Mucociliary clearance refers to the clearance of the mucus and substances that are adsorbed in the nasal cavity or dissolved in the mucus, pushing it into the gastrointestinal tract through nasopharynx (Ozsoy et al., 2009). Such physiological clearance mediated by the hair-like cilia on the surface of epithelial cells with their fluctuating movement is responsible for the clearing of nasally administered drugs from the nose with a half-life of 21 min (Ozsoy et al., 2009). Compared to other routes of drug administration, nasal mucosa has a relatively porous and thin basement membrane (Bajracharya et al., 2019). Nasal mucosa offers a large surface area for the molecular absorption because of its highly vascularized subepithelial tissues and can direct the delivery of nasally administered molecules into the systemic circulation (Khafagy et al., 2009a). Moreover, intranasal pathway can avoid the first-pass metabolism by the liver, thereby enhancing the bioavailability of drugs (Erdő et al., 2018). Rapid nasal absorption of drugs facilitates quick onset of pharmacological effects (Montegiove et al., 2022). The nasal route has the advantage of self-medication by patients due to the easy accessibility of the nasal cavity (Khafagy et al., 2009a). Most intranasal drugs used to treat local inflammation of the nasal mucosa are small molecules such as steroids. Some small peptides, such as oxytocin and vasopressin, as well as some vaccines are administered via nasal routes for systemic effects (Ozsoy et al., 2009). Currently, nasal route has attracted attention as an alternative to the invasive administration of insulin injection (Pontiroli et al., 1989; Ozsoy et al., 2009; Chen et al., 2018).

The choice of nasal route for drug administration is mainly determined by considering the physicochemical properties of drugs, including hydrophobicity and molecular size (Behl et al., 1998). The bioavailability of nasally administered hydrophilic peptides and proteins, particularly for those with a size of >1,000 Da, is very low because of their weak permeability and susceptibility to proteolysis in the nasal mucosa (McMartin et al., 1987). The inverse relationship between the molecular weights of proteins or peptides and efficiency of nasal absorption is well established (Pires et al., 2009). Pseudo-first-pass effects that degrade drugs in the nasal mucosa by the action of diverse enzymes also hinder the nasal absorption (Sarkar, 1992). Taken together, the major hindrances for nasal delivery include low absorption of large molecules, possible enzymatic degradation in the nasal mucosa, and the clearance by rapid action of mucociliary transport (Khafagy et al., 2009a; Ozsoy et al., 2009), all of which contribute to the low bioavailability of drugs. In addition, the limit of volumes (25–250 μl) for nasal administration, influence of nasal pathologic conditions such as rhinitis and common cold on the absorption, and potential for nasal irritation by drug administration in the nasal cavity, are significant factors governing in the choice of the nasal route for drug application (Ozsoy et al., 2009; Kashyap and Shukla, 2019).

To overcome the above-mentioned hurdles of intranasal drug delivery, peptides and proteins can be modified to increase their membrane permeability and stability, by supplementing the drug with enzyme inhibitors that perturb the pseudo-first-pass effects (Pires et al., 2009). Absorption enhancers may also promote the uptake of drugs through the nasal mucosa and advanced formulations using liposome, lipid emulsions, and nano- and micro-particles can be used to optimize intranasal drug delivery (Ozsoy et al., 2009). Nano- and micro-particle systems, matrix systems where the molecules are distributed in the polymeric structures, facilitate sufficient retention of drugs in the nasal cavity and the protection against enzymatic activity. It also mediates tight contact between drug and nasal mucosa and even promotes the opening of tight junctions (Ozsoy et al., 2009). To attain sufficient bioavailability for nasally administered proteins and peptides, absorption enhancers, including cyclodextrin, surfactants, and mucoadhesive polymers have been applied in many formulation studies (Khafagy et al., 2009a).

