Plant-derived glucose transport inhibitors with potential antitumor activity

Abstract

Glucose, a key nutrient utilized by human cells to provide cellular energy and a carbon source for biomass synthesis, is internalized in cells via glucose transporters that regulate glucose homeostasis throughout the human body. Glucose transporters have been used as important targets for the discovery of new drugs to treat cancer, diabetes, and heart disease, owing to their abnormal expression during these disease conditions. Thus far, several glucose transport inhibitors have been used in clinical trials, and increasing numbers of natural products have been characterized as potential anticancer agents targeting glucose transport. The present review focuses on natural product glucose transport inhibitors of plant origin, including alkaloids, flavonoids and other phenolic compounds, and isoprenoids, with their potential antitumor properties also discussed.

1. INTRODUCTION AND SIGNIFICANCE

Glucose is consumed in our daily diet and involved in several biological processes as a precursor for the biosynthesis of the carbon skeleton (Warburg, 1956). It is metabolized initially to pyruvate that can either be converted to lactate or enter the tricarboxylic acid (TCA) cycle [TCA cycle, also named as citric acid cycle (CAC) or Krebs cycle] in mitochondria to produce adenosine triphosphate (ATP) (Fothergill-Gilmore & Michels, 1993; Saraste, 1999). Glucose is the primary source for producing chemical energy in the form of ATP in the majority of human cells, and glucose uptake plays a critical role in cellular survival and growth (Zierler, 1999). However, glucose cannot diffuse passively via the plasma membrane, and it has to be internalized via plasma membrane spanning proteins, namely, transporters of glucose. These include glucose transporters (GLUTs), sodium glucose cotransporters (SGLTs), and the recently discovered sugar will eventually be exported transporters (SWEETs) (Deng & Yan, 2016).

GLUTs, the uniporters encoded by solute carrier 2 (SLC2) to facilitate sugar transport along a concentration gradient, belong to a family of 14 transmembrane proteins. Of these, 11 are glucose transporters, with GLUTs1–4 having been investigated the most comprehensively (Mueckler & Thorens, 2013). GLUT1 is a basal glucose transporter expressed throughout the human body and the primary transporter in human erythrocytes. It is expressed in the majority of human cells to intake glucose, regardless of the intracellular glucose concentration (Burant & Bell, 1992). GLUT2 is expressed mainly in the liver, pancreas, small intestine, and kidney. It has a higher capacity and lower affinity than GLUT1 to transport glucose. The liver takes glucose up rapidly only when the blood glucose level is high, while the pancreas secretes insulin with a lower affinity (Uldry, Ibberson, Hosokawa, & Thorens, 2002). GLUT3 is expressed mainly in the neurons to ensure a constant rate of glucose uptake irrespective of blood glucose levels (Burant & Bell, 1992; Simpson et al., 2008). As the sole member sensitive to insulin in the GLUT family, GLUT4 is expressed primarily in fat tissue and skeletal muscles and stored in intracellular GLUT4 storage vesicles (Huang & Czech, 2007). When insulin binds to and activates insulin receptors in the target cells, GLUT4 is translocated to the plasma membrane to increase the glucose uptake, and then glucose is stored as glycogen in the skeletal muscle (Govers, 2014).

SGLTs are another group of glucose transporters that contain 12 proteins, of which SGLTs1, 2, 4, and 5 are sugar transporters, but SGLT3 is a glucose sensor (Wright, Loo, & Hirayama, 2011). SGLTs are expressed by intestinal brush border cells (enterocytes) and by the cells of proximal tubules in the kidney, and the ratio of sodium coupling to glucose for SGLT1 was found to be different from that for SGLT2 (Hummel et al., 2010). It was evidenced that glucose binds to a glucose sensor that resides on the external face of the enterocyte luminal membrane and generates an intracellular signal to enhance the expression of SGLT1 (Dyer, Vayro, & Shirazi-Beechev, 2003). Inhibition of SGLT2 provides potentially a new option for the treatment of diabetes (Jabbour & Goldstein, 2008).

The ubiquitous SWEETs are new members in the family of glucose transporters. Compared to humans who have one SWEET gene, plants have around 20 different SWEET genes. SWEETs in plants are sugar translocators and play a critical role in suspecting pathogens (Chen et al., 2010; Feng & Frommer, 2015). However, based on our best knowledge, the function of SWEETs in humans is not yet reported.

It is well known that glucose transport plays a critical role in maintaining glucose homeostasis in the heart (Sohn et al., 2013; Ware et al., 2011), and several natural SGLT2 inhibitors have been characterized to benefit to the treatment of cardiovascular and diabetes (Andrianesis, Glykofridi, & Doupis, 2016; Hsia, Grove, & Cefalu, 2017). Moreover, glucose is used to produce energy and to synthesize biomass (Warburg, Wind, & Negelein, 1927), and glucose-uptake is found to be increased during malignancies. Also, glucose-deprivation is toxic to tumors and sensitizes cancer cells to chemotherapy (Schroll, LaBonia, Ludwig, & Hummon, 2017), and cancer cells take up a majority of their glucose in via GLUTs. Thus, these proteins have become attractive targets for the potential treatment or diagnosis of cancer (Granchi, Fancelli, & Minutolo, 2014). For example, a GLUT-specific radioactive analog of glucose, 2-[¹⁸F] fluoro-2-deoxy-D-glucose, namely, F-18 FDG, has been used in PET scans, for the diagnosis of cancers (Kim et al., 1992; Ben-Haim & Ell, 2009). GLUT1 has been found to be upregulated in the majority of cancer types, and GLUT2 expression is increased drastically in liver, pancreatic, gastric and colon cancers. In addition, GLUT3 is upregulated in lung, neck, head, ovarian, breast, and bladder cancers, while GLUT4 is overexpressed in colon, lymphoid, breast, and pancreatic tumors (Macheda, Rogers, & Best, 2005; Qian, Wang, & Chen, 2014).

