Oral Medications Using GLP-1 and Similar Compounds for Treating Obesity

Nowadays, it is known that multiple loci in the human genome are associated with adipogenesis and the physiological function of the adipose tissue. It is well established that the peroxisome proliferator-activated receptor gamma (PPAR-γ) is the master regulator of adipogenesis and the differentiation of pre-adipocytes to adipocytes. Worth noting is that there are several other factors (such as the C/EBPs, KLFs, PRDM16, and PGC-1α) steering the development of the adipose tissue. The expression levels of each factor determine the phenotypical characteristics of the adipose tissue [7].

Morphologically, the mammalian adipose tissue can be divided into two major types: white adipose tissue (WAT) and brown adipose tissue (BAT) [8]. White adipocytes are characterized as large spherical cells that store energy-providing lipids in the form of lipid droplets and are the main components of the WAT. This type of tissue is distributed in two anatomical compartments of the body: the subcutaneous and the visceral [9]. Obesity is characterized with ectopically expressed WAT, in areas such as mesenteric, omental and retroperitoneal fat. BAT, on the other hand, is more metabolically active than WAT due to the high mitochondrial activity, and functions by dissipating energy through heat [10]. It is known that the activity of BAT slows down with aging. Nevertheless, there is a third type of adipocytes, the so-called beige (or brite) adipocytes. These are formed from white adipocytes upon cold exposure or induced by activation of the PPARγ receptor and they demonstrate a similar energy-dissipating activity as BAT. The balance of both types of adipose tissue is crucial for maintaining homeostasis.

The adipose tissue, however, is more complex than this—it is composed of more than just adipocytes, and it also contains endothelial cells, macrophages, neutrophils, lymphocytes, and fibroblasts. This makes the adipose tissue metabolically active, which is visible through the production of various biologically active substances (called adipokines) important for cellular signaling and maintenance of homeostasis: enzymes, lipid transport molecules, cytokines, and complement-related proteins. Some of the most important substances secreted by the adipose tissue are hormones, such as adiponectin, an insulin-sensitivity regulator [11].

Impaired adipogenesis and dysfunction of BAT are well-established contributors to obesity and its associated comorbidities, including insulin resistance, type 2 diabetes mellitus (T2DM), inflammation, and cardiovascular disease. In obesity, adipose cells lose their ability to store excess lipids effectively, leading to ectopic fat deposition in organs such as the liver and muscles. This ectopic fat accumulation impairs cellular function and exacerbates metabolic dysregulation. Subcutaneous adipose tissue in obesity often experiences reduced vascularization, resulting in hypoxia. This hypoxic environment promotes fibrogenesis and the recruitment of macrophages, further perpetuating tissue inflammation [7,12]. The chronic inflammatory state, driven by the secretion of tumor necrosis factor α (TNFα) and other pro-inflammatory cytokines, disrupts insulin signaling pathways, causing insulin resistance. Over time, these metabolic and inflammatory changes increase the risk of T2DM, atherosclerosis, and other obesity-related conditions [2,7].

The primary objective of obesity treatment is weight loss, and this parameter is also established as a first measure of therapy efficacy when comparing different treatments. However, it is important to note that weight loss alone does not necessarily result in health benefits. Currently, more importance is given to loss of excessive fat specifically, while preserving or increasing muscle mass in obese individuals [13]. The aforementioned parameters, BMI and waist circumference, are the main readily available parameters. However, a whole range of other metabolic parameters, such as plasma triglycerides, cholesterol, Hb1Ac, inflammation status etc. can also be taken into account when evaluating a particular obesity treatment and its benefits [13,14].

As outlined by WHO, the main strategies in the obesity management encompass healthy diet, physical activity, and medical and surgical interventions [15]. The initial two measures belong to conservative management of obesity, which entails dietary interventions (such as lowering the caloric intake) together with physical activity and behavioral therapy, and they necessitate the intervention of clinicians, psychologists, and dietitians [5]. The latter two interventions encompass (1) bariatric and metabolic procedures (bariatric surgery and endoscopic procedures) and (2) pharmacotherapy of obesity. These interventions will be discussed briefly in the following text.

Besides the GLP-1 receptor, one additional target is the GIP receptor. The GIPR, same as the GLP-1R, belongs to the glucagon subfamily of class B G-protein coupled receptors (GPCRs) and is present on pancreatic cells (pancreatic α-, β-, δ- cells and pancreatic polypeptide cells), in the central nervous system, blood vessels, immune system, heart, BAT (unlike GLP-1), and bone cells [57,96]. Unlike the GLP-1 pathway, which remains preserved even in the diabetogenic environment, the GIP response is impaired in diabetes [98,99]. It is also interesting to note that weight loss is observed in both pharmacological combinations: GLP-1- and GIP- receptor agonism, as is the case tirzepatide, but also in combination with a GIP receptor antagonism, which is case in AMG 133 [96].

AMG 133 or Maridebart Cafraglutide, developed by company Amgen Amgen (Amgen, Thousand Oaks, CA, USA), is an optimized drug conjugate containing two GLP-1 analogs coupled to human monoclonal antibody against the GIP receptor. It has recently successfully completed phase 2 clinical trials for obesity (NCT05669599) and has proceeded to Phase 3 for once-monthly subcutaneous administration [133,134].

The exact mechanism of weight loss for both combinations is still unclear, but there are several hypotheses. For tirzepatide, three concurrent mechanisms are provided as explanation. Firstly, it is believed that weight loss is based on the biased signaling on GLP-1 receptor towards diminished receptor desensitization. Secondly, the central antiemetic effect of GIP receptor agonism is believed to increase the tolerable dose of GLP-1 component. Lastly, the fatty acid component in tirzepatide is a biochemical tool that could enhance brain delivery and exert higher central appetite-suppressing effect driven by both, GLP-1- and GIP- receptor agonism [99]. On the other hand, the success of GIPR antagonist/GLP-1 agonist combination origins on the theory that GIP receptor agonists induce receptor internalization upon chronic activation and, therefore, function in a similar manner as GIPR antagonists [99,135,136]. However, there are opposing theories suggesting also that the exposure to an antagonist re-sensitizes the GIP receptors, resulting in the increased number of receptors on the cell surface and in restoration of the GIP response [137].

GLP-1/glucagon co-agonism has also a synergistic effect in the weight loss. Due to the paradoxical catabolic-anabolic actions of glucagon, a combination with a GLP-1 analog manages to achieve a negative energy balance. Centrally, both glucagon and GLP-1 promote satiety, mainly through hypothalamus-vagus-liver axis, but glucagon increases energy expenditure additionally. The increased energy expenditure seems to be result of increased thermogenesis and of various catabolic effects with a subsequent counter-regulatory anabolic responses trying to restore the exhausted metabolites (e. g. hepatic gluconeogenesis). Glucagon also exerts an insulinotropic effect on β-cells in the postprandial state. The appetite-suppressing effect of GLP-1 alone seems to be largely independent of energy expenditure, while antihyperglycemic effects of GLP-1 can effectively counteract the hyperglycemic potential of glucagon. This could provide a plausible explanation for the strong synergistic potential of GLP-1R/GCGR co-agonists and their subsequent success in clinical trials [138,139]. Some of the dual GLP-1R/GCGR agonists, such as Survodutide (Boehringer Ingelheim, Ingelheim am Rhein, Germany) and Efinopegdutide (Merck & Co., Kenilworth, NJ, USA), have already obtained the FDA fast-track designation for treatment of MASH (Survodutide: LiverageTM programme, NCT06632444; Efinopegdutide: NCT05877547) due to their high efficiency in liver fat reduction. In addition to the benefits on liver, GLP-1R/GCGR co-agonists showed a significant improvement of the HbA1c value in diabetes patients [132]. Survodutide has also advanced to phase 3 clinical studies for treatment of obesity (SYNCHRONIZE™ programme). In addition, the phase 3 studies on survodutide are evaluating alternative subcutaneous formulations (NCT07221591) as well as its cardiovascular safety (NCT06077864).

The GLP-1R/GIPR/GCGR co-agonists are also currently under investigation for their capacity in treatment of T2DM, obesity, and MASH. Retatrutide (Eli Lilly and Company, Indianapolis, IN, USA) is currently in phase 3 of clinical trials (TRIUMPH programme), where its antiobesity effect is being compared with placebo (NCT06383390 and NCT05929066) and with tirzepatide (NCT06662383). Several other ongoing clinical trials are investigating the effect of retatrutide in patients with obesity and other comorbidities such as diabetes, cardiovascular disease, osteoarthritis of the knee, chronic kidney disease (CKD), and chronic low back pain. Another GLP-1R/GIPR/GCGR triple agonists, Efocipegtrutide (HM15211, Hanmi Pharmaceutical, Seoul, South Korea), has completed phase 1 for treatment of obesity (NCT03374241) and is currently in phase 2 for treatment of MASH (NCT04505436).

Further combinations that focus on the synergistic effect with GLP-1 include those with peptide YY (PYY) and amylin analogs, which both promote appetite suppression in the central nervous system. Amylin is a peptide hormone co-secreted with insulin from the pancreatic ß-cells. Its analog cagrilintide is being tested in combination with semaglutide (Cagrisema, Novo Nordisk, Bagsværd, Denmark: semaglutide 2.4 mg/cagrilintide 2.4 mg), which is currently in Phase 3 clinical trials. The resulted weight loss of 17.1% after 20 weeks is significantly higher than the weight loss of 9.5% for the same time period with same dose of semaglutide alone [140].

