Epithelial ovarian cancer (EOC) accounts for approximately 95% of ovarian cancer incidence, and is a leading cause of gynecologic cancer mortality worldwide (1, 2). Current standard-of-care treatment for newly diagnosed patients is cytoreductive debulking surgery plus neoadjuvant or post-operative platinum-based chemotherapy. Most patients initially respond to chemotherapy, but unfortunately up to 80% will eventually relapse leading to patient demise (3). Thus, platinum resistance presents a major clinical challenge. Angiogenesis inhibitor (bevacizumab) and the poly (ADP-ribose) polymerase inhibitors (olaparib, rucaparib and niraparib) provide some benefits for a subset of patients, but can only delay the relapse of platinum-resistant EOC (4, 5). Notably, recent large-scale clinical trials using immune-checkpoint inhibitors (anti-PD1/L1 monoclonal antibodies) failed to provide clinical benefit in EOC. In the past decades, the 5-year relative survival rates of ovarian cancer have only been moderately improved, from 43% in 1995 to 50% in 2018 in the USA (6, 7). Thus, treatment options for platinum-resistant EOC patients are limited, and present a major unmet clinical need.

Folate receptor alpha (FRα), encoded by the FOLR1 gene, has attracted considerable interest due to its high expression in several cancer types including those of lung and breast. FRα shows restricted tissue expression on the plasma membrane of epithelial cells in kidney, lung, ovary, fallopian tube, uterus, cervix, epididymis and placenta, and is highly expressed in approximately 80% of EOC. Additionally, the ability of FRα to internalize relatively large molecules renders it suitable for developing targeted therapies (8, 9). Despite their anti-tumor effects in preclinical models, folate-cytotoxic drug conjugates and no conjugated humanized antibody have yet to demonstrate clinical efficacies (10). In contrast, mirvetuximab soravtansine (MIRV), or Elahere (ImmunoGen), the first FRα-targeting antibody-drug conjugate (ADC), has recently been approved by the US FDA to treat platinum-resistant ovarian cancer (11). Here, we summarize the biology of folate receptors, review different strategies to target FRα, and discuss potential mechanisms of ocular adverse events associated with MIRV. The approval of MIRV has renewed interest to develop other FRα-targeting therapeutics for treatment beyond EOC.

Humans cannot synthesize folate, an essential vitamin for eukaryotic cell proliferation and differentiation, and must obtain folate from dietary sources (12). The uptake of extracellular folate is achieved mainly through three types of folate transporters, including the reduced folate carrier, RFC (encoded by the SLC19A1 gene), the proton-coupled folate transporter, PCFT(encoded by the SLC46A1 gene), and folate receptors (FRs) (13). Ubiquitously expressed RFC serves as the major route of folate transport into systemic tissues (12), whereas PCFT is a proton-coupled transporter responsible for dietary folate absorption in the small intestine (14). Both RFC and PCFT are low-affinity, high-throughput transporters. In contrast, FRs are high affinity, low-throughput transporters that transfer folate through endocytosis in selected tissues ( Figure 1 ).

Folate trafficking via FRα is considered to proceed via potocytosis, a lipid raft-mediated endocytosis mechanism (15). Folate binds specifically to FRα, forming a receptor-ligand complex, and subsequently intracellular vesicles are generated by invagination and budding off. Once internalized, the vesicles join together to from early endosomes, which acidify and fuse with lysosomes to release folates for the one-carbon metabolic reaction (16, 17).

