Substance Delivery across the Blood-Brain Barrier or the Blood-Retinal Barrier Using Organic Cation Transporter Novel Type 2 (OCTN2)

Abstract

The membrane impermeability of a drug poses a significant challenge in drug research and development, preventing effective drug delivery to the target site. Specifically, the blood-brain barrier (BBB) presents a formidable obstacle to the delivery of drugs targeting the central nervous system (CNS) into the brain, whereas the blood-retinal barrier (BRB) presents a tremendous obstacle to the delivery of drugs targeting the ocular diseases into the eyes. The development of drugs for Alzheimer’s or Parkinson’s disease targeting the CNS and for diabetic retinopathy and age-related macular degeneration targeting the eyes remains an unmet medical need for patients. Transporters play a crucial physiological role in maintaining homeostasis in metabolic organs. Various types of solute carrier (SLC) transporters are expressed in the capillary endothelial cells of the BBB, facilitating the delivery of nutrients from the blood flow to the brain. Therefore, carrier-mediated transport across the BBB can be achieved using SLC transporters present in capillary endothelial cells. It is well-known that CNS drugs typically incorporate N-containing groups, indicating that cation transporters facilitate their transport into the brain. In fact, carrier-mediated transport across the BBB can be accomplished using glucose transporter type 1 (Glut1) as a glucose transporter, L-type amino acid transporter 1 (LAT1) as a large neutral amino acid transporter, and H⁺/cation antiporter as a cation transporter. Surprisingly, although organic cation transporter novel type 2 (OCTN2) is expressed in the capillary endothelial cells, there has been limited investigation into OCTN2-mediated substance delivery into the brain across the BBB. Furthermore, it is suggested that OCTN2 is expressed at the BRB. In this prospective review, I present the advantages and possibilities of substance delivery into the brain across the BBB or into the eyes across the BRB, mediated by OCTN2 via carrier-mediated transport or receptor-mediated transcytosis.

1. Introduction

A drug holds no significance unless it effectively exerts its effects at the intended target site. However, the membrane impermeability poses a significant challenge in the development of specific types of pharmaceutical agents, particularly central nervous system (CNS) drugs used in the treatment of neurodegenerative diseases such as Alzheimer’s disease (AD) or Parkinson’s disease (PD) due to the blood-brain barrier (BBB) and ocular disease drugs used in the treatment of diabetic retinopathy and age-related macular degeneration due to the blood-retinal barrier (BRB), as well as cancer drugs. Notably, clinical trials for AD drugs have experienced repeated failures, leading some pharmaceutical companies to abandon the development of AD drugs [1]. Recently, there have been notable developments in the field. The anti-Aβ monoclonal antibody aducanumab [2] received FDA approval in 2021. Furthermore, the anti-Aβ protofibril monoclonal antibody lecanemab [3] obtained clinical approval from the FDA in 2023. Additionally, the anti-Aβ monoclonal antibody donanemab [4] successfully completed a phase 3 clinical trial for early AD in 2023 (NCT04437511), yielding favorable results. Aducanumab disrupted the BBB and entered the brain through such fenestration [2]. However, in the context of AD, it is hypothesized that BBB disruption occurs in the early stages of the disease [5]. Therefore, pharmaceutical agents, including monoclonal antibodies like aducanumab, lecanemab, and donanemab, may penetrate the brain through the compromised BBB and subsequently exert their activity there. However, disease-induced BBB disruption varies depending on conditions such as disease stages, daily situations, and individual variability and is not consistent throughout a patient’s life. Thus, CNS drugs must traverse the BBB [6] [Figure 1], which is composed of (i) a biological barrier involving efflux by multiple drug resistance 1 (MDR1, P-glycoprotein), a representative ATP binding cassette (ABC) transporter expressed as a transmembrane protein at the apical membrane of the capillary endothelial cells; (ii) a physical barrier based on the hydrophobic lipid bilayer membrane of the capillary endothelial cells; (iii) a physical barrier formed by tight junctions between the capillary endothelial cells due to adhesion molecules such as claudin; and (iv) a physical and biological barrier lined with pericytes and astrocytes. The BBB represents a multitiered obstacle to CNS drugs, drawing on the principles of biophysical structuralism advocated by Dr. Lévi-Strauss [7,8]. Overcoming the BBB presents a challenging task in drug development. It is well-established that existing CNS drugs and substances capable of penetrating the brain contain nitrogen atoms in N-containing groups. This observation suggests that cation transporters at the BBB transport low-molecular-weight substances with N-containing groups as their recognition units. Therefore, cation transporter-mediated transport emerges as a promising solution to traverse the BBB without being exported by MDR1 or repelled by the lipid bilayer membrane [Figure 1]. On the other hand, unlike low-molecular-weight compounds, high-molecular-weight substances, including monoclonal antibodies or nanoparticles, are subject to alternative strategies such as receptor-mediated transcytosis at the BBB [9]. There are several types of cation transporters, including organic cation transporter novel type 2 (OCTN2) (SLC22A5) [10], H⁺/cation antiporter [11,12], organic cation transporter 1 (OCT1), and OCT2 [10,13]. Drug delivery across the BBB into the brain using the H⁺/cation antiporter [14,15] has been previously discussed. Furthermore, it is suggested that OCTN2 is expressed in human corneal and conjunctival epithelial cells [16] and in rat retinal capillary endothelial cells at the inner BRB [17]. OCTN2 facilitates the transport of zwitterionic carnitine through unique features. Transporters facilitate the movement of their specific substrates based on their structure, hydrophobicity, and size, although some transporters can share the same substrates. Thus, the strategy of each transporter complements that of others, including OCTN2 [10] and the H⁺/cation antiporter [11,12]. In this concise prospective review, I present the benefits and possibilities of substance delivery into the brain or the eyes mediated by OCTN2.