Peptide YY1-36 is an enteric hormone co-secreted from intestinal L-cells together with GLP-1 and oxyntomodulin. After being cleaved by DPP-4 to PYY3-36, it activates preferentially the Y2 and Y5 receptors in arcuate nucleus of hypothalamus and has a satiating effect due to the Y2 agonism [141]. It was observed that the postprandial concentrations of PYY and GLP-1 increase 10- to 20-fold following the RYGB. This may explain the mechanism of reduced food intake after the surgery, as it was shown that the concomitant inhibition of both GLP-1 and PYY3-36 pathway had an opposite effect, namely increasing the food intake after RYGB [142]. The GLP-1/PYY combination seems promising, as the combination of PYY analog NNC0165-1875 (Novo Nordisk, Bagsværd, Denmark) together with semaglutide (2.4 mg) have already completed the phase 2 clinical trials. However, the project has been terminated without specifying the reason [143].

Bimagrumab is a monoclonal antibody which blocks activin type II receptor and is also being tested in combinations with GLP-1 agonists. It promotes muscle hypertrophy by blocking the atrophic effect of the endogenous ligands, such as activin and myostatin [144]. It is in phase 2 clinical studies for treatment of sarcopenia (NCT02333331). The effects of bimagrumab are, however, not only restricted to myoblasts, but they expand further to adipocytes, concretely to BAT. For instance, lineage tracing show that BAT, unlike WAT, shares common progenitor cells with skeletal muscle tissue. In addition, myostatin is proven to be a potent inhibitor of brown adipocyte differentiation [145]. Bimagrumab, as a myostatin inhibitor, increases the energy expenditure and promotes the lean muscle mass growth even under conditions of a negative energy balance [146]. The combinations of bimagrumab with both semaglutide and tirzepatide are currently being investigated in clinical trials sponsored by Eli Lilly (Eli Lilly and Company, Indianapolis, IN, USA). The phase 2 clinical trial evaluating the combination of semaglutide and bimagrumab (NCT05616013) reported results after 48 weeks of treatment, showing that 63.9% of participants achieved a body weight reduction greater than 15%. Results from the 72-week analysis are still pending. A similar study is underway to assess the combination of bimagrumab and tirzepatide for the treatment of obesity in patients without T2DM (NCT06643728). A parallel trial in patients with T2DM (NCT06901349) was also initiated but has since been withdrawn for strategic reasons by the sponsor.

The term "small-molecule drugs" refers generally to chemically synthesized compounds with a low molecular weight (usually under 500 Da) [147]. The synthetically designed, small-molecule GLP-1 analogs could be a prominent alternative to the peptide drugs, primarily due to the economic reasons.

Orforglipron (LY3502970; Eli Lilly and Company, Indianapolis, IN, USA) is currently being evaluated in several phase 3 clinical studies for the treatment of obesity, each focusing on distinct aspects of disease management. These include investigations into its effects on obesity-related comorbidities (NCT06672939), obstructive sleep apnea (NCT06649045), weight maintenance (NCT06584916), hypertension in obese patients (NCT06948422), obesity in patients with T2DM (NCT06972459), osteoarthritis (NCT07153471), urinary incontinence (NCT07202884), and its efficacy and safety in pediatric participants aged 6–17 years (NCT06672549).

Pfizer had initiated phase 1 and phase 2 clinical trials with another small-molecule GLP-1 receptor agonist, danuglipron (Pfizer Inc., New York, NY, USA). However, during dose-optimization studies (NCT06568731 and NCT06567327), a single asymptomatic participant experienced a potential case of drug-induced liver injury, which resolved upon discontinuation of danuglipron. Following a comprehensive review of data from previous clinical studies, the company decided to discontinue the development of danuglipron. Although the phase 2 study evaluating the combination of danuglipron with the small-molecule GIPR antagonist PF-07976016 was also terminated, Pfizer announced its intention to continue developing PF-07976016 independently. Results from the ongoing phase 2 study in adults with obesity (NCT06717425) are expected in December 2025.

The company Genentech (Genentech, South San Francisco, CA, USA), which is part of Roche group, announced after the acquisition of Carmot Therapeutics (Berkeley, CA, USA) the positive results of phase 1 clinical trials for CT-996 (NCT05814107) [148]. The CT-996 is the same as orforglipron and danuglipron, a small-molecule GLP-1 analog designed to be a biased GLP-1 receptor agonist that preferentially activates cAMP signaling over β-arrestin recruitment [149]. Of note, with this acquisition, Roche also obtained access to portfolio of two other small-molecule drugs: CT-388 and CT-868, which are both GLP-1/GIP receptor co-agonists [150]. Roche (Basel, Switzerland) has recently initiated phase 2 studies on RO7795081 (CT-996) in patients with obesity (NCT07081958) and patients with T2DM (NCT07112872).

The surge of newly initiated clinical studies, together with the intensifying efforts of pharmaceutical companies to keep pace with the rapidly expanding obesity market, underscores the value of small-molecule approaches in overcoming many of the challenges associated with oral peptide administration. However, the question remains to what extent they are specific for target receptors, and what the consequences of potential off-targeting might be [151]. Therefore, the ongoing clinical trials will seek to provide the answers.

To summarize the current approaches for obesity treatment based on GLP-1 analogs, when it comes to the drug design, there are two main strategies:(1)Multiple receptors targeting: This strategy involves the peptide drugs which are binding two or more (usually incretin) GPCRs. The rational peptide design and the conserved key residues within binding pockets of GPCRs both enable that a created peptide molecule or peptide conjugate bind specifically to multiple targets at once [152]. Since all of the abovementioned incretin mimetics are agonists to GPCRs, this kind of poly-pharmacology will be very attractive for their future drug development [81].(2)Biased signaling: Some GLP-1-biased agonists mentioned above, such as orforglipron and danuglipron, activate a biased signal cascade, in which GPCR binds preferentially or almost exclusively to G-protein instead to another associated effector protein, in this case, concretely with the ß-arrestin 2. In this manner, the GLP-1 receptor internalization can be avoided. This concept offers the possibility of selecting the signaling pathway, which will have an increased therapeutic effect, while the undesired downstream signal cascades and consequent side-effects could be minimalized [81]. However, it is important to note that biased signaling can be achieved with peptide analogs as well, and that small-molecule drugs potentially show less receptor specificity than peptide analogs [153].

In September 2019, as the FDA and EMA approved the first oral dosage form of semaglutide (Rybelsus^®), SNAC was brought into the public eye as a gastrointestinal PE. Here, the delivery of semaglutide was achieved by a tablet that remains and completely erodes in the lower stomach region [164]. The SNAC was already granted GRAS status in 2015, when it was used as PE in the first oral formulation of vitamin B12 designed to reach the efficacy of an injection (Eligen B12™, Emisphere Technologies Inc., Roseland, NJ, USA) [169]. It is evident that this helped introducing SNAC as drug excipient in Rybelsus^® years later. Novo Nordisk also acquired Emisphere in 2020 to, as the company announced, further develop Eligen SNAC technologies in its current and future drug pipeline assets for biologic medicines [170,171].

Acting as a local buffer, SNAC, a synthetic N-acylated amino acid derivative of salicylic acid, increases pH in the microenvironment, and so it protects the semaglutide from acidic degradation by gastric enzymes. The absorption-enhancing mechanisms of SNAC are also based on its formation of non-covalent complexes with semaglutide, leading to the augmented lipophilicity of the peptide and consequent facilitated transmembrane transport. Moreover, it also facilitates the peptide absorption by fluidizing the biomembranes without affecting the tight junctions and epithelial integrity of mucosa, allowing the peptide to predominantly follow the passive transcellular transport. Finally, SNAC also promotes the monomerization of semaglutide, thus preventing its oligomerization. The oligomerization is an unfavorable process that leads to an increased dynamic size and molecular weight of the drug. This, in turn, has a detrimental effect on the rate of absorption [165,168,169,172].

Chen et al. described a new delivery system for semaglutide that combines physical (microneedle) and non-physical (PE) modes of delivery in a capsule device. Here, they used SNAC as PE. The device was tested in Yorkshire swines as a large animal model. It was demonstrated that the self-triggered microneedle system provides improved tissue interaction and prolonged fixation to the gastric mucosa. This is crucial for the permeation-enhancing activity of SNAC and the resulting local delivery of semaglutide [173,174].

Another example of an "old hand" PE is chitosan [175]. Based on its remarkable permeation-enhancing properties due to its charge, chitosan has emerged as an interesting substance for oral peptide delivery [176,177]. The positive charges of glucosamine residues in chitosan biopolymers under mildly acidic conditions promote electrostatic interactions with the negatively charged mucus layer reducing mucus clearance. Furthermore, the positively charged amino groups can interact with negatively charged integrin α_vβ₃ on the apical membrane of the intestinal mucosa. This initiates integrin clustering along the cell border, activating the downstream signaling cascade that transiently disrupts tight junction integrity [165,178,179].

For oral delivery of antidiabetic drugs (e.g., GLP-1 and its analogs), various nanoparticle-based formulations have been combined with chitosan and this enhanced the drug uptake and the therapeutic efficiency due to its physicochemical properties [163,180,181,182]. An overview about this nanoparticle-based delivery for GLP-1 peptides is given in the chapter about nanostructured drug delivery systems (see Section 6.4 for more details).