To reach the bloodstream on the capillaries, drugs have to pass through the mucus barrier. Subsequently, drugs are known to follow the normal pathways for transepithelial transport (Figure 1). Transepithelial transport can occur by energy-independent direct uptake or paracellular diffusion, and by energy-dependent endocytic pathways and transcytosis, receptor-aided transport, or carrier-medicated translocation (Kristensen and Nielsen, 2016).

Transcellular transport mediated by either passive diffusion or active transport is a cellular process primarily used for the transport of lipophilic molecules (Ozsoy et al., 2009). Paracellular transport, a slow and passive aqueous pathway (Fortuna et al., 2014) that is accomplished by the regulation of tight junctions and passage through the intercellular spaces between adjacent cells, is important for the transepithelial absorption of drugs (Lemmer and Hamman, 2013). The localized high concentration of CPPs may influence the integrity of tight junctions, subsequently affecting the transduction of cargos through the paracellular cleft (Kristensen and Nielsen, 2016). It was found that some CPPs, such as transportan, MAP (model amphipathic peptide), and L-R5, reversibly modulate tight junctions in the transepithelial transport (Lindgren et al., 2004; Bocsik et al., 2019; Ragupathy et al., 2021). The transcytotic pathway is used for the vesicular transport of macromolecules from the apical to basolateral or the reverse, depending on the cargos and cellular processes in epithelial cells (Tuma and Hubbard, 2003). After the endocytic uptake of cargo with CPPs, transepithelial transport can be accomplished via transcytosis, avoiding lysosomal degradation (Kristensen and Nielsen, 2016).

Among the passage routes of drugs via nasal mucosa, large molecules like peptides, proteins, and nucleic acids are often reported to cross the nasal epithelium following the transcellular pathway or endocytosis (Illum, 2003; Ozsoy et al., 2009). In contrast, small molecules internalize the epithelium via the paracellular pathway through intercellular tight junctions (Ozsoy et al., 2009). Also found is that transepithelial transport of macromolecules such as protein and peptide therapeutics is highly limited (Bruno et al., 2013).

Although CPP-driven nasal delivery is regarded as a promising tool in facilitating non-injectable delivery of drugs, there are some challenges. First, CPPs in the nasal cavity are susceptible to enzymatic degradation in the mucus and the surface of the epithelium because of the inherent physicochemical characteristics of peptides (Kaur et al., 2016). The chemical modification of CPPs, peptide cyclization, and its PEGylation can improve the metabolic stability of CPPs (Szabó et al., 2022). In addition, as for CPP-mediated nose-to-brain delivery, the systemic exposure of nasally administered drugs might also occur because nasal coadministration of peptides with CPPs is shown to transport the peptide drugs to both systemic and brain compartments. Pharmacokinetic studies can further develop the specified methodologies for targeted delivery of drugs from the nasal cavity to specific compartments.

Recently, innovative nasal drug delivery systems have been developed to maximize its bioavailability by enhancing the solubility, permeability, and retention time at the nasal mucosa. For example, specific formulations have been investigated for nasal drug delivery, such as soft nano-vesicular systems, microsphere/inert carrier dry powder platform for brain-targeted delivery, and binary composite solid microparticles as nasal permeation enhancers (Hafner, 2022). Unlike these strategies, CPPs per se are able to internalize live cells and are peptide enhancers that mediate the transepithelial transport of large molecules without affecting cellular integrity. In addition, CPPs can be applied in the combination of the above-mentioned drug delivery platforms, enhancing the feasibility of CPP-aided nasal drug delivery.

Nasal administration of drugs can direct the rapid systemic absorption of drugs by circumventing the hepatic first-pass metabolism and gastric degradation, allowing fast onset of pharmacological action (Tai et al., 2022). In general, larger particles more than 10 μm accumulate in the respiratory area, whereas particles smaller than 5 μm can reach the lungs (van Rijt et al., 2014). Most administered drugs by the nasal route can be absorbed in the respiratory area of the nasal cavity enriched with vasculature throughout the extensive surface area (Ozsoy et al., 2009). The respiratory region around the inferior turbinate is the major site for systemic entry of nasal drugs due to its greatest surface area in the nasal cavity (Fortuna et al., 2014). Blood supply to the nasal mucosa and its flow rate are critical determinants for the systemic absorption of drugs to the blood circulation via diffusion (Fortuna et al., 2014).

