- Email: [email protected]
Carbohydrate-based inducers of cellular stress for targeting cancer cells
Bioorganic & Medicinal Chemistry Letters 26 (2016) 1452–1456
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl
Carbohydrate-based inducers of cellular stress for targeting cancer cells Fidelis T. Ndombera, Garrett C. VanHecke, Shima Nagi, Young-Hoon Ahn ⇑ Department of Chemistry, Wayne State University, 5101 Cass Ave, Detroit, MI 48202, USA
a r t i c l e
i n f o
Article history: Received 17 December 2015 Revised 20 January 2016 Accepted 21 January 2016 Available online 22 January 2016 Keywords: N-Aryl glycoside library Reactive oxygen species inducer Cytotoxicity 2-Deoxy-D-glucose
a b s t r a c t Small molecules that block the altered metabolism in cancer or increase the production of reactive oxygen species (ROS) are emerging as potential anti-cancer agents. Considering that various carbohydrates can be used for cellular energetics or protein N-glycosylation of which interruption can lead to cellular stress, we have synthesized and evaluated a library of N-aryl glycosides for induction of ROS and cytotoxicity in H1299 cancer cell line. Two N-aryl glycosides (K8 and H8) were identified that induce about 2-fold induction of ROS and cytotoxicity in H1299 cells. We further showed that the acetylated form of K8 (K8A) activates AMPK, and stabilizes p53 in HEK293 cells, and induce a higher cytotoxicity than 2-deoxy-D-glucose in H1299 cell line. Ó 2016 Elsevier Ltd. All rights reserved.
Metabolic reprogramming of cancer cells is emerging as a key process for their survival and proliferation.1 Most cancer cells are programmed to increase glycolysis, a phenomenon known as Warburg effect, without maximizing ATP synthesis in mitochondria.1,2 The increased glycolytic intermediates, such as glucose 6-phosphate or 3-phosphoglycerate, are directed to the pentose phosphate pathway (PPP) and serine biosynthesis, respectively, to provide ribose, NADPH, and other building blocks for anabolism of proliferating cells.1,2 The Krebs cycle intermediates are replenished by uptake of glutamine, which leads to increased production of NADPH for lipid biosynthesis and precursors of amino acids.2 Importantly, this metabolic reprogramming is essential for maintaining the redox homeostasis of cancer cells.3 NADPH produced by the above pathways is an essential reducing equivalent for protecting against reactive oxygen species (ROS) in cancer cells.1 Furthermore, serine biosynthesis increases glycine availability, which is necessary for glutathione biosynthesis.1,3 In addition, mutation of oncogenes, such as MYC and KRAS, is closely interconnected with metabolic reprogramming and redox regulation.3,17 This indicates the intricate control of metabolism and redox balance for survival of cancer cells. This metabolic tuning and redox control of cancer cells is vital for their proliferation, but also provides selective strategies for treating cancer.1,4,5 First, a high glycolytic activity of cancer cells was targeted by antiglycolytic drugs, including 2-deoxy-D-glucose (2-DG). 2-DG, a glucose analog that is devoid of 2-hydroxyl group, is converted to 2-DG 6-phosphate, which blocks glycolysis by ⇑ Corresponding author. http://dx.doi.org/10.1016/j.bmcl.2016.01.063 0960-894X/Ó 2016 Elsevier Ltd. All rights reserved.
inhibiting hexokinase and phosphoglucose isomerase.4 2-DG also interferes with N-glycosylation of proteins in endoplasmic reticulum (ER).4 Therefore, treatment of 2-DG decreases ATP synthesis, increases cellular ROS, causes ER stress, and increases the ratio of oxidized glutathione over reduced glutathione, resulting in cancer cell death.4 2-DG has been evaluated in clinical trials of patients with several tumors, mostly in combinatorial therapy with cytotoxic drugs.4,14,23 In addition, 3-bromopyruvate is a promising anti-glycolytic drug that blocks pyruvate kinase in glycolysis.5 In a similar context, glucose starvation induced the selective cytotoxicity in several cancer cells relative to their isogenic non-transformed cells.6 Secondly, increased generation of ROS and an intricate control of redox status in cancer cells relative to normal cells provide a basis of designing ROS-inducing or ROS-activated electrophilic anticancer agents.7 Phenethyl isothiocyanate (PEITC), piperlongumine (PL), sulforaphane, and N-(2,5-dihydroxyphenyl) acetamide-based agents that retain potential electrophilic reactive groups have shown the selective cytotoxicity to malignant cells relative to non-transformed cells.7–11,15,21 This is due to vulnerability of cancer cells to reach lethal ROS levels above the antioxidant protective threshold.7,12 Recent library screening identified several ROS enhancing compounds among which mostly electrophilic compounds were found to induce cytotoxicity.8 Overall, this suggests that compounds with the property of blocking metabolic activity and/or enhancing ROS may be a potential therapeutic strategy targeting cancer cells. Here we report N-aryl glycoside small molecules that enhance ROS, perturb the cellular energetics, and induce the cytotoxicity in H1299, a non-small lung cancer cell line.