To develop oral peptide formulations, MCFAs, such as sodium caprylate (C8) and sodium caprate (C10), are studied in depth as PEs with surfactant-like actions. Both fatty acids are FDA-approved food additives, which makes them desirable PEs due to relatively low regulatory hurdles. So far, Rybelsus^® and Mycapssa^® (Amryt Pharma, Boston, MA, USA) are the only marketed oral peptide-based drugs utilizing PE-based absorption processes. While Rybelsus^® tablets use SNAC as PE for semaglutide, Mycapssa^® uses C8 in the oral octreotide formulation of a lipidic enteric-coated capsule [164,166].

In detail, C8 is an eight-carbon saturated fatty acid with a molecular weight of 166.19 g/mol and a pKa of 5.19, while C10 is a sodium salt of a saturated fatty acid composed of 10 carbons with a molecular weight of 194.25 g/mol and a pKa value of 4.95 [164]. With respect to the absorption-enhancing effects of fatty acids, the most widely accepted mechanisms for C8 and C10 include the modulation of tight junctions (paracellular) and the perturbation and fluidization of cell membranes (transcellular). Up to now, their absorption-enhancing pathways remain unclear. A possible explanation combines multiple mechanisms depending on the physicochemical properties of PEs, the physiological conditions (cell type and pH) and the concentration of PE and drug at the absorption site [164,165,183,184].

In 2022, Tran et al. (Eli Lilly, Lilly Research) demonstrated that amphiphilic C10 was superior to other tested PEs (e.g., SNAC) in increasing oral absorption of a fatty acid acylated GLP-1R/GCGR co-agonist peptide (4.5 kDa) in a rat jejunal closed loop model [185]. A subsequent study of Tran et al. published in 2023 evaluated the intestinal absorption of GLP-1R/GIPR co-agonist (here termed as LY) and it compared the effects of C10 as permeation enhancer in combination with two different peptide drugs. For comparison, they used C10 in combination with semaglutide and with a GLP-1/GIP analog (here also termed as LY). The data clearly revealed higher oral absorption of the GLP-1/GIP analog in rats compared to semaglutide. This effect was explained by enhanced proteolytic stability of the GLP-1R/GIPR co-agonist and the more pronounced monomeric state of the GLP-1R/GIPR co-agonist in solution. The group also investigated the uptake mechanisms in vivo in rats and minipigs. It was showed that C10 promotes both para- and transcellular pathways of the peptide uptake [186].

The studies on C10 as permeation enhancer continued and the company (Eli Lilly) posted in January 2025 the results of phase 1 clinical studies, where the safety and tolerability of the subcutaneously and intravenously administrated LY3493269 (NCT04178733), as well as the orally given drug (Part 1A: NCT04498390; Part 1B: NCT05794243) were evaluated. Of note, LY3493269 is a GLP-1R/GIPR co-agonist and could be a similar or even the same drug as previously mentioned GLP-1/GIP analog. The study part 1B was terminated by sponsor as enough data was obtained from study part 1A. The orally administered drug was co-administered with C10 in various doses. The drug was administered in doses ranging from 8 mg to 48 mg in a test capsule together with C10 that was co-administered in a separate capsule in one of the two dosages: 250 and 500 mg. The latter (1B) study was designed to have 3 arms and to compare two different test capsules (capsule test 1 and 2, each 4 mg peptide), both co-administered with 280 mg C10. The third arm or the reference group was a tablet (4 mg drug), co-administered with 300 mg SNAC. Of note, a previous another study (NCT04682106) on the same peptide tested up to 600 mg SNAC. The results indicated that the AUC of test capsule 2 was approximately 30% relative to that of the reference tablet formulation. Moreover, the results from recent preclinical studies on monkeys which compared erodible tablets for gastric delivery containing different PEs (C10, SNAC, and 5-CNAC). This study highlighted that the influence of the dissolution profile in combination with a particular PEs on the peptide bioavailability. The study confirmed that both SNAC and C10 are suitable PEs for gastric delivery of the investigated peptide (LY, a GLP-1/GIP analog). The absolute bioavailability for the particular peptide was around 4% in both cases. Of note, the reference formulation containing semaglutide and SNAC showed lower bioavailability [187]. In January 2025, Tran et al. also published a study on mucoadhesive tablets designed for gastric delivery a GIP/GLP-1 analog peptide. The peptide was co-formulated with SNAC in a formulation layer that was surrounded on one or on both sides with mucoadhesive sodium alginate layer. Although the gastric retention was successful, it did not result in increased bioavailability. This can be explained by slower dissolution rate of the peptide in vivo due to the swelling sodium alginate polymer [188].

An interesting study was performed by Niu et al. [167] (Novo Nordisk): they combined the two PEs, SNAC, and C10 in one tablet formulation. Permeability tests were performed in a gastric organoids-based cell model. The combination of SNAC and C10 was superior to either SNAC or C10 alone. Further optimization of tablet formulation was performed by adjusting certain parameters such as PE amount, relative ratio between C10 and SNAC, and the peptide dose. The oral formulation for the GLP-1 analog was tested in beagle dogs, in which the combination of SNAC and C10 was again superior to either SNAC or C10 alone with respect to the bioavailability results. Finally, oral administration was performed in humans, where the combination of SNAC and C10 displayed similar bioavailability compared to the formulation with SNAC only. These results suggested translatability challenges of the permeation enhancement capability of C10 in the stomach between dogs and humans differing in physiology, gastric pH environment as well as concentration of C10 in the gastric lumen. To exploit the full potential of oral peptide therapeutics, this study highlights that more clinically relevant studies are of utmost importance in bridging the gap between preclinical research and clinical application.

Ionic liquids (ILs) represent an alternative permeation-enhancing approach regarding oral peptide delivery development. In brief, ILs represent a class of organic salts being liquid at temperatures below 100 °C. Their low melting point is attained due to bulky and asymmetric organic cations and organic/inorganic anions being present in the structure of ILs. This anion–cation salt structure of ILs shapes their special physicochemical properties including non-volatility, high thermal, and chemical stability, as well as solubility improvement of a wide range of insoluble substances and good water miscibility [189,190,191]. Based on their highly tunable physicochemical properties, ILs allow specific design for delivery systems to fulfill drug-dependent requirements and optimizations. For example, ILs are capable of enhancing solubility of poorly soluble drugs, improving the stability and permeability of drugs in the intestinal mucus as well as promoting penetration across the intestinal epithelial cell barrier [192,193].

Among ILs, choline and geranic acid (CAGE) is an auspicious delivery vehicle for oral drug applications [194]. Banjeree et al. described a highly effective oral delivery system for insulin by incorporating CAGE administered in enterically coated capsules to non-diabetic rats via oral gavage [195]. The study of Agatemor et al. reported beneficial effects of CAGE in subcutaneous GLP-1 delivery [196]. Another approach focused on the preparation of imidazole-based ionic liquids that were functionalized to silica nanoparticles for oral drug delivery of insulin. By integrating ILs into nanoparticles, a controlled and prolonged release profile can be achieved [197]. The described drug delivery approaches might serve as an inspiration to use ILs for future developmental strategies in oral peptide delivery, including GLP-1 applications.

Another interesting aspect of CAGE is reported in the study of Nurunnabi et al. They discussed a body weight reducing effect of orally administered CAGE by decreasing fat absorption through the intestine and concomitantly, providing a feeling of satiety in rats [198]. These observations might provoke the idea of a combinatorial treatment with CAGE and GLP-1 for obesity. Further, recently published data of Rebollo et al. described a salcaprozate-based IL formulation, namely CHONAC for choline salcaprozate. In this work, they have developed this CHONAC IL-based oral formulation for gastric delivery of a GLP-1 analog [199].

Although a lot of beneficial effects of ILs for (oral) drug delivery are reported, it is worth keeping in mind that the permeation-enhancing properties of ILs might correlate with toxicity events. Therefore, a rational design of ILs to guarantee high safety and performance standards is of utmost importance [190,193].

To conclude this section, the use of PEs is a common strategy to improve absorption and bioavailability of oral GLP-1 peptides. PEs with different mechanisms of actions can be combined in formulation development followed by a synergistic enhancement of peptide absorption. PEs should be simple and easy to manufacture while possessing a high safety profile as well as favorable absorption-enhancing effects. The "one-size-fits-all" approach may not be effective for oral peptide formulations because it is a complex interplay between various physiological and physicochemical properties of the tested organism and the applied PEs, respectively [165,185].

The main function of the nanocarrier is to protect the active substance from degradation by intestinal fluids and enzymes; however, it can also serve for the purpose of sustained drug release. This is especially relevant for achieving a practical dosing pattern. Furthermore, the surface modifications can be employed for one or more of the following purposes: (1) to provide the means of guidance, (2) to increase the cellular uptake, and (3) if possible, to facilitate navigation of the nanocarrier within the cell and protect it from lysosomal degradation. In other words, the ligands on the surface of nanocarriers can affect not just the manner of uptake into the cell, but also the subsequent transportation route and fate of the nanocarrier. Nevertheless, the primary challenge in the oral delivery of peptides is still overcoming the intestinal mucosa. In this regard, there is potential for targeting of several receptors in the gastrointestinal mucosa.