Other GLUTs have been recently found to be upregulated in several cancer types, and many GLUT-inhibitors have shown cancer-related potency, even though no GLUT-inhibitor has been approved as an anticancer agent by the U.S. FDA thus far (Barron, Bilan, Tsakiridis, & Tsiani, 2016; Granchi, Fortunato, & Minutolo, 2014; Granchi, Fortunato, & Minutolo, 2016). In our search for glucose transport inhibitors, the synthetic compound WZB115 was found to exhibit glucose transport inhibitory activity and selective cytotoxicity against MCF7 human breast cancer cells (Liu, Zhang, Cao, Liu, Bergmeier, & Chen, 2010; Zhang, Liu, Chen, & Bergmeier, 2010), and WZB117 has been characterized (Liu et al., 2012). Interestingly, WZB117 reduced the self-renewing capability of pancreatic Panc1, ovarian A2780, and GS-Y03 cancer stem cells (CSC) (Shibuya et al., 2015). It targets GLUT1, GLUT3, and GLUT4 and binds to GLUT4 and the endofacial site of GLUT1 to compete with cytochalasin B (Ojelabi, Lloyd, Simon, De Zutter, & Carruthers, 2016). Recently, SGLT1 and SGLT2 have been found to be overexpressed in mouse models of pancreatic and prostate adenocarcinomas, and canaglifozin, a FDA-approved synthetic SGLT2 inhibitor, was found to potentiate the antitumor efficacy of gemcitabine and to increase tumor necrosis (Scafoglio et al., 2015).

In previous review articles, the overall information included on glucose transport and metabolism in cancer cells and the progress of new therapeutic developments for glucose transport inhibitors focused on small molecules, with only a limit number of natural products, were summarized (Qian, Wang, & Chen, 2014). Natural product sodium glucose cotransporter (SGLT) inhibitors and their potential antidiabetic activity have been reviewed (Blaschek, 2017; Choi, 2016). In addition, synthetic and naturally derived glucose transport inhibitors and their potential anticancer activities have been discussed (Granchi, Fortunato, & Minutolo, 2014; Granchi, Fortunato, & Minutolo, 2016). As an extension of these previous reviews, in the present contribution, plant-derived constituents showing antitumor potential mediated in part through glucose transport inhibition are included, with their plant origin, structures, activities, and mechanisms of action discussed.

2. GLUCOSE TRANSPORT INHIBITORS FROM EDIBLE PLANTS

Increasing numbers of plant-derived natural products showing clinical implications in cancer, cardiovascular diseases, and diabetes have been characterized as GLUT and SGLT inhibitors. Of these, phlorizin, the first naturally occurring sodium glucose co-transport (SGLT) inhibitor, has been used to identify the SGLT mechanism (Vick, Diedrich, & Bauman, 1973). Following this, many natural product glucose transport inhibitors have been characterized, of which several compounds have been tested in vitro and in vivo, and in some cases have reached clinical trials as potential cancer chemotherapeutic agents (Blaschek, 2017; Choi, 2016; Granchi et al., 2016, Qian, Wang, & Chen, 2014).

Edible plants are well known to contribute to the improvement of overall human health, and many edible plant extracts have been investigated recently for their glucose transport inhibitory activity. After an in vitro intestinal glucose transport system was established, in which glucose uptake was measured between apical and basolateral sides of Caco-2 human colon cancer cells, guava [Psidium guajava L. (Myrtaceae)] fruit and leaf extracts were identified as sodium-dependent and -independent glucose transport inhibitors (Müller et al., 2018). An apple [Malus domestica Borkh. (Rosaceae)] extract was found to inhibit methyl-α-D-glucopyranoside (αMDG) transport via hSGLT1 in a dose-dependent manner. Glucose transport was found to be inhibited when everted sacs, segments of the small intestine of male C57BL/6N mice, were treated with radioactive αMDG followed by an apple (M. domestica) extract (Schulze et al., 2014). Similarly, glucose uptake was reduced when the everted gut sacs obtained from male albino rats were treated with an aqueous extract (3.6 mg/ml) of bitter melon [Momordica charantia L. (Cucurbitaceae) (Mahomoodally, Fakim, & Subratty, 2004). Plant phenol-containing seed extracts from the legumes, Vicia faba L. var. equina and Vicia faba L. var. minor (Fabaceae), were found to reduce intestinal glucose transport in male white Wistar rats (Sobrini, Martinez, Ilundain, & Larralde, 1983), and a similar seed extract from the common bean [Phaseolus vulgaris L. (Fabaceae)] reduced glucose transport in the rat ileum (Motilva, Martinez, Ilundain, & Larralde, 1983). Delphinol^®, a standardized extract of maqui berries [Aristotelia chilensis (Molina) Stuntz (Elaeocarpaceae)], was found to decrease glucose uptake in sections of the mouse jejunum by inhibition of a sodium glucose transporter, which suppressed glucose increase in the post-prandial blood of individuals who suffered from impaired glucose regulation (Hidalgo et al., 2014). Also, extracts of both Matricaria recutita L. (Asteraceae) (chamomile) and Camellia sinensis (L.) Kuntze (Theaceae) (green tea) reduced glucose uptake in Caco-2-TC7 differentiated cells, with the M. recutita extract found to target GLUT2 in Na⁺-free conditions and GLUT5-mediated fructose transport (Villa-Rodriguez et al., 2017). The aqueous extract of black tea (C. sinensis) was found to decrease mucosal glucose uptake in male Sprague-Dawley rats by an average of 44%, over a 45-minute period (Kreydiyyeh, Baydoun, & Churukian, 1994).

3. ALKALOID GLUCOSE TRANSPORT INHIBITORS

Alkaloids are found widely in bacteria, fungi, plants, and animals, and exhibit many different types of biological activities (de Sousa Falcão, 2008). Several alkaloids have been found to exhibit glucose transport inhibitory activity. For example, vinblastine (1) (FIGURE 1), a bisindole alkaloid obtained from the Madagascar periwinkle plant, Catharanthus roseus (L.) G.Don (Apocynaceae), was approved by the U.S. FDA in the 1960s as an anticancer drug for the treatment of breast cancer and Hodgkin’s and non-Hodgkin’s lymphomas. It targets β-tubulin to prevent tubulin congregation and suppresses microtubule dynamics at the mitotic spindle leading to M-phase arrest during cell cycle progression (Moudi, Go, Yien, & Nazre, 2013). After a phase I study conducted with 38 patients (17–68 years old) who suffered from an early stage of Hodgkin’s lymphoma (HL) and were treated with injection of ABVD [adriamycin (doxorubicin, 25 mg/m²), bleomycin (10 mg/m²), vinblastine (6 mg/m²), and dacarbazine (375 mg/m²)] on days 1 and 15 (of a 28-day cycle) for 173 cycles, 35 of the patients were in a state of complete remission from HL (Boleti and Maed, 2007). Also, a combination of mitomycin C, vinblastine, and cisplatin (MVP) has been used effectively to treat stage III NSCLC (Ellis et al., 1995). Interestingly, vinblastine was found to inhibit glucose transport through reducing 2-DG uptake in glioma C6 cells (Singh, Gao, Singh, Kunapuli, & Ravindra, 1998).