Blood-brain barrier (BBB) is constructed through the networking of brain capillary endothelial cells that are held together by tight junctions, which segregates the blood perfusion into the CNS (Hallschmid, 2021). It serves as a physiological or metabolic barrier that imposes highly selective permeation of molecules into the CNS (Daneman and Prat, 2015; Ross et al., 2018), although it perturbs the delivery of neurotherapeutics (Pardridge, 2003). The factors that restrict the influx of drugs include membrane barrier, inter-endothelial cell tight junctions, efflux transporters, including p-glycoprotein (Cioni et al., 2012) and degrading enzymes of BBB (Ross et al., 2018).

Currently, intranasal delivery is suggested as a promising alternative to invasive strategies because it can directly transport the molecules to the brain bypassing the BBB (Crowe et al., 2018; Erdő et al., 2018; Kashyap and Shukla, 2019). The intranasal route has been found to be effective in delivering the therapeutic molecules by avoiding liver metabolism, renal filtration, elimination in the gastrointestinal tract, and degradation in the serum (Mistry et al., 2009; Malhotra et al., 2013). Anatomically, nasal mucosa is directly connected to the cerebrospinal fluid (CSF) and parenchymal regions of the brain (Dhuria et al., 2010; Lochhead and Thorne, 2012; Crowe and Hsu, 2022). Nose-to-brain pathway was found to be efficient in the delivery of bioactive peptides and growth factors, such as exendin-4, leptin, orexin, vasoactive intestinal peptide (VIP), galanin-like peptide (GALP), interferon-β, insulin-like growth factor (IGF), nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), and basic fibroblast growth factor (bFGF) (Kamei et al., 2017). Furthermore, nose-to-brain transport of small molecular drugs, proteins, siRNA, and stem cells in vivo, and of peptide drugs, including insulin, vasopressin, melanocortin, angiotensin II, and oxytocin in humans was successfully accomplished (Lalatsa et al., 2014; Kanazawa, 2015).

The transport of molecules after the absorption in the nasal cavity is known to follow multiple pathways (Dhuria et al., 2010). The olfactory region, more specifically the olfactory bulb, is the only compartment throughout the human body where peripheral regions keep in direct contract with the CNS (Selvaraj et al., 2018). This olfactory epithelium directs the brain delivery of drugs via olfactory and trigeminal nerve pathways circumventing the BBB (Trevino et al., 2020). The axonal transport though the trigeminal or olfactory nerves is the major pathway for nose-to-brain transport and additionally, diffusional influx into the brain also contributes to this transport (Dhuria et al., 2010; Kamei et al., 2017). The olfactory pathway is occured in the upper region of the nasal route, the olfactory region, and is stretched to the olfactory bulb while trigeminal transport pathway extends to the brain stem (Dhuria et al., 2010). Following olfactory pathway, drugs travel along the axon of the neuron and pass through the nerve bundle that crosses the cribriform plate, and eventually enter the olfactory bulb located on the surface of the brain (Selvaraj et al., 2018). Nasally administered drugs can across the epithelium in the lateral respiratory region and then to the pons through the trigeminal nerve (Thorne et al., 2004; Crowe and Hsu, 2022) through the transportation along the axonal route of trigeminal nerves (Selvaraj et al., 2018).

After the passage through the mucus, paracellular and transcellular transport and extracellular diffusion can mediate the access to the CNS, which can occur in the epithelium of respiratory and olfactory mucosa (Dhuria et al., 2010). Drugs that pass through the lamina propria can diffuse into the perineural space, and then eventually be transported to the CSF (Lochhead and Davis, 2019). After the localization of molecules in the CSF, olfactory bulb, and brain stem, they are transferred to the parenchymal regions of the brain (Dhuria et al., 2010).

