Antifolates in cancer therapy: Structure, activity and mechanisms of drug resistance
Introduction
Folates are B9 vitamins that serve as one-carbon donors in multiple crucial biosynthetic pathways including de novo biosynthesis of purines and thymidylate, amino acid metabolism and methylation reactions (Stockstad, 1990). Folates are composed of three chemical components: pteridine ring, p-aminobenzoic acid (PABA) and a glutamate residue (Fig. 1). Folates can be found in an oxidized form, folic acid, or as the naturally occurring reduced folates. Folic acid can be synthesized and reduced by the normal bacterial flora that is resident in the small intestine. Reduced folates are found as the partially reduced form 7,8-dihydrofolate (DHF) or the reduced species 5,6,7,8-tetrahydrofolate (THF). THF cofactors are the biologically active congeners of folates. In humans, the primary circulating reduced folate is 5-methyl-THF (5-CH3-THF) (Fig. 1) which is found at low, yet sufficient physiological concentrations of ∼5–30 nM in the blood (Ifergan and Assaraf, 2008). Unlike bacteria and plants that possess an autonomous biosynthetic capacity to generate their own folate cofactors, metazoan organisms including mammals lack this enzymatic capacity for folate biosynthesis. Mammals must therefore obtain their folates from the diet; green leafy vegetables serve as the major source of our dietary folate intake.
In eukaryotes, folate metabolism is compartmentalized in two main subcellular compartments: the cytosol and mitochondria (Tibbetts and Appling, 2010). 5,10-CH2-THF is the cofactor necessary for the reductive methylation activity of thymidylate synthase (TS) that catalyzes the conversion of 2′-deoxyuridine monophosphate (dUMP) to 2′-deoxythymidine monophosphate (dTMP). The byproduct of this reaction is DHF, which is efficiently recycled to THF by dihydrofolate reductase (DHFR), an abundant key cytosolic enzyme that maintains the cellular THF cofactor pool (Fig. 2). The reduced folate cofactor, 10-CHO-THF is first used by the enzyme glycineamide ribonucleotide formyltransferase (GARFT), for the formation of the imidazole ring of purines; the second enzyme that utilizes 10-CHO-THF as a cofactor is 5-aminoimidazole-4-carboxamide ribonucleotide formyltransferase (AICARFT), a more downstream enzyme in the purine biosynthetic pathway which eventually generates the purine intermediate inosine 5′-monophosphate (IMP) (Fig. 2). Another cellular form of THF cofactor is 5-methyl-THF (5-CH3-THF) (Fig. 1); 5-CH3-THF serves as a cofactor, along with vitamin B12, for the catalytic activity of methionine synthase (MS) which mediates the conversion of homocysteine to methionine. ATP can be conjugated to methionine, thereby forming S-adenosylmethionine (AdoMet); the latter serves as the universal methyl group donor in multiple methylation reactions such as methylation of neurotransmitters, phospholipids, RNA, cytosine nucleotide residues within CpG islands in DNA, as well as proteins including histones (Reviewed in Fox and Stover, 2008) (Fig. 2).
THF cofactors that originate in the cytosol can be transported into mitochondria in their monoglutamate form via the mitochondrial folate transporter (MFT/SLC25A32) (Lawrence et al., 2011, Titus and Moran, 2000). Inside mitochondria, folates are predominantly used for the biosynthesis of formate, which is later used for cytoplasmic one-carbon reactions, as well as for mitochondrial glycine biosynthesis (Fig. 2). Mitochondria also serve as a prominent folate reservoir accumulating as much as 40% of total cellular folates, which does not exchange with the cytosol (Fox and Stover, 2008, Lin et al., 1993). However, some communication does occur between mitochondria and cytosol that is facilitated by the transport of one-carbon donor substrates such as formate, glycine and serine (Tibbetts and Appling, 2010) (Fig. 2).