There are four members in FRs family, including FRα (257aa, 30kDa), FRβ (255aa, 29kDa), FRγ (245aa, 28kDa) and FRδ (250aa, 28.6kDa), encoded by FOLR1 (Gene ID: 2348), FOLR2 (Gene ID: 2350), FOLR3 (Gene ID: 2352) and FOLR4 (Gene ID: 390243), respectively. FRs, also known as the folate binding proteins (FBPs), bind folic acid (FA) and 5-mTHF as well as folate-conjugated compounds with high affinity, and transport them inside cells by receptor-mediated endocytosis. FRα, FRβ and FRδ are all glycophosphatidylinositol (GPI) anchored cell-membrane proteins, whereas FRγ is a secreted protein lack of a GPI anchored region (18). FRα is the most studied family member, and is the focus of this Review. FRβ is mainly expressed in placental and myeloid leukocytes, including activated macrophages, tumor-infiltrating macrophages and acute as well as chronic myelogenous leukemia (19–21). FRβ-null mice are apparently normal, indicating that its function is dispensable to maintain organismal homeostasis (22). FRγ is expressed in neutrophil granulocytes and monocytes. FRδ, also named JUNO, is highly expressed in regulatory T cells and mammalian eggs. FRδ lacks the folate-binding pocket, and is unable to bind folate (23). The interaction between FRδ on the egg surface and IZUMO1 on the sperm surface is critical for mammalian fertilization as FRδ knockout eggs are unable to fuse with sperm (24).

FRα is mainly expressed on the plasma membrane of epithelial cells in several tissues, in particular the apical brush-border membrane of proximal renal tubular cells, retinal pigment epithelium, the choroid plexus (25), type1 and 2 pneumocytes in the lung, ovary, fallopian tube, uterus, cervix, epididymis, submandibular salivary gland, bronchial glands and trophoblasts in the placenta (26). FRα has a high affinity for reduced folates, such as tetrahydrofolate (THF), 5-mTHF and FA.

FA is a nutrient essential for embryonic development. Folate deficiency can cause embryonic lethality with neural tube defects and orofacial anomalies (27, 28). FRα and its cargo FA are essential for proper mammalian embryogenesis. Knockout of the Folr1 gene is embryonic lethal in mice around the time of neural tube closure (22). Reduced FRα expression and function is associated with craniofacial anomalies, abnormal heart development, and neural tube defects (29). Consistently, daily maternal folate supplementation, before and during pregnancy markedly decreased embryonic mortality. Hundreds of genes were differentially expressed at the gestational day 9.5 between Folr1^-/- and wild-type embryos. These genes are implicated in the regulation of digestive and cardiovascular system development (27). In the placenta, FRα transports folates from the mother to the fetus (30, 31). Folate deficiency in pregnancy is associated with neural tube defects, restricted fetal growth and fetal programming of diseases later in life (32–34). Importantly, the risk of abnormal pregnancy outcomes is increased in pregnant women taking folate antagonists to treat cancer and other diseases.

FRα is also required to maintains functionalities of several organs in adult animal. Adult mice lacking Folr1 had lower blood folate levels and higher renal folate clearance rate (35). This is because kidneys maintain folate homeostasis in the body through glomerular filtration and tubular reabsorption process. The primary transporter for folate reabsorption in the kidneys is FRα, expressed on the apical surface of proximal tubular cells. FRα transports folate from the tubule lumens into tubular cells via receptor-mediated endocytosis (36). Kidney ischemia-reperfusion injury significantly reduces the expression of FRα and RFC, contributing to low folate level in acute kidney injury (AKI) (37). In spontaneously hypertensive rat (SHR), a deletion variant in the Folr1 promoter region results in impaired folate reabsorption in the renal tubules, and increased risk for diabetes mellitus and cardiovascular disease (38). Within the brain, FRα is selectively expressed in the choroid plexus, and promotes a vesicular transport of 5-mTHF across the choroid plexus (39). It has been reported that mutations in the FOLR1 gene cause cerebral folate transport deficiency resulting in a childhood onset neurodegenerative disease (40–42).