FIGURE 1.

Structures of alkaloids showing glucose transport inhibitory and potential antitumor activities

Several other indole alkaloids derived from the leaves of Alstonia macrophylla Wall. (Apocynaceae) were found to inhibit SGLT1 and SGLT2 in COS-1 African green monkey kidney fibroblast-like cells, of which 10-methoxy-N(1)-methylburnamine-17-O-veratrate (2) and alstiphyllanine D (3) showed the most potent activity, with IC₅₀ values of 4.0 μM (2) and 5.0 μM (3) against SGLT1 and 0.5 μM (2) and 2.0 μM (3) against SGLT2 (Arai et al., 2010). For these indole alkaloids, substitutions at N1, N4, and C-17 proved to be important for the mediation of SGLT inhibitory activity. Introducing a methyl group at N1 and an aromatic ester unit at the C-17 position enhanced the inhibitory potency, but changing 2 or 3 to a N(4)-oxide resulted in the activity being abolished.

4. FLAVONOID GLUCOSE TRANSPORT INHIBITORS

Flavonoids are a very large group of secondary metabolites found in fruits, vegetables and flowers (Yao et al., 2004). Substitutions in the flavan structure by hydroxy and methyl groups and sugar units create a wide range of flavonoid derivatives (Nijveldt et al., 2001), which may be classified into several major structural subtypes. The flavonoids are well- known for their anti-inflammation and effects on the reduction of nitric oxide synthase (thus reducing ischemia-reperfusion injury) (Nijveldt et al., 2001), and several of these compounds have been identified as GLUT and/or SGLT inhibitors.

4.1. Anthocyanins

Anthocyanins are phenolic compounds found in various berry fruits, of which pelargonidin-3-O-glucoside (4) (FIGURE 2), characterized from strawberries [Fragaria virginiana Duchesne (Rosaceae)], was found to exhibit glucose-uptake inhibitory activity in Caco-2 human colon cancer cells, with an IC₅₀ value of 705 μM (Manzano & Williamson, 2010).

FIGURE 2.

Structures of an anthocyanin, chalcones, and a modified chalcone lactone showing glucose transport inhibitory and potential antitumor activities

4.2. Chalcones

Many chalcones (1,3-diaryl-2-propen-1-ones) isolated from edible plants exhibit potential antimalarial, antiviral, and antiinflammatory activities (Nowakowska, 2007). Of these, phloretin (5) (Figure 2), a dihydrochalcone derived from the apple tree [Malus domestica Borkh. (Rosaceae)] (Gosch, Halbwirth, & Stich, 2010), was found to inhibit competitively 3-O-methyl-D-glucose uptake via GLUT1 in human erythrocytes (Martin, Kornmann, & Fuhrmann, 2003). Phlorizin (6), a glucoside of phloretin (5), was obtained as the first SGLT2 inhibitor from the apple tree (Malus domestica) (Andrianesis, Glykofridi, & Doupis, 2016). Both phloretin (5) and phlorizin (6) inhibited GLUT4-mediated glucose transport, with IC₅₀ values of 9.4 μM (5) and 140 μM (6), indicating that introducing a glucose unit results in the GLUT4 inhibitory activity being decreased (Kasahara & Kasahara, 1997). The potential antitumor activity of phloretin (5) has been reviewed recently (Choi, 2019). Also, it potentiated the antiproliferative effect of paclitaxel in HepG2 human liver cancer cells. Tumor growth was inhibited when six–seven-week-old NOD.CB17-PRKDC(SCID)/J (NOD-SCID) mice were inoculated with HepG2 cells and treated (i.p.) with 5 (10 mg/kg) plus paclitaxel (1 mg/kg) thrice a week for six weeks (Yang et al., 2009).

Xanthohumol (7), a prenylated chalcone isolated from hops [Humulus lupulus L. (Cannabaceae)], reduced [³H-2-DG] uptake in HTR-8/SVneo human first-trimester extravillous trophoblast cells, with an IC₅₀ value of 3.6 μM. This activity was proposed to be mediated through three major intracellular signaling pathways, namely, the mTOR, tyrosine kinases (TKs), and c-Jun N-terminal kinases (JNK) pathways (Correia-Branco et al., 2015). The potential antitumor activity of xanthohumol (7) has been reviewed recently (Jiang, Sun, Xiang, Wei, & Li, 2018). For example, pancreatic tumor growth was inhibited when nude mice were inoculated with Panc1 human pancreatic cancer cells and treated (i.p.) daily with 7 (25 mg/kg) for 27 days (Jiang et al., 2015).

(+)-Cryptocaryone (8), a modified chalcone lactone isolated from Cryptocarya rubra C.R. Skeels. (Lauraceae), was found to exhibit potent cytotoxicity against HT-29 human colon cancer cells, with an IC₅₀ value of 0.32 μM. At a concentration of 30 μM, this compound inhibited significantly glucose transport in H1299 human lung cancer cells, indicating that it may mediate its cytotoxicity at least in part through interaction with glucose transporters (Ren et al., 2014).

4.3. Flavan Derivatives

Flavanones are distributed widely in citrus fruits and exhibit inhibitory activity towards chemically induced colon cancer progression (Nijveldt et al., 2001; Yao et al., 2004). Naringenin (9) (FIGURE 3), a flavanone isolated from grapefruit [Citrus paradisa Macfad. (Rutaceae)], was found to decrease the 2-DG uptake in differentiated 3T3-L1 cells, with IC₅₀ values of 61 μM and 71 μM under basal and insulin stimulated conditions, respectively (Claussnitzer, Skurk, Hauner, Daniel, & Rist, 2011). The potential antitumor activity of naringenin (9) has been discussed in a recent review article (Salehi et al., 2019). In an in vivo investigation, tumor metastasis was inhibited when four-week-old female BALB/c mice were inoculated with 4T1 mouse breast cancer cells that were transduced with a TGF-β1 generating 4T1/TGF-β1 cells and treated (orally) daily with 9 (200 mg/kg suspended in 1% carboxyl methyl cellulose) for 30 days (Zhang et al., 2016).

FIGURE 3.

