Around 90% of patients who respond will have disease recurrence, which usually progresses to the development of chemoresistance [3]. The 5-year survival is also at around 49% [26]. Therefore, development of alternative antitumour drugs is warranted.
Claudin antibody therapy
Presently, antibody therapy targeting claudins has been an important area of research, with a few entering the phase I/II clinical trials, particularly those targeting CLDN18.2 [27, 28] and CLDN6 [29]. In early developmental research currently, human anti-CLDN3 IgG1 (IgGH6) antibodies have been developed. They have very high specificity to CLDN3 molecules. Confocal microscopy has shown IgGH6 to be actively internalised in the tumour cells, following native CLDN3 binding, and co-localised, probably within intracellular vesicles, with a Clostridium perfringens enterotoxin (CPE) peptide. Therefore, this selective uptake into tumour cells indicates a potential use for antibody-drug conjugate for therapeutic ovarian cancer applications [30]. To date, there are no clinical trials on anti-CLDN3/4 antibodies for ovarian or other cancers. However, the principle of exploiting claudin overexpression using monoclonal antibodies was utilised in a phase I/II clinical trial [NCT02054351]. Preclinical models showed IMAB027 (anti-CLDN6) to induce potent antitumour activity by antibody- and complement-dependent cytotoxicity whilst ensuring there are no off-target effects. Preliminary phase I trial data showed that IMAB027 presents a safe and well-tolerated treatment for recurrent, advanced ovarian cancer patients [31].
Clostridium perfringens enterotoxin
Both CLDN3 and CLDN4 are the natural receptors for CPE. This potent toxin induces rapid cytolysis through increasing membrane permeability. They form small complexes that oligomerise and create a hexameric pore on the membranous surface, which leads to calcium influx and cell death [32]. It is also understood that occludin also participates in the breaking down of the TJ. A 30-amino acid long peptide on the C-terminus of the CPE allows it to bind to the relatively large extracellular loops on the claudins (Fig. 4) [32]. Therefore, considering CLDN3 and CLDN4 abundance in ovarian cancer, several studies have highlighted CPE as a potential antitumour drug. Proof-of-principle demonstration for CPE use as a chemotherapeutic agent has been successful in in vitro studies [33]. Through in situ mRNA hybridisation, CLDN3 expression was found in vivo in human prostate carcinoma epithelium. The level of cytotoxicity correlated with CLDN3 overexpression in a primary culture of metastatic prostatic adenocarcinoma treated with CPE [34]. This toxicity suggests CPE to likely be cytotoxic in vivo and therefore act as an alternative to chemotherapy.
Further cell culture work has found CLDN4 also to be the main target for CPE. In another study using CLDN4 overexpressing human pancreatic cancer cell lines, an acute dose-dependent cytotoxic effect was discovered when treated with CPE [36]. This was interestingly restricted to CLDN4 expressing cells, and the strength of CLDN4 expression determined the magnitude of the effects. The toxicity was determined in vitro by trypan blue exclusion and the 86Rb-release assay. This study also went further and assessed CPE’s activity in vivo, where CPE intramural injection on nude mice with PANC-1 cell line xenografts demonstrated extensive tumour cell area necrosis and tumour growth reduction.
The effects of CPE are further highlighted in another study utilising chemosensitive and chemoresistant tumour samples. Fresh human ovarian cancer cell lines and established cancer cell lines were used to evaluate CLDN3 and CLDN4 expression by real-time qRT-PCR. Established ovarian cancer cell line OVA-1 and a fresh OSPC, found to be chemoresistant, were used to establish ovarian xenografts in severe combined immunodeficient (SCID) mice [13]. The study demonstrated that multiple intraperitoneal CPE administrations at sublethal doses lead to significant tumour growth inhibition in all SCID mouse xenografts, thereby circumventing the initially chemoresistant properties of the xenograft tissue. Furthermore, tumour progression inhibition and survival extension were also induced by CPE. This phenomenon, however, occurred regardless of chemoresistance or chemosensitivity. Moreover, most of the mice that harboured OVA-1 xenografts and treated with repeated intraperitoneal CPE injections remained alive and free of any detectable tumour for over 120 days. Therefore, the findings of this study indicate that CPE-based therapy is an effective and beneficial treatment to ovarian cancer patients refractory to conventional treatment modalities. An exciting phenomenon seen in this study [13] is that CLDN3 and CLDN4 were also overexpressed at significantly higher levels in chemoresistant/recurrent tumours in contrast to chemosensitive tumours, concordant with previous studies [12, 15].
Chemotherapy and CPE
While more studies are required to clear the controversy between the expression patterns of CLDN3, CLDN4, and chemotherapy efficacy, the use of CPE to sensitise chemoresistant cells was explored. Gao et al. [35] showed that CPE infusion in a three-dimensional epithelial ovarian cancer culture model, developed using SKOV-3 and RMUG-L cell lines, downregulated CLDN4 and translocated it to the cytoplasm (Fig. 4). This was also evident in Madin-Darby canine kidney and Caco-2 colorectal carcinoma cells, where CLDN4 was specifically disintegrated and relocated after CPE administration, thereby diminishing TJ function. Analysis through qRT-PCR showed that the cell line OVCA-429 expressed the highest levels of CLDN4 mRNA, followed by SKOV-3, RMUG-L, and TOV112D expressing the lowest. Tumour growth inhibition and sensitisation, as measured by MTT assay, in response to Taxol and carboplatin, combined with CPE, was greatest in the OVCA-429 cell line [35]. This may be in response to the reduced barrier function due to reduced CLDN4 expression after CPE injection and therefore increased drug penetration and accumulation in the tumour core. This theory is further supported by the fact that SKOV-3 cell lines responded less, and TOV112D was completely resistant, as measured by cell growth and viability assays. Interestingly, for SKOV-3, there was no enhanced antitumour effect with carboplatin, despite a significant antitumour response with Taxol. This may be due to reduced CLDN4 levels and carboplatin being less potent than Taxol. Furthermore, the study using mice bearing SKOV-3 xenografts also showed that repeated intraperitoneal CPE injections could sensitise epithelial ovarian cancer cells to low-dose Taxol, thereby suppressing large tumour burdens in vivo [35]. This is concordant to Santin et al. [13] findings where multiple intraperitoneal administrations lead to a decreased tumour burden. Moreover, oligonucleotide microarray analysis between CPE-treated and control SKOV-3 cells showed that CPE treatment induced upregulation of genes such as NADH dehydrogenase (ubiquinone) 1 beta subcomplex and glutaminyl-peptide cyclotransferase-like, both important for intracellular protein degradation, receptor signalling regulation, proliferation, angiogenesis, and apoptosis. They have also been identified to attenuate molecules such as phosphoglucomutase 1, important for cellular metabolism. Therefore, it is understood that stimulation of the ubiquitin-proteasome pathways by CPE may contribute to the increased sensitivity of the tumours to chemotherapy.