Going back in history, the first published work about small scale carrier systems for GLP-1 and its analogs focused on microparticles (MPs) rather than on nanoparticles. It was in the year 2000, when Joseph et al. encapsulated a DPP-4-resistant analog of GLP-1 (D-ala2-GLP-1) in microspheres consisting of poly(lactide-co-glycolide)-COOH (PLGA-COOH) and olive oil (Table 1). The -COOH end group of PLGA was meant to prolong the gastrointestinal passage time and olive oil to enhance the release rate. To deliver therapeutic concentrations of D-ala2-GLP-1 to normal CD1 and diabetic db/db mice by the discussed microspheres, a very high dosage of 250 µg/mouse was used. The study showed that the microparticles lower the glycemic response to orally given glucose significantly compared to pure D-ala2-GLP-1, which lost its activity within 4 h. An extended effect over at least 6 h was achieved by the orally administered microencapsulated D-ala2-GLP-1 [221].

Another type of microparticles was developed by Zhang et al. in 2014 [222]. Exenatide was encapsulated in cross-linked alginate-hyaluronate microspheres, which keep compact in acidic conditions but swell in alkaline medium. Two hours after oral administration, a glucose lowering effect was measured, which lasted for a moderately prolonged period of another 4 h. The relative bioavailability was calculated to be around 10%, meaning that the 10-fold dosing was needed for the oral exenatide microspheres (500 µg exenatide/kg) to achieve a similar effect as the subcutaneously administered drug (50 µg/kg).

One year later, Soudry et al. prepared self-assembled exenatide nanoparticles made of bovine serum albumin (BSA) and dextran, which can be cross-linked via its reactive OH-groups and protect the delivery system from degradation. To further enhance the stability of these primary nanoparticles, reduce the drug release but enhance its absorption in the intestine, they were embedded in gastro-resistant microparticles, consisting of hydroxypropyl methylcellulose (HPMC) and Eudragit^® L100-55 (Evonik Operations GmbH, Darmstadt, Germany), which promote the passage through the enterocytes. Drug loading and encapsulation efficiency of the particles stayed rather low. After internalization of the microparticles into enterocytes, they were transported via the lymphatic pathway to the systemic blood circulation as dextran is augmenting the lymphatic uptake. By this, the first pass effect was circumvented and the relative bioavailability of oral microparticles (dosing: 165 µg exenatide/kg) compared to s.c. injection of the product Byetta^® and free exenatide solution (dosing: 65 µg/kg) reached stunning 77% and 89%, respectively [204]. Pharmacodynamical investigations in diabetic mice revealed that the formulation (100 µg exenatide/kg twice daily) significantly improved different glycemic parameters to a similar extent as the s.c. exenatide solution (40 µg/kg twice daily) and stopped weight gain [200].

In 2021, Ren et al. had the idea to load exenatide-PLGA nanoparticles into yeast cell wall particles resulting in an irregularly shaped nano-in-micro carrier with adequate sustained release properties. Yeast cell wall particles are porous carriers targeting intestinal phagocytic cells. Their implementation enhanced the gastrointestinal stability of exenatide and improved the blood glucose lowering effect significantly over pure exenatide-PLGA nanoparticles [223].

Polylactic acid (PLA) and poly lactic-co-glycolic acid (PLGA) are FDA and EMA approved biodegradable polymers which are widely used for the preparation of drug delivery systems for proteins and peptides. They show a high biocompatibility, good stability, and the physical properties of the polymer can be adjusted to achieve sustained release [267].

In 2013, Wang et al. developed the first exendin-4 encapsulating PLGA nanoparticles. These were coated with positively charged, mucoadhesive chitosan, which enhanced their transport through membranes and cellular uptake compared to the free drug and the uncoated particles [224]. One year later, GLP-1 loaded PLGA nanoparticles were compared according to their in vitro permeability with solid lipid nanoparticles (SLNs) and porous silicon (PSi) nanoparticles in a study of Araujo et al. As the described carrier systems all exhibited negative charge due to the primary materials used, they have been coated with chitosan and demonstrated a stronger contact with cells afterwards, as already described by Wang et al. [224]. Coated PLGA nanoparticles, which were the only ones avoiding the release at pH 1.2, showed the best sustained release properties, followed by PSi nanoparticles. The SLNs caused a burst and sole release of about 40% at low pH in the simulated stomach. The sustained release properties of the PLGA nanoparticles were also proven in permeability studies, where they showed the lowest amounts of GLP-1 permeating across intestinal cell monolayers in 3 h [206].

Subsequently, the same working group around Santos and Sarmento studied the PLGA-chitosan and PSi-chitosan nanoparticles in more detail and added the cell penetrating peptide (CPP) polyarginine R9 in order to improve transcellular transport. CPP are small, polycationic peptides which enhance the drug transport across the endothelial membrane and its oral bioavailability without damaging the cells or provoking a cellular response. Furthermore, these nanoparticles were coated with an enteric resistant polymer, to protect the drug and the carrier system from the acidic environment in the stomach, and were loaded with a DPP4-inhibitor to reduce enzymatic degradation of GLP-1. For these modifications, a microfluidic flow-focusing glass device was used. The final size of the spherical shaped particles was pretty large with values around 60 µm. Only insignificant amounts of GLP-1 were released at acidic conditions. After changing the pH to 6.8, a burst release was observed for the PSi nanoparticles followed by nearly no further release, while the PLGA nanoparticles demonstrated a constant sustained release profile. This can be explained by the different binding of GLP-1, which is loaded into the pores of the PSi nanoparticles through physical adsorption, whereas it is encapsulated inside the polymer matrix of the PLGA nanoparticles, resulting in a longer release time. The integration of the CPP into the nanosystem led to an intensification of the cell-particle interactions [225].

In a further study, in vivo tests on diabetic rats have been performed by Araujo et al. for the PLGA nanoparticles. Two hours after oral administration of the system (200 µg GLP-1/kg) the blood glucose levels decreased while the plasmatic insulin levels increased significantly compared to the control group. Blood glucose stayed low for further 6 h. The combination of GLP-1 and DPP4-inhibitor in the nanoparticles resulted in enhanced hypoglycemic effects over the system without DPP4-inhibitor due to the reduced enzymatic degradation of the drug [226].

The working group around Sun K. coupled PLGA with polyethylene glycol (PEG) to circumvent the enzymatic degradation and prolong the circulation time of exenatide loaded PLGA nanoparticles. The PEG-PLGA nanocarriers were conjugated with Fc-groups as targeting ligand to Fc-receptors (FcRn) on epithelial cells. They achieved a longer residence time in the GI-tract, permeated cells faster and to a greater extent, showing a stronger hypoglycemic effect than unmodified nanoparticles. Additionally, extended hypoglycemic effects by oral administration of the Fc-PEG-PLGA nanoparticles (100 µg exenatide/kg) compared to s.c. injection of exenatide (10 µg exenatide/kg) in diabetic mice were proven [227].

In a further study, the invented nanoparticles were loaded with exenatide complexed with zinc ions to improve the stability of the peptide. Here, the NPs have been decorated with transferrin (Tf) instead of Fc-groups. Tf binds to Tf-receptors on enterocytes, thereby the intestinal permeability and cell uptake of the spherical Tf-PEG-PLGA NPs compared to the non-targeted NPs was enhanced (for more details, see Section 6.1.2). The receptor-mediated uptake was proven by a competitive receptor binding of the targeted NPs and free Tf. The relative bioavailability compared to s.c. injection of exenatide was 6.45% [268]. The same working group loaded exenatide-Zn²⁺ into PEG-PLGA NPs functionalized with low molecular weight protamine (LMWP), which significantly enhanced the penetration of the intestinal epithelium compared to NPs without protamine. A quite high release of nearly 50% of the drug occurred at gastric pH within 2 h. Nevertheless, the oral application of these nanoparticles (100 µg exenatide/kg) to diabetic rats led to a better hypoglycemic effect compared to non-functionalized NPs. The effect lasted for 24 h and a relative bioavailability (vs. s.c. injection of 10 µg exenatide/kg) of 7.44% was achieved [228].

The LMWP-PLGA nanoparticles have also been coupled with dextran (DEX) instead of PEG. Positively charged dextran is a hydrophilic, non-toxic polymer, improving drug encapsulation in nanoparticles and enhancing interactions between NPs and GI tract. However, drug loading and encapsulation efficiency were equal to that of LMWP-PEG-PLGA NPs. LMWP enhanced the affinity of the NPs to epithelial cells and the cell internalization as well as transmembrane transport efficiency. The relative bioavailability after oral administration of the NPs (100 µg exenatide/kg) was 8.4% compared to s.c. injection (10 µg exenatide/kg), with a significant and extended hypoglycemic effect on diabetic mice [229]. The LMWP-DEX-PLGA nanoparticles resulted in a 3% higher relative bioavailability than particles without LMWP but could not nearly reach the relative bioavailability of 89% achieved by the BSA-DEX microparticles developed by Soudry et al. [204].

Furthermore, Song et al. invented a block copolymer, CSKSSDYQC-dextran-PLGA (CSK-DEX-PLGA), and prepared exenatide-zinc loaded nanoparticles out of it. CSK shows a special affinity to goblet cells (see Section 6.1.3 for more details) and has already been used by the same research group four years before for the preparation of exenatide loaded chitosan nanoparticles [239]. The now discussed CSK-DEX-PLGA nanoparticles exhibited enhanced mucus permeation and significantly improved the absorption efficiency in the small intestine compared to DEX-PLGA nanoparticles without CSK. Although a high drug release of 60% occurred at gastric pH within 2 h, a prolonged hypoglycemic response and a relative bioavailability of 9.2% (100 µg exenatide/kg oral vs. 10 µg exenatide/kg s.c.) in diabetic rats after oral administration was measured [213].