Structures of flavan derivatives showing glucose transport inhibitory and potential antitumor activities

Epigallocatechin gallate (EGCG, 10), a flavan ester found in green tea [Camellia sinensis (L.) Kuntze (Theaceae)], inhibited hGLUTs 1, 3, and 4 by binding to their exofacial site (Ojelabi, Lloyd, De Zutter, & Carruthers, 2018) and decreased glucose and fructose transport via GLUT2. It also inhibited fructose uptake via GLUT5 in Xenopus laevis oocytes that were overexpressed with these GLUTs (Gauer, Tumova, Lippiat, Kerimi, & Williamson, 2018). As reviewed previously, EGCG (10) shows potential antitumor activity (Gan, Li, Sui, & Corke, 2018). In an in vivo study, tumor growth was inhibited when six–eight-week-old female BALB/c nude mice were inoculated with SGC-7901 human gastric cancer cells and treated (i.p.) daily with 10 (1.5 mg/per mice) for four weeks (Zhu et al., 2007).

(−)–Kurarinone (11) and sophoraflavanone G (12) (FIGURE 3) are prenylated flavanones characterized as potent Na⁺-glucose cotransporter (SGLT) inhibitors from the roots of Sophora flavescens Aiton (Fabaceae). In COS-1 cells, both 11 and 12 inhibited SGLT1, with IC₅₀ values of 10.4 and 18.7 μM, respectively. They also showed SGLT2 inhibitory activity, with IC₅₀ values of 1.7 μM (11) and 4.1 μM (12) (Sato, Takeo, Aoyama, & Kawahara, 2007). Investigation of the activity of these flavans and their analogues (Sato, Takeo, Aoyama, & Kawahara, 2007) indicates that the SGLT inhibitory effect decreases slightly when the C-5 methoxy group is replaced by a hydroxy group or when the dihydropyrone C ring opens to convert the molecule to an analogous chalcone. In addition, the activity is weakened when a hydroxy group is introduced at the C-3 position, and the activity declines greatly when the C-4a/5a double bond is saturated followed by introduction of a hydroxy group at the C-5a position.

Interestingly, both (−)–kurarinone (11) and sophoraflavanone G (12) were found to exhibit cytotoxicity toward a panel of human cancer cell lines, with IC₅₀ values being in the range 2–27 μg/mL, of which the potency was weakened slightly by methylation of the hydoxy group at the C-5 or C-2’ position (Sun et al., 2007). In an in vivo study, lung tumor growth was inhibited when four–six-week-old athymic nu/nu (BALB/c) mice were inoculated with A549 human lung cancer cells and treated (i.p.) daily with 11 (20 or 40 mg/kg) for 27 days (Yang et al., 2018). Mechanistically, (−)–kurarinone (11) was found to mediate its cytotoxicity toward H1688 human SCLC cells through the mitochondrial- and receptor-mediated apoptotic pathways (Chung, Lin, Lin, Chan, & Yang, 2019).

Hesperidin (13), a flavanone glycoside identified from the orange [Citrus sinensis (L.) Osbeck (Rutaceae)], was found to reduce the uptake of ¹⁴C-glucose in Caco-2-TC7 cells at concentrations 80 and 800 μM, and transport of both ¹⁴C-glucose (via GLUT2) and ¹⁴C-fructose (via GLUT5) was inhibited by 13 in Xenopus laevis oocytes that were microinjected with GLUT2 and GLUT5 mRNAs (Kerimi et al., 2019). Also, hesperidin (13) induced significantly a reduced mouse hepatic GLUT2 expression and an elevated the mouse adipocyte GLUT4 level when male five-week-old C57BL/KsJ-db/db mice were fed with food supplemented with 13 (0.2 g/kg) for five weeks (Jung, Lee, Park, Kang, & Choi, 2006).

As reviewed previously, hesperidin (13) showed potential antitumor efficacy (Devi et at., 2015). In a phase clinical study conducted on healthy human subjects (18–75 years old) in three independent investigations using different amounts of orange juice supplemented with hesperidin, hesperidin-containing orange juice was found to modulate postprandial blood glucose levels by partially inhibiting intestinal GLUTs (Kerimi et al., 2019). Also, this compound was shown cytotoxicity toward various types of human cancer cells and other evidences of potential antitumor activity, as summarized in a previous review (Devi et al., 2015).

4.4. Flavones

Apigenin (14) (FIGURE 4), one of the flavone components of apple skin [Malus domestica Borkh. (Rosaceae)] (Nijveldt et al., 2001), inhibited 2-DG uptake in CD18 and S2–013 human pancreatic cells at a concentration of 25 μM, with both GLUT1 gene and protein expression reduced in these cells (Melstrom et al., 2008). The cytotoxicity against Hep-2 human laryngeal carcinoma cells of cisplatin was enhanced by 14, owing to the reduced expression of GLUT1 and p-Aκt [p-protein kinase B (PKB)] proteins in Hep-2 cells (Xu et al., 2014). Also, [¹⁴C]-glucose uptake via GLUT2 and GLUT7 was inhibited when Xenopus laevis oocytes were transfected with mRNAs of these GLUTs and treated with 14 (Gauer, Tumova, Lippiat, Kerimi, & Williamson, 2018).

FIGURE 4.

Structures of flavones showing glucose transport inhibitory and potential antitumor activities

Luteolin (15), a flavone component of celery [Apium graveolens L. (Apiaceae)] (Nijveldt et al., 2001), reduced 2-DG uptake at a concentration of 10 μM in differentiated mouse MC3T3-G2/PA6 adipose cells via GLUT4 and decreased the insulin stimulated phosphorylation of insulin receptor B and Aκt activation, to lower the levels of plasma membrane-translocated GLUT4. Structure-activity relationship studies indicated that the C-2 and C-3 double bond and 4’- or 3’,4’-hydroxy groups are critical for 15 to mediate its GLUT4 inhibitory activity (Nomura et al., 2008).

As concluded in a recent review article, luteolin (15) is a promising antitumor agent (Imran et al., 2019). It inhibited MDA-MB-231 human breast cancer cell growth and reduced cell migratory and invasive capabilities at a concentration of 100 μM. Both mRNA and protein expression of Notch-1, MMP2, MMP9, VEGF in MDA-MB-231 cells were reduced after 48 hour-treatment with 15 (Sun et al., 2015). Tumor growth was inhibited when six–eight-week-old BALB/c mice were inoculated with MDA-MB-231 cells and treated daily with 15 (20 mg/kg) via tail vein injection for 14 consecutive days (Sun et al., 2015), while tumor growth was inhibited when six–eight-week-old BALB/c mice were inoculated with LoVo human colon cancer cells and treated (i.p.) daily with 15 (20 mg/kg, alternate days) for one month (Chen, Zhang, Gao, & Shi, 2018).