F. T. Ndombera et al. / Bioorg. Med. Chem. Lett. 26 (2016) 1452–1456
Because various carbohydrates can be used for cellular bioenergetics or glycosylation in proteins, we sought to identify a carbohydrate derivative, such as 2-DG, that induces the cellular stress by increasing the level of ROS and inducing cytotoxicity in cancer cells. Interestingly, the cellular entry of various carbohydrates other than glucose can be facilitated by glucose transporter, GLUT-1, that is highly expressed for enhanced glycolysis in cancer cells.13 The substitution at the anomeric carbon (C1), carbon 2 (C2) and carbon 6 (C6) of glucose is highly tolerated by GLUT-1.13 To increase the potential candidate compounds, we synthesized a library of N-aryl-glycosides that are assembled by simple N-glycosylation of 12 carbohydrates and 8 aniline derivatives (Scheme 1). N-Glycosides are more stable than O-glycosides, and N-glycosylation would provide diverse structural modifications. Previously, b-N-aryl-glycosides were synthesized by simple heating at low temperatures (40 °C) in a slightly acidic buffer at pH 6.5 in one step in which excess amount of carbohydrate over aniline led to selective synthesis of b-isomers.16 To streamline the method, we modified this one step N-glycosylation in aqueous buffer (20% DMSO) at higher temperatures (70 °C). After heating individual carbohydrate (A–L) and aniline (1–8) (Scheme 1) in a ratio of 16:1 in phosphate buffer (pH 6.5), the mixture was directly injected to a semi-preparative HPLC column for purification of products.24 After LC–MS analysis of individual products, 46 compounds that had a high purity (>90%) were selected for the subsequent screening assay. The purity of individual compounds is shown in Table S1. It is noted that many of reactions, including
Scheme 1. Synthesis and building blocks of N-aryl glycosides.
1453
one with 2-deoxy-D-ribose (E) and 4-chloro-aniline (2), results in only one major peak of N-aryl glycoside (Fig. 1A), but a few reactions, including one with D-ribose (I) and 2-aminocoumarin (4), gave two peaks of identical mass in HPLC (Fig. 1B). Two peaks are likely to correspond to a- and b-anomers of given products, which were designated as a and b, respectively, without further characterization at this stage. To identify N-aryl glycosides that increase the ROS level, we used a cell-permeable fluorescent probe, 20 ,70 -dichlorodihydrofluorescein diacetate (DCF-DA), for detection of ROS increase in nonsmall cell lung carcinoma H1299 cells. DCF-DA readily crosses cell membranes and is retained in cells upon deacetylation by esterases.18,19 It is then oxidized by various species of ROS to emit fluorescence.18 With DCF assay, we screened a library of N-aryl glycosides (46 compounds in 100 lM, in duplicates) for an increase of ROS in H1299 cells. The fluorescence increases of individual N-aryl glycosides in comparison to a non-treated control are summarized in Figure 2A. A majority of compounds induced no significant change of ROS. However, six compounds (H8, G8, B8, F6, C4b, and K8) induced 2-fold or higher increases (Fig. 2A). Interestingly, it is notable that several of 1-amino-naphthalene (8)-containing compounds were identified to increase ROS levels. For example, H7a and H7b that contain aniline (7) did not induce ROS significantly in comparison to H8 that contains 1-amino-nathpthalene (8). In comparison, about 1.5-fold increase of ROS level was observed by treatment of 2-DG (10 mM) (Fig. S1). We then re-synthesized several 8-series compounds, including A8, C8, H8, and K8, for further evaluation.24,25 After re-synthesis, we were able to determine the anomeric configuration of compounds by NMR analysis (Figs. S4–S7).22 These four compounds were then evaluated for an increase of ROS levels in a dose-dependent manner (Fig. 3A). H8 and K8 showed a dose-dependent 2-fold increase of ROS level within 100 lM concentration. Notably, H8 and K8 contain a similar carbohydrate structure with D-arabinose (H) and D-xylose (K), respectively, which are aldopentoses with a reversed stereochemistry at 2-OH and 3-OH positions. On the other
Figure 1. HPLC chromatograms of selected N-aryl glycosides. (A) The reaction with 2-deoxy-D-ribose (E) and 4-chloro-aniline (2) results in one peak of an expected mass. (B) The reaction with D-ribose (I) and 2-amino-coumarin (4) resulted in two peaks of an expected mass, which were designated as a and b.
1454
F. T. Ndombera et al. / Bioorg. Med. Chem. Lett. 26 (2016) 1452–1456
Figure 2. Evaluation of N-aryl glycosides for ROS increase and cytotoxicity in H1299 cells. (A) DCF-DA assay was used for measuring the fluorescence intensity after incubation of individual N-aryl glycoside (100 lM) (Fi) versus non-treated control (Fo) after 5-h incubation in H1299 cells. (B) MTT assay was used for analyzing the cytotoxicity of individual N-aryl glycoside (100 lM) (Vi) versus non-treated control (Vo) after 48-h incubation in H1299 cells.