THF cofactors are negatively charged under physiological pH due to the ionization of the dicarboxylic glutamic acid moiety. Therefore, folate cofactors cannot cross the plasma membrane by passive diffusion and rely on specific uptake systems for their entry into the cell. Accordingly, three transport systems are currently known to accommodate folate uptake:
(a) The reduced folate carrier (RFC/SLC19A1): RFC was the first transporter to be described at the kinetic level (Goldman et al., 1968, Zhao et al., 2009a). RFC belongs to the major facilitator superfamily (MFS) of transporters, and more specifically to the solute carrier (SLC) family of facilitative carriers (Pao et al., 1998, Saier, 1999). Within the SLC19A sub-family resides RFC (SLC19A1) and its close homologues, thiamine transporters THTR1 (SLC19A2) and THTR2 (SLC19A3) (Diaz et al., 1999, Dutta et al., 1999, Fleming et al., 1999, Labay et al., 1999, Rajgopal et al., 2001). RFC is a 591 amino acids transmembrane protein with a molecular mass of ∼85 kDa and 12 hydrophobic transmembrane domains (TMD) with short hydrophilic N-terminus and a long hydrophilic C-terminus, both of which reside in the cytoplasm (Cao and Matherly, 2004, Ferguson and Flintoff, 1999, Liu and Matherly, 2002). RFC contains a large cytoplasmic loop connecting TMD6 and TMD7. The human RFC is N-glycosylated at a single conserved consensus N-glycosylation site located at the loop connecting TMD1 and TMD2 (Asn58) (Wong et al., 1998). Although harboring a single N-glycosylation site, RFC undergoes heavy glycosylation at Asn58, which results in a ∼20 kDa increase in its molecular mass (i.e. core MW of ∼65 kDa vs ∼85 kDa of the N-glycosylated form) (Wong et al., 1998).
RFC is devoid of an ATP-binding domain and hence its transport activity is not driven by direct ATP hydrolysis. Instead, it is a bidirectional antiporter that facilitates the exchange of folates with organic phosphates such as adenine nucleotides as well as thiamine mono- and pyrophosphate, which are largely retained within the cell (Fig. 3) (Goldman, 1971, Henderson and Zevely, 1983, Zhao et al., 2002, Zhao et al., 2001b). This major asymmetry in the concentrations of organic phosphates across the plasma membrane constitutes the driving force for the uphill transport of folates into the cell via RFC (Goldman, 1971, Zhao et al., 2002, Zhao et al., 2001b). Consistent with RFC activity as a folate/organic phosphate exchanger, impaired folate efflux was demonstrated under depletion of extracellular anions. This effect could be reversed upon restoration of extracellular anions including AMP, phosphate or thiamine pyrophosphate (Henderson and Zevely, 1983). Moreover, it was recently shown that ectopic overexpression of RFC resulted in an approximately 15-fold decline in cellular viability in medium lacking folates but not in folate-containing medium, further supporting the bidirectional transport activity of RFC (Ifergan et al., 2008). As reflected in its name, RFC exhibits relatively high transport affinity for reduced folates (Km = 1–3 μM), but poor affinity for the oxidized folate, folic acid (Km = 200–400 μM) (Sirotnak, 1985).
In humans, RFC is ubiquitously expressed in a variety of normal tissues and malignant tumors, including bone marrow, breast, lung, heart, small intestine and lymphocytes (Liu et al., 2005, Matherly and Angeles, 1994, Matherly et al., 1992, Whetstine et al., 2002a). Multiple cDNA isoforms were described for the human RFC, differing in their 5′-untranslated regions (UTRs) (Gong et al., 1999, Tolner et al., 1998, Whetstine et al., 2002a). Human RFC gene expression is driven by two minimal promoters (A and B) and a linker region that consists of 3 additional binding elements (Whetstine and Matherly, 2001). Minimal promoter A is localized within a 47 bp region (positions −3912 to −3959) and is regulated by an inducible cAMP-response element (CRE)/AP1-like element (CRE/AP1). Minimal promoter B resides within a 46 bp region, located upstream to the transcription start site (positions −4508 to −4550). This promoter contains a constitutive element and is regulated by a GC-box. Apart from the constitutive GC-box and inducible CRE/AP-1 element, additional promoter elements including AP-2, myeloid zinc finger 1 (Mzf-1) and E-box are contained within or near four tandemly repeated sequences upstream of promoter A (Whetstine et al., 2002b).