Several CPPs, including TAT-PTD, polyarginine, penetratin, transportan, and their derivatives, have been employed for the drug delivery through the nasal and intestinal epithelium. TAT-PTD, penetratin and polyarginine are generally categorized as cationic CPPs (Kristensen and Nielsen, 2016). More recently, our research group identified a novel hydrophobic CPP derived from the human TCTP (TCTP-PTD) and demonstrated its potential applicability for transepithelial delivery by nasal administration (Kim et al., 2011a; Bae and Lee, 2013; Bae et al., 2018a). TAT-PTD was the first CPP that was used as a peptide vector for in vitro insulin delivery across the intestinal epithelium (Liang and Yang, 2005). Insulin chemically linked to TAT-PTD (CGGGYGRKKRRQRRR) showed 6-8 times higher absorption of insulin compared to insulin in the Caco-2 monolayer system, via active and transcytosis-like mechanisms (Liang and Yang, 2005). However, some studies reported the inability of TAT-PTD to translocate monolayers of Caco-2 and MDCK epithelial cells (Violini et al., 2002), and in intestinal insulin absorption from the ileum in rats (Kamei et al., 2008). Therefore, polyarginine and penetratin and their analogs have often applied for epithelial drug delivery. An in vitro study of epithelial transport of insulin by D-form of arginine octamer (D-R8) indicated that cellular uptake is the critical step for efficient epithelial absorption and showed that D-R8 mediates intestinal epithelial transport of insulin by energy-independent unsaturable uptake (Kamei et al., 2013b).

Biotinylated transportan and its analog transportan 10 were reported to pass through epithelial Caco-2 monolayer cells mainly via transcellular transport and partly via paracellular pathway, while penetratin passes across cell monolayers in low yields, possibly due to partial degradation (Lindgren et al., 2004). Fluorescein-labeled D-R6 translocates into the rat ileal membranes via energy-dependent pathway, involving electrostatic interactions between anionic proteoglycans in membrane and cationic D-R6 peptide. Such interactions initiate endocytic pathways (Kamei et al., 2009), suggesting that intermolecular binding between CPP and cargo is important for CPP-aided intestinal absorption of proteins and peptides. D-R6 enhanced the transepithelial uptake of insulin but not of leuprolide because insulin is negatively charged and associates with cationic R-D6 (Kamei et al., 2009). Importantly, the strength of interaction between CPP and cargo positively correlates with the delivery efficiency across the nasal mucosa of rats (Khafagy et al., 2010b).

Studies showed that chain length, amphipathicity, hydrophobicity, and basicity of CPPs influence intestinal permeation of insulin (Kamei et al., 2013a). Specifically, not only hydrophobic Trp but also positively charged Arg and Lys residues are important for transmucosal delivery of peptides and proteins (Kamei et al., 2013a). For the complex of penetratin with insulin in transepithelial transduction through the intestinal epithelium in rats, the presence of Arg residue in the peptide was found to be a prerequisite for facilitating transepithelial insulin transport (Kristensen et al., 2015b). Also, the stereochemistry change of penetratin from L- to D-form showed differential effects according to the cargos because IFN-β with D-penetratin showed elevated intestinal and nasal mucosal absorption compared to that of L-penetratin but this phenomenon was not observed in the cases of exendin-4 and glucagon-like peptide 1(GLP-1) delivery in rats (Khafagy et al., 2009b). In most cases, CPP-mediated nasal delivery of insulin and exendin-4 utilized the L-form of penetratin, TCTP-PTD, and their variants because it showed better efficiency than D-forms.