2. Discussion

2.1. Transporters

Transporters [18] are tissue-specific transmembrane proteins that selectively transport substances such as nutrients or metabolites across membranes, driven by differences in concentration between the inside and outside or energy derived from ATP. Solute carrier (SLC) transporters [19] facilitate the influx of nutrients and other endogenous materials, exhibiting substrate selectivity not strictly determined by structural similarity. On the other hand, MDR1 is an efflux transporter [20] that shows broad substrate selectivity for hydrophobic materials, such as exogenous toxins and drugs. MDR1, when expressed in capillary endothelial cells, induces drug impermeability in the BBB or the inner BRB. Overexpression of MDR1 induces drug resistance in cancer cells [21]. The development of CNS drugs for neurodegenerative diseases, such as AD or PD, presents a significant challenge for many pharmaceutical companies. In drug discovery or development, membrane impermeability is a substantial concern. Substrates of various SLC transporters could be transported into the brain across the BBB or into the eyes across the BRB without being exported by MDR1. Therefore, carrier-mediated transport can offer a solution to the impermeability based on the BBB or the BRB.

It is well-known that most CNS drugs contain N-containing groups [11,12]. This discovery implies that the N-containing groups serve as a recognition unit for transporters. Comprehensive verification has not been conducted, but some have verified the substrate specificity of the transporter. Memantine [Figure 2], clinically employed for treating AD as an antagonist of the N-methyl-D-aspartate (NMDA) receptor, forms a positively charged salt under physiological pH in the blood flow, making it unable to pass through the apical membrane of capillary endothelial cells via passive diffusion. Indeed, memantine exhibited concentration-dependent membrane permeation, with brain uptake saturating above a certain concentration [22]. Therefore, memantine, featuring an N-containing group, crosses the membrane through carrier-mediated transport. This suggests that CNS drugs containing N-containing groups are transported from the blood flow into the brain via carrier-mediated transport.

Transporters dynamically convey their structurally compatible substrates through their pore, transitioning from an outward state to an inward state or from an inward state to an outward state in an alternative access mechanism [23]. The precise transport mechanism remains elusive. Computational transport simulations can elucidate the mode of interaction among a transporter protein, its substrate, and other molecules, such as ions, depending on the case of antiporters and symporters, with the exception of uniporters. High-molecular-weight compounds, such as nanoparticles or monoclonal antibodies, cannot physically pass through their narrow pores due to size limitations dictated by the principles of biophysical structuralism. However, high-molecular-weight compounds can still bind to them through covalent or non-covalent interactions. Transporters exhibit relaxed substrate specificity. Thus, compounds mimicking their authentic substrates could be recognized by them as a transporter recognition unit. Incorporating the characteristics of brain-penetrating compounds allows designed compounds to enter the brain through carrier-mediated transport, utilizing cation transporters that recognize the N-containing groups. Nonetheless, compounds with the N-containing groups can be recognized by several cation transporters including OCTN2, H⁺/cation antiporter, OCT1, and OCT2 simultaneously due to their ambiguous substrate recognition ability. In addition, it is true that positively charged compounds such as cell-penetrating peptides (CPPs), including TAT, bind to negatively charged heparan sulfate proteoglycans (HSPGs) on the cell surface to induce endocytosis [24]. The clustering of positively charged compounds and HSPGs induces endocytosis in some cases. On the other hand, the internalization across the membrane into cells by low-molecular-weight compounds with N-containing groups is carrier-mediated transport using SLC transporters. Low-molecular-weight compounds do not form the clusters with HSPGs and subsequently induce endocytosis, although they happen to be endocytosed as by-standers. However, low-molecular-weight substrates covering nanoparticles can form such clusters that are endocytosed.