hand, A8 and C8 did not induce a significant increase of ROS level (Fig. 3A). A8 contains D-glucose, and it did not induce a significant increase of ROS level, suggesting that a pentose rather than a hexose may be important. Also, there was almost no increase of ROS level by C8 that contain L-arabinose, showing the differential effect of the stereochemistry of D-arabinose in H8 versus L-arabinose in C8. In control experiments, D-arabinose (H), D-xylose (K) or 1naphthylamine (8) alone did not induce an increase of ROS level (Fig. 3B). We then measured the cytotoxicity of the same 46 compounds on cell growth and viability of H1299 cells using MTT assay. N-Aryl glycosides (100 lM in duplicate) were incubated in H1299 for 48 h, and MTT assay results are summarized in Figure 2B. The comparison between the ROS enhancing property and the cytotoxicity indicated that these data do not appear to be significantly correlated (Fig. 2). Out of five compounds with highest cytotoxicity (H8, K8, H7a, H7b, C7) (Fig. 2B), two compounds (H8 and K8) induced a high level of ROS while three compounds (H7a, H7b, and C7) induce almost no increase of ROS (Fig. 2A). We then reevaluated cytotoxicity of H8 and K8 in comparison to hydrogen peroxide and 2-DG. Hydrogen peroxide reduced viability with IC50 value at 60 lM after a 48 h incubation (Fig. 4A). 2-DG showed about 80% viability when treated at 500 lM concentration (Fig. 4A), which is similar to previous Letters.23 In identical conditions, H8 and K8 reduced the viability with IC50 values at about 75 lM (Fig. 4B), which is similar or much higher cytotoxicity versus hydrogen peroxide or 2-DG, respectively. Because the IC50 values of H8 and K8 are relatively high, we further modified K8 by acetylating all hydroxyl groups.25 We thought that acetylated derivatives of K8 may enter the cells more efficiently than their non-acetylated forms, which may increase the potency of compounds. Previously, acetylated 18F-2-deoxy-D-glucose rapidly penetrated the cell membrane as a result of its high lipophilicity and was metabolized to non-acetylated 2-DG in cancer cells.20
Interestingly, the acetylated derivative of K8 (K8A) reduced the viability of H1299 with IC50 at 30 lM, which is about 2.5-fold increase versus K8 (Fig. 4C). In controls, D-xylose (K) and its acetylated form (KA) did not induce a significant cytotoxicity in H1299 cells (Fig. 4D). K8A also induced about 1.5-fold increase of ROS level within 50 lM (Fig. S2). In order to see the potential mechanism of K8A for cytotoxicity, we evaluated the activation of AMP-activated protein kinase (AMPK) upon treatment of K8A (Fig. 5). AMPK is a stress-sensing kinase that is activated by phosphorylation when AMP level increases. AMPK phosphorylation halts protein synthesis and cell cycle while inducing p53 stabilization.23 2-DG that decrease ATP synthesis and increase cellular ROS was shown to phosphorylate AMPK and stabilize p53 for apoptosis. The treatment of 2-DG to HEK293 cells increased AMPK phosphorylation in a dose-dependent manner (Fig. 5). In a similar manner, the incubation of K8A within 100 lM concentration led to a dose-dependent increase of AMPK phosphorylation (Fig. 5). In addition, p53, which is downstream of AMPK, is stabilized upon treatment of K8A (Fig. 5). Because 2-DG is also known to induce ER stress by interfering with N-glycosylation of proteins, we further evaluated K8 and K8A for induction of glucose-regulated protein 78 (GRP78/BiP), which is a hallmark of ER stress.26 Incubation of K8A and K8 at 200 lM concentration in HEK293 cells showed the induction of GRP78 (Fig. S3). This indicates that K8A may be involved in disrupting cellular energetics and inducing ER stress for its cytotoxicity. In summary, we have developed a library of simple N-aryl glycoside derivatives. Cellular ROS and cytotoxicity assays of this Naryl-glycoside library led to identifying the potential candidate (K8A) that activates AMPK, and stabilizes p53 in HEK293 cells, and induce cytotoxicity with a higher potency than 2-DG in H1299 cells. Evaluation of K8A derivatives for cytotoxicity, the mechanism of action, and target identification will be further investigated in future studies.
F. T. Ndombera et al. / Bioorg. Med. Chem. Lett. 26 (2016) 1452–1456
1455
Figure 3. ROS induction assays of N-aryl glycosides containing 1-amino-naphthalene (8). DCF-DA assay was used for measuring the fluorescence intensity (Fi) after 5-h incubation of A8, C8, H8, and K8 (A) or H, K, and 8 (B) in H1299 cells with an increasing concentration (0–100 lM) versus a non-treated control (Fo).
Figure 4. The cytotoxicity of N-aryl glycosides H8, K8 and K8A. MTT assay was used for measuring the cytotoxicity (Vi) of hydrogen peroxide and 2-DG (A), H8 and K8 (B), K8A (C), K and KA (D) versus a non-treated control (Vo) by incubating individual compounds for 48 h in H1299 cells.
1456
F. T. Ndombera et al. / Bioorg. Med. Chem. Lett. 26 (2016) 1452–1456
Figure 5. K8A induces a dose-dependent induction of AMPK phosphorylation and p53 stabilization similar to 2-DG. K8A or 2-DG was incubated for 16 h in HEK293 cells, and lysates were probed for Western blotting with indicated antibodies.