Quercetin (16) is a flavone found in lettuce [Lactuca sativa L. (Asteraceae)], olives [Olea europaea L. (Oleaceae)], onions [Allium cepa L. (Amaryllidaceae)], and parsley [Petroselinum crispum (Mill.) Fuss (Apiaceae)] (Nijveldt et al., 2001). It inhibited GLUT1-, 3-, and 4-mediated 2-DG uptake in HEK 293 human embryonic kidney cells, with respective IC₅₀ values of 2.0 μM (GLUT1), 17.7 μM (GLUT3), and 1.7 μM (GLUT4) (Ojelabi et al., 2018). It also inhibited 2-DG transport in L929 mouse fibroblasts by binding to an exofacial site on GLUT1, with an IC₅₀ value of 8.5 μM (Hamilton et al., 2018).

The potential antitumor activity of quercetin (16) has been reviewed recently (Rauf et al., 2018). Earlier, it has been tested as a potential anticancer drug in a phase I clinical trial. Fifty-one patients (18–75 years old) who were suffering from different cancers were selected for the study. Quercetin dehydrate powder was dissolved in DMSO at 50 mg/ml for doses up to 945 mg/m² or 100 mg/ml for higher doses (up to 1700 mg/m²). The rate of intravenous (i.v.) injection was rapid (in 30 seconds) for the initial 60 mg/m² dose, but the doses above 945 mg/m² were given in five minutes. Dose levels from the level 1 (60 mg/m²) to the level 10 (1700 mg/m²) were decided based on the amount of quercetin (16) used. A bolus dose of 1400 mg/m² was proposed either weekly or at a three-week interval for a possible phase II clinical trial (Ferry et al., 1996).

Fisetin (17), a flavone derived from citrus fruits (Nijveldt et al., 2001), inhibited 2-DG uptake in myelocytic U937 and lymphocytic Jurkat cells in a dose-dependent manner when tested at 1–100 μM concentrations, which is more potent than either apigenin (14) or quercetin (16) (Park, 1999). As reviewed previously, fisetin (17) showed antitumor efficacy against several different types of animal models (Lall, Adhami, & Mukhtar, 2016). Fisetin (17) suppressed 451Lu human melanoma cell growth, with the IC₅₀ values being estimated as 80 μM, 37.2 μM, and 17.5 μM for the 24-, 48-, and 72-h treatments, respectively. Tumor growth was inhibited when athymic (nu/nu) female nude mice were inoculated with 451Lu cells and treated (i.p.) with 17 (45 mg/kg, twice a week) for 45 days (Syed et al., 2011). Mechanistically, fisetin (17) mediates its cytotoxicity toward 451Lu cells through reducing the key G1 phase cell cycle regulatory proteins cyclin-dependent-kinases (cdk-2, −4, and −6) and down-regulating proteins in the Wnt pathway (Syed et al., 2011). In addition, fisetin (17) showed a synergistic effect with sorafenib when five-week-old BALB/c female nude mice were inoculated with HeLa human cervical cancer cells and treated orally with sorafenib (10 mg/kg) or sorafenib (10 mg/kg) plus 17 (4 mg/kg) twice a week for five weeks (Lin et al., 2016).

The blood brain barrier (BBB) is a key regulator of glucose availability to glial and neuronal cells in the brain, and hCMEC/D3, an endothelial BBB cell line, is a good model to investigate the BBB mediated glucose transport in vitro. Myricetin (18), a flavone derived from fruit peels (Nijveldt et al., 2001), decreased 2-DG uptake in hCMEC/D3 cells as an inhibitor of glucose transport in a concentration-dependent manner when tested at concentration of 30, 100, and 300 μM (Meireles et al., 2013).

As summarized previously, myricetin (18) is a potential antitumor lead compound (Devi, Rajavel, Habtemariam, Nabavi, & Nabavi, 2015). It exhibited cytotoxicity toward T24 human bladder cancer cells (IC₅₀ 85 μM with 24 h treatment) by increasing the number of cells in the G2/M phase of the cell cycle and decreased the migratory capacity of these cells. Tumor growth was suppressed when four-week-old female BALB/c nude mice were inoculated with T24 cells and treated (i.p.) daily with 18 (5 mg/kg) for five days (Sun et al., 2012).

4.5. Isoflavones

Isoflavones are found especially in soy [Glycine max (L.) Merr. (Fabaceae)] and in other leguminous seeds. They function as a suppressive agent for the treatment of chemically induced mammary cancer without apparent reproductive or endocrinological toxicities (Zaheer & Akhtar, 2017). Two isoflavonoids, genistein (19) and daidzein (20) (FIGURE 5) derived from soy (Glycine max) (Taylor, Levy, Elliot, & Burnett, 2009), were found to inhibit glucose uptake competitively via GLUT1 in human erythrocytes (Martin, Kornmann, & Fuhrmann, 2003).

FIGURE 5.

Structures of isoflavones, a stilbenoid, an arylheptanoid, a flavonolignan, and an isoprenoid showing glucose transport inhibitory and potential antitumor activities

The potential antitumor efficacy of genistein (19) has been reviewed previously (Taylor, Levy, Elliot, & Burnett, 2009). It has also been tested in clinical trials either alone or with certain FDA-approved anticancer drugs. For example, in a block-randomized double blind phase II clinical trial to determine the efficacy and safety of short-term genistein intervention in patients with localized prostate cancer, 23 patients were selected and treated with genistein (19), and possible therapeutic effects in the early stage localized prostate cancer were predicted for this isoflavone (Lazarevic et al., 2011).

In another phase II clinical trial, 60 patients (median age 71 years) with bladder cancer were selected and treated (orally) daily with genistein (19) in capsules (G-2535) for 14 or 21 days or up to 30 days if surgery was delayed. G-2535 was found to be well-tolerated, and no significant differences were found in the treatment group (G-2535) and the placebo in terms of adverse effects. Also, the phosphorylated epidermal growth factor receptor was significantly reduced in 19-treated group compared to placebo (Messing et al., 2012).