CPPs can be linked to cargos via covalent conjugation or noncovalently using a simple mixing method or can be expressed as a fusion protein. For D-R9 and insulin complexation, covalent conjugation of D-R9 to insulin, not their coadministration, is necessary for the translocation through rat pulmonary alveolar epithelium (Patel et al., 2009). D-R9-insulin conjugates showed enhanced hypoglycemic effects in diabetic rats following pulmonary administration. However, coadministration of D-R9 and insulin facilitated the intestinal absorption of insulin (Morishita et al., 2007). In addition, the efficiency of epithelial transport can also be influenced whether CPP is conjugated to the N- or C-terminus in the complex. For the delivery of the active fragment of parathyroid hormone (PTH 1–34), conjugation at N-terminus showed better efficiency than C-terminal conjugation, but coadministration method was more efficient for the delivery of PTH 1-34 in the Caco-2 monolayer (Kristensen et al., 2015a).

Coadministration of CPPs with cargos sometimes forms CPP-cargo complexes via intermolecular interactions (Khafagy et al., 2010b). Such interactions include hydrophobic and electrostatic interactions, and hydrogen bonding, which are determined by the factors, including the molecular nature of CPP and cargo. It is generally accepted that not only the binding efficiency between CPP and cargo but other factors collectively contribute to and determine the absorption efficacy (Khafagy et al., 2010b). Other studies warn that covalent linking or fusion of cargo with CPP has the potential to interfere with the natural functions of the cargo (Kristensen et al., 2015a). Most studies on the nasal delivery of antidiabetic peptides, such as insulin, exendin-4, and GLP-1 with CPPs confirmed the effectiveness of the simple physical mixing of CPP with cargos in vivo.

For physical mixing methods, the extent of intermolecular interaction between CPP and cargo is a key determinant for efficient delivery, as shown for penetratin and its analogs (Khafagy et al., 2010b). The strategies for promoting the permeation of the transepithelial layer include conjugation of additional moieties to CPP that enhance the intermolecular interactions, mucus permeation, and adhesion. Conjugation of penetratin with bis-β-cyclodextrin (bis-CD), a cyclic oligosaccharide, improved assembly of insulin into nanocomplexes, and penetratin-bis-CD enhanced epithelial translocation of coadministered insulin compared with that of penetratin (Zhu et al., 2014). Such enhancement was due to the elevated hydrophobicity by bis-CD conjugation, which reinforced electrostatic and hydrophobic interactions between insulin and nanocomplexes (Zhu et al., 2014). Furthermore, penetratin-bis-CD tends to retard the proteolysis of insulin compared to that of penetratin alone (Zhu et al., 2014). Similarly, the conjugation of stearic acid to the R6EW (SA-R6EW), a R7E and R8W substituent of R8, considerably enhanced intestinal delivery of insulin in rats compared to that of R8 via enhanced intermolecular interactions in the complex (Zhang et al., 2015).

Poly (N-(2-hydroxypropyl) methacrylamide polymer) (pHPMA)-coated penetratin-insulin complex that has almost neutral surface charge showed enhanced mucus diffusion and insulin delivery in a diabetic rat model compared to nanoparticles whose surface charge is negative (Shan et al., 2015). pHPMA-coating induced the facilitated mucus permeation and adhesion by reducing the potential electrostatic interaction with mucus constituents (Shan et al., 2015). Also, nanoparticles combined with anionic chitosan showed enhanced paracellular penetration in the mucosal delivery by facilitating mucosal adhesion (Ozsoy et al., 2009; Casettari and Illum, 2014). Chitosan has been widely used as an intranasal absorption enhancer for various therapeutics and is a mucoadhesive agent because amines in chitosan mediate its interaction with sialic acid expressed on the mucosal layer (Ozsoy et al., 2009; Casettari and Illum, 2014). Chitosan, a polycationic polymer that has positive charge, enhances paracellular transport by inducing the structural reorganization of tight junction-associated proteins (Schipper et al., 1997). It was found to reduce mucociliary clearance from the nasal cavity and to open tight junctions in the olfactory epithelium (Ozsoy et al., 2009; Casettari and Illum, 2014).