2.2. Carrier-Mediated Transport Using OCTN2

OCTNs are organic ion transporters, including OCTN1, OCTN2, and OCTN3 [25]. Specifically, OCTs and OATs have evolved molecularly from OCTN2, as zwitterionic carnitine is a substrate of OCNT2 [26]. Consequently, OCTN2 may exhibit broad substrate tolerance. OCTN2 is composed of 557 amino acids and contains 12 transmembrane spanning domains, with domains 1–7 responsible for cation transport.

Amisulpride [Figure 2], an antipsychotic agent that blocks the dopamine D2 receptor, contains N-containing groups and may therefore be a substrate of cation transporters, including amine transporters at the BBB. It was reported that amisulpride was transported into cells that overexpressed 5.9-fold greater OCTN2 than the control cell line transfected with an empty pcDNA5 vector. This uptake was inhibited by L-carnitine, an OCTN2 substrate [27]. The rank order of expressed mRNA levels in human CMEC/D3 cells was OCTN2 >> OCTN1 > PMAT >> OCT3 > OCT1 [28]. CMEC/D3 cells are brain microvascular endothelial cell lines. Therefore, utilizing OCTN2-mediated transport presents an excellent strategy for delivering substances into the brain across the BBB. Substrates of OCTN2 include L-carnitine (the Michaelis-Menten constant (Km) 8–87 μM [29]), acetyl-L-carnitine (Km 8.5 μM [30]), acetylcholine, dopamine, norepinephrine, thiamine, quinidine, verapamil, TEA (Km 0.3–74 mM [29]), 1-methyl-4-phenylpyridinium, pyrilamine (mouse kidney slice; Km 236 μM [30]), diphenhydramine, procainamide, lidocaine [31], spironolactone, mildronate (Km 26 μM [30]) [32], etoposide (independent of Na⁺; Km 150 ± 34.1 µM), cephaloridine, ipratropium (Km 53 μM [30]), tiotropium, emetine, entecavir, imatinib, oxaliplatin, colistin [33], and amisulpride (Km 185.3 ± 68.0 μM) [27] [Figure 3]. Interestingly, [¹⁴C]-colistin was not transported by OCT1 and OCTN1, whereas it was transported by OCTN2 [34].

It is well-known that H⁺/cation antiporters transport N-containing low-molecular-weight compounds such as pyrilamine at the apical membrane of capillary endothelial cells [11,12,14,15]. Evaluations for drug delivery through carrier-mediated transport using H⁺/cation antiporters [14,15] have been conducted more extensively than those involving OCTN2. Currently, research on OCTN2-mediated transport has primarily focused on tissues such as the small intestine, kidneys, or lungs, which are distinct from the brain.

(i) Acetyl-N-[³H]-methyl-L-carnitine uptake was inhibited by D-carnitine or verapamil (an OCTN2 substrate), reducing it to approximately 30% of the control in an in vitro assay using human pulmonary epithelial cells [35]. Thus, acetyl-L-carnitine was transported into human pulmonary epithelial cells via OCTN2-mediated transport.

(ii) Moreover, gemcitabine is an anti-cancer drug that shows nucleoside metabolic inhibition and has low membrane permeability. Four conjugates of L-carnitine and gemcitabine as prodrugs [Figure 4] exhibited 3-fold stability and 5-fold transportation, dependent on Na⁺ and temperature, in an in vitro assay using human OCTN2-transfected HRPE cells, compared to gemcitabine alone as a control. This transportation was inhibited by L-carnitine alone [36]. Therefore, L-carnitine conjugates are suggested to be substrates for OCTN2.

(iii) Three renal-targeting prodrugs, namely ketoprofen–L-carnitine, ketoprofen–glycolic acid–L-carnitine, and ketoprofen–glycine–L-carnitine [Figure 5], were assessed for concentration-dependent uptake by OCTN2 using human OCTN2–MDCK monolayers. The dissociation constant Ki values were 82.2 ± 5.3 μM for ketoprofen–L-carnitine and 14.4 ± 1.4 μM for ketoprofen–glycine–L-carnitine. The Km values were 77.0 ± 4.0 μM for ketoprofen–L-carnitine and 58.5 ± 8.7 μM for ketoprofen–glycine–L-carnitine. J_max values of L-carnitine were 0.412 ± 0.015 pmol/(s cm²) for ketoprofen–L-carnitine and 0.0789 ± 0.0037 pmol/(s cm²) for ketoprofen–glycine–L-carnitine. Ketoprofen–glycolic acid–L-carnitine was unstable in metabolic and chemical buffers. Ketoprofen–glycine–L-carnitine demonstrated higher affinity for OCTN2 than ketoprofen–L-carnitine in an inhibitory assay [37]. This finding suggests that the structures of designed compounds should be fine-tuned through repeated evaluations.