Acknowledgments This work was supported by Wayne State University Start-up fund and WSU University Research Grant. Supplementary data Supplementary data (NMR spectra and other characterization of compounds) associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bmcl.2016.01.063. References and notes 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
Schulze, A.; Harris, L. A. Nature 2012, 491, 364. Vander Heiden, G. W.; Cantley, C. L.; Thompson, B. C. Science 2009, 324, 1029. Ward, S. P.; Thompson, B. Cancer Cell 2012, 21, 297. Zhang, D.; Li, J.; Wang, F.; Hu, J.; Wang, S.; Sun, Y. Cancer Lett. 2014, 355, 176. Galluzzi, L.; Kepp, O.; Vander Heiden, G. W.; Kroemer, G. Nat. Rev. Drug Disc. 2013, 12, 829. Palorini, R.; Cammarata, F. P.; Balestrieri, C.; Monestiroli, A.; Vasso1, M.; Gelfi, C.; Alberghina, L.; Chiaradonna, F. Cell Death Dis. 2013, 4, e732. Pelicano, H.; Carney, D.; Huang, P. Drug Resist. Update 2004, 7, 97. Adams, J. D.; Boskovic, V. Z.; Theriault, R. J.; Wang, J. A.; Stern, M. A.; Wagner, K. B.; Shamji, F. A.; Schreiber, L. S. ACS Chem. Biol. 2013, 8, 923. Gorrini, C.; Harris, S. I.; Mak, W. T. Nat. Rev. Drug Disc. 2013, 12, 931. Wang, J.; Yi, J. Cancer Biol. Ther. 2008, 12, 1875. Ivanova, D.; Bakalova, R.; Lazarova, D.; Gadjeva, V.; Zhelev, Z. Adv. Clin. Exp. Med. 2013, 22, 899. Trachootham, D.; Alexandre, J.; Huang, P. Nat. Rev. Drug Disc. 2009, 8, 579. Calvaresi, C. E.; Hegenrother, J. P. Chem. Sci. 2013, 4, 2319.
14. Raez, E. L.; Papadopoulus, K.; Ricart, D. A.; Chioreau, G. E.; Dipaola, S. R.; Stein, N. M.; Lima, R. M. C.; Schlesselman, J. J.; Tolba, K.; Langmuir, K. L.; Kroll, S.; Jung, T. D.; Kurtoglu, M.; Rosenbl, J. Cancer Chemother. Pharmacol. 2013, 71, 523. 15. Vadukoot, K. A.; Abdulsalam, F. S.; Wunderlich, M.; Pullen, D. E.; LanderoFigueroa, J.; Mulloy, C. J.; Merino, J. E. Bioorg. Med. Chem. 2014, 22, 6885. 16. Bridiau, N.; Benmansour, M.; Legoy, D. M.; Maugard, T. Tetrahedron 2007, 63, 4178. 17. Shen, L.; O’Shea, M. J.; Kaadige, R. M.; Cunha, S.; Wilde, R. B.; Cohen, L. A.; Welm, L. A.; Ayer, A. D. Proc. Natl. Acad. Sci. U.S.A. 2015, 112, 5425. 18. Wang, H.; Joseph, A. J. Free Radical Biol. Med. 1999, 27, 612. 19. Winterbourn, C. C. Biochim. Biophys. Acta 2014, 1840, 730. 20. Waki, A.; Fujibayashi, Y.; Magata, Y.; Yokohama, A.; Sadato, N.; Tsuchida, T.; Ishii,; Yonekura, Y. J. Nucl. Med. 1998, 2, 245. 21. Ahn, Y. H.; Hwang, Y.; Liu, H.; Wang, X. J.; Zhang, Y.; Stephenson, K. K.; Boronina, T. N.; Cole, R. N.; Dinkova-Kostova, A. T.; Talalay, P.; Cole, P. A. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 9590. 22. NMR spectra of A8, C8, H8, and K8 are shown in Figures S4–S7. The structures are shown in Figure 3. The anomeric stereochemistry was assigned by examining the coupling constant between anomeric proton and adjacent C– H: A8 (4.80 ppm, J = 7.9 Hz), C8 (4.73 ppm, J = 8.2 Hz), H8 (4.70 ppm, J = 8.0 Hz), K8 (4.67 ppm, J = 8.2 Hz). The structure of C8, H8, and K8 are shown as pyranoses, given that arabinose prefers pyranose over furanose in water. 23. Sahra, B. I.; Laurent, K.; Giuliano, S.; Larbret, F.; Ponzio, G.; Gounon, P.; Marchand-Brustel, L. Y.; Giorgetti-Peraldi, S.; Cormont, M.; Bertolotto, C.; Deckert, M.; Auberger, P.; Tanti, J.; Bost, F. Cancer Res. 2010, 70, 2465. 24. To prepare the N-aryl glycosides, a reaction mixture of 1 equiv aniline derivative (1–8) in 20% dimethyl sulfoxide (DMSO) was added to 16 equiv of preheated carbohydrate (A–L) in potassium phosphate buffer (pH 6.5) in a 4 mL glass vial and incubated at 70 °C for 16 h. The reactant mixture was filtered twice using cotton wool to remove any solid particles and a HPLC injection was made to purify the product. Products were purified by a semipreparative C-18 reversed-phase column HPLC with a gradient elution method (a mobile phase composition of water and acetonitrile, a flow rate of 10 mL/ min, injection volume of 3–4 mL, run time of 75 min and detection wavelengths of 214 nm and 254 nm). The yields after HPLC purification are in the range of 5–15%. Products were eluted at retention times between 30– 35 min and collected, flash-frozen in liquid nitrogen, and lyophilized under vacuum. Identity of individual product was confirmed by LC–MS ESI (Table S1). The products were then dissolved in DMSO and stored in 20 °C freezer for assays. 