Insulin is a polypeptide hormone that circulates in the blood stream to regulate the glucose homeostasis by binding to its cell surface receptor and facilitates glucose uptake into cells (Rahman et al., 2021). Insulin is administered subcutaneously to the patients with diabetes mellitus who have insulin deficiency. It is addressed that pharmacokinetic profile of intranasal insulin mimics that of the endogenous ‘pulsatile’ insulin (Illum and Davis, 1992; Khafagy et al., 2009a). TCTP-PTD and its variants TCTP-PTD 13, 13M2, as well as penetratin and its analogs shuffle (R,K fix) 2 and PenetraMax were tested for the systemic delivery of intranasal insulin.

Glucagon-like peptide-1 (GLP-1), an endogenous 30-amino acid peptide hormone that is mainly produced by intestinal L-cells in response to ingestion of nutrients, plays a role in the regulation of glucose homeostasis (Runge et al., 2007). Exendin-4, a homolog of GLP-1 acting as an agonist for the GLP-1 receptor, is a 39-amino acid peptide that was initially identified in the venomous lizard Heloderma suspectum (Runge et al., 2007). Both GLP-1 and exendin-4 are antidiabetic peptides useful for treating type 2 diabetes that bind to and activate GLP-1 receptor thereby facilitating glucose-induced insulin secretion. Endogenous GLP-1 is susceptible to cleavage by dipeptidyl peptidase IV (DPP-IV) whereas exendin-4 resists DPP-IV-induced peptide degradation (Runge et al., 2007). TCTP-PTD variants, such as TCTP-PTD 13M2 and 13M3, and penetratin and its analog shuffle (R,K fix) 2 were used in nasal administration of exendin-4 and GLP-1 via physical mixing method and were verified to be effective in normal or diabetic murine models.

Insulin delivery through nose-to-brain pathway is aimed to improve memory by enhancing synaptic plasticity and glucose uptake, thereby attenuating the neuropathy associated with AD (Hallschmid, 2021). Mounting evidence has shown that acute and chronic intranasal insulin administration facilitates memory and recognition functions (Brünner et al., 2013) both in healthy individuals and patients with mild cognitive impairment of AD (Hallschmid, 2021). Because insulin cannot be passively transported into the brain due to its high molecular weight of 5,800 Da (Hallschmid, 2021) and the efficiency of delivery to the brain by intranasal administration is low, CPPs are considered as the ideal potential vehicles (Kamei and Takeda-Morishita, 2015).

Takeda-Morishita group employed the noncovalent method for nose-to-brain delivery of penetratin that was found to be better than octaarginine (R8) in intestinal and nasal delivery (Kamei et al., 2008; Khafagy et al., 2009a). When they compared the absorption and distribution profiles of systemic (i.v.) and intranasal (i.n.) administration of insulin in mice, they found that i.v. insulin was mainly found in the systemic circulation with slight distribution in the brain (Kamei and Takeda-Morishita, 2015). In contrast, the intranasal route enhanced the distribution of insulin to the olfactory bulb with minimal systemic exposure of insulin. As i.n. insulin requires administration of more than 10 times dose than i.v. route for the comparable levels of insulin in the brain, this group employed L- and D-penetratin to maximize the nose-to-brain insulin delivery (Kamei and Takeda-Morishita, 2015). L-penetratin exhibited a more efficient transport of insulin to the mouse brain following i.n. administration than D-penetratin, whereas D-penetratin showed less systemic insulin exposure (Kamei and Takeda-Morishita, 2015).

Higher insulin in the blood did not necessarily correlate with elevated insulin levels in the brain, implying that blood-to-brain delivery is not the main contributor to nasal L-penetratin/insulin delivery (Kamei and Takeda-Morishita, 2015). Coadministration of nasal insulin with L- or D-penetratin mediated the delivery of insulin to the distal regions of the mouse brain, such as the cerebral cortex, brain stem, and cerebellum, possibly from the nasal cavity to CSF and then to the brain regions (Kamei and Takeda-Morishita, 2015). The pathway of peptides from the nasal cavity to the brain regions involves their epithelial uptake through the lamina propria, and influx to CSF passing through the cribriform plate or through the olfactory/trigeminal pathway in the axonal transport pathway (Dhuria et al., 2010). Nasal coadministration of insulin with D- or L-penetratin elevated the hippocampal insulin level both 15 and 60 min after administration, for subsequent delivery into deeper brain regions (Kamei and Takeda-Morishita, 2015). The distribution of insulin in the hippocampal region might be beneficial for improving memory and recognition in the hippocampus (Akhtar and Sah, 2020).