(iv) Two L-carnitine ester derivatives of prednisolone for asthma treatment, namely prednisolone-L-carnitine (PDC) and prednisolone succinate-L-carnitine (PDSC) [Figure 6], were assessed for OCTN2-mediated transportation using human bronchial epithelial BEAS-2B cells. The rank order of transport at 37 °C was PDSC > prednisolone > PDC. PDSC exhibited 1.79-fold greater transport than prednisolone. This transportation was conducted in a temperature-, time-, and Na⁺-dependent manner, suggesting OCTN2-mediated transport due to the mechanism of OCTN2 [38].

Based on the above, it seems unlikely that studies on OCTN2-mediated transport to cross the blood-brain barrier (BBB) have been conducted. Therefore, it is worth exploring this aspect.

2.3. OCTN2-Mediated Endocytosis/Transcytosis

Interestingly, OCTN2 mediates not only carrier-mediated transport but also receptor-mediated endocytosis/transcytosis. A strategy for drug delivery via endocytosis/transcytosis using nanoparticles covered with transporter substrates is currently being pursued [39].

(i) L-Carnitine-conjugated nanoparticles, with or without polyethylene glycol (PEG) linkers, were taken up by Caco-2 cells as an in vitro model for intestinal uptake, compared to bare nanoparticles without L-carnitine modification. The uptake mechanism was suggested to be OCTN2-dependent endocytosis in an in vitro assay using selective inhibitors of different modes of endocytosis. For instance, indomethacin (an inhibitor of caveolin-dependent endocytosis) showed more than 80% uptake inhibition (100 μM), and chlorpromazine (an inhibitor of clathrin-dependent endocytosis) showed approximately 50% uptake inhibition (50 μM). Additionally, using a low temperature of 4 °C (a stopper for endocytosis machinery operation) showed more than 80% uptake inhibition. The in vivo uptake of oral L-carnitine-conjugated nanoparticles in the intestinal tracts of rats was evaluated. Nanoparticles without PEG linkers were taken up in the duodenum, jejunum, or ileum to a greater extent than nanoparticles with PEG linkers, respectively. Nanoparticles with zwitterionic L-carnitine at the tip of PEG linkers might have electrostatically interacted with the negatively charged mucin in the mucus layers. This formulation was applied for paclitaxel delivery [40].

(ii) L-Carnitine-conjugated poly(lactic-co-glycolic acid) (PLGA) nanoparticles (LC-PLGA NPs) encapsulating therapeutic drugs, such as paclitaxel, were developed. These nanoparticles were taken up by Caco-2 cells via OCTN2-mediated endocytosis. In the process of endocytosis, multipoint binding could increase the interaction and thereby enhance nanoparticle uptake. Paclitaxel in plasma was detected at a higher level for LC-PLGA NPs compared to bare PLGA NPs in in vivo pharmacokinetics after oral administration in rats [41].

(iii) L-Carnitine is a substrate not only for OCTN2 but also for the Na⁺/Cl^–-coupled transporter ATB^0,+ (SLC6A14), which transports neutral and basic amino acids. LC-PLGA NPs targeting OCTN2 and ATB^0,+ were developed. The expression levels of OCTN2 and ATB^0,+ are higher in colon cancer cells than in normal colon cells. Indeed, 5-FU-loaded LC-PLGA NPs increased uptake and enhanced anti-tumor activity in colon cancer cells positive for OCTN2 and ATB^0,+ compared to 5-FU alone. Drug release was suggested to occur due to lysosomal escape facilitated by the fusion between nanoparticles and lysosomes. In this lysosomal process, OCTN2 proteins and ATB^0,+ proteins were likely to be degraded by lysosomal enzymes without being recycled to the plasma membrane of colon cancer cells [42]. Polyvalent interactions between the cell-surface ligands on the nanoparticles and cell receptors such as OCTN2 and ATB^0,+ might induce endocytosis. Nanoparticles cannot penetrate through the pores of transporters due to their size.