Ismail et al. were the first who encapsulated liraglutide in PLGA nanoparticles which achieved a 1.5-fold higher cell permeation than liraglutide solution [230,231]. To provide enteric resistance at low pH, Senduran et al. coated liraglutide loaded PLGA nanoparticles with Eudragit^® L30D (Evonik Operations GmbH, Darmstadt, Germany). Gastric resistance and sustained release over 48 h, attributed to the PLGA polymer matrix, were observed. Oral treatment of diabetic mice with the invented nanoparticles (75 µg liraglutide/kg and 150 µg liraglutide/kg, respectively) twice a day over 9 weeks showed a significant decrease in blood sugar levels 5 weeks after treatment start for the higher dosed NPs. Interestingly, the body weight was not influenced during the test period [232].

Polylactic acid (PLA) nanoparticles loaded with liraglutide were investigated by Uhl et al. in 2020 [238]. To overcome the poor mucosal penetration, they coated the NPs with a cyclic, arginine-rich cell penetrating peptide (cyclic R9-CPP), similar to that used by Araujo et al. 5 years before [225]. Cyclic CPPs additionally enhance enzymatical stability. An increased cell binding, previously also reported by Araujo et al., resulting in a prolonged retention time, significantly enhanced uptake and bioavailability of the NPs modified with CPP in comparison to unmodified NPs. The relative bioavailability compared to subcutaneously injected liraglutide was not tested. As liraglutide leaked out of the NPs within a few hours during storage at 2–8 °C, a freeze-dried, stable formulation for long-time storage was prepared.

Combining liraglutide-PLGA nanoparticles with a cholesterol derivative in a microfluidic chip was reported to enhance the oral absorption and unidirectional transcytosis of the intestinal epithelium by Wu et al. Additionally, a phospholipid-PEG was included to shield the NPs and permit mucus penetration. The NPs demonstrated rather high drug loading and encapsulation efficiency as well as sustained release properties over 3 h. Daily oral administration over 28 days led to the effective control of hyperglycemia. The effect on glycemic parameters of these oral NPs in a very high dosage of 5 mg liraglutide/kg was comparable to s.c. injection of free liraglutide (0.2 mg/kg) and partially eliminated fat deposition [233].

To achieve smart oral delivery of liraglutide, the working group around Sun K. developed self-ablating zwitterionic dilauroyl phosphatidylcholine-hemagglutinin-PLGA NPs showing an irregular core-shell structure and compared them to self-assembled nanoparticles. Zwitterionic carrier systems facilitate overcoming the mucus and epithelial barriers. The lipid shell of the self-ablating NPs degrades when the NPs are captured by lysosomes and then the inner core is exposed. Hemagglutinin facilitates the lysosomal escape of the PLGA NPs by disruption of the lysosomal membrane. The self-ablating NPs demonstrated higher encapsulation efficiency and cell uptake, better stability, bioavailability, and hypoglycemic effect than self-assembled nanoparticles. The relative bioavailability of the orally administered NPs compared to s.c. injection of free liraglutide solution reached approximately 10% [234].

In recent years, the research group around Santos and Sarmento returned to the investigation of PLGA nanoparticles for the oral transport of semaglutide. This time they focused on PLGA-PEG NPs targeting the FcRn on enterocytes as the working group around Sun K. did in 2018, but they improved the specificity of the targeting system by the use of a small FcRn-targeted peptide (FcBP2) and an affibody (ZFcRn), respectively, instead of the IgG Fc-fragment. To improve the encapsulation efficiency, semaglutide was ion-paired with the polyamine spermidine before loading into FcBP2- and ZFcRn-NPs. Due to the complexation with spermidine, only 5% of the drug was set free under gastric conditions and no further drug was released in the intestinal medium. Cell interactions of the invented NPs were improved by about two-fold compared to non-targeted NPs. The targeted NPs revealed a distinct and enhanced glucose-lowering effect in a humanized FcRn transgenic mice model after single oral application (rather high dosage of 3 mg semaglutide/kg) in comparison to the oral free drug, but no significant difference was observed compared to the non-targeted NPs. Daily oral administration of the FcBP2- and ZFcRn-NPs for 7 days resulted in a similar to higher glucoregulatory effect compared to the s.c. injection of semaglutide (60 µg/mL). These pharmacodynamic results are contradictory to the aforementioned release test, where no drug was set free in intestinal medium [235,236].

Due to concerns that PEG is not biodegradable under normal in vivo conditions and might accumulate in the body, Zhao et al. proposed using amphoteric sulfobetaine as an alternative for coating PLGA NPs loaded with liraglutide. Sulfobetaine is a biodegradable, hydrophilic, and electrically neutral zwitterion, offering better stealth properties than PEG to protect peptide drugs from degradation. After self-assembling on the surface of PLGA-NPs, sulfobetaine enhances the circulation half-life, mucus penetration, cell affinity, and uptake by shielding the negative charge and hydrophobic properties of PLGA. Indeed, the prepared NPs improved mucus and intestinal permeability as well as cell uptake, and the oral relative bioavailability was increased by 4% compared to the free drug. In comparison to s.c. injected liraglutide (54 µg/kg), the orally administered sulfobetaine PLGA NPs (540 µg liraglutide/kg) achieved a relative bioavailability of 9%. Single and multiple oral doses of these NPs lowered the blood glucose level in diabetic mice significantly and to a similar extent as s.c. injected liraglutide [237,269].

Chitosan is a cationic, natural polysaccharide with a special ability to adhere to mucus and to transiently open tight junctions of epithelial cells, thereby increasing the paracellular permeation. Because of its safety, stability, bio-adhesiveness, pH-sensitivity, and absorption enhancing effect, it is widely used in pharmaceutics [270]. (Further details are provided in Section 5.2).

In 2011, the first exendin-4 loaded chitosan NPs were prepared by Nguyen et al. These self-assembled NPs, consisting of positively charged chitosan and the negatively charged peptide poly(γ-glutamic acid) (γ-PGA), were freeze-dried and filled into hard gelatin capsules coated with enteric resistant polymer (Eudragit^® L100-55, Evonik Operations GmbH, Darmstadt, Germany). After reconstitution the particles showed sizes around 260 nm and a positive surface charge. The capsules, orally administered to diabetic rats, exhibited a prolonged hypoglycemic effect. The relative bioavailability of the oral capsule (300 µg exendin-4/kg) compared to s.c. injection of free exendin-4 (50 µg/kg) was 14% [203].

Two years later, Ahn et al. conjugated low-molecular-weight chitosan and cysteinylated exendin-4 via disulfide bonds, which are cleavable in plasma and GI mucus, containing cysteine-rich glycoproteins, to maintain a maximum of active drug. These conjugates, formed by electrostatic interactions, revealed a relative bioavailability (400 µg/kg orally vs. 50 µg/kg s.c.) of rather low 6.4% [205].

The already previously mentioned research group around Sun K. used the CSK peptide to prepare goblet cell-targeting chitosan nanoparticles encapsulating exenatide. The CSK-chitosan NPs achieved a relative bioavailability (50 µg exenatide/kg orally vs. 5 µg exenatide/kg s.c.) of similarly low 6.56% as it was reported by Ahn et al. [205,239]. These nanoparticles have been further refined by the development of a liraglutide loaded core-shell system consisting of chitosan, CSK, hemagglutinin to enhance lysosomal escape and poly-N-(2-hydroxypropyl)-methacrylamide (pHPMA), a hydrophilic and mucus-inert substance for mucosal delivery of nanocarriers. After the nanocapsules pass the mucus, the pHPMA shell is degraded and the particle core is exposed. CSK mediates the cellular uptake by goblet cells where the nanocapsules are partially destroyed, forming smaller spheres which expose hemagglutinin, helping the nanocapsules to escape lysosomes. The invented core-shell nanoconstruct exhibited higher stability and cellular uptake than all of the control groups. The relative bioavailability of the oral nanocapsules compared to the s.c. injection of free liraglutide was 10.12%, 2.25 times higher than that of oral liraglutide solution [240,241].

In 2023, Jakhar et al. prepared liraglutide chitosan NPs coated with the gastric resistant polymer Eudragit^® S100 (Evonik Operations GmbH, Darmstadt, Germany). Sustained release was observed over 2 days. The oral treatment of obese mice for two weeks with 100, 200, and 400 µg liraglutide/kg twice a day significantly reduced body weight compared to treatment with 200 µg free liraglutide/kg administered via oral route [242].

Inorganic silica nanoparticles offer unique features like chemical stability, safety for pharmaceutical use, and large surface areas, making them ideal for hosting guest molecules of various sizes, shapes, and functions. Porous silicon (PSi) nanoparticles have already been successfully used for the oral administration of biomacromolecules. They present several advantages such as top-down production, high surface-to-volume ratio, adjustable particle and pore size, high drug loading capacity, biodegradability, and biocompatibility. Moreover, effortless surface modification of the PSi, either by physical adsorption or chemical conjugation, make their properties easily tunable. The drug loading is usually a simple process under minimal harsh conditions, avoiding possible degradation of the drugs. The drug is retained inside the mesopores, typically showing pore sizes between 2 and 3 nm, by physical adsorption or electrostatic interactions [180,225,243].

Kaasalainen et al. have studied the GLP-1 loading and desorption of PSi NPs in detail. They concluded that electrostatic interactions between peptide and PSi NPs play a minor role compared to hydrophobic interactions as the highest drug loading was achieved when the pH was close to the isoelectric point of the peptide and its charge was nearly zero. About 14% of the loaded drug was bound irreversible to the NP surface, building a fixed layer that facilitates the desorption of the following peptides bound to the NPs [271]. Reviewing the studies of this chapter revealed that all of the drug loaded silica nanoparticles exhibited rather large sizes between 300 and 750 nm. Additionally, at least half of them showed a rapid and rather low release of 25–40% in the intestinal medium within 5–30 min (Table 3).