4.6. Stilbenoid

Stilbenes are 1,2-diarylethene derivatives found exclusively in liverworts and higher plants, with ring A substituted with two hydroxy groups in the meta-position and with ring B substituted with hydroxy and methoxy groups in the ortho-, meta- and/or para-positions (Cassidy, Hanley, & Lamuela-Raventos, 2000). trans-Resveratrol (21), a stilbene occurring in peanuts [Arachis hypogaea L. (Fabaceae)] and grapes [Vitis vinifera L. (Vitaceae)], exhibits antibacterial, antifungal, and antitumor activities (Aluyen et al., 2012). As a GLUT1 and GLUT3 inhibitor, trans-resveratrol (21) inhibited competitively 2-DG uptake in human U937 histiocytic lymphoma and HL-60 leukemia cells, both in a dose-dependent manner (20–120 μM) (Park, 2001). It binds to the endofacial site of GLUT1 to reduce the amount of this glucose transporter bound to cytochalasin B (Salas, et al., 2013). It reduced [³H-2-DG] uptake in A2780 human ovarian cancer cells and further reduced ATP and lactate production, owing to its induction of autophagy and decrease of cellular metabolic activities (Kueck et al., 2007). Importantly, this stilbene was found to inhibit insulin-stimulated glucose uptake in human fat cell suspensions (Gomez-Zorita, Tréguer, Mercader, & Carpéné, 2013).

trans-Resveratrol (21) also showed potential antitumor efficacy and has been tested in several cancer clinical trial studies (Berman, Motechin, Wiesenfeld, & Hold, 2017). In a phase I randomized double-blind clinical trial, nine subjects who had confirmed stage IV colorectal cancer and hepatic metastases were selected and administered with trans-resveratrol (21) at 5.0 g as SRT501, a mixture of 21 in an aliquot of 4 ml sodium docusate solution, for six weeks. The dose was deemed to be safe and tolerated, and the further immunohistochemistry analysis of tumor tissue revealed that caspase-3 was increased in the SRT501 treated group compared to placebo (Howells et al., 2011).

5. OTHER OXYGEN HETEROCYCLIC AND PHENOLIC COMPOUDS

Several plant-derived oxygen heterocyclic and phenolic compounds exhibit both glucose transport inhibitory and potential antitumor activities, including curcumin (22) and silibinin (23). These two compounds showed promising in vivo antitumor efficacy, and silibinin (23) has also been evaluated in clinical trials as a potential anticancer drug for the treatment of prostate and hepatocellular cancers.

5.1. Diarylheptanoid

Turmeric [Curcuma longa L. (Zingiberaceae)] is a major source of various curcuminoids, of which curcumin (22) shows antitumor, radioprotective, and cardioprotective effects (Amalraj, Pius, Gopi, & Gopi, 2017). It was found to reduce GLUT1 gene and protein expression in NSCLC A549 cells transfected with pcDNA3.1-GLUT1 vector, and it also decreased the invasive capability of A549 cells by reducing expression of matrix metalloproteinases (MMPs) 1 and 2. This ultimately resulted in reduction of proliferation rates of untransfected and GLUT1 transfected A549 cells (Liao, Wang, Deng, Ren, & Li, 2015). Curcumin (22) has been subjected to many in vivo studies to date relative to its potential antitumor efficacy (Amalraj, Pius, Gopi, & Gopi, 2017). For example, it reduced metastasis when five–six-week-old BALB/c nude mice were inoculated with untransfected A549 cells and treated (i.p.) daily with 22 (200 μg/kg) for four weeks (Liao, Wang, Deng, Ren, & Li, 2015).

5.2. Flavonolignan

As a competitive inhibitor of GLUT4, silibinin (23, also known as silibin or silybin) (FIGURE 5), a flavonolignan obtained from the extract of the milk thistle Silybum marianum (L.) Gaertn. (Asteraceae), reduced basal and insulin-dependent 2-DG uptake in differentiated 3T3-L1 adipocytes at a concentration of 40 μM (Zhan, Digel, Küch, Stremmel, & Füllekrug, 2011). In addition, silymarin, an extract of S. marianum with 23 as a key active component, inhibited cell growth through increasing cell cycle regulation proteins in DU145 human prostate cancer cells (Zi, Grasso, Kung, & Agarwal, 1998). Tumor growth was inhibited when athymic nude (nu/nu) male mice were inoculated with DU145 cells and fed with 0.05% or 0.1% of 23 (w/w) containing diet for 60 days, or when athymic nude mice were fed with 0.05% or 0.1% of 23 (w/w) for three weeks, inoculated by DU145-cells, and then fed with 0.05% or 0.1% of 23 (w/w) for additional six weeks. No apparent sign of toxicity was observed in mice, and insulin-like growth factor-binding protein-3 (IGFBP-3) levels were found increased in mouse plasma (Singh et al., 2002). These in vivo studies indicate that silibinin (23) exhibits both antitumor and tumor preventive properties, and these activities might be mediated mechanistically through an IGFBP-related pathway, such as cell survival (antiapoptotic) signaling via IGFBP-IGF-1/IGF-1R pathway (Singh et al., 2002).

In a phase I clinical trial for toxicity of high-dose silibinin in a phytosome [silybin-phytosome (Siliphos®)] and a dose recommended for a phase II study, 13 prostate cancer patients with the median age of 70 years were selected and administered orally with 2.5, 5, 10, 13, 15, and 20 g of Siliphos® for a total of 91 courses (four weeks for each course). Dose limiting toxicity was defined as grade 3 or 4 non-hematologic toxicity, or grade 4 hematologic toxicity, and a daily oral dosage of 13 g in three divided doses of Siliphos® was recommended for a phase II clinical trial for the treatment of prostate cancer (Flaig et al., 2007). However, a phase II clinical trial that was conducted on 12 prostate cancer patients with the median age of 57 years who were given orally with 13 g of Siliphos® for 14–31 days was not continued further, owing to the low tissue penetration of 23 (Flaig et al., 2010).

Silibinin (23) was also tested in a phase I clinical trial for advanced hepatocellular carcinoma (AHC). Three male patients (mean age 53 years) suffering from AHC were enrolled for the treatment during a span of 12 weeks with Siliphos®. However, the trial ended without conclusions, owing to deaths of all patients (Siegel et al., 2014).