(iv) Stearoyl-L-carnitine-conjugated poly(lactic-co-glycolic acid) nanoparticles (SC-LC-PLGA NPs) were evaluated for the treatment of muscular pathologies such as muscular dystrophies due to OCTN2 expression in skeletal muscle cells. Indeed, SC-LC-PLGA NPs demonstrated increased uptake compared to nontargeted carriers in myotubes. Thus, the SC-NPs have the potential to be candidates for delivering therapeutic agents against muscular pathologies [43].

(v) Furthermore, LC-PLGA NPs through PEG crossed the BBB via OCTN2-mediated transcytosis in the BBB endothelial cell line hCMEC/D3. Additionally, LC-PLGA NPs entered glioma cells through OCTN2-mediated endocytosis in the glioma cell line T98G. LC-PLGA NPs containing paclitaxel as a payload demonstrated greater anti-glioma efficacy in both 2D-cell and 3D-spheroid models compared to paclitaxel alone or paclitaxel-loaded unmodified PLGA NPs [44]. Nanoparticles cannot pass through the pores of OCTN2 due to their size, unlike low-molecular-weight compounds. Relatively low-molecular-weight paclitaxel can penetrate the membrane via passive diffusion [45]. Thus, endosomal escape of paclitaxel in glioma cells might occur through passive diffusion across the endosomal membrane. Paclitaxel is a substrate of MDR1 [46], although it is anticipated that glioma cells adjacent to killed glioma cells might be affected by bystander effects due to paclitaxel’s passive diffusion across the plasma membrane. It was revealed that the treatment resistance of glioblastoma (GBM) cells was associated with cytoprotection based on the increase of OCTN2 and its substrate L-carnitine [47]. L-Carnitine facilitates the transport of middle- and long-chain acids as an energy resource into mitochondria for β-oxidation. Thus, OCTN2-mediated endocytosis at the BBB and subsequently at GBM cells might be effective for GBM therapy. Moreover, the expression of ATB^0,+, which transports L-carnitine, was identified in glioma stem cells [48].

Receptor-mediated transcytosis using monoclonal antibodies targeting the transferrin receptor or insulin receptor is often employed when delivering substances across the BBB into the brain [9]. The affinity of an anti-transferrin receptor antibody for the transferrin receptor must be moderate. If the affinity is too high, such complexes in endosomes are degraded in the degradation pathway, leading to insufficient results. On the other hand, if the affinity is moderate, such complexes in endosomes are separated during acidification and are exocytosed via the secretory pathway. Thus, affinity tuning is important [9]. It is a matter of curiosity whether monoclonal antibodies targeting OCTN2 induce receptor-mediated transcytosis or not. Ligand-receptor clustering induces endocytosis [49,50,51]. Nanoparticles covered with several L-carnitine molecules might form clusters based on the cross-linking of ligands on a nanoparticle and receptors on the cell surface, subsequently undergoing endocytosis. The mechanisms of endocytosis in this process based on ligand-receptor clustering remain unknown, although there are many types of endocytosis. If nanoparticles are used as a carrier, L-carnitine is more preferable as a vector than anti-OCTN2 monoclonal antibodies in terms of manufacturing, cost, and preservation. Intriguingly, nanoparticles covered with several transferrin molecules induced endocytosis instead of anti-transferrin receptor monoclonal antibodies [52,53]. It is useful to know that even ONTN2 ligands mediated endocytosis through clustering. However, the high selectivity of monoclonal antibodies might induce endocytosis more effectively than L-carnitine or transferrin, overcoming such disadvantages. Thus, receptor-mediated transcytosis using monoclonal antibodies is also an attractive method. It is currently unknown which protein shows a higher expression level at the BBB, OCTN2 or the transferrin receptor. The comparable mRNA amounts of OCTN2 and the transferrin receptor in human CMEC/D3 cells are not clear. Accordingly, the drug design of nanoparticles to cross the membrane via endocytosis might be subject to reconsideration regarding which receptor protein to use or which vector to employ. Antibodies administered through the parenteral route do not impede L-carnitine transportation in the small intestine and may not induce serious carnitine deficiency.

