet.^[99]

In one of the above investigations led by Imbernon and Saponaro et al., the team utilized fluorescently labeled Liraglutide and reported that endothelial cells of BBB do not express GLP-1R. This was further supported by employing in vitro BBB models and radiolabeled compounds ¹²⁵I-GLP1 and ¹²⁵I-Liraglutide. Instead, they showed that Liraglutide enters the ME (one of seven CVOs) via tanycyte-mediated transcytosis as early as 60 seconds following iv injection. Inhibiting tanycytic transcytosis by botulinum toxin or selectively knock down of Glp1r expression in tanycytes by AAV-mediated shRNA delivery indicated the loss of Liraglutide transport into the brain. This impaired transport also led to the diminished metabolic benefits mediated by Liraglutide.^[96] The other two studies failed in the detection of GLP-1R immunoreactivity on brain endothelial cells using a highly selective GLP-1R antibody with confocal or electron microscopy.^[97,98] However, the contrary opinion still holds. Smith and colleagues mated the Glp 1r-Cre mouse line with the ROSA26EYFP transgenic mouse line. They hence introduced the fluorescent reporter into GLP-1R expressing cells. Brain GLP-1R⁺ cells were then FACS-purified and sequenced using single-cell RNA sequencing. They were able to provide the first profile of GLP-1R expression cells in the brain and further showed the detection of a population of arterial endothelial GLP-1R⁺ cells.^[100]

In addition to the above controversies presented in PubMed by different research groups, a fundamental question that remains to be addressed is: although GLP-1 or GLP-1RAs may get access to the brain via a rapid and passive receptor-independent mechanism,^[88,89] or by a receptor-mediated event,^[95] does such access physiologically and pharmacologically relevant to their brain functions? It is also worth noting that the controversy generated could be because the detection method with fluorescence or radioisotope labelling being over-sensitive. Should such over-sensitive methods generate false positive results, technology advancement is required for our future investigations.

Can gut produced native GLP-1 or peripherally administrated GLP-1RAs access the brain via BCB

A few studies have shown that fluorescence labeled GLP-1RAs can interact with the choroid plexus, as well as certain CVOs.^{[97,98,101,102]} However, clear labeling was virtually absent in mouse ependymal cells of the choroid plexus when high-resolution microscopy was applied.^[103] Furthermore, choroid plexus labeling with peripheral fluorescence labelled GLP-1RA administration can occur in GLP-1R knockout mice as well, indicating that this event does not require GLP-1R, could be a false positive result as well. Finally, it is unlikely that fluorescence labeled GLP-1RA can enter the rat CSF.^[95] Thus, peripheral GLP-1 and GLP-1RAs may not cross the rodent BCB, commented very recently by Buller and Blouet in their 2024 review article.^[99] Consistently, a clinical study has reported that transfer of Liraglutide from blood to CSF was very minimal in patients with T2D who had received a long-term Liraglutide treatment.^[104] In the study conducted by Christensen and colleagues, patients were treated with Liraglutide for 5–22 months, and the treatment has effectively induced the weight loss of 7–11 kg.^[104] They reported that plasma Liraglutide level but not CSF Liraglutide level (at extremely low level) that tended to correlate with body weight loss in patients who has received the long-term Liraglutide treatment.^[104]

Can gut produced native GLP-1 or peripherally administrated GLP-1RAs access brain via CVOS

A few recent studies have suggested that GLP-1RAs can access certain CVOs following their peripheral administration.^[101,105] This may not occur for gut produced native GLP-1, as its half-life is very short. GLP-1R is likely expressed in ME in the hypothalamus and the AP in the hindbrain (Figure 1), while exact cell population or cell linage within these two CVOs that express GLP-1R needs further investigation and clarification.^[102,106] The utilization of electron microscopy allowed the detection of GLP-1R immune-reactivity on ME dendrites and on the surface of axon varicosities and axon terminals.^[106] However, functional consequences on ME or AP GLP-1/GLP-1R signaling activation are remaining controversial at the current stage. AP GLP-1 signaling is necessary and sufficient for producing taste avoidance following peripheral GLP-1RA administration, suggested by Zhang and colleagues recently.^[107] However, an early study reported that electrolytic ablation of the AP, or another CVO, the SFO, or AP plus SFO, did not prevent the anorectic effects of Ex-4 peripheral administration.^[108] Further cell linage studies are required to clarify GLP-1R expressing neurons in different CVOs, followed by the utilization of cell linage specific knockout approach to clarify their role in mediating the anorexic and other brain effect of peripheral GLP-1RA administration. It is worth mentioning that ME or AP GLP-1R-expressing neurons may mediate the anorexic effect of peripheral GLP-1RA administration directly, without allowing the drugs entering the brain region furtherly.