6. Isoprenoids

Isoprenoids, including terpenoids as a major group, are a large and diverse class of natural products composed of two or more isoprenyl groups that connect each other in different modes. Among these compounds, gossypol is a dimeric sesquiterpene derived from species of the Malvaceae family, including Thespesia populnea (L.) Sol. ex Corrêa (Boonsri, Karalai, Ponglimanont, Chantrapromma, & Kanjana-opas, 2008) and Gossypium barbadense L. (extra-long staple cotton) (Malvaceae) (Dowd, & Pelitire, 2006). Gossypol is an axial chiral compound and exists naturally as an enantiomeric mixture, owing to the restricted rotation of its internaphthyl 2,2’-bond, which results in helical M [(–)-gossypol] and P [(+)-gossypol (24)] isomers (Freedman, Cao, Oliveira, Cass, & Nafie, 2003). Gossypol showed GLUT1 inhibitory potency to reduce 2-DG in HL-60 human leukemia cells and human erythrocytes (Pérez et al., 2009), and it also exhibits a broad spectrum of cytotoxicity toward various human cancer cells. For example, all of (±)-gossypol, (+)-gossypol, and (–)-gossypol were found to show non-selective antiproliferative activity against a panel of human cancer cell lines, with IC₅₀ values being in the range 0.3–6.1 μg/mL, which was attributable primarily to the content of (–)-gossypol (Band et al., 1989). Apoptosis was induced in HL60 human promyelocytic leukemia cells when cells were treated with gossypol acetic acid at concentrations of 50 μM and 100 μM (Balci, Sahin, & Ekmekci, 1999). Interestingly, (+)-gossypol (24), isolated from the wood of Thespesia populnea, showed potent cytotoxicity toward human HeLa cervical and KB oral epidermoid cancer cells, with IC₅₀ values of 80 and 40 ng/mL, respectively. However, an analogue of 24, (+)-6,6′-methoxygossypol, did not show such activity, indicating that methylation of hydroxy group at the C-6 and C-6’ position of 24 results in its cytotoxic potency against HeLa and KB cells being greatly decreased (Boonsri, Karalai, Ponglimanont, Chantrapromma, & Kanjana-Opas, 2008).

In a phase I/II clinical trial study, 20 women patients who suffered with metastatic breast cancer refractory to doxorubicin and paclitaxel received oral gossypol daily for four weeks, using doses between 30 and 50 mg. The maximal tolerated dose (MTD) of 40 mg/day with a median serum gossypol concentration of 271 ng/ml was found to be tolerated, but no therapeutic responses were observed. However, alterations in cyclin D1 and Rb expression and a decrease in serial serum tumor marker values (CEA, BR2729 or CA15–3) observed in these patients suggest a potential role for gossypol to be used in conjunction with other cell cycle-specific compounds (Van Poznak et al., 2001).

Oridonin (25), a diterpene derived from the whole tea plant of Rabdosia rubescens (Hemsl.) H.Hara (Lamiaceae) (Yang, Lin, & Wei, 2017), was found to decrease the glucose uptake and subsequently reduced the expression of GLUT1 mRNA and protein in SW480 human colorectal cancer cells (Yao et al., 2017). It inhibited the growth of multiple myeloma, acute lymphoblastic T-cell leukemia (Jurkat), and adult T-cell leukemia (MT-1) cells, with an ED₅₀ value being in the range 0.75–2.7 μg/mL (Ikezoe et al., 2005). In addition, oridonin (25) reduced expression of p-AMPK (an ATP-sensor), induced autophagy, and inhibited the proliferation of SW480 cells. Tumor growth was inhibited when six-week-old BALB/c nude mice were inoculated with SW480 cells and treated (i.p.) with 25 (15 mg/kg, every alternate day) for two weeks (Yao et al., 2017). Mechanistically, oridonin (25) caused apoptosis of MT-1 cells, down-regulated levels of Mcl-1 and BCL-xL, but not Bcl-2 protein, in both MT-1 and RPMI8226 cells, and inhibited NF-κB activity in MT-1 cells, indicating that this diterpenoid might be useful as adjunct therapy for individuals with lymphoid malignancies, including the lethal disease, adult T-cell leukemia (Ikezoe et al., 2005).

7. Conclusions

Glucose is a key nutrient utilized for the production of cellular energy in the form of ATP and for the synthesis of biomass, and glucose transporter proteins, including GLUTs, SGLTs, and SWEETs, are responsible for the uptake of glucose in human cells and have become attractive targets for the discovery of potential anticancer agents. Thus far, many plant-derived inhibitors of glucose transport have been discovered, which target mainly GLUTs 1–4, 5, and 7 and SGLTs 1 and 2 (TABLE 1). Interestingly, most of these glucose cotransport inhibitors are either major components of edible plants or analogues of these plant-derived compounds, indicating that these natural inhibitors may be used to address the management of the toxicities observed from the currently-used anticancer drugs. Thus, characterization of novel small-molecule naturally occurring glucose transport inhibitors, especially SWEET inhibitors, should be a promising strategy for the development of new anticancer drugs.

TABLE 1.