2.4. A Promising Method for Delivering Substances into the Brain across the BBB

To prevent serious off-target side effects, achieving highly selective distribution into the brain is crucial. OCTN2 is expressed in various tissues and plays a vital role in carnitine homeostasis [54]. Carnitine influences the mitochondrial membrane transport of long-chain fatty acids [55]. It is abundant in the kidneys and thyroid, while also being expressed to some extent in the heart, brain, liver, small intestine, and lungs [21]. Therefore, off-target side effects due to incorrect distribution might occur. In the kidneys, OCTN2 is localized on the apical membrane of renal tubular epithelial cells and plays a physiological role in the reabsorption of glomerular filtered L-carnitine across the renal tubular epithelial cells [56]. Thus, designed OCTN2 substrates would be reabsorbed in the kidneys after glomerular filtration or exported in urine without causing severe adverse effects. In the thyroid, it is uncertain how transporters function in the literature, although OCTN2 is abundantly expressed there. The thyroid gland is a smaller organ than the brain. Few designed OCTN2 substrates might be transported via OCTN2-mediated transport [36,37,38,40]. In the small intestine, OCTN2 is localized on the apical membrane of enterocytes. The uptake of designed OCTN2 substrates might be enhanced via carrier-mediated transport. This is advantageous for oral CNS agents in terms of bioavailability. The amount of oral OCTN2 substrate uptake is enhanced by OCTN2 in the small intestine. Oral drugs are easier for patients to take, compared to intravenous drugs. Oral low-molecular-weight drugs have several merits, such as manufacturing cost, storage at room temperature, and ease of taking them. In the liver, OCTN2 is localized on the apical membrane of hepatocytes. Designed OCTN2 substrates might be uptaken in the liver and subsequently enzymatically metabolized or exported in bile. In the lungs, OCTN2 is localized on alveolar epithelial cells [35]. Designed OCTN2 substrates in the blood flow would not interact with OCTN2 directly. In the heart, immunoreactivity for OCTN2 was observed in the plasma membrane of cardiac muscle cells and in the membranes of the intercalated disc of mouse ventricle cardiomyocytes. L-[³H]-carnitine was taken up into the cardiac muscle cells via OCTN2-mediated transport [57]. Nonetheless, it was suggested that OCTN2 might contribute to the cardiac uptake of cardiovascular drugs in the human heart [28]. Designed OCTN2 substrates in the blood flow would not interact with OCTN2 directly. Therefore, comprehensively, orally or intravenously administered designed OCTN2 substrates targeting CNS diseases might be transported to the liver to be metabolized instead of the brain. However, a sufficient amount of them might still be transported into the brain across the blood-brain barrier via OCTN2-mediated transport, avoiding incorrect distribution. Charged compounds such as carnitine derivatives cannot go through the membrane and therefore will not be exported by MDR1, which captures the hydrophobic compounds passing through the membrane based on a hydrophobic vacuum cleaner model [58,59] [Figure 1].

According to transporter-conscious drug design, based on ligand-based drug design, the design of CNS drugs transported by OCTN2 mimics the molecular structure of existing OCTN2 substrates. Two approaches are conceivable. One involves compounds that structurally contain a transporter recognition unit within the molecule itself through covalent bonds. The other involves prodrugs that connect a parent drug and a transporter recognition unit through a cleavable bond. In the latter approach, it is illuminated that L-carnitine esters of drugs such as GSC, GHC, GOC, and GDC [34] are promising compounds transported by OCTN2. A topology model of OCTN2 was established [60]. However, no results matching OCTN2 were found in the RCSB PDB (accessed on Friday, 13 October 2023). Thus, the X-ray structure of the OCTN2-L-carnitine complex, which is essential for structure-based drug design, is unknown. The optimization of OCTN2-mediated membrane permeability can be achieved through repeated experiments involving structural modifications. Competitive studies using [³H]-L-carnitine would reveal the possibility of OCTN2 substrates, although it does not matter which transporter mediates as long as the designed compounds pass through the membrane. It is known that some compounds are substrates for both OCTN2 and the H⁺/cation antiporter [11,12,14,15]. Ideally, when designed compounds are transported by OCTN2 and the H⁺/cation antiporter in the capillary endothelial cells at the BBB, they would be transported into the brain synergistically and effectively. Ultimate compound designs [Figure 7] possess both L-carnitine as an OCTN2 recognition unit and a 2-(dimethylamino)ethyl group as an H⁺/cation antiporter recognition unit. More ultimate compound designs possess only a 2-(dimethylamino)ethyl group as an H⁺/cation antiporter recognition unit. This is because some substrates, such as quinidine, pyrilamine, and diphenhydramine, due to relaxed substrate specificity, are recognized by both OCTN2 and the H⁺/cation antiporter. Additionally, H⁺/cation antiporter might not transport anion moieties. The X-ray crystal structures of OCTN2 and the H⁺/cation antiporter should be clarified to enable transporter-conscious drug designs based on computer-aided drug design in a computational approach, in addition to manual approaches.