In an in vitro study using human ovarian cancer cell lines, small-hairpin RNA and CLDN4 mimic peptide were used to silence CLDN4 gene expression and inhibit CLDN4 activity, respectively [15]. This interestingly improved the apoptotic response to paclitaxel in human-derived OVCAR-3 and PEO-4 ovarian tumour cells. OVCAR-3 cells with reduced CLDN4 proliferated more slowly with enhanced mitotic arrests when compared to their CLDN4 overexpressing cells. Furthermore, they identified that in OVCAR-3 cells, CLDN4 seemed to interact with the tubulin, thereby having a profound effect on the microtubular network polymerisation and structure. Consequently, reducing CLDN4 activity influenced the cells’ increased response to paclitaxel. Since the TJ is maintained throughout the cell cycle and cell division, CPE can deliver anticancer drugs at all times of the cell cycle. This is clinically important, as many chemotherapeutic agents require the cell to be at a specific cell cycle phase, such as paclitaxel’s action during the M phase. Therefore, the increased sensitivity to chemotherapy after CLDN3 and CLDN4 dysregulation highlights their potential as therapeutic targets.
CPE and imaging
In addition to its potential as an anticancer therapy, CPE has also been shown to be a promising tool for fluorescence imaging systems, important for patient management when receiving neoadjuvant chemotherapy [37]. This has been shown to play a vital role during interval debulking. Fluorescein isothiocyanate (FITC) can conjugate with CPE (FITC-c-CPE); this is capable of binding and internalising into many CLDN3 and CLDN4 expressing ovarian carcinoma cells both in vitro and in vivo. The conjugate also binds rapidly to tumours. This method is highly sensitive to the visualisation of peritoneal micrometastatic tumour implants and identifying ovarian tumour spheroids in malignant ascites in vivo in real time that can otherwise be missed by conventional visual observation. Furthermore, such optical methods have allowed for the conjugation of gold nanoparticles to CPE, which subsequently binds specifically to claudins. In an in vitro study, through utilising gold nanoparticle-mediated laser perforation techniques, ablation of cells derived from human and canine tumour cell lines was possible, eliminating more than 75% of claudin overexpressing cells and not majorly interfering with claudin non-expressing cells [38]. Therefore, both studies highlight the role of CPE in developing a practical optical approach in primary debulking surgery and identification of residual disease after neoadjuvant chemotherapy treatment.
Limitations of CPE therapy
Therapeutic approaches under development that utilise CLDN3 and CLDN4 seem primarily focused on CPE; its direct cytotoxic effect, ability to sensitise chemoresistant cancers, and screening/image-directed therapy capabilities justify the necessity to investigate this molecule further. Two studies administered CPE multiple times [13, 35], which may encourage the formation of neutralising antibodies, thereby reducing efficacy of CPE [39]. However, considering the immune dysregulation within the peritoneal cavity of advanced stage ovarian cancer patients, this may be overcome [39]. Also, the anti-enterotoxin antibodies are not made and released in time to prevent the consequences of CPE ingestion [39]. Therefore, these findings support the notion that the immune system may not impair the activity of CPE [39]. However, more clinical studies are warranted to accurately determine the immune response and, therefore, the efficacy against CPE in human patients.
Generally, intraperitoneal administration is preferred over intravenous administration due to significantly fewer adverse outcomes. However, this route requires the even distribution of the toxin throughout the abdominal cavity to reach the tumour tissue. Since many elderly ovarian cancer patients would have undergone surgery and subsequent adhesions, these may prevent the homogenous distribution of CPE therapy. This will reduce its localised efficacy [13]. Moreover, CPE will likely favour intraperitoneal ovarian tumour plaques through passive diffusion, as the distance is only a few millimetres. This means that local CPE administration in patients with a significant tumour burden will have reduced efficacy, due to its inability to deeply penetrate large tumour masses. This suggests that local CPE administration would mainly benefit patients with either microscopic residual disease or small-volume macroscopic cancers that are resistant to standard chemotherapy [13]. This, therefore, promotes the idea of downregulating CLDN3 and CLDN4 to sensitise the cells to chemotherapeutic agents using CPE. This is further supported by using CPE at high doses for a short period of time, leading to better efficacy and fewer adverse events. Furthermore, as this therapy is not reliant upon the immune system, it is beneficial for elderly ovarian cancer patients undergoing immunosuppressive chemotherapy [13]. Consequently, whilst CPE usage seems promising as an ovarian cancer therapy, it still requires phase I and phase II clinical trials, necessary to determine the feasibility of this therapeutic approach.