Plant-derived glucose transport inhibitors with potential antitumor activity

No	Compound	Plant Origin	Biological Potency	Reference
1	Vinblastine	Catharanthus roseus	Reduced 2-DG uptake in glioma C6 cells and has been used effectively to treat stage III NSCLC	Boleti et al. (2007), Moudi et al. (2013), Ellis et al. (1995), Singh et al. (1998)
2	10-methoxy-N(1)-methylburnamine-17-O-veratrate	Alstonia macrophylla	Inhibited SGLT1 (IC50 4.0 μM) and SGLT2 (IC50 0.5 μM) in COS-1 cells	Arai et al., 2010
3	Alstiphyllanine D	Alstonia macrophylla	Inhibited SGLT1 (IC50 5.0 μM) and SGLT2 (IC50 2.0 μM) in COS-1 cells	Arai et al. (2010)
4	Pelargonidin-3-O-glucoside	Fragaria virginiana	Inhibited glucose uptake in Caco-2 cells (IC50 705 μM)	Manzano et al. (2010)
5	Phloretin	Malus domestica	Inhibited GLUT4-mediated glucose transport (IC50 9.4 μM) in GLUT4 cells and suppressed HepG2 xenograft tumor growth (10 mg/kg)	Andrianesis et al. (2016), Gosch et al. (2010), Kasahara et al. (1997), Yang et al. (2009)
6	Phlorizin	Malus domestica	Inhibited GLUT4-mediated glucose transport (IC50 140 μM) in GLUT4 cells	Andrianesis et al. (2016), Kasahara et al. (1997)
7	Xanthohumol	Humulus lupulus	Reduced [3H-2-DG] uptake in HTR-8/SVneo cells (IC50 3.6 μM) and suppressed pancreatic tumor growth (25 mg/kg)	Correia-Branco et al. (2015), Jiang et al. (2015)
8	(+)-Cryptocaryone	Cryptocarya rubra	Reduced glucose uptake in H1299 cells and showed cytotoxicity against HT-29 cells (IC50 0.32 μM)	Ren et al. (2014)
9	Naringenin	Citrus paradisa	Decreased the 2-DG uptake in differentiated 3T3-L1 cells and inhibited 4T1 breast mouse tumor metastasis (200 mg/kg)	Claussnitzer et al. (2011), Salehi et al. (2019), Zhang et al. (2016)
10	Epigallocatechin gallate	Camellia sinensis	Inhibited GLUT2 and GLUT5 in Xenopus ooctyes and suppressed SGC-7901 gastric xenograft tumor growth (1.5 mg/kg)	Ojelabi et al. (2018), Gauer et al. (2018), Zhu et al. (2007)
11	(−)-Kurarinone	Sophora flavescens	Inhibited SGLT1 (IC50 10.4 μM) and SGLT2 (IC50 1.7 μM) in COS-1 cells and showed cytotoxicity against a panel of human cancer cell lines (IC50 2–27 μg/mL)	Sato et al. (2007), Sun et al. (2007)
12	Sophoraflavanone G	Sophora flavescens	Inhibited SGLT1 (IC50 18.7 μM) and SGLT2 (IC50 4.1 μM) in COS-1 cells and showed cytotoxicity against a panel of human cancer cell lines (IC50 2–27 μg/mL)	Sato et al. (2007), Sun et al. (2007)
13	Hesperidin	Citrus sinensis	Inhibited GLUT2 and GLUT5 in Xenopus ooctyes and showed potential antitumor activity	Kerimi et al. (2019), Jung et al. (2006)
14	Apigenin	Malus domestica	Inhibited GLUT2 and GLUT7 in Xenopus ooctyes and enhanced cytotoxicity against Hep-2 cells of cisplatin	Nijveldt et al. (2001), Melstrom et al. (2008)
15	Luteolin	Apium graveolens	Inhibited GLUT4 in MC3T3-G2/PA6 adipose cells and invasive capability of MDA-MB-231 cells	Imran et al. (2019), Nijveldt et al. (2001), Nomura et al. (2008), Sun et al. (2015)
16	Quercetin	Allium cepa Lactuca sativa Petroselinum crispum Olea europaea	Inhibited GLUT1 (IC50 2.0 μM), GLUT3 (IC50 17.7 μM), and GLUT4 (IC50 1.7 μM) in HEK293 cells and showed anticancer potentials in a phase I clinical trial (1400 mg/m2)	Nijveldt et al. (2001), Ojelabi et al. (2018), Rauf et al. (2018), Ferry et al. (1996)
17	Fisetin	Citrus fruits	Reduced 2-DG uptake in U937 and Jurkat cells and showed cytotoxicity against 451Lu cells	Nijveldt et al. (2001), Park (1999), Lall et al. (2016), Syed et al. (2011)
18	Myricetin	Fruit peels	Decreased 2-DG uptake in hCMEC/D3 cells, showed cytotoxicity against T24 cells (IC50 85 μM), and suppressed T24 bladder tumor growth (5 mg/kg)	Nijveldt et al. (2001), Meireles et al. (2013), Devi et al. (2015), Sun et al. (2012)
19	Genistein	Glycine max	Inhibited GLUT1 in human erythrocytes and showed anticancer potential in a phase I clinical trial for the treatment of prostate cancer and in a phase II clinical trial for the treatment of bladder cancer	Lazarevic et al. (2011), Martin et al. (2003), Messing et al. (2012), Taylor et al. (2009)
20	Daidzein	Glycine max	Inhibited GLUT1 in human erythrocytes	Taylor et al. (2009), Martin et al. (2003)
21	trans-Resveratrol	Arachis hypogaea Vitis vinifera	Inhibited 2-DG uptake in U937 and HL-60 cells and showed anticancer potential in a phase I clinical trial for the treatment of colon cancer	Aluyen et al. (2012), Salas, et al. (2013), Howells et al. (2011)
22	Curcumin	Curcuma longa	Inhibited GLUT1 in A549 cells and suppressed A549 xenograft lung tumor growth (200 μg/kg)	Amalraj et al. (2017), Liao et al. (2015)
23	Silibinin	Silybum marianum	Reduce basal and insulin-dependent 2-DG uptake in differentiated 3T3-L1 adipocytes at 40 μM and suppressed DU145 xenograft tumor growth	Singh et al. (2002), Zhan et al. (2011), Zi et al. (1998)
24	(+)-Gossypol	Thespesia populnea Gossypium barbadense	Inhibited GLUT1 in HL-60 cells and in human erythrocytes and showed cytotoxicity against a panel of human cancer cell lines (IC50 0.3–6.1 μg/mL)	Band et al. (1989), Balci et al. (1999), Boonsri et al. (2008), Freedman et al. (2003), Pérez et al. (2009)
25	Oridonin	Rabdosia rubescens	Inhibited GLUT1 in human SW480 cells and showed cytotoxicity against Jurkat/MT-1 cells (ED50 0.75–2.7 μg/mL)	Ikezoe et al. (2005), Yang et al. (2017), Yao et al. (2017)

Abbreviations

αMDG

Methyl-α-D-glucopyranoside

Aκt

Protein kinase B (PKB)

ATP

Adenosine triphosphate

BBB

Blood brain barrier

Caco-2 cells

human colon cancer cells

CHO cells

Chinese hamster ovary cells

COS-1 cells

African green monkey kidney fibroblast-like cells

CSC

Cancer stem cells

2-DG

2-Deoxy-glucose

GLUT

Glucose transporter

HL

Hodgkin’s lymphoma

IC₅₀

The concentration of a compound required for 50% inhibition of cell viability

NSCLC

Non-small cell lung cancer

SGLT1

Sodium glucose cotransporter-1

SLC2

Solute carrier 2

SWEET

Sugar will eventually be exported transporter

T2DM

Type 2 diabetes mellitus

TCA

Tricarboxylic acid