On the other hand, in the former approach, most brain-penetrant compounds possess N-containing groups. Many compound designs are carried out largely as a result of empirical or incidental processes. Antihistamine drugs enter the brain and induce drowsiness as a side effect. In fact, it is uncertain which transporters transport brain-penetrant compounds at the BBB. Recently, it was clarified that amisulpride was a potential substrate of OCTN2 [27]. Compounds with N-containing groups would be recognized by certain types of cation transporters, including OCTN2 or H⁺/cation antiporter. Certainly, as a rational design, L-carnitine esters of CNS drugs are potent as OCTN2 substrates to cross the BBB. Nonetheless, the tissue-to-blood concentration ratio of [³H]-L-carnitine (250 ng/kg) was 0.38 ± 0.05 in the brain, 6.92 ± 1.25 in the lungs, 7.89 ± 1.58 in the heart, 8.71 ± 0.60 in the liver, 9.82 ± 2.10 in the kidneys, 4.98 ± 0.19 in the digestive tract, 10.4 ± 3.5 in the pancreas, and 1.05 ± 0.20 in the muscle, 4 h after intravenous injection in an in vivo assay using mice [61]. There is no need for pessimism because unmodified compounds do not enter the brain largely due to the presence of the BBB [62]. However, wrong distribution should be avoided. Nanoparticles (40–50 nm in diameter), appropriately covered with several types of ligands to the receptors such as OCTN2, H⁺/cation antiporter, transferrin receptor, insulin receptor, or low-density lipoprotein (LDL) receptor that are expressed on the surface of the capillary endothelial cells at the BBB, would form cross-linkages to induce transcytosis across the BBB and elicit activity in the brain without off-target side effects [9].

2.5. Possibilities of Carrier-Mediated Transport into the Eyes Using OCTN2

A large number of patients suffer from ocular diseases such as diabetic retinopathy and age-related macular degeneration. Delivering drugs into the eyes is challenging due to the blood-retinal barrier (BRB) [63], which is anatomically divided into the inner BRB, composed of retinal capillary endothelial cells [64], and the outer BRB, composed of retinal pigment epithelial cells. It has been revealed that OCTN2 is expressed in rat retinal capillary endothelial cells at the inner BRB [17] and in human corneal and conjunctival epithelial cells [16]. Thus, OCTN2-mediated substance delivery into the eyes across the BRB might be established in a manner similar to OCTN2-mediated substance delivery into the brain across the BBB or H⁺/cation antiporter-mediated substance delivery into the eyes across the BRB [15]. However, studies on substance delivery into the eyes using OCTN2 have not been actively conducted. Currently, it is uncertain whether OCTN2 is expressed at the human inner BRB. Eye drops are considered a suitable formulation for substance delivery using OCTN2 expressed in corneal and conjunctival epithelial cells. Nonetheless, active ingredients in eye drops might pass through the nasolacrimal duct before OCTN2 transports them. Therefore, breakthrough ideas are required for OCTN2-mediated drug delivery into the eyes.

3. Conclusions

CNS drug development for neurodegenerative diseases faces challenges in crossing the BBB into the brain. It had been believed that hydrophobic brain-penetrating compounds were transported into the brain via passive diffusion. Currently, drug modalities have shifted from relatively low-molecular-weight compounds to high-molecular-weight compounds such as antibody drugs. Consequently, this wrong perception that low-molecular-weight substances would cross the membrane across the BBB via passive diffusion has delayed CNS drug development. The fact that hydrophobic brain-penetrating compounds possess N-containing groups in their molecules suggests the execution of carrier-mediated transport using cation transporters. Actually, clinically available CNS drugs such as memantine (H⁺/cation antiporter substrate) [11] and amisulpride (OCTN2 substrate) [27] are substrates of cation transporters. The characterization of OCTN2, such as the amino acid sequence and topology, is known, although that of the H⁺/cation antiporter remains unknown. Nonetheless, the design of most brain-penetrating compounds with N-containing groups was carried out empirically or incidentally. Therefore, rational transporter-conscious drug design using OCTN2 can be conducted, although this strategy has not been trialed much [

Table 1. Summary of OCTN2-mediated drug delivery into cells including CNS diseases and other diseases discussed in this review.