The first attempt to load silica nanoparticles with GLP-1 was taken by Qu et al. in 2012 [243]. They prepared a rather simple, pH-sensitive nanosystem consisting of Aerosil^® 200 (Evonik Industries AG, Essen, Germany), adsorbed GLP-1, and enteric resistant coating with Eudragit^® L100 (Evonik Operations GmbH, Darmstadt, Germany). Nonetheless, about 35% of GLP-1 were released at pH 1.2 within 2 h. A burst release of 80% of the drug within 1 h at pH 7.4 is assumed to be the reason for a quite short increase in its plasma concentration for only 2 h. After oral application of the NPs in a rather high dose (1 mg GLP-1/kg), they demonstrated a significant hypoglycemic effect compared to the orally administered pure drug. A relative bioavailability of 35.67% versus intraperitoneal injection of GLP-1 (0.33 mg/kg) was observed. No relative bioavailability compared to s.c. injection was tested.

As already mentioned in the Section 6.3, Araujo et al. compared GLP-1 loaded PSi nanoparticles with SLNs and PLGA nanoparticles according to their in vitro permeability in 2014/15 [206,225]. Subsequently the same working group around Santos and Sarmento investigated these mucoadhesive PSi nanoparticles, whose pores were loaded with GLP-1, in more detail. The mucoadhesiveness was reached again by surface modification with chitosan. As before, these NPs were embedded in an enteric resistant polymer matrix including a DPP-4 inhibitor by an advanced aerosol flow reactor. Only the cell penetrating peptide was not included this time. The NPs showed controlled drug delivery, enhanced mucoadhesion over NPs without chitosan, and improved intestinal permeability compared to the free drug and NPs without DPP-4 inhibitor. Blood glucose levels were significantly improved in diabetic mice compared to the control groups treated with GLP-1/DPP4 solution or empty NPs [180,244]. Three years later, the research group functionalized the previously described NPs with the Fc-fragment of immunoglobulin to target the Fc-receptor on epithelial cells in the intestine. As described above, the working group around Sun K. did the same for exenatide-PLGA NPs at the same time [227]. The DPP-4 inhibitor was not included this time. For the entrapment in the pH-sensitive polymer, matrix glass capillary microfluidics were used, as already conducted before by Araujo et al. [225]. The drug loading and encapsulation efficiency of the final NPs were rather low. The drug was set free in a pH-responsive, sustained manner. The interaction of the NPs with the intestinal cells was increased due to Fc-receptor targeting, enhancing the GLP-1 absorption compared to free drug and NPs without Fc-fragment [245].

Large pore dendritic silica NPs with a pore size of 10 nm were loaded with exenatide and coated with chitosan by Abeer et al. High drug loading, due to the large and dendritic pore structure of the NPs, was achieved. Functionalizing of the NPs with phosphonate optimized the surface charge and drug loading efficiency [246]. In 2020, Lamson et al. described small (<100 nm), negatively charged silica nanoparticles acting as physiochemical PEs that facilitate the oral delivery of exenatide. Specifically, these nanoparticles work not by moving across the intestinal epithelium as delivery vehicles, but by binding intestinal surface receptors that mediate the opening of tight junctions. Enterically coated exenatide capsules (1 mg/kg) were co-administered with the silica NP suspension (200 mg/kg). Compared to the same dose of subcutaneously administered exenatide, the silica-assisted, orally delivered peptide achieved a 10% relative bioavailability in treated mice [228].

One type of lipid-based nanoparticles are the so-called solid lipid nanoparticles (SLNs), consisting of crystalline lipids and showing a diameter between 50 and 1000 nm. SLNs can be used for controlled drug release and drug targeting, they protect their cargo against degradation, are biocompatible, and can be produced easily in large-scale [272]. In 2014, Araujo et al. made an attempt to prepare GLP-1 loaded oral SLNs comparing them to PLGA and PSi nanoparticles as described in Section 6.3, but they had trouble with the drug release from the SLNs (Table 3) [206]. Nanostructured lipid carriers (NLCs), consisting of crystalline and liquid lipids, surfactants, and water, are the successors of SLNs [273]. Additionally, to the aforementioned advantages of SLNs, NLCs show the ability to stimulate the endogenous GLP-1 secretion from enteroendocrine cells by mimicking the effect of lipid nutrients. Interestingly, the single lipid components alone do not affect the GLP-1 secretion, only the composition of lipids within an NLC is effective [274]. In 2018, Shrestha et al. encapsulated exenatide as well as liraglutide in NLCs. A very high encapsulation efficiency around 90% was reached and about 20% of the drug was released under gastric conditions, whereas 60–100% of the drug was set free by a burst release within 30 min in the intestinal medium. Although in vitro studies proved an enhanced permeation of NLCs compared to free drug solution across intestinal cells in the absence of mucus, in vivo and ex vivo studies revealed that the NLCs became stuck in the mucosa and could not reach the epithelial cells in the intestine. Further modifications to overcome the mucosal barrier were required.

As the developed NLCs additionally showed a GLP-1 secretory effect on enteroendocrine cells, they were promising dual-action oral delivery systems that were worth being further researched [247]. Therefore, the working group changed the carrier system to lipid nanocapsules, containing reverse micelles (RM-LNCs) loaded with exenatide. The drug was entrapped with a remarkable efficiency of 85%. The final RM-LNCs demonstrated gastric resistance and prolonged release in the intestine. The endogenous secretion of GLP-1 was additionally triggered by the empty RM-LNCs with a size of approximately 200 nm. In a former study, it was postulated that smaller sizes do not achieve a GLP-1 secretory effect in in vitro cell tests, but this effect was not observed in human ileal organoids in a present study of the research group [275,276]. The relative bioavailability after oral administration of exenatide-loaded RM-LNCs (500 µg exenatide/kg) compared to s.c. injection of the free drug solution (50 µg/kg) was low, revealing values around 4%. Nevertheless, it was approximately three-fold higher than that of the orally administered exenatide solution (500 µg/kg). The invented nanosystem demonstrated its dual-action of endogenously secreted GLP-1 and intestinal delivered exenatide following acute and chronic treatment of obese/diabetic mice [248].

To further enhance the endogenous GLP-1 secretion and prolong the glycemic effects of the invented nanocapsules, the research group proposed to use PEGylated RM-LNCs loaded with exenatide instead. The improved formulation enabled an administration every second day, showing similar effects to the daily applied formulation and equal effects to the daily applied PEGylated RM-LNCs without exenatide loading in diabetic mice [249]. In 2024, the RM-LNCs were also tested for the delivery of semaglutide. Again, a very high encapsulation efficiency of 90% was reached and a better trend for glucose reduction was observed during a 4-week daily oral treatment than with exenatide RM-LNCs, pure RM-LNCs, and free drug, respectively [250]. Finally, the research group around Sun K. published a study about exenatide-loaded folic acid (FA)-modified RM-LNCs, targeting enterocytes as folate receptors are abundant in gastrointestinal tract. The folic acid was combined with PEG and phospholipids (DSPE) before the inclusion in RM-LNCs. The cell uptake and transcellular transport of these nanocapsules (FA-RM-LNCs) was significantly improved over that of unmodified RM-LNCs, resulting in a 1.28-fold increased bioavailability. The relative bioavailability of the orally given FA-RM-LNCs (100 µg exenatide/kg) compared to s.c. injection (10 µg/kg) reached 7.5%. The hypoglycemic effect of a single oral dose of FA-RM-LNCs was significantly higher than after oral application of exenatide solution and showed prolonged characteristics over the subcutaneously injected drug. After multiple dosages of FA-RM-LNCs over 5 days, blood glucose levels were reduced by approximately 30% while the unmodified RM-LNCs and the subcutaneously injected drug reached reductions about 20% [217].

A further relevant group of nanoparticles made of lipids are self-assembled, spherical phospholipid bilayer vesicles called liposomes. Liposomes are highly biocompatible and biodegradable. Their physical and chemical properties can be easily modified to design effective drug carriers [252]. Exendin-4-liposomes coated with chondroitin sulfate-g-glycocholic acid (GCA) were described by Suzuki et al. GCA is an important part of the human bile salts and was chosen as coating to enhance the intestinal absorption. By these orally applied liposomes (200 µg exendin-4/kg), a comparably high relative bioavailability versus s.c. administration of the drug solution (20 µg exendin-4/kg) of 19.5% was reached. A prolonged release of the drug was proven in the intestine. Unfortunately, gastric release was not tested. The oral application of the exendin-4-liposomes (300 µg/kg daily) over 4 weeks revealed a significant reduction in body weight. At this point, it is important to highlight that this reduction in body weight was even stronger than that achieved by daily s.c. injections of 20 µg/kg free exendin-4 solution [251]. Liraglutide loaded corona cationic liposomes were recently published by Ding et al. Bovine serum albumin (BSA) was adsorbed on the surface, leading to enhanced mucus permeation because of its hydrophilic properties and the nearly neutral surface charge. During the penetration process, the outer shell of the liposomes gets degraded, exposing the cationic core of the liposomes. The liposomes were additionally linked to AT-1002, a hexamer synthetic peptide that increases intestinal permeability by opening the tight junctions between the cells. The liposomes showed prolonged release, good physiological safety, and enhanced absorption via intra- and transcellular transport. The relative bioavailability after intra-jejunal application of the liposomes (540 µg liraglutide/kg) was 9% compared to s.c. injection of the free drug (54 µg/kg), an improvement of 5% over free liraglutide applied into the jejunum. Further studies should focus on enabling oral administration of the formulation [252,253].