Interestingly, a recent study suggested that RFC is crucial for the epigenetic regulation of neural crest development in Xenopus (Li et al., 2011). The authors demonstrated that inhibition of Xenopus RFC, blocked expression of a series of neural crest marker genes while overexpression of RFC or injection of 5-CH3-THF expanded the neural crest territories. They also found that mono- and trimethyl-Histone 3-K4 levels were dramatically lower under RFC knockdown. These results provide a possible explanation for the neural tube defect phenotype seen in embryos of pregnant women consuming a folate-deficient diet. One could speculate that decreased synthesis of AdoMet could contribute to this reduced methylation effect.
(b) The proton-coupled folate transporter (PCFT/SLC46A1): Another route of folate uptake occurs via the PCFT gene product (Qiu et al., 2006) (Fig. 3). PCFT is a 459 amino acids transmembrane protein with a molecular mass of 55 kDa (Unal et al., 2008). Based on hydropathy analysis, the predicted membrane topology of PCFT reveals 12 TMDs with a large intracellular loop after TMD 6; the N- and C-termini reside in the cytoplasm. The large predicted extracellular loop between the first and second TMDs contains two consensus N-glycosylation sites which were further confirmed by in vitro experiments (Qiu et al., 2007, Unal et al., 2008). Targeted disruption of the consensus N-glycosylation sites decreased influx activity by 40%; however, lack of N-glycosylation had no effect on the plasma membrane targeting of PCFT (Unal et al., 2008).
PCFT was initially described as a heme transporter (influx Km = 125 μM) and was therefore named heme carrier protein-1 (HCP1) (Shayeghi et al., 2005). However, it was subsequently shown that the hemin transport capacity of PCFT is lower by two orders of magnitude compared to folate transport (Qiu et al., 2006); this was recently confirmed by the group of McKie who first cloned HCP-1 (Laftah et al., 2009). Hence, this gene is now referred to as PCFT due to its high transport affinity towards oxidized folates and reduced folates.
PCFT functions as a unidirectional symporter that co-transports folates along with protons into the cell (Qiu et al., 2006) (Fig. 3). The high concentration of protons in the acidic microclimate of the upper small intestine, which is partly facilitated by the activity of Na+/H+ exchangers, is the driving force for the uptake of folates (Zhao et al., 2009a). As predicted from its proton-coupled transport capacity, PCFT operates optimally at an acidic pH of 5.5, with folate transport activity increasing as the pH decreases, as shown in mammalian cells and Xenopus oocytes (Qiu et al., 2006). Unlike RFC, PCFT exhibits high affinity for both folic acid and 5-CH3-THF at pH 5.5 (Km ∼ 0.5–1 μM). Although the highest transport activity of PCFT was documented at pH 5.5, some transport activity was also observed at physiological pH (i.e. pH 7.4; Lasry et al., 2008, Zhao et al., 2009a). PCFT was recently reported to form homo-oligomers that appear to possess functional transport activity (Hou et al., 2011).
A recent study addressed the extent to which PCFT contributes to the transport and pharmacological activity of the folic acid antagonist (i.e. antifolate) pemetrexed as well as other antifolates, relative to the contribution of the ubiquitously expressed RFC at physiological pH (Zhao et al., 2008). Either PCFT or RFC cDNAs were stably transfected into a folate transporter-null HeLa cell variant to achieve transport activities similar to their endogenous function in parental HeLa cells. PCFT and RFC produced comparable increases in pemetrexed cytotoxicity in growth medium containing 40 nM 5-formyl-THF. However, PCFT had little or no effect on the activities of MTX, raltitrexed and PT523 in comparison with RFC, irrespective of the folate growth source that was either folic acid or 5-formyl-THF. Hence, when cells were grown with 5-formyl-THF as the sole folate source, RFC alone could fully restore activity to the above antifolates. In contrast, pemetrexed was the only antifolate for which PCFT fully restored pharmacological activity. However, it is possible that at the acidic microenvironment of solid tumors, PCFT may have a more significant cytotoxic effect for antifolates other than pemetrexed. It was further found that PCFT, expressed at high levels in Xenopus laevis oocytes and in folate transporter-competent HepG2 cells, exhibited a high affinity for pemetrexed, with remarkable influx Km values of 0.2–0.8 μM at pH 5.5. PCFT markedly increased the growth inhibitory activity of pemetrexed, but not that of the other antifolates in HepG2 cells grown with 5-formyl-THF at physiological pH. These findings illustrate the unique role that PCFT plays in the transport and pharmacological activity of pemetrexed.