25. To prepare K8A, D-xylose (500 mg, 3.33 mmol) and 1-amino-napthalene (525 mg, 3.66 mmol) was refluxed in ethanol for 8 h to obtain brown sparkly crystalline solid (730 mg). The crystalline solid (500 mg) was reacted with acetic anhydride (8.17 mmol) in pyridine (0.79 mL, 9.81 mmol) in an ice bath for 3 h and in the cold room for another 13 h. The reactant mixture was then quenched by pouring into ice/water, extracted, washed with aq sodium bicarbonate, water and brine, and purified by column chromatography (1:4 ethyl acetate/hexane solvent conditions). After purification, a yellow solid K8A (622 mg, 1.55 mmol) was obtained and it was further recrystallized from ethanol to provide the white K8A crystals. 1H NMR (400 MHz, CDCl3) d 7.80 (t, J = 4.8 Hz, 1H), 7.72 (t, J = 4.8 Hz, 1H), 7.46 (t, J = 4.4 Hz, 2H), 7.37 (q, J = 8 Hz, 2H), 6.79 (d, J = 6.8 Hz, 1H), 5.44 (t, J = 9.6 Hz, 2H), 5.17 (t, J = 9.2 Hz, 1H), 5.13– 5.03 (m, 1H), 4.84 (d, J = 9.2 Hz, 1H), 4.16 (q, J = 5.6 Hz, 1H), 3.51 (t, J = 10.8 Hz, 1H), 2.1 (s, 3H), 2.07 (s, 3H), 2.06 (s, 3H). 26. Lee, A. S. Methods 2005, 35, 373.
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl
Carbohydrate-based inducers of cellular stress for targeting cancer cells Fidelis T. Ndombera, Garrett C. VanHecke, Shima Nagi, Young-Hoon Ahn ⇑ Department of Chemistry, Wayne State University, 5101 Cass Ave, Detroit, MI 48202, USA
a r t i c l e
i n f o
Article history: Received 17 December 2015 Revised 20 January 2016 Accepted 21 January 2016 Available online 22 January 2016 Keywords: N-Aryl glycoside library Reactive oxygen species inducer Cytotoxicity 2-Deoxy-D-glucose
a b s t r a c t Small molecules that block the altered metabolism in cancer or increase the production of reactive oxygen species (ROS) are emerging as potential anti-cancer agents. Considering that various carbohydrates can be used for cellular energetics or protein N-glycosylation of which interruption can lead to cellular stress, we have synthesized and evaluated a library of N-aryl glycosides for induction of ROS and cytotoxicity in H1299 cancer cell line. Two N-aryl glycosides (K8 and H8) were identified that induce about 2-fold induction of ROS and cytotoxicity in H1299 cells. We further showed that the acetylated form of K8 (K8A) activates AMPK, and stabilizes p53 in HEK293 cells, and induce a higher cytotoxicity than 2-deoxy-D-glucose in H1299 cell line. Ó 2016 Elsevier Ltd. All rights reserved.
Metabolic reprogramming of cancer cells is emerging as a key process for their survival and proliferation.1 Most cancer cells are programmed to increase glycolysis, a phenomenon known as Warburg effect, without maximizing ATP synthesis in mitochondria.1,2 The increased glycolytic intermediates, such as glucose 6-phosphate or 3-phosphoglycerate, are directed to the pentose phosphate pathway (PPP) and serine biosynthesis, respectively, to provide ribose, NADPH, and other building blocks for anabolism of proliferating cells.1,2 The Krebs cycle intermediates are replenished by uptake of glutamine, which leads to increased production of NADPH for lipid biosynthesis and precursors of amino acids.2 Importantly, this metabolic reprogramming is essential for maintaining the redox homeostasis of cancer cells.3 NADPH produced by the above pathways is an essential reducing equivalent for protecting against reactive oxygen species (ROS) in cancer cells.1 Furthermore, serine biosynthesis increases glycine availability, which is necessary for glutathione biosynthesis.1,3 In addition, mutation of oncogenes, such as MYC and KRAS, is closely interconnected with metabolic reprogramming and redox regulation.3,17 This indicates the intricate control of metabolism and redox balance for survival of cancer cells. This metabolic tuning and redox control of cancer cells is vital for their proliferation, but also provides selective strategies for treating cancer.1,4,5 First, a high glycolytic activity of cancer cells was targeted by antiglycolytic drugs, including 2-deoxy-D-glucose (2-DG). 2-DG, a glucose analog that is devoid of 2-hydroxyl group, is converted to 2-DG 6-phosphate, which blocks glycolysis by ⇑ Corresponding author. http://dx.doi.org/10.1016/j.bmcl.2016.01.063 0960-894X/Ó 2016 Elsevier Ltd. All rights reserved.