Finally, self-emulsifying drug delivery systems (SEDDSs) and nanoemulsions belong to the lipid-based nanoformulations. SEDDSs consist of lipids, surfactants, and co-solvents, which spontaneously form oil-in-water nano- or microemulsions with the gastric fluids after application. They can help to overcome the low oral absorption of peptide drugs by excluding hydrolyzing enzymes from the lipophilic SEDDS droplets, enhance mucus permeation, due to their slippery surface and their good absorption membrane permeating properties, and control drug release [254,255]. Nearly all of the herein reviewed articles about SEDDSs and nanoemulsions demonstrated a comparatively elevated relative bioavailability over other nanosystems.

In 2018, the first paper about exenatide SEDDS was published by Menzel et al. To efficiently incorporate exenatide in the lipid SEDDS droplets, the working group increased the lipophilicity of exenatide by hydrophobic ion pairing with the anionic surfactant sodium docusate (DOC). A 20% (m/v) dilution of the SEDDS in phosphate buffer pH 6.8 resulted in stable and relatively low droplet sizes around 45 nm. The SEDDS enhanced mucus permeation and reached a relative bioavailability of approximately 15% compared to s.c. injection of exenatide (SEDDS: 150 µg exenatide/rat vs. 20 µg exenatide/rat s.c.) [254]. Two years later, these exenatide-DOC SEDDSs were compared to exenatide SEDDSs ion paired with the stronger anionic sodium n-octadecyl sulfate (SOS). All of the SEDDSs demonstrated sizes below 30 nm. The membrane permeability was enhanced by exenatide-SOS SEDDS 3.5-fold over the exenatide-DOC SEDDS and 6.4-fold over the exenatide solution. Compared with the s.c. injection of the exenatide solution (50 µg/kg), exenatide-SOS SEDDS (300 µg exenatide/kg) achieved a relative bioavailability of notable 19.6% and exenatide-DOC SEDDS 15.2% [255].

The same working group around Bernkop-Schnürch additionally compared hydrophobic ion pairing of exenatide with the cationic surfactant tetraheptylammonium bromide (THA) to exenatide-DOC, respectively. The prepared exenatide-THA and -DOC nanoemulsions exhibited enhanced intestinal membrane permeability and an even higher relative bioavailability of remarkable 28% and 16%, respectively, after oral administration (300 µg exenatide/kg) compared to s.c. injection of the free drug (50 µg/kg). The huge difference in the bioavailability of the described nanoemulsions can be explained by the different surface charge, which was positive for the exenatide-THA NCs and negative for the exenatide-DOC system. The invented drug delivery systems showed droplet sizes below 30 nm [256]. Lin et al. developed an exenatide loaded oil-in-water nanoemulsion with an amazing drug encapsulation efficiency of 96%. The lipid droplets of this system are liquid at 37 °C, semi-liquid at room temperature, and solid at a storage temperature of 4 °C, providing long-term stability. The oil droplets effectively protected exenatide from degradation in the gastrointestinal tract and were expected to possess higher cell surface binding, cellular association, and consequently enhanced cell uptake due to their liquid and deformable properties. The authors postulated an increased intestinal lymphatic uptake due to the negative charge of the droplets and a resultant pancreas targeting. The relative bioavailability after oral application of the nanoemulsion (300 µg exenatide/kg) compared to s.c. injection of exenatide (50 µg/kg) was quite high (23.8%) and the drug release was prolonged [257].

In recent years, Lu et al. invented reverse micelle-SEDDSs (RM-SEDDSs), with exenatide loaded in the hydrophilic core of phospholipids. The SEDDSs were coated with the natural, anionic polysaccharide hyaluronic acid, leading to enhanced nanoparticle adhesion and cellular uptake. The surface potential of the double-layered nanoparticles stood close to zero and sustained release properties were observed. An enhanced hypoglycemic response over the uncoated NPs for at least 12 h after oral application was proven, but the orally applied dosage was rather high (1 mg/kg) [215].

RM-SEDDSs have also been tested by the research group around Bernkop-Schnürch in 2024, when they studied their potential as carriers for semaglutide in comparison to hydrophobically ion paired SEDDSs. Their results presented a higher lipophilicity for RM, which resulted in a significantly enhanced encapsulation efficiency in RM-SEDDSs, better stability, and prolonged release. Both types of SEDDSs significantly improved cell permeation, demonstrating the beneficial effect of lipophilic complexes for the delivery of semaglutide [277].

Exenatide-Zn²⁺ was combined with deoxycholic acid-low molecular weight protamine to prepare nanocomplexes by Han et al. in 2021 [258]. Low molecular weight protamine is a non-toxic, cationic peptide, that enhances cell penetration and prevents exenatide from intracellular degradation. To target apical sodium-dependent cholic acid transporters in the ileum, hydrophilic, negatively charged glycocholic acid-poly (ethylene glycol)-b-polysialic acid copolymers were introduced in the drug delivery system. Polysialic acid is an endogenous polysaccharide that enhances stability and transport of NPs. The dual cholic acid-functionalized NPs were filled into enteric-coated capsules for their oral administration. In vitro tests in simulated intestinal fluid containing trypsin revealed a high protection of exenatide, loaded into the invented drug delivery system, against degradation. The release profile showed prolonged characteristics, cell uptake and transport were increased and ileum targeting was realized. After oral gavage of the NPs (300 µg exenatide/kg) to diabetic rats, a relative bioavailability of 12.32% compared to s.c. injection of 20 µg exenatide/kg was measured and a distinct loss in body weight was recognized.

Cyclodextrins have been successfully used as excipients for oral drug delivery systems with low toxicity. Presas et al. have used amphiphilic cyclodextrins, functionalized with cationic groups and PEGylated-phospholipid chains (mPEG-DSPE 2000), to form electrostatic complexes with liraglutide and further self-assemble to NPs, which protect the drug from enzymatic degradation. Finally, a dextran sulphate coating was added to the nanocomplexes. Liraglutide was encapsulated to a striking 100%. After intra-jejunal application of the NPs (200 µg liraglutide/kg), they reduced the glycemic response in rats significantly and to a similar extent as s.c. injection of 200 µg liraglutide/kg. The oral administration of the invented nanocomplexes has not been tested so far [259].

Metal organic framework (MOF) NPs, consisting of inorganic clusters bridged by organic ligands, possess a high encapsulation capacity for exendin-4 because of their large pore size and surface area. Zhou et al. used iron MOF nanoparticles equipped with amino groups which allow trapping of negatively charged peptides by electrostatic interactions and provide the opportunity for further modifications. A remarkable encapsulation efficiency of 87% was realized. The working group coated the NPs with a hydrophobic, zwitterionic hydrogel layer to enhance mucus permeation and internalization by epithelial cells. The NPs were loaded together with the effervescent components, citric acid and sodium bicarbonate, into pH-sensitive capsules to protect them from degradation in gastric acid. The capsule shell dissolves in the intestine and CO₂ bubbles are set free, because of the acid and bicarbonate, which promote the movement of the NPs. Increased mucus permeation and endothelial cell uptake was proven due to the CO₂ propulsion and the hydrophobic, zwitterionic surface of the NPs. The exendin-4 release from the invented MOF NPs showed a prolonged characteristic in the intestine and a relative pharmacological bioavailability of 17% was reported after their oral application (400 µg exendin-4/kg) to diabetic rats compared to s.c. injection of exendin-4 (80 µg/kg) [260,278].

Bao et al. used a special corn protein, called zein, as basic material for the preparation of exenatide embedded nanoparticles. Zein is a prolamin with a high amount of hydrophobic amino acid residues. Therefore, it is not soluble in water but in ethanol and binds other ingredients by hydrophobic interaction. Additionally, phosphatidylcholine was incorporated to enhance the stability of the peptide drug and the transportation over the intestinal epithelium. Cholic acid was used to promote intestinal absorption via bile acid transporters, as performed one year before by Suzuki et al. for liposomes [251]. Casein acted as stabilizer by suppressing hydrophobic aggregation of the NPs and hypromellose phthalate protected exenatide against degradation in the gastric fluids. Two variants of the zein NPs were tested: one was prepared by dispersing all of the single ingredients in zein (DIS NPs) and for the second one, exenatide was complexed to micelles with phosphatidylcholine and cholic acid first (COM NPs) to increase the loading efficiency from 53.6% to 79.7%. Nearly no drug release occurred in the simulated stomach, followed by an initial burst release and subsequent sustained release in the intestinal medium. COM-NPs exhibited better sustained release properties, protection of exenatide against enzymatical digestion, and intestinal absorption than DIS NPs. The achieved relative bioavailability compared to s.c. exenatide injection (60 µg/kg) for DIS NPs and COM NPs (600 µg exenatide/kg) were 5.6% and 10.9%, respectively [261].