The cloning of the human PCFT gene paved the way for the discovery that inactivating mutations within the PCFT coding region are the genetic cause for the rare autosomal recessive disorder, hereditary folate malabsorption (HFM; OMIM 229050) (Atabay et al., 2010, Lasry et al., 2008, Meyer et al., 2010, Qiu et al., 2006, Shin et al., 2011, Shin et al., 2012, Zhao et al., 2007). HFM is a folate deficiency syndrome caused by impaired intestinal folate absorption. Patients with HFM present with low folate levels in the blood and cerebrospinal fluid (CSF). HFM manifests within the first few months of life with anemia, diarrhea, hypogammaglobulinemia, severe infections and failure to thrive. Moreover, due to the poor folate levels in the CSF and hence the impaired folate uptake into the central nervous system (CNS), neurological abnormalities and deficits occur including seizures and mental retardation. Unless the condition is diagnosed early and treated with high doses of oral folates, HFM is fatal, or neurological deficits become permanent.
Consistent with its main role in intestinal folate absorption, the highest PCFT expression can be observed in the upper small intestine, mainly in the duodenum and the upper jejunum. PCFT is also expressed in a wide variety of tissues including the kidney, liver, placenta and spleen, and to a lesser extent, colon and testis (Qiu et al., 2006). PCFT gene expression in human tumor cell lines was also examined; it was found that the colorectal cancer cell line, Caco-2, shows the highest level of PCFT gene expression. High expression was also documented in HeLa-R5 (a derivative of HeLa cells with a genomic deletion of the RFC locus), SW620 and A549 tumor cells. HeLa and MCF-7 cells exhibited moderate levels of PCFT expression. In contrast, PCFT mRNA levels in the two T-cell leukemia lines CCRF-CEM and Jurkat were essentially undetectable (Gonen et al., 2008). This has led to the identification of a dense, 1085 bp CpG island, present in the PCFT promoter and the coding region (nucleotides −600 to +485) which upon methylation plays a key role in the silencing of the PCFT gene (Gonen et al., 2008). It was specifically shown that this CpG island is densely methylated in the above human leukemia cell lines which fail to express PCFT; however, this CpG island was shown to lack methylation in PCFT-expressing cell lines (Gonen et al., 2008). Interestingly, these results were in complete concordance with a previous study which found that a spectrum of human carcinoma cell lines displays a prominent folate transport activity at acidic pH (5.5) that exceeded the influx obtained at pH 7.4 (Zhao et al., 2004b). In contrast, folate transport activity at pH 7.4 in CCRF-CEM leukemia cells exceeded the influx obtained at pH 5.5. These results suggested that whereas PCFT is expressed at substantial levels in various carcinoma cell lines, it is poorly expressed in leukemia cells. Furthermore, a recent study which examined PCFT gene expression in a wide array of 53 human tumor cell lines found similar results (Kugel Desmoulin et al., 2011). PCFT was found to be expressed in the majority of human solid tumor cell lines of different origins, however, low PCFT transcript levels exist in human leukemias, including ALL and acute myeloid leukemia (Kugel Desmoulin et al., 2011). The highest levels of PCFT transcripts were observed in Caco-2 (colorectal adenocarcinoma), SKOV3 (ovarian carcinoma), HepG2 (hepatoma), and H69 (small cell lung cancer) cells (Kugel Desmoulin et al., 2011).
The promoter of the PCFT gene was characterized using a sequential deletion analysis which identified a 271 bp fragment upstream to the first ATG that drives the same promoter activity obtained with a large 3.1 kb fragment. Further refinement showed that the minimal PCFT promoter localizes to only 157 bp (Stark et al., 2009a); these results were corroborated in an independent study (Diop-Bove et al., 2009). The basal promoter was shown to be rich in functional GC-box sites which play an important role in regulation of PCFT gene expression (Stark et al., 2009a). PCFT was also found to be regulated by a more selective transcription factor in certain tissues. In this respect, a vitamin D(3) and vitamin D receptor (VDR) response element was shown to increase intestinal PCFT gene expression, resulting in enhanced cellular folate uptake (Eloranta et al., 2009). Moreover, treatment with vitamin D(3) resulted in increased PCFT mRNA levels, both in vitro and ex vivo. The VDR response element in the PCFT promoter was identified and localized to nucleotides −1694 to −1680 (Eloranta et al., 2009).