inhibiting hexokinase and phosphoglucose isomerase.4 2-DG also interferes with N-glycosylation of proteins in endoplasmic reticulum (ER).4 Therefore, treatment of 2-DG decreases ATP synthesis, increases cellular ROS, causes ER stress, and increases the ratio of oxidized glutathione over reduced glutathione, resulting in cancer cell death.4 2-DG has been evaluated in clinical trials of patients with several tumors, mostly in combinatorial therapy with cytotoxic drugs.4,14,23 In addition, 3-bromopyruvate is a promising anti-glycolytic drug that blocks pyruvate kinase in glycolysis.5 In a similar context, glucose starvation induced the selective cytotoxicity in several cancer cells relative to their isogenic non-transformed cells.6 Secondly, increased generation of ROS and an intricate control of redox status in cancer cells relative to normal cells provide a basis of designing ROS-inducing or ROS-activated electrophilic anticancer agents.7 Phenethyl isothiocyanate (PEITC), piperlongumine (PL), sulforaphane, and N-(2,5-dihydroxyphenyl) acetamide-based agents that retain potential electrophilic reactive groups have shown the selective cytotoxicity to malignant cells relative to non-transformed cells.7–11,15,21 This is due to vulnerability of cancer cells to reach lethal ROS levels above the antioxidant protective threshold.7,12 Recent library screening identified several ROS enhancing compounds among which mostly electrophilic compounds were found to induce cytotoxicity.8 Overall, this suggests that compounds with the property of blocking metabolic activity and/or enhancing ROS may be a potential therapeutic strategy targeting cancer cells. Here we report N-aryl glycoside small molecules that enhance ROS, perturb the cellular energetics, and induce the cytotoxicity in H1299, a non-small lung cancer cell line.
F. T. Ndombera et al. / Bioorg. Med. Chem. Lett. 26 (2016) 1452–1456
Because various carbohydrates can be used for cellular bioenergetics or glycosylation in proteins, we sought to identify a carbohydrate derivative, such as 2-DG, that induces the cellular stress by increasing the level of ROS and inducing cytotoxicity in cancer cells. Interestingly, the cellular entry of various carbohydrates other than glucose can be facilitated by glucose transporter, GLUT-1, that is highly expressed for enhanced glycolysis in cancer cells.13 The substitution at the anomeric carbon (C1), carbon 2 (C2) and carbon 6 (C6) of glucose is highly tolerated by GLUT-1.13 To increase the potential candidate compounds, we synthesized a library of N-aryl-glycosides that are assembled by simple N-glycosylation of 12 carbohydrates and 8 aniline derivatives (Scheme 1). N-Glycosides are more stable than O-glycosides, and N-glycosylation would provide diverse structural modifications. Previously, b-N-aryl-glycosides were synthesized by simple heating at low temperatures (40 °C) in a slightly acidic buffer at pH 6.5 in one step in which excess amount of carbohydrate over aniline led to selective synthesis of b-isomers.16 To streamline the method, we modified this one step N-glycosylation in aqueous buffer (20% DMSO) at higher temperatures (70 °C). After heating individual carbohydrate (A–L) and aniline (1–8) (Scheme 1) in a ratio of 16:1 in phosphate buffer (pH 6.5), the mixture was directly injected to a semi-preparative HPLC column for purification of products.24 After LC–MS analysis of individual products, 46 compounds that had a high purity (>90%) were selected for the subsequent screening assay. The purity of individual compounds is shown in Table S1. It is noted that many of reactions, including
Scheme 1. Synthesis and building blocks of N-aryl glycosides.
1453
one with 2-deoxy-D-ribose (E) and 4-chloro-aniline (2), results in only one major peak of N-aryl glycoside (Fig. 1A), but a few reactions, including one with D-ribose (I) and 2-aminocoumarin (4), gave two peaks of identical mass in HPLC (Fig. 1B). Two peaks are likely to correspond to a- and b-anomers of given products, which were designated as a and b, respectively, without further characterization at this stage. To identify N-aryl glycosides that increase the ROS level, we used a cell-permeable fluorescent probe, 20 ,70 -dichlorodihydrofluorescein diacetate (DCF-DA), for detection of ROS increase in nonsmall cell lung carcinoma H1299 cells. DCF-DA readily crosses cell membranes and is retained in cells upon deacetylation by esterases.18,19 It is then oxidized by various species of ROS to emit fluorescence.18 With DCF assay, we screened a library of N-aryl glycosides (46 compounds in 100 lM, in duplicates) for an increase of ROS in H1299 cells. The fluorescence increases of individual N-aryl glycosides in comparison to a non-treated control are summarized in Figure 2A. A majority of compounds induced no significant change of ROS. However, six compounds (H8, G8, B8, F6, C4b, and K8) induced 2-fold or higher increases (Fig. 2A). Interestingly, it is notable that several of 1-amino-naphthalene (8)-containing compounds were identified to increase ROS levels. For example, H7a and H7b that contain aniline (7) did not induce ROS significantly in comparison to H8 that contains 1-amino-nathpthalene (8). In comparison, about 1.5-fold increase of ROS level was observed by treatment of 2-DG (10 mM) (Fig. S1). We then re-synthesized several 8-series compounds, including A8, C8, H8, and K8, for further evaluation.24,25 After re-synthesis, we were able to determine the anomeric configuration of compounds by NMR analysis (Figs. S4–S7).22 These four compounds were then evaluated for an increase of ROS levels in a dose-dependent manner (Fig. 3A). H8 and K8 showed a dose-dependent 2-fold increase of ROS level within 100 lM concentration. Notably, H8 and K8 contain a similar carbohydrate structure with D-arabinose (H) and D-xylose (K), respectively, which are aldopentoses with a reversed stereochemistry at 2-OH and 3-OH positions. On the other
Figure 1. HPLC chromatograms of selected N-aryl glycosides. (A) The reaction with 2-deoxy-D-ribose (E) and 4-chloro-aniline (2) results in one peak of an expected mass. (B) The reaction with D-ribose (I) and 2-amino-coumarin (4) resulted in two peaks of an expected mass, which were designated as a and b.
1454
F. T. Ndombera et al. / Bioorg. Med. Chem. Lett. 26 (2016) 1452–1456
Figure 2. Evaluation of N-aryl glycosides for ROS increase and cytotoxicity in H1299 cells. (A) DCF-DA assay was used for measuring the fluorescence intensity after incubation of individual N-aryl glycoside (100 lM) (Fi) versus non-treated control (Fo) after 5-h incubation in H1299 cells. (B) MTT assay was used for analyzing the cytotoxicity of individual N-aryl glycoside (100 lM) (Vi) versus non-treated control (Vo) after 48-h incubation in H1299 cells.
hand, A8 and C8 did not induce a significant increase of ROS level (Fig. 3A). A8 contains D-glucose, and it did not induce a significant increase of ROS level, suggesting that a pentose rather than a hexose may be important. Also, there was almost no increase of ROS level by C8 that contain L-arabinose, showing the differential effect of the stereochemistry of D-arabinose in H8 versus L-arabinose in C8. In control experiments, D-arabinose (H), D-xylose (K) or 1naphthylamine (8) alone did not induce an increase of ROS level (Fig. 3B). We then measured the cytotoxicity of the same 46 compounds on cell growth and viability of H1299 cells using MTT assay. N-Aryl glycosides (100 lM in duplicate) were incubated in H1299 for 48 h, and MTT assay results are summarized in Figure 2B. The comparison between the ROS enhancing property and the cytotoxicity indicated that these data do not appear to be significantly correlated (Fig. 2). Out of five compounds with highest cytotoxicity (H8, K8, H7a, H7b, C7) (Fig. 2B), two compounds (H8 and K8) induced a high level of ROS while three compounds (H7a, H7b, and C7) induce almost no increase of ROS (Fig. 2A). We then reevaluated cytotoxicity of H8 and K8 in comparison to hydrogen peroxide and 2-DG. Hydrogen peroxide reduced viability with IC50 value at 60 lM after a 48 h incubation (Fig. 4A). 2-DG showed about 80% viability when treated at 500 lM concentration (Fig. 4A), which is similar to previous Letters.23 In identical conditions, H8 and K8 reduced the viability with IC50 values at about 75 lM (Fig. 4B), which is similar or much higher cytotoxicity versus hydrogen peroxide or 2-DG, respectively. Because the IC50 values of H8 and K8 are relatively high, we further modified K8 by acetylating all hydroxyl groups.25 We thought that acetylated derivatives of K8 may enter the cells more efficiently than their non-acetylated forms, which may increase the potency of compounds. Previously, acetylated 18F-2-deoxy-D-glucose rapidly penetrated the cell membrane as a result of its high lipophilicity and was metabolized to non-acetylated 2-DG in cancer cells.20
Interestingly, the acetylated derivative of K8 (K8A) reduced the viability of H1299 with IC50 at 30 lM, which is about 2.5-fold increase versus K8 (Fig. 4C). In controls, D-xylose (K) and its acetylated form (KA) did not induce a significant cytotoxicity in H1299 cells (Fig. 4D). K8A also induced about 1.5-fold increase of ROS level within 50 lM (Fig. S2). In order to see the potential mechanism of K8A for cytotoxicity, we evaluated the activation of AMP-activated protein kinase (AMPK) upon treatment of K8A (Fig. 5). AMPK is a stress-sensing kinase that is activated by phosphorylation when AMP level increases. AMPK phosphorylation halts protein synthesis and cell cycle while inducing p53 stabilization.23 2-DG that decrease ATP synthesis and increase cellular ROS was shown to phosphorylate AMPK and stabilize p53 for apoptosis. The treatment of 2-DG to HEK293 cells increased AMPK phosphorylation in a dose-dependent manner (Fig. 5). In a similar manner, the incubation of K8A within 100 lM concentration led to a dose-dependent increase of AMPK phosphorylation (Fig. 5). In addition, p53, which is downstream of AMPK, is stabilized upon treatment of K8A (Fig. 5). Because 2-DG is also known to induce ER stress by interfering with N-glycosylation of proteins, we further evaluated K8 and K8A for induction of glucose-regulated protein 78 (GRP78/BiP), which is a hallmark of ER stress.26 Incubation of K8A and K8 at 200 lM concentration in HEK293 cells showed the induction of GRP78 (Fig. S3). This indicates that K8A may be involved in disrupting cellular energetics and inducing ER stress for its cytotoxicity. In summary, we have developed a library of simple N-aryl glycoside derivatives. Cellular ROS and cytotoxicity assays of this Naryl-glycoside library led to identifying the potential candidate (K8A) that activates AMPK, and stabilizes p53 in HEK293 cells, and induce cytotoxicity with a higher potency than 2-DG in H1299 cells. Evaluation of K8A derivatives for cytotoxicity, the mechanism of action, and target identification will be further investigated in future studies.