Three years later, the same working group proposed similar NPs, replacing phosphatidylcholine and casein by rhamnolipids, surface-active glycolipids that improve aggregation stability by reducing the hydrophobicity of zein NPs. This time, liraglutide complexed with cholic acid was used as incorporated drug and hypromellose phthalate was not contained anymore. Sustained release was observed. Rhamnolipids and cholic acid improved the mucus and transcellular permeation of liraglutide as well as its oral absorption compared to NPs without these excipients. The orally applied nanoparticles, in a high dosage of 4 mg liraglutide/kg, reached a relative pharmacological bioavailability of 9.6% compared to s.c. injection of free liraglutide solution (400 µg/kg) [262].

Another type of zein nanoparticles for the transport of liraglutide was created by Ziebarth et al. The NPs were stabilized by pH-sensitive Eudragit^® RS100 (Evonik Operations GmbH, Darmstadt, Germany) and mucoadhesive chitosan. They demonstrated a rather good persistence in gastric and intestinal fluids and prolonged release properties. The orally applied NPs achieved an area under the curve (AUC) of approximately 693 ng*h/mL in rats. Compared to the s.c. injection of free liraglutide, which was applied in the same dosage (200 µg liraglutide/kg) and achieved an AUC of approximately 978 ng*h/mL, a relative bioavailability of impressive 71% can be calculated [182]. As an additional remark, it is important to mention that oral zein NPs have been proven to stimulate the endogenous secretion of GLP-1 by Reboredo et al. even without drug loading, mimicking the effect of protein nutrients [279]. In this respect, a combinatory effect of stimulated endogenous production of GLP-1 and delivered drug is to be expected for zein NPs loaded with GLP-1 or its analogs.

A rather new approach to deliver liraglutide and semaglutide is the encapsulation in milk exosomes. These are extracellular vesicles with sizes about 30–150 nm, low immunogenicity, efficient cargo protection, and stable gastrointestinal properties. The restrictions of milk exosomes are their limited mucus permeation, because of repulsive forces and steric effects, their low drug loading efficiency, and the possible degradation of membrane proteins on their surface, which aid their internalization. To overcome these challenges, Xiao et al. have designed hybrid vesicles with self-adaptive surface properties. Therefore, they fused milk exosomes with liposomes and introduced a highly hydrophilic zwitterionic polymer linked via pH-sensitive hydrazone bonds to phospholipids (DSPE-Hyd-PMPC) during the preparation of liposomes. Due to these modifications, the encapsulation efficiency of semaglutide was enhanced over natural milk exosomes, membrane proteins were preserved, and the mucus was traversed efficiently by the invented delivery system. The drug loading ranged around remarkable 30%. The freeze-dried vesicles were loaded into enteric resistant capsules and bypassed the stomach after oral administration. In the intestine, the zwitterionic hydration layer improved the mucus penetration because of its strong hydrophilicity and neutral charge. Approaching the intestinal cell surface, the hydration layer was detached under the weakly acidic conditions and cationic lipids as well as membrane proteins of the vesicles were revealed, improving cellular uptake and transport, respectively. After oral administration of the vesicles (300 µg semaglutide/kg) a relative bioavailability of 9% was determined in comparison to s.c. injection of the free drug (30 µg/kg). Their blood glucose lowering effect was similar to subcutaneously injected semaglutide for 6 h; within the following 6 h, the hybrid vesicles showed enhanced effects over the s.c. applied drug and significantly intensified hypoglycemic effects compared to orally administered free drug and natural milk exosomes, respectively. Blood lipid-related indicators showed improved values in diabetic rats after oral treatment with the developed hybrid vesicles over 6 weeks, comparable to that of animals treated with s.c. injected semaglutide over the same period of time [263].

Aminoclay is a 3-aminopropyl-functionalized magnesium phyllosilicate, which is positively charged and, therefore, provides the ability of electrostatic interactions with negatively charged drugs. It shows a low risk of toxicity and can improve thermal stability of proteins. As aminoclay is facilitating trans- and paracellular transport, it can consequently enhance cell uptake. In 2023, Song et al. published their research on aminoclay liraglutide nanocomplexes prepared by spontaneous self-assembly. After reversing the negative surface charge by citric acid, the NPs were coated with negatively charged, pH-sensitive Eudragit^® S100 (Evonik Operations GmbH, Darmstadt, Germany), leading to particle sizes of about 370 nm. The drug entrapment efficiency with a value of >90% was exceptionally high. The drug release at pH 1.2 remained below 30%, whereas there was a rapid and extensive drug release of about 70% observed within 1 h at pH 7.4. The nanocomplexes revealed enhanced cell permeability compared to pure liraglutide. After oral application of the nanocomplexes in a very high dosage of 15 mg liraglutide/kg to diabetic mice for 30 days, significant loss in body weight was observed compared to the oral application of the pure drug in the same dosage. Compared to s.c. injection of liraglutide (0.3 mg/kg), approximately half of the body weight reduction was achieved [264]. The same transport system was additionally tested for the delivery of semaglutide, leading to similar characteristics as the liraglutide aminoclay NPs (Table 3). Only the zeta potential was strongly negative this time and the drug release at pH 7.4 was less pronounced. Cell uptake was enhanced again by the NPs and significant therapeutic effects concerning glycemic control and body weight were measured [265].

Bile acid derivatives are potent, biocompatible drug carriers, which boost the uptake and oral bioavailability of pharmaceuticals because of their binding affinity to apical sodium-dependent bile acid transporter and a type of G-protein coupled receptor at the epithelial surface. They influence lipid metabolism and, consequently, are of interest for the treatment of obesity. Kweon et al. have recently developed nanoparticles by electrostatic interactions of liraglutide, taurolithocholate (TLC), and lysine-conjugated deoxycholic acid (LDC). These particles have been stabilized by the addition of PEG and the surfactant Labrasol^® (Gattefossé, Saint-Priest, France). The drug encapsulation efficiency reached an outstanding 98%. After high-dosed oral application of the NPs (10 mg liraglutide/kg), they achieved a quite low relative bioavailability of 4% compared to s.c. injection of liraglutide (0.5 mg/kg), which was determined to be 0% for free oral liraglutide (10 mg/kg). The glucose lowering effect in diabetic mice after a high single dose of NPs (10 mg liraglutide/kg) as well as after daily administration for 28 days was equal to or even slightly stronger than that of s.c. applied liraglutide (0.1 mg/kg). Application over 28 days resulted in lower body weight and food intake as the s.c. administered drug [202]. Due to trouble with uncontrolled precipitation during the preparation and low production yield, the formulation had to be changed. Instead of taurolithocholate, the negatively charged bile acid sodium deoxycholate (DC) was used for the ion-pairing with liraglutide and LDC. The non-ionic surfactant n-dodecyl-β-D-maltoside (DDM) replaced Labrasol^®, leading to the formation of nanomicelles. PEG was not included anymore. The new nanomicelles demonstrated similar characteristics as the former NPs (see Table 3) and aggregation, precipitation as well as oxidative degradation were prevented. The freeze-dried product was filled into enteric resistant capsules. After daily oral administration of the invented nanomicelles (10 mg liraglutide/kg) over a period of 12 weeks, they exhibited a similar glucose-lowering effect as s.c. injected liraglutide (1 mg/kg) and higher body weight reduction at a dose of 5 mg liraglutide/kg. The weight of adipose tissue was significantly decreased in diabetic mice compared to diabetic mice without treatment. Oral treatment with nanomicelles in a dosage of 20 mg liraglutide/kg reduced the adipose tissue weight to a stronger extent than s.c. injected drug. Additionally, the oral nanomicelles and s.c. injected drug reduced the adipocyte size, and superior regulation of lipid metabolism was obtained than without treatment [266].

In summary, several attempts to enhance the oral bioavailability of GLP-1 and its analogs by nanosized drug carriers have been made during the last 25 years, reaching from widely applied PLGA NPs over lipid-based NPs to innovative approaches like the use of milk exosomes as transport systems [263,280]. Main aspects that have to be considered are the protection of the peptide drug against acidic and enzymatic degradation, enhanced mucus adhesion, and permeation as well as cell interaction and cell permeability. Finally, increased cell uptake, unidirectional transport through cells, and storage stability are important features to produce most effective formulations. It can be postulated that most of the tested nanoparticles improved the oral bioavailability of GLP-1 and its analogs compared to the free drug, showing prolonged release and extended effects as well as low cell toxicity. Several studies reported about oral nano delivery systems for GLP-1 and its analogs which reached the same or even better glucoregulatory and weight loss effects than a s.c. injection of drug [200,202,233,236,237,251,259,263,266]. Nevertheless, to the best of our knowledge, none of these nano- or microparticulate systems have been studied in clinical trials by now for oral delivery, much less reached an approval by the FDA or EMA. Reasons are assumed to be financial aspects, lack of the detailed toxicity studies, and of profound proofs for prolonged effects in humans. Finally, upscaling of the manufacturing process, particularly the peptide synthesis as the main bottleneck, is still the primary challenge for manufacturers. The complexity of required facilities along with surging demand underscore the importance of developing new, forward-looking strategies that will align capital spending with future infrastructure needs [281].

Advanced microbiome therapeutics (AMTs) represent an innovative in situ strategy for the oral delivery of peptide-based medicines using engineered microbes [300,301]. As summarized in the review by Vazquez-Uribe et al. [300], preclinical data on AMTs in animal models of metabolic disorders have shown promising results. More specifically, two genetically engineered probiotic strains were highlighted as delivery systems for recombinant GLP-1 and exendin-4, using Escherichia coli Nissle [302] and Saccharomyces boulardii [303], respectively. Remarkably, these plasmid-based AMT interventions lead to a significant decrease in food intake and weight gain in mouse models of obesity.