These investigators also assessed the distribution of nasally administered physical mixtures of insulin and L- or D-penetratin and confirmed the direct transport of insulin from the nasal cavity to the brain parenchyma in rats (Kamei et al., 2016). Analysis using a gamma counter and autoradiography of the brain showed the greatest radioactivity in the olfactory bulb and the brain regions close to the administration site at 15 min after nasal administration of radiolabeled insulin with L- or D- penetratin (Kamei et al., 2016). Noncovalent insulin/penetratin complexes effectively deliver insulin to the olfactory bulb or CSF and distal regions of the brain, including the hippocampus through the pathways, the cerebral cortex, cerebellum, and brainstem, possibly by bulk flow or passive diffusion (Kamei et al., 2017). Quantitative analysis of brain CSF samples at 15 min after nasal administration of insulin/L-penetratin showed that insulin levels were increased only in the anterior portion of the CSF, indicating that insulin was distributed to the CSF from the anterior part of the brain near the olfactory bulb (Kamei et al., 2017). These results collectively indicate that penetratin enables direct insulin transport from the nasal cavity to the parenchymal region of the brain (Kamei et al., 2017).

More recently, Kamei et al. clarified the transport route of L-penetratin-aided nose-to-brain delivery of insulin and exendin-4 (Kamei et al., 2021). Here, L-penetratin did not facilitate the transport of fluorescence-labeled exendin-4 and insulin via the trigeminal pathway. Rather, it shifted their distribution to the olfactory mucosal pathway. The fractions of the drugs that were systemically absorbed or transported through trigeminal pathway did not contribute to the distribution of peptides to the brain. Insulin with L-penetratin diffused to the surface of the olfactory bulb as well as the lower region of the brain around the hypothalamus, possibly from the CSF. Coadministered L-penetratin enabled the peptides to penetrate through the nasal epithelial membrane. Then, the peptides crossed the lamina propria and cribriform plate, thereby reaching the CSF. The peptides were transported from the surface of the olfactory bulb or bottom region of the brain around the hypothalamus to the deeper part of the brain through the CSF or perivascular spaces of the cerebral artery (Kamei et al., 2021).

Intranasal coadministration of noncovalently bound L-penetratin with insulin increased insulin levels in the brain and its pharmacological effects were tested at two differential stages of memory impairment using a senescence-accelerated mouse-prone 8 (SAMP8) model that exhibits spontaneous aging and senescence (Kamei et al., 2017). SAMP8 mice showed age-related cognitive dysfunction with reduced learning ability score as well as elevated amyloid beta (Aβ) concentration in the circulation (Kamei et al., 2017). Chronic nasal insulin/L-penetratin treatment retarded the progression of mild cognitive dysfunction and memory loss in SAMP8 mice at 16–24 weeks of age, an early stage of memory loss (Kamei et al., 2017). However, the severe recognitive dysfunction seen at the progressive stage associated with the deposition of Aβ was not ameliorated following nasal L-penetratin/insulin. Instead, it unexpectedly elevated Aβ plaque accumulation in the hippocampal region of the mice. Likewise, nose-to-brain delivery of L-penetratin/insulin complex was differentially affected according to severity of AD (Kamei et al., 2017). At the early stages of dementia devoid of Aβ accumulation in the brain, nose-to-brain delivery of L-penetratin/insulin appeared to have the therapeutic effect by preventing the mild cognitive dysfunction of early dementia, whereas it did not prevent severe recognitive dysfunction (Kamei et al., 2017).