Another interesting level of regulation of PCFT gene expression emerged with the discovery that PCFT is regulated by Nuclear Respiratory Factor-1 (NRF-1), the dominant transcription factor regulating mitochondrial biogenesis and respiration (Gonen and Assaraf, 2010). In this study, three functional NRF-1 binding sites residing within the basal promoter of the PCFT gene were characterized. These consensus NRF-1 binding sites proved to be functional as NRF-1 was found to bind to these sites. Targeted mutational inactivation of these NRF-1 binding sites resulted in 60% decrease in promoter activity. Consistently, overexpression of NRF-1 or a constitutively active NRF-1 VP-16 construct resulted in increased reporter activity and increased PCFT mRNA levels. Conversely, introduction of a dominant-negative NRF-1 construct markedly repressed reporter activity and PCFT mRNA levels; likewise, introduction of NRF-1 siRNA duplexes to cells resulted in decreased PCFT transcript levels. These findings provide a molecular and functional linkage between mitochondria biogenesis and folate metabolism, thereby revealing that PCFT is an NRF1-responsive gene that behaves like a mitochondrial gene.
(c) Folate receptors (FRs): FRα, FRβ and FRγ: The third route of folate uptake proceeds via the family of folate receptors (FRs). FRs are high-affinity folate-binding membrane glycoproteins, encoded by three different genomic loci including FRα, FRβ and FRγ (Fig. 3). There is ∼70–80% amino acid homology between the FRs and they contain 245–257 amino acids as well as several N-glycosylation sites (Reviewed in Elnakat and Ratnam, 2004, Elnakat and Ratnam, 2006). FRα and FRβ are glycosylphosphatidylinositol (GPI)-anchored cell surface glycoproteins, involved in folate transport, whereas FRγ is a secreted protein (Shen et al., 1995).
Among folate transport routes, FRα and FRβ display the highest documented binding affinity for folic acid (Kd = 0.1–1 nM) (Wang et al., 1992). Folate uptake mediated by FRs proceeds via receptor-mediated endocytosis (Kamen et al., 1988, Lu and Low, 2002, Rothberg et al., 1990, Salazar and Ratnam, 2007). In this respect, it was recently suggested that PCFT plays a role in FRα-mediated endocytosis by exporting folates from acidic folate-containing endosomes into the cytoplasm (Zhao et al., 2009b). From a physiological point of view, it is less clear what the role of FRs is when co-expressed with RFC and/or PCFT as the folate transport rate via RFC is 100-fold faster than that via FRs. It is possible that under folate-deficiency conditions, or in tissues with low RFC and PCFT expression, the activity of FRs becomes more significant.
FRα (FOLR1) expression is mostly observed in epithelial cells of the uterus, placenta, choroid plexus, retina and kidney (Parker et al., 2005, Ross et al., 1994, Salazar and Ratnam, 2007). FRα is also expressed in a variety of cancers, mainly of epithelial origin, including adenocarcinomas of the ovary, cervix, uterus, endometrium, kidney, lung, breast, bladder and pancreas (Elnakat and Ratnam, 2004, Elnakat and Ratnam, 2006, Parker et al., 2005, Ross et al., 1994). The FRα gene is regulated by two promoter regions, one that resides upstream to exon 1 and is named P1, whereas the second, located upstream of exon 4 is termed P4 (Elwood et al., 1997, Kelemen, 2006, Saikawa et al., 1995). in vitro studies have shown that FRα is positively regulated by extracellular folate depletion (Kane et al., 1988) and elevated levels of homocysteine (Antony et al., 2004). In addition, steroid hormone regulation of FRα gene expression was also demonstrated; in this respect, FRα was shown to be down-regulated by the estrogen receptor (Kelley et al., 2003, Rochman et al., 1985), while being up-regulated by both retinoic acid (Bolton et al., 1999) and the glucocorticoid receptor (Kelemen, 2006, Tran et al., 2005). Similarly, FRβ (FOLR2) also displays a restricted pattern of tissue-specific gene expression in placenta, thymus, spleen and various malignant cells of myelomonocytic lineage (Elnakat and Ratnam, 2004, Ratnam et al., 1989, Ross et al., 1994, Ross et al., 1999, Shen et al., 1994). Unlike FRα, the FRβ gene is regulated and expressed from a single promoter that encodes a single transcript, by the regulation of Sp1 and GA-binding protein (GABP) (Sadasivan et al., 1994). It was demonstrated that FRβ can be highly up-regulated (20-fold) by all-trans retinoic acid in a dose-dependent and reversible manner in the absence of terminal differentiation or cell growth inhibition (Wang et al., 2000). Furthermore, the molecular mechanism of transcriptional induction of FRβ by all-trans retinoic acid was also established (Hao et al., 2003).