FRα is normally expressed in fallopian tube but not the ovary, consistent with EOC originating from the fallopian tube fimbriae rather than from ovary epithelial cells (20, 43). The expression of FRα can be regulated by folate levels. Folate deficiency increases FRα expression in vivo and in vitro ( 44). Intracellular folate deficiency is associated with increased homocysteine. Homocysteine can promote the binding of heterogeneous nuclear ribonucleoprotein E1 (hnRNP E1) to the 5’ end of FOLR1 mRNA, upregulating FOLR1 expression at the level of translation (45). Folate deficiency also decreases DNA methylation, and global DNA hypomethylation may account for elevated of FRα expression in highly aggressive EOC (46). FRα levels correlate with histological stage and grade (47). A soluble form of FRα, known as soluble folate receptor (sFR), outperforms CA125 as a EOC recurrence marker, even when the CA125 level remains low (18, 48, 49).

The high expression of FRα in malignant tumors makes it a potential target for anti-tumors drug development. Various strategies have been explored, including monoclonal antibodies, antibody-drug conjugate (ADC), FRα-specific CAR T, vaccines, small molecules, and folate-drug conjugate ( Figure 2 ) (17, 58). Several clinical trials involving FRα-targeted agents are currently ongoing ( Table 1 ). Notably, an ADC drug has recently been approved by US FDA.

The understanding of the molecular characteristics of EOC have advanced in the past decade. However, platinum resistance remains a major clinical challenge, and renders EOC the most fatal gynecological malignancy. Angiogenesis inhibitors (bevacizumab) and PARP inhibitors (olaparib, rucaparib, and niraparib) have not significantly increased overall survival in most patients. Innovative and effective therapeutic strategies are urgently needed. In this regard, FRα has emerged as an appealing and clinically verified candidate for the development of targeted therapies. The relatively enriched expression of FRα on the surface of cancer cells and the ability of FRα to transport cytotoxic payloads into cancer cells have inspired the development of various therapeutic modalities including antibodies, ADCs, CAR T, vaccines, small molecules, and folate-drug conjugate. Notably, MIRV, a FRα-targeting ADC, has recently been approved by US FDA to treat adult patients with PROC, fallopian tube cancer or primary peritoneal cancer. Several promising FRα-targeting modalities are under clinical evaluation. It will be of interest to see their efficacy on EOC and other FRα-expressing cancer types.

It is also interesting that ADC is the only FRα-targeting modality that has achieved clinical efficacy so far. We speculate that the inhibition of FRα function via monoclonal antibodies may not be enough to inhibit tumor growth. This is because FRα is not a major survival signaling pathway even in FRα-high tumors. In addition, RFC is the major folate transporter and co-expressed with FRα. Although folate is an essential vitamin, suppressing FRα activity is not sufficient to block folate transport into cells. On the other hand, folate-drug conjugates can act similarly as FRα-targeting ADCs to deliver toxic payload into FRα-high cells. However, considering that RFC and PCFT are major folate transporters in many tissues, folate-drug conjugates likely can enter any cells expressing RFC and PCFT. Thus, folate-drug conjugates likely have less targeting specificity and therapeutical index than FRα-targeting ADCs.

In our opinion, further basic and clinical investigations are warranted to maximize the clinical efficacy of MIRV. MIRV is currently only approved for ovarian cancers with high expression of FRα. Considering that FRα is highly expressed in several cancer types, MIRV may be effective in these contexts. In addition, MIRV is known for its bystander effect. Therefore, MIRV may benefit patients with cancers expressing low to moderate level of FRα, analogous to the situation of HER2-targeting ADC, DS-8201a. Lastly, blurred vision occurs in 50-60% of patients treated with MIRV (115, 116). This peculiar high prevalence of ocular toxicity is uncommon in other ADCs, and can be debilitating for patients in our experience. The exact pathological mechanism is yet to be elucidated to improve the prophylactic treatment. Undoubtedly, the landmark approval of MIRV will fuel the interest to develop novel FRα-targeting diagnostic and therapeutic approaches to treat cancer.

JM collected the related paper and drafted the manuscript. LW and LY created the figures. TS, XL, RY, YJ and JL revised this manuscript. QL conceived the structure of manuscript and revised the manuscript. All authors read and approved the final manuscript. All authors contributed to the article and approved the submitted version.