#	Compounds	Disease	OCTN2 Substrate	Cargo	Linker between Cargo and Carrier	Uptake Mechanism	Status	References
1	Conjugates of L-carnitine and gemcitabine	Cancer	L-Carnitine	Gemcitabine	Cleavable alkyl chain	Carrier-mediated transport	Basic research	[36], [Figure 4]
2	Conjugates of L-carnitine and ketoprofen	Inflammation	L-Carnitine	Ketoprofen	Cleavable alkyl chain	Carrier-mediated transport	Basic research	[37], [Figure 5]
3	Conjugates of L-carnitine and prednisolone	Asthma	L-Carnitine	Prednisolone	Cleavable alkyl chain	Carrier-mediated transport	Basic research	[38], [Figure 6]
4	Paclitaxel-encapsulated poly(lactic-co-glycolic acid) nanoparticles (PLGA NPs) conjugated to L-carnitine	Cancer	L-Carnitine	Paclitaxel	Non-covalent bond	Receptor-mediated endocytosis	Basic research	[40]
5	Paclitaxel-encapsulated LC-PLGA NPs conjugated to L-carnitine	Cancer	L-Carnitine	Paclitaxel	Non-covalent bond	Receptor-mediated endocytosis	Basic research	[41]
6	5-FU-encapsulatied PLGA NPs conjugated to L-carnitine	Colon cancer	L-Carnitine and ATB0,+	5-FU	Non-covalent bond			[42]
7	Stearoyl-L-carnitine conjugated poly(lactic-co-glycolic acid) nanoparticles (SC-LC-PLGA NPs)	Muscular dystrophies	L-Carnitine	-	Non-covalent bond	Receptor-mediated endocytosis	Basic research	[43]
8	Paclitaxel-encapsulated PLGA NPs conjugated to L-carnitine	Glioma	L-Carnitine	Paclitaxel	Non-covalent bond	Receptor-mediated endocytosis	Basic research	[44]
9	Conjugates of L-carnitine and CNS drug	CNS disease	L-Carnitine	CNS drug	Cleavable alkyl chain	Carrier-mediated transport	Under analysis in Tashima lab	[Figure 7]
10	Conjugates of L-carnitine and CNS drug	CNS disease	N-containing group	CNS drug	Cleavable alkyl chain/non-Cleavable alkyl chain	Carrier-mediated transport	Under analysis in Tashima lab	-
11	CNS drug-encapsulated nanoparticle conjugated to L-carnitine	CNS disease	L-Carnitine	CNS drug	Non-covalent bond	Receptor-mediated endocytosis	Under analysis in Tashima lab	-

]. Furthermore, OCTN2-mediated drug delivery into the eyes has not been investigated yet. Unmet medical needs persist in serious eye diseases such as diabetic retinopathy and age-related macular degeneration [65]. Various transporters, such as OCTN2 and H⁺/cation antiporter [11,12,14,15], are expressed in the eyes. In general, transporters recognize their substrate structure when transporting based on an alternating access model. It was reported that prodrugs with L-carnitine as an OCTN2 substrate for tissues other than the brain were transported across the cell membrane. Hence, the addition of L-carnitine as a transporter recognition unit to the parent compound through a suitable cleavable linker can establish drug delivery into the brain across the BBB. Moreover, nanoparticles covered with L-carnitine were transported via receptor-mediated endocytosis, probably due to ligand-receptor clustering, different from carrier-mediated transport. Therefore, OCTN2 is a promising transporter for brain-targeted delivery. Inspired readers are encouraged to submit better ideas for effective brain drug delivery across the BBB or the BRB.

Drug design should be interdisciplinary. The target tissue, cell, or protein is a part of the living body. It is necessary to know the essence of living organisms. The decoding of the human genome was completed in April 2003. At that time, genome drug discovery, utilizing high throughput screening and combinatorial chemistry, was optimistically expected to generate innovative drugs. However, this expectation did not materialize, and not many new drugs have been launched in the market. This pessimistic outcome is attributed to a lack of seeds and technologies, as well as the commodification of existing modalities. Human genome decoding revealed that the number of protein-coding genes is 20,000 to 25,000 in humans and 30,000 in mice [66]. It is true that genotype-based drug discovery has yielded innovative results, such as the discovery of receptors among orphan receptors, but it might have limitations. The number of long non-coding RNAs (96,308) in humans is larger than that in mice (87,774) [67]. Thus, long non-coding RNAs are transcripts that are suggested to play a role in regulating humanity as a whole under epigenetic regulation. The trajectory of OCTN2 substrates is controlled by the OCTN2 machinery systems that are constructed by the human genome and epigenetic regulation based on long non-coding RNAs. Phenotype-based drug discovery is considered to be carried out based on tissues, cells, organelles, and materials regulated by the physical and biological machinery systems, following the structuralism advocated by Dr. Lévi-Strauss [7,8]. Therefore, medical chemists and pharmaceutical scientists should comprehensively engage in drug development to create innovative pharmaceutical agents.