The restricted pattern of tissue-specific expression of both FRα and FRβ renders them attractive candidates for diagnostic purposes and selective delivery vehicles of therapeutic agents in malignant and non-malignant disorders, hence minimizing toxic side effects in non-target healthy tissues, as will be discussed below (Hilgenbrink and Low, 2005, Paulos et al., 2004, Xia and Low, 2010; for further FR-based applications see Section 4).
Forty nine ATP-binding cassette (ABC) transporters exist in humans, all of which are transmembrane proteins that couple ATP hydrolysis to the transport of endogenous and exogenous substrates across cellular membranes. ABC transporters are classified into seven families, one of which is the important family of multidrug resistance (MDR) proteins (MRP/ABCC), that is comprised of 13 members (reviewed in Borst and Elferink, 2002, Deeley et al., 2006, Gottesman et al., 2002, Szakacs et al., 2006). Members of the MRP/ABCC family confer MDR to various hydrophobic and hydrophilic cytotoxic compounds by acting as low affinity, high capacity ATP-dependent drug efflux pumps (Reviewed in Assaraf, 2006).
Folates are among the various transport substrates of MRP1-5 (Fig. 3); it has been demonstrated that these efflux pumps have the ability to export THF, 5-CH3-THF, 5,10-CH2-THF and 10-CHO-THF out of the cell (Assaraf, 2006, Chen et al., 2002, Hooijberg et al., 2003, Kusuhara et al., 1998, Zeng et al., 2001, Zeng et al., 2000). MRP1-5 are also capable of exporting antifolates (Fig. 4 and Table 1). Overexpression of MRP1-5 may result in antifolate resistance as will be elaborated in section 3 (Assaraf, 2006). Another ATP-driven efflux transporter that is capable of exporting folates and antifolates is the breast cancer resistance protein (BCRP/ABCG2), which also belongs to the ABC superfamily of efflux pumps (Fig. 3). Like the MRPs, BCRP functions as a high capacity, low-affinity exporter of multiple large hydrophobic or amphiphilic substrates, harboring some positive or negative charge. Among its substrates are anticancer drugs including anthracyclines as mitoxantrone, irinotecan (SN-38), topotecan, antifolates as well as vitamins such as folates and riboflavin (reviewed in Assaraf, 2006, Polgar et al., 2008).
Following cellular uptake of folates via the influx routes described above, THF cofactors and antifolates undergo a unique metabolism known as folylpolyglutamylation catalyzed by folylpoly-γ-glutamate synthetase (FPGS) (Fig. 3). This enzyme catalyzes the MgATP-dependent, sequential addition, of multiple equivalents of glutamic acid to the γ-carboxyl chain of THF cofactors and glutamate-containing antifolates (Baugh et al., 1973, McBurney and Whitmore, 1974, McGuire et al., 1980, Moran, 1999, Moran et al., 1976).
FPGS has three known functions in mammalian cells; (a) enhanced cellular retention; polyglutamate congeners of THF and antifolates are polyanions and therefore are incapable of traversing the lipid bilayer by passive diffusion. Additionally, long chain (>3) polyglutamate derivatives are no longer substrates of efflux transport systems including the ATP-driven efflux transporters of the MRP/ABCC family (MRP1-5) (Wielinga et al., 2005, Zeng et al., 2001) and ABCG2 (ABCG2/BCRP) (Volk and Schneider, 2003); consistently, polyglutamate conjugates of THF and antifolates are not transport substrates of RFC (Matherly and Goldman, 2003). Consequently, polyglutamate conjugates of THF cofactors and antifolates are efficiently retained within cells. (b) FPGS exists in two compartmentalized isoforms, i.e. cytosolic (cFPGS) and mitochondrial (mFPGS) (Fig. 3). Polyglutamylation of folates by the mFPGS results in mitochondrial accumulation of folate polyglutamates, a process crucial for glycine biosynthesis which occurs exclusively in mitochondria (Lin et al., 1993). (c) The third function of FPGS concerns the bioactivity of polyglutamates of THF cofactors and antifolates; it is well established that most polyglutamate THF cofactors are much better substrates than their parent monoglutamate forms for various folate-dependent enzymes. This can result from either increased affinity and/or enhanced Vmax (Schirch and Strong, 1989). This is also true for antifolates where it has been shown that polyglutamate conjugates of antifolates display a markedly enhanced inhibitory capacity of their target enzymes (Allegra et al., 1985, Allegra et al., 1987, Baggott et al., 1986, Jackman et al., 1991, Shih et al., 1997).