Conclusion

As we have discussed above whether gut-produced native GLP-1 or peripherally administrated GLP-1RA therapeutic agents access BBB remains controversial to date. The receptor-independent passive access theory presented by Kastin and colleagues^[88,89] challenges the classical dogma that BBB only allows its paracellular access of molecules that are smaller than 500 Da, while native gut-produced GLP-1 and GLP-1RA therapeutic agents are much larger than 3000 Da. The receptor-mediated theory via a transcellular mechanism by Fu and colleagues^[95] is not convincing yet as a few recent studies failed in the detection of GLP-1R in brain endothelia cells, the foundation of BBB.^{[97, 98, 99]} Clear interaction of fluorescence labeled GLP-1RA with mouse ependymal cells of the choroid plexus was virtually absent when high-resolution microscopy was applied.^[103] Importantly, the detected interaction also occurred in GLP-1R knockout mice. We hence intend to suggest that gut-produced native GLP-1 and peripherally administrated GLP- 1RA as therapeutic agent do not access the brain via BCB. Indeed, CSF GLP-1RA level is very minimal in patients with long-term Liraglutide treatment. Peripherally administrated GLP-1RAs may access the brain via ME and AP, as well as other selective CVOs. Key questions that remain to be addressed include exactly what neuronal GLP-1R producing cell population that mediates the anorexic function of peripheral GLP-1RA administration? The study by Smith and colleagues in profiling GLP-1R expression brain cell paved an avenue for our further investigations.^[100] Will the access be sufficient in triggering PKA activity as well as c-FOS activation in the AP, the nucleus of the solitary tract (NTS), and elsewhere? As we have mentioned above that endogenous GLP-1 has a very short half-life (around 2 min) due to the existence of the degradation enzyme DPP-4.^[2,109] A study has shown that only 25% of gut produced endogenous GLP-1 reaches the portal circulation intact, whereas further cleavage happens in the liver causes an additional 40%–50% reduction of intact GLP-1, leading to only 10%–15% of those enters the systemic circulation.^[10] As illustrated in Figure 2, peripheral GLP-1 may send the signal to the brain through a neuroendocrine signaling mechanism, without entering the brain. This could be mediated by GLP-1R that are expressed on the portal vein that is heavily innervated by the vagal afferent nerve. This signal can be further sent to the nodose ganglion at the base of the skull, where a collection of neuronal cell bodies is located.^[110] Portal vein GLP-1R expression has been demonstrated by several previous investigations.^{[111, 112, 113]}

Figure 2

Proposed mechanism of brain access of endogenous GLP-1. Followed by nutritional stimulations, endogenous GLP-1 is secreted by enteroendocrine L cells located mainly in the distal intestine. Due to the existence of the degradation enzyme DPP-4 in the circulation, only 10%–15% of active GLP-1 will reach the systemic circulation and possibly exerts its effects centrally. Thus, it is possible that central effects of GLP-1 is mediated by a neuroendocrine signalling mechanism via the vagus nerve. Gut-produced GLP-1 activates GLP-1R in the portal vein where its concentration is the highest and sends the projection via vagal afferent neurons to their cell bodies called nodose ganglion, which also expresses GLP-1R. The nodose ganglion terminates in the nucleus tractus solitarius (NTS) in the brainstem and further transmit the signals to other brain regions to regulate metabolic and other responses. This evolutionarily developed mechanism via organ-organ communication, if exists, should also participate in mediating the anorexic and other brain functions of peripherally administrated GLP-1RAs.

The portal vein and vagus afferent nerve mediated effect of peripheral native GLP-1_{7–36 amide} administration on attenuating intestinal fat absorption has been reported recently by Hoffman and colleagues in two rodent models¹¹⁰. Briefly, Hoffman and colleagues have demonstrated that in obesogenic diet induced obesity and insulin resistance hamster as well as mouse models, portal vein injection of native GLP-1_{7–36 amide} effectively improved lipid homeostasis, reflected by reduced postprandial and fasting plasma lipid levels. Such inhibitory effects of portal vein GLP-1_{7–36 amide} injection were then shown, with the application of pharmacological and surgical denervation, to be dependent on intact afferent vagal singling cascade. Furthermore, such inhibitory effects were lost in the GLP-1R knockout hamster model.^[110] It remains to be determined whether gut produced native GLP-1 or peripherally administrated GLP-1RAs exert their anorexic and other brain functions involves such organ-organ communication.

One may argue that with extended half-life, therapeutic agents GLP-1RAs are resistant to DPP-4 mediated degradation, allowing them to reach to the brain to exert their anorexic and other brain functions directly. However, if the portal vein system is evolutionarily developed for peripheral-central communications, including sending the signal to the brain for native GLP-1 to exert its anorexic and other brain functions, this existing system likely also mediates the pharmacological function of various GLP-1RAs. Finally, we have learned that brain not only express GLP-1R but also the hormone GLP-1 per se. The same proglucagon gene is expressed in rodent gut, pancreatic islet α-cells and the brain.^[114] The same post-transcriptional machinery exists for processing the pre-hormone proglucagon in the gut and the brain. We are far away from understanding exactly why brain needs to express its own GLP-1 and how brain GLP-1, gut-produced GLP-1, as well as their receptor GLP-1R expressed in selected cell linages, orchestrate the anorectic program physiologically and patho-physiologically. Biomedical research enters the new era as we are now paying more and more attention to inter-organ crosstalk.^[115] It appears that fundamental breakthroughs on organ-organ communications are needed in this area before we can advance our understanding on the anorectic effect and other brain functions of peripherally administrated GLP-1RAs.