The investigators suggested that the phenomenon might occur due to the competition between insulin and Aβ for the degradative pathway by insulin-degrading enzyme (IDE), an enzyme involved in the clearance of extracellular Aβ from the brain (Shiiki et al., 2004). Because of such competition for IDE, IDE-mediated Aβ degradation can be inhibited thereby leading to Aβ accumulation in the brain (Kamei et al., 2018). Accumulation of Aβ whether as soluble or insoluble matter inhibits insulin receptor expression on the neurons by increasing the internalization and degradation of the receptors. Therefore, excessive insulin delivery to the brain is not recommended as it may lead to severe recognition impairment due to Aβ plaque formation, corresponding to 33–38 weeks of SMAP8 mice.

Having established that intranasal insulin/L-penetratin administration was effective at mild stage of cognitive dysfunction but not at severe cognitive dysfunction in senescence-accelerated mice (Kamei et al., 2018), this group investigated the effectiveness of GLP-1 receptor agonist, exendin-4, in modulating severe cognitive impairment. It was previously established that GLP-1 receptor agonists play a role in memory and learning in the brain by stimulating receptor binding and promoting downstream insulin signaling (Kamei et al., 2018). Exendin-4 exhibited a lower risk of hypoglycemia than insulin following excessive systemic exposure.

Intranasal coadministration of exendin-4 with L-penetratin but not with D-penetratin increased exendin-4 distribution throughout the various brain regions, including the olfactory bulb, the entry point for nose-to-brain delivery (Lochhead and Thorne, 2012) and hippocampus, where exendin-4 mediates memory and learning functions after administration in normal ddY mice (Kamei et al., 2018). These observations imply that direct nose-to-brain transport is responsible for the elevated exendin-4 levels in the brain following intranasal coadministration of L-penetratin/exendin-4. However, coadministration of GLP-1 and L-penetratin showed negligible distribution in the olfactory bulb and the whole brain due to elevated systemic absorption (Kamei et al., 2018). Because D-penetratin did not enhance the nose-to-brain delivery of exendin-4 and GLP-1, they suggested that D-penetratin/cargo complex that resists proteolysis, possibly remains in the cytoplasm compared to L-penetratin/cargo complex, and is released less to lamina propria located in the middle part of the olfactory bulb (Kamei et al., 2018). It was also found that nose-to-brain delivery of L-penetratin/exendin-4 can promote the insulin signaling pathway only in the presence of adequate levels of insulin to stimulate insulin receptors in the brain (Kamei et al., 2018). Brain delivery of exendin-4 and insulin by L-penetratin stimulated the downstream signaling of insulin, as shown by Akt phosphorylation in the hippocampus in normal ddY mice (Lalatsa et al., 2014; Kamei et al., 2018).

To test the therapeutic efficacy of nose-to-brain delivery of insulin with exendin-4, exendin-4 and L-penetratin were nasally administered for 30 days with or without insulin at a dose that is adequate to avoid Aβ accumulation in SAMP8 mice (Kamei et al., 2018). Intranasal exendin-4 administration without CPP contributed to learning in the SAMP8 mouse model and coadministration of L-penetratin and insulin improved cognitive function by facilitating the brain delivery of exendin-4 and insulin signaling activation in the brain (Kamei et al., 2018). In addition, nasal administration of exendin-4 contributed to the decrease in Aβ accumulation (Kamei et al., 2018). They speculated that the amount of insulin that does not induce Aβ deposition may induce the limited expression of insulin receptors on the hippocampal neurons under pathological conditions (Kamei et al., 2018). In this situation, exendin-4-mediated GLP-1 receptor stimulation appears to further enhance the initiation of insulin signaling by activating its downstream player, IRS-1 (Kamei et al., 2018). Collectively, these findings suggest that nose-to-brain delivery of L-penetratin/insulin was effective in the early stage of cognitive dysfunction and that nasal administration of L-penetratin/exendin-4 with insulin boosted therapeutic effect of exendin-4 in the progressive memory dysfunction in vivo.