From a physiological point of view, the process of polyglutamylation is reversed by the counteracting activity of the lysosomal enzyme γ-glutamyl hydrolase (GGH). GGH is a secreted glycoprotein enzyme located in the lysosome. GGH catalyzes the hydrolysis of the γ-glutamyl tail of folate and antifolate polyglutamates (Fig. 3). GGH exhibits a Km value in the micromolar range and is a high-turnover enzyme that possesses a catalytic cysteine residue at the active site (Reviewed in Galivan et al., 2000, Schneider and Ryan, 2006).
Section snippets
Antifolates and their mode of action
Due to the major role that folates play in the de novo biosynthesis of purines and thymidylate, it was realized in the 1940s that antifolates are effective in the treatment of childhood acute leukemia (Farber et al., 1948). Decades afterwards, this discovery led to the development of a large group of rationally designed, new generation antifolates (Walling, 2006) (Fig. 4 and Table 1). These antifolates were designed to target and inhibit key folate-dependent enzymes, thereby leading to
Molecular mechanisms of antifolate resistance in cancer
Mechanisms of resistance to antifolates frequently emerge that hinder their clinical efficacy. Intensive molecular research provided deep mechanistic insights into the various mechanisms underlying resistance to various antifolates (Fig. 5). Notably, many of the mechanisms that were characterized in vitro were further identified in vivo, in patients who either presented with (i.e. inherent drug resistance) or developed resistance (acquired drug resistance) to antifolate-containing chemotherapy.
Future perspectives: emerging antifolates and small molecule folate-drug conjugates for personalized medicine
For some antifolates, tumor selectivity may not be optimal since many antifolates are taken up by RFC which is ubiquitously expressed in normal cells from healthy tissues, while poorly expressed or mutated in drug resistant tumor cells (Assaraf, 2007, Assaraf, 2006). This may consequently result in untoward toxicity to various proliferating healthy tissues. For example, clinical studies with lometrexol and AG2034 were discontinued due to extensive accumulation and long-term retention of
References (417)
Regulation of human dihydrofolate reductase activity and expression
Vitamins & Hormones
(2008)Enhanced inhibition of thymidylate synthase by methotrexate polyglutamates
Journal of Biological Chemistry
(1985)Evidence for direct inhibition of de novo purine synthesis in human MCF-7 breast cells as a principal mode of metabolic inhibition by methotrexate
Journal of Biological Chemistry
(1987)Selective multiplication of dihydrofolate reductase genes in methotrexate-resistant variants of cultured murine cells
Journal of Biological Chemistry
(1978)Folate receptors: reflections on a personal odyssey and a perspective on unfolding truth
Advance Drug Delivery Reviews
(2004)Kinetics of the formation and isomerization of methotrexate complexes of recombinant human dihydrofolate reductase
Journal of Biological Chemistry
(1988)The role of multidrug resistance efflux transporters in antifolate resistance and folate homeostasis
Drug Resistance Updates: Reviews and Commentaries in Antimicrobial and Anticancer Chemotherapy
(2006)Characterization of the coexisting multiple mechanisms of methotrexate resistance in mouse 3T6 R50 fibroblasts
Journal of Biological Chemistry
(1992)- et al.
Loss of folic acid exporter function with markedly augmented folate accumulation in lipophilic antifolate-resistant mammalian cells
Journal of Biological Chemistry
(1997) Computer modelling of antifolate inhibition of folate metabolism using hybrid functional petri nets
Journal of Theoretical Biology
(2006)