Abstract
Claudin (CLDN) proteins, a tetra-transmembrane family containing over 20 members, have been identified as key structural and functional components of intercellular seals, tight junctions (TJs). CLDNs are involved in the barrier and fence functions of TJs. Loosening the TJ barrier is one strategy for increasing drug absorption and delivery to the brain. Due to aberrant CLDN expression, the TJ fence function is frequently dysregulated in carcinogenesis. In addition, CLDN-1 is a co-receptor for the hepatitis C virus. Together these characteristics indicate CLDNs as promising targets for drug development, and CLDN binders are potential candidates for delivering drugs, treating cancer, and preventing viral infection. Before 2008, a receptor-binding fragment of Clostridium perfringens enterotoxin was the only CLDN binder available. Since then, several challenges regarding the generation of monoclonal antibodies against CLDNs have been surmounted, leading to breakthroughs in CLDN-targeted drug development. Here, we provide an overview of the recent progress in technology using created CLDN binders–anti-CLDN monoclonal antibodies.
Introduction
The paracellular spaces of epithelial and endothelial cells are sealed by tight junctions (TJs) that regulate the movement of ions and solutes from intercellular spaces across the epithelium and endothelium (Tsukita and Furuse, 2000). In addition, TJs act as “fences” that maintain cellular polarity by separating the apical and basolateral compartments of membranes and preventing the free movement of lipids and proteins on the membrane. Freeze-fracture replica electron microscopy revealed that TJs are composed of a meshwork of continuous, anastomosing intramembranous strands, called TJ strands (Farquhar and Palade, 1963; Staehelin, 1973). TJ strands contain three families of integral membrane proteins. Those in the TJ-associated MARVEL protein (TAMP) family have four transmembrane domains and two extracellular loops (ECLs); the TAMP family includes occludin, tricellulin, and marvelD3 (Raleigh et al., 2010). The junctional adhesion molecule (JAM) family is a member of the immunoglobulin super family (Van Itallie and Anderson, 2014) and contains JAM-A, JAM-B, JAM-C, coxsackie and adenovirus receptor, endothelial cell-selective adhesion molecule, JAM-L, and JAM-4 (Martìn-Padura et al., 1998; Bazzoni, 2003). The third protein family in TJ strands is the claudins (CLDNs), comprising of 27 members in mammals (Mineta et al., 2011).
CLDNs are critical components of TJ seals (Furuse et al., 1998). There are many barrier-forming, but also several channel forming CLDNs (Günzel and Fromm, 2012). A series of studies on these integral membrane proteins has revealed that CLDNs are promising targets for drug delivery, cancer treatment, mucosal vaccination, and antiviral protection (Kondoh et al., 2008; Hashimoto et al., 2016c, 2017a). Agents that bind the ECL domain typically are the first choice for drug development against membrane proteins. The first CLDN binder was Clostridium perfringens enterotoxin (CPE) (Sonoda et al., 1999). C-terminal receptor binding fragment of CPE (C-CPE) had been widely used for proof of concept for CLDN-mediated cancer therapy, enhancement of mucosal absorption, and mucosal vaccination (Kondoh et al., 2005; Saeki et al., 2009; Kakutani et al., 2010). Because C-CPE is immunogenic, the direct clinical application of C-CPE is limited (Suzuki et al., 2011). Several recent studies involving anti-CLDN antibodies (Abs)—the so-called second-generation CLDN binders—provided new insights into CLDN-directed drug development (Table 1). The ECL peptides of CLDNs are beyond the scope of this paper because the ECL peptides are not CLDN binders but just antagonists of trans interaction of ECLs. Here, we overview various CLDN binders and their application in proof-of-concept studies for CLDN-directed drug development.
List of anti-claudin monoclonal antibodies and their applications
CLDN Family
CLDNs comprise two ECL domains, four transmembrane (TM) domains, and two intercellular domains (Fig. 1). The ECLs of CLDNs form an antiparallel β-sheet structure composed of five β-strands (Suzuki et al., 2014). ECL1 is approximately 50 amino acids and contains four β-strands and a single extracellular helix. In contrast, ECL2 is approximately 25 amino acids, making up a single β-strand and the cell-surface-exposed region of TM3. The C-terminal end of almost all CLDNs is associated with zonula occludens (ZO)-1, -2, and -3 (Itoh et al., 1999a,b). To build TJ strands, CLDN family members interact with each other through their transmembrane domains within a single cell (cis interaction) and through their ECLs between adjacent cells (trans interaction) (Furuse et al., 1999; Suzuki et al., 2017). The expression profiles of CLDNs differ among tissues, and the resulting variety of CLDN member combinations results in the tissue-specific characteristics of the TJ-barrier and -channel. Therefore, the generation of member-specific molecules that uniquely recognize and bind the various ECLs is crucial for the development of CLDN-directed drugs.
Schematic illustration of claudin structure. CLDNs are 20- to 27-kDa molecules comprising four transmembrane (TM) domains with two extracellular loops (ECLs). ECL1 and ECL2 have approximately 50 and 25 amino acids, respectively. Secondary structural elements are indicated by rectangles (α-helices) and arrows (β-strands). The α-helix in TM3 is much longer than in the other TM domains. The thickness of the TM domains, except for TM3, is the same as that of the lipid bilayer. There is an extracellular helix between β4 strand and TM2.
The First CLDN Binder: Clostridium perfringens Enterotoxin
CPE is a food poison in humans (Shrestha et al., 2016). In 1997, the receptors for CPE were identified; 2 years later, the receptors were found to be identical to CLDN-3 and -4 (Katahira et al., 1997; Morita et al., 1999). CPE has two domains: the N-terminal cytotoxic domain and the C-terminal receptor-binding domain, which binds several CLDNs including CLDN-3 and -4 (Shrestha et al., 2016). Accordingly, CPE and C-CPE have been used therapeutically as a CLDN-targeted toxin and binder, respectively. In particular, CPE was used for CLDN-targeted cancer therapy (Michl et al., 2001; Gao and McClane, 2012), and C-CPE served as a delivery ligand for cancer-directed cytotoxic molecules and mucosa-targeted antigen (Saeki et al., 2009; Kakutani et al., 2010). The mucosal absorption-enhancing activity of C-CPE is 400-fold higher than that of a clinically used absorption enhancer, sodium caprate, thus establishing the proof of concept for CLDN-targeted enhancement of drug absorption (Kondoh et al., 2005).
In addition, C-CPE enhances the nasal, pulmonary, and intestinal absorption of biologics (Uchida et al., 2010). Because CPE and C-CPE both originate from a toxin, their antigenicity hampers their clinical application. Antigenic determination assays of CPE revealed that the C-terminal fragment corresponding to amino acids 286–305 was immunogenic (Sugii, 1994), and repeated mucosal administration of C-CPE induced the production of anti-C-CPE IgG (Suzuki et al., 2011). The C-terminal fragment corresponding to amino acids 290–319 constitutes the receptor-binding domain of CPE (Hanna et al., 1991). Using this 30-amino-acid fragment to prepare a claudin-binding peptide with low antigenicity and high CLDN affinity might yield a lead molecule for CLDN binders.
The Second-Generation CLDN Binders: Antibodies
Another potential direction for the development of CLDN binders is the generation of Abs, because high CLDN-specificity is preferable in order to minimize unintended side effects. Created in 2005, the first Abs targeting CLDN ECLs were polyclonal Abs generated by immunizing chickens with peptides corresponding to the ECL of CLDN-1, -3, or -4 (Offner et al., 2005). The first monoclonal Ab (mAb) targeting the ECL of a CLDN was developed through intramuscular immunization of mice with plasmid DNA encoding the ECL of CLDN-18.2 (Sahin et al., 2008). A series of studies focusing on the generation of anti-CLDN mAbs provided proof of concept for CLDN-directed drug development (Table 1).
CDLN-1 as Targets for Enhancement of Epidermal Absorption and to Prevent Hepatitis C Virus Infection.
The barrier function of the epidermis protects mammals from invading pathogens and materials. Epidermal administration would be an ideal route for drugs because it is noninvasive, easily discontinued, and avoids first-pass metabolism in the liver (Bäsler et al., 2016). However, owing to its barrier function, the epidermis prevents the influx of xenobiotics (including drugs) from the outer environment into the body. The epidermis has two major barrier structures: the stratum corneum and TJs in the stratum granulosum. The stratum corneum is the top layer of epidermis and constitutes the hydrophobic layer. One strategy to increase the epidermal absorption of drugs has focused on chemical and physical disruption of the stratum corneum (Asbill et al., 2000).
Another strategy to promote epidermal absorption is to enhance permeation through the TJs in the stratum granulosum. Small interfering RNA suppression of the CLDN-1 and -4 in these TJs (Brandner et al., 2015) in vitro loosened their integrity (Kirschner et al., 2013). Whereas CLDN-4-deficient mice showed no impairment of the epidermal barrier (Fujita et al., 2012), CLDN-1-deficient mice exhibited normal TJ morphology but dysregulation of the epidermal barrier against small molecules (∼600 Da) (Furuse et al., 2002). In addition, treatment of normal human epidermal keratinocytes—a model of the human epidermis—with an anti-CLDN-1 mAb (clone 7A5) or anti-CLDN-4 mAb (clone 3B11) weakened TJ integrity (Nakajima et al., 2015). Of note, treatment of normal human epidermal keratinocytes with anti-CLDN-1 mAb but not anti-CLDN-4 mAb increased the paracellular permeability of the macromolecule tracer 4-kDa fluorescein isothiocyanate–dextran (Nakajima et al., 2015). Together, these findings recommend CLDN-1 as a promising target for the development of an enhancer of epidermal absorption.
More than 184 million people worldwide are infected with hepatitis C virus (HCV), which is a major cause of liver cirrhosis and hepatocellular carcinoma (Thrift et al., 2017). Recent progress in direct-acting antiviral agents, such as sofosbuvir and simeprevir, that target HCV enzymes have improved treatment efficacy for HCV infection (Ahmed, 2018). Although preventing viral entry into hepatocytes is an attractive target for anti-HCV agents, the development of inhibitors of HCV entry remains challenging.
Numerous receptors regulate the entry of HCV into hepatocytes (Douam et al., 2015). Host factors involved in the initiation of HCV infection include heparin sulfate (Barth et al., 2006), low-density lipoprotein receptor (Monazahian et al., 1999), CD81 (Pileri et al., 1998), scavenger receptor class B type I (Scarselli et al., 2002), occludin (Ploss et al., 2009), epidermal growth factor receptor (Lupberger et al., 2011), Niemann-Pick C1-like 1 (Sainz et al., 2012), and CLDN-1 (Evans et al., 2007). To achieve infection, after various receptor-associated processes, HCV interacts with scavenger receptor class B type I on the membrane of hepatocytes, and then the viral E2 envelope glycoproteins bind to CD81. In response to signaling through activated epidermal growth factor receptor, the E2-CD81 complex diffuses to the membrane and, subsequently, CD81 interacts with CLDN-1. Finally, the E2-CD81–CLDN-1 complex is endocytosed with occludin. Formation of the CLDN-1 and CD81 complex is essential for HCV entry (Colpitts and Baumert, 2017). A concept of inhibiting the complex formation by anti-CLDN-1 mAb was confirmed by a Förster resonance energy transfer system (Mailly et al., 2015). These findings indicate that interference between CLDN-1 and CD81 may be a pathway for preventing HCV entry. An in silico model revealed that CD81 ECL2 interacted with the Q63 to V66 of CLDN-1 and I33 to Y35 of CLDN-1, which are responsible for the steric formation of CLDN-1 ECL (Davis et al., 2012). Anti-CLDN-1 mAb (clone, OM-7D3-B3), but not the anti-CLDN-1 mAb (clones, 3A2 and 7A5), was bound to the interface (Fig. 2). Anti-CLDN-1 mAbs (3A2 and 7A5) inhibited HCV infections in human liver chimeric mice (Fukasawa et al., 2015). In addition, an anti-CLDN-1 mAb (OM-7D3-B3) eliminated chronic HCV infection by preventing the de novo infections (Mailly et al., 2015).
Estimated binding regions and affinity of claudin (CLDN)-1 binders. Estimated binding regions and affinity of CLDN-1 binders are illustrated. Amino acid of CLDN-1 extracellular loops at each position shows one letter code. Estimated binding region of each CLDN-1 binder is annotated using dashed circles, and affinity of each binder is indicated in parentheses. Binding regions of each CLDN-1 binder were estimated from the literature: 3A2 and 7A5 (Fukasawa et al., 2015; Nakajima et al., 2015), and OM-7D3-B3 (Fofana et al., 2010). Interfaces of CLDN-1 and CD81 ECL2 are annotated as dashed rectangles. ECL1, extracellular loop 1; ECL2, extracellular loop 2.
CLDN-2 as a Target for Inflammatory Bowel Disease.
Inflammatory bowel diseases (IBDs), including ulcerative colitis and Crohn’s disease, are characterized by chronic relapsing idiopathic inflammation of the gastrointestinal tract that is associated with long-term and sometimes irreversible impairment of gastrointestinal structure and function (Podolsky, 2002). Patients with IBDs experience inflammation-induced “leaky flux” diarrhea resulting from the passive loss of ions and water from the circulation into the intestinal lumen as a consequence of an impaired intestinal barrier (Hering et al., 2012). The chronic intestinal inflammation is induced through numerous exogenous and endogenous signals and is mediated by various immune cells (de Souza and Fiocchi, 2016). Therefore, one therapeutic strategy for IBDs is to attenuate inflammation, and anti-inflammatory agents including steroids, thiopurines, methotrexate, and mAbs to tumor necrosis factor (TNF)-α have been used clinically (de Souza and Fiocchi, 2016). Because epithelial translocation of exogenous substances activates intestinal immune responses in IBDs, another treatment strategy is to attenuate this translocation by tightening the leaky intestinal epithelium.
In this context, CLDN-1, -3, -4, -5, -7, and -8 are frequently observed in tight epithelia, whereas CLDN-2 is primarily expressed in leaky epithelia (Luettig et al., 2015). Immunohistochemical analysis of colon biopsy samples from patients with IBDs revealed that CLDN-3, -5, and -8 were decreased in expression, whereas CLDN-2 expression was increased (Zeissig et al., 2007). Therefore, inhibition of CLDN-2 expression might increase TJ integrity and ameliorate intestinal inflammation.
Anti-CLDN-2 mAb (clone 1A2) was the first CLDN-2-specific binder developed; 1A2 dose- and time-dependently increased TJ integrity in human Caco-2 cells, which are frequently used as a monolayer model of gastrointestinal cells (Takigawa et al., 2017). Treatment of Caco-2 cells with TNF-α decreased TJ integrity, whereas treatment of the TNF-α-treated Caco-2 cells with 1A2 recovered it. Interestingly, co-treatment with 1A2 and an anti-TNF-α mAb (infliximab or adalimumab) exhibited additive enhancing activity in the recovery of TJ integrity. These findings suggest that inhibitors of CLDN-2-induced attenuation of TJ integrity are novel therapeutic drugs that are effective in combination therapy with anti-inflammatory agents.
CLDN-5 as a Target for Drug Delivery to the Brain.
The number of drugs that successfully complete clinical trial evaluation as central nervous system (CNS) drugs is less than half that of non-CNS drugs (Kaitlin, 2014). A current bottleneck in the development of CNS drugs is the lack of drug delivery systems to efficiently deliver drugs to the CNS (Pardridge, 2005; Gribkoff and Kaczmarek, 2017). The blood-brain barrier (BBB) is composed of endothelial cells, pericytes, and astrocytes, and BBB prevents toxic substances circulating in the blood from entering the brain (Rubin and Staddon, 1999). The intercellular spaces between adjacent endothelial cells are sealed by TJ seals, and knockout analysis revealed that CLDN-5 is a critical component of the TJ seals that prevents the influx of solutes (<800 Da) into the mouse brain (Nitta et al., 2003). The concept of enhancing drug delivery to the brain by modulating CLDN-5-based TJ seals in the BBB was first proposed more than a decade ago (Nitta et al., 2003).
The immunization of rats with plasmid DNA encoding CLDN-5 or their treatment with liposomes containing CLDN-5 led to the generation of CLDN-5-specific mAbs (Hashimoto et al., 2017b, 2018b). In a popular in vitro BBB model, treatment of the cells with anti-CLDN-5 mAbs loosened TJ integrity and thus enhanced the permeability of solutes (sodium fluorescein, 376 Da; dextran, 4 kDa) (Hashimoto et al., 2017b), thus demonstrating CLDN-5 binder as a promising lead for drug delivery to the brain. However, CLDN-5-deficient mice die within 10 hours after birth (Nitta et al., 2003). In addition, a method is available for injecting mannitol into patients with brain tumors; mannitol opens the TJ seal and enables the delivery of anticancer drugs directly to lesions. This therapy is associated with transient adverse effects only, such as the leakage of albumin into surrounding brain tissue, and is free of clinically significant adverse effects (Rapoport, 2000). Therefore, safety estimation and evaluation of a CLDN-5-directed delivery system is a critical point in translational research. In this regard, the injection of small interfering RNA to knock down CLDN-5 increased the permeability of solutes to the brain without any apparent adverse effects (Campbell et al., 2008; Keaney et al., 2015). TJs of the BBB comprise CLDN-1, -3, -5, and -12 as well as occludin and tricellulin (Reinhold and Rittner, 2017); electron microscopy analysis showed that knockout of CLDN-5 did not affect the morphology of TJ seals (Nitta et al., 2003). Taken together, these findings support CLDN-5 as a promising target in the development of a system for drug delivery to the brain.
CLDNs as Targets for Cancer Therapy.
Most malignant tumors are derived from epithelial cells (Jemal et al., 2008). Normal epithelial cells are polarized and spatially asymmetric owing to the fence function of TJs, which separates the apical and basolateral membrane domains by preventing free movement of protein and lipid on the cell membrane. Malignant transformation of epithelium is associated with the loss of both cellular polarity and epithelial integrity, thus leading to abnormal cell growth, the detachment of malignant cells from the primary tumor site, and the formation of distant metastasis (Wodarz and Näthke, 2007). In addition, cancer-associated alterations in the expression profiles of TJ proteins lead to dysregulation of TJ function. In particular, CLDNs are aberrantly upregulated or downregulated during the carcinogenesis of various malignant tumors (Osanai et al., 2017). As a feature of disrupted cell polarity and epithelial architecture, CLDNs translocate from TJs to the apical and basolateral domains of the cell membrane (Oliveira and Morgado-Diaz, 2007). Consequently the delocalized CLDNs on the cell membrane might be targets for cancer therapy.
Among the 27 CLDN family members, CLDN-3 and -4 are the CLDNs most frequently overexpressed in malignant tumors; in addition, their expression levels are correlated with tumor malignancy (Ding et al., 2013; Leech et al., 2015). Both CLDN-3 and -4 have been considered as useful cancer markers owing to their frequent overexpression in numerous cancers, especially prostate (Long et al., 2001), breast (Kominsky et al., 2004), pancreatic (Michl et al., 2001), and ovarian cancers (Hough et al., 2000). An anti-CLDN-4 mAb (KM3934) inhibited tumor growth in xenograft models of pancreatic and ovarian cancer (Suzuki et al., 2009). Another anti-CLDN-4 mAb (5D12) attenuated tumor growth in gastric and colon cancer without apparent adverse effects (Hashimoto et al., 2016a). In addition, anti-CLDN-3 and -4 bispecific mAbs (KM 3907 and 5A5) showed antitumor activity (Kato-Nakano et al., 2010; Li et al., 2014a). Because TJs retain their ability to seal intercellular spaces in cancer cells, modulation of TJs is a useful strategy for promoting the penetration of anticancer drugs into tumor tissues (Gao et al., 2011; Kuwada et al., 2015). In this context, an anti-CLDN-4 mAb (4D3) enhanced the permeability of anti-cancer drugs into tumor tissue, and combined treatment with 4D3 and anti-cancer drugs led to additive antitumor effects (Kuwada et al., 2015). Furthermore, treatment with C-CPE sensitized CLDN-4-expressing cancers to anticancer agents (Gao et al., 2011).
In addition to their overexpression in cancers, CLDN-3 and -4 are expressed in numerous normal tissues. For example, the first CLDN binder, C-CPE, accumulated in liver, intestine, and thyroid tissues, and injection of C-CPE-fused toxin led to liver toxicity in mice (Li et al., 2014b). CLDN-3-deficient mice developed gallstone diseases (Tanaka et al., 2018), and CLDN-4-deficient mice showed urothelial hyperplasia and lethal hydronephrosis (Fujita et al., 2012). The expression of CLDN-3 and -4 was age-dependently decreased in the liver, kidney, and pancreas (D’Souza et al., 2009). Together, these studies demonstrate that safety evaluation of CLDN-targeted cancer therapy is critical for the clinical use of anti-CLDN-3 and -4 mAbs.
Similar to the association of CLDN-3 and -4 and various cancers, CLDN-1 and -2 are frequently overexpressed in colorectal cancers (Miwa et al., 2001; Kinugasa et al., 2007), where CLDN-1 and -2 promoted tumor proliferation and invasiveness (Dhawan et al., 2005, 2011). In contrast, treatment with an anti-CLDN-1 mAb (6F6) suppressed tumor growth in colorectal cancer xenograft mice (Cherradi et al., 2017). In addition, anti-CLDN-2 mAb (1A2) attenuated tumor growth in xenograft mice (Hashimoto et al., 2018a). However, CLDN-1-deficient mice died within 24 hours after birth, and CLDN-2-deficient mice showed decreased reabsorption of Na+, Cl–, and water in proximal renal tubules, compared with wild-type mice (Muto et al., 2010). Furthermore, renal expression of CLDN-2 decreased during aging (D’Souza et al., 2009). Whether CLDN-1 and -2 are efficient and safe targets for cancer therapy remains to be clarified.
Clinical Research on CLDN-Targeted Cancer Therapy.
Currently, two anti-CLDN mAbs—the anti-CLDN-18.2 IMAB362 and the anti-CLDN-6 IMAB027 mouse–human chimeric mAb—are entered in clinical research (Sahin et al., 2015; Singh et al., 2017). The expression of CLDN-18.2 in normal tissues is strictly confined to differentiated gastric mucosa, and it is absent from the gastric stem cell zone (Sahin et al., 2008). In primary gastric cancers and their metastases, CLDN-18.2 is expressed on the outer cell membrane and thus is available for mAb binding (Sahin et al., 2008) (Table 1). CLDN-18.2 was expressed in ≥80% of advanced gastroesophageal cancers and overexpressed in ≥40% (Singh et al., 2017); in addition, CLDN-18.2 was expressed in 54.6% of pancreatic cancers (Wöll et al., 2014). A phase III trial of IMAB362 for advanced gastroesophageal cancer (NCT03504397) commenced in 2017. To date, none of the IMAB362-treated patients has shown CLDN mAb-related adverse effects.
CLDN-6 expression is limited to tissues during embryonic development but is activated in various cancers, including gastric adenocarcinoma, embryonic carcinoma, and choriocarcinomas (Micke et al., 2014) (Table 1). In particular, CLDN-6 was expressed in 69.4% of ovarian carcinomas (Wang et al., 2013). A phase I/II trial of IMAB027 for recurrent advanced ovarian cancer (NCT02054351) is ongoing; patients who received IMAB027 have not demonstrated CLDN mAb-specific adverse effects.
Conclusions
Since their identification in 1998, CLDNs have been important elements in epithelium-targeted drug development. Their roles have expanded from simply increasing drug absorption by modulating mucosal epithelium and now include epithelium-targeted epidermal absorption of drugs, drug delivery to the brain, cancer therapy, HCV therapy, mucosal vaccination, and the prevention of inflammatory diseases. However, only two anti-CLDN mAbs, both for cancer therapy, have entered clinical trials to date (Sahin et al., 2015; Singh et al., 2017). These anti-CLDN mAbs target CLDN-18.2, whose expression is strictly limited to the gastric mucosa, and CLDN-6, which is expressed only during embryonic development. CLDNs are expressed in normal tissues, and the analysis of CLDN-knockout mice has revealed specific abnormal phenotypes, including lethality, liver failure, and kidney failure (Furuse, 2009). Consequently, safety is the major concern regarding the clinical application of CLDN-directed drugs.
The abnormal phenotypes in CLDN-knockout mice are often surmised to result from dysregulation of the TJ seal function or the paracellular transport function of the CLDNs in TJs. Furthermore, the CLDNs that are drug targets in various scenarios—including mucosal vaccination, HCV therapy, cancer therapy, and IBDs—are localized in the cell membrane (Kakutani et al., 2010; Fukasawa et al., 2015; Takigawa et al., 2017). The cut-off size for permeability in paracellular spaces probably is less than 10 nm (Günzel and Fromm, 2012), whereas IgG molecules typically measure approximately 14.5 nm × 8.5 nm × 4.0 nm (Tan et al., 2008). These characteristics suggest that IgG-type CLDN binders will have minimal adverse effects because they will not enter into paracellular spaces between adjacent cells as they do not interact with the CLDNs incorporated into TJs. Therefore, bigger may be better, in the case of CLDNs, and anti-CLDNs mAbs are exciting potential lead compounds for CLDN-directed drug development.
Acknowledgments
We thank all members of the various laboratories for their useful comments and discussions.
Authorship Contributions
Wrote or contributed to the writing of the manuscript: Hashimoto, Okada, Shirakura, Tachibana, Sawada, Yagi, Doi, Kondoh.
Footnotes
- Received August 8, 2018.
- Accepted December 6, 2018.
This study was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (18K19400, 18H03190, 17K19487, 16K13044, and 24390042); a grant from the Japan Agency for Medical Research and Development (AMED); a grant from the Adaptable and Seamless Technology Transfer Program through Target-driven R&D, the Japan Science and Technology Agency (JST); a grant from the Platform for Drug Discovery, Informatics, and Structural Life Science of the Ministry of Education, Culture, Sports, Science and Technology, Japan; a grant from the JST CREST (JPMJCR17H3); and the Takeda Science Foundation. Y.H. was supported by a postdoctoral fellowship from the Japan Society for the Promotion of Science (JSPS) (17J03990).
Abbreviations
- Ab
- antibody
- BBB
- blood-brain barrier
- C-CPE
- C-terminal receptor-binding domain of CPE
- CLDN
- claudin
- CNS
- central nervous system
- CPE
- Clostridium perfringens enterotoxin
- ECL
- extracellular loop
- HCV
- hepatitis C virus
- IBD
- inflammatory bowel disease
- JAM
- junctional adhesion molecule
- mAb
- monoclonal antibody
- TJ
- tight junction
- TNF-α
- tumor necrosis factor-α
- Copyright © 2019 by The American Society for Pharmacology and Experimental Therapeutics