F. T. Ndombera et al. / Bioorg. Med. Chem. Lett. 26 (2016) 1452–1456
1455
Figure 3. ROS induction assays of N-aryl glycosides containing 1-amino-naphthalene (8). DCF-DA assay was used for measuring the fluorescence intensity (Fi) after 5-h incubation of A8, C8, H8, and K8 (A) or H, K, and 8 (B) in H1299 cells with an increasing concentration (0–100 lM) versus a non-treated control (Fo).
Figure 4. The cytotoxicity of N-aryl glycosides H8, K8 and K8A. MTT assay was used for measuring the cytotoxicity (Vi) of hydrogen peroxide and 2-DG (A), H8 and K8 (B), K8A (C), K and KA (D) versus a non-treated control (Vo) by incubating individual compounds for 48 h in H1299 cells.
1456
F. T. Ndombera et al. / Bioorg. Med. Chem. Lett. 26 (2016) 1452–1456
Figure 5. K8A induces a dose-dependent induction of AMPK phosphorylation and p53 stabilization similar to 2-DG. K8A or 2-DG was incubated for 16 h in HEK293 cells, and lysates were probed for Western blotting with indicated antibodies.
Acknowledgments This work was supported by Wayne State University Start-up fund and WSU University Research Grant. Supplementary data Supplementary data (NMR spectra and other characterization of compounds) associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bmcl.2016.01.063. References and notes 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
Schulze, A.; Harris, L. A. Nature 2012, 491, 364. Vander Heiden, G. W.; Cantley, C. L.; Thompson, B. C. Science 2009, 324, 1029. Ward, S. P.; Thompson, B. Cancer Cell 2012, 21, 297. Zhang, D.; Li, J.; Wang, F.; Hu, J.; Wang, S.; Sun, Y. Cancer Lett. 2014, 355, 176. Galluzzi, L.; Kepp, O.; Vander Heiden, G. W.; Kroemer, G. Nat. Rev. Drug Disc. 2013, 12, 829. Palorini, R.; Cammarata, F. P.; Balestrieri, C.; Monestiroli, A.; Vasso1, M.; Gelfi, C.; Alberghina, L.; Chiaradonna, F. Cell Death Dis. 2013, 4, e732. Pelicano, H.; Carney, D.; Huang, P. Drug Resist. Update 2004, 7, 97. Adams, J. D.; Boskovic, V. Z.; Theriault, R. J.; Wang, J. A.; Stern, M. A.; Wagner, K. B.; Shamji, F. A.; Schreiber, L. S. ACS Chem. Biol. 2013, 8, 923. Gorrini, C.; Harris, S. I.; Mak, W. T. Nat. Rev. Drug Disc. 2013, 12, 931. Wang, J.; Yi, J. Cancer Biol. Ther. 2008, 12, 1875. Ivanova, D.; Bakalova, R.; Lazarova, D.; Gadjeva, V.; Zhelev, Z. Adv. Clin. Exp. Med. 2013, 22, 899. Trachootham, D.; Alexandre, J.; Huang, P. Nat. Rev. Drug Disc. 2009, 8, 579. Calvaresi, C. E.; Hegenrother, J. P. Chem. Sci. 2013, 4, 2319.
14. Raez, E. L.; Papadopoulus, K.; Ricart, D. A.; Chioreau, G. E.; Dipaola, S. R.; Stein, N. M.; Lima, R. M. C.; Schlesselman, J. J.; Tolba, K.; Langmuir, K. L.; Kroll, S.; Jung, T. D.; Kurtoglu, M.; Rosenbl, J. Cancer Chemother. Pharmacol. 2013, 71, 523. 15. Vadukoot, K. A.; Abdulsalam, F. S.; Wunderlich, M.; Pullen, D. E.; LanderoFigueroa, J.; Mulloy, C. J.; Merino, J. E. Bioorg. Med. Chem. 2014, 22, 6885. 16. Bridiau, N.; Benmansour, M.; Legoy, D. M.; Maugard, T. Tetrahedron 2007, 63, 4178. 17. Shen, L.; O’Shea, M. J.; Kaadige, R. M.; Cunha, S.; Wilde, R. B.; Cohen, L. A.; Welm, L. A.; Ayer, A. D. Proc. Natl. Acad. Sci. U.S.A. 2015, 112, 5425. 18. Wang, H.; Joseph, A. J. Free Radical Biol. Med. 1999, 27, 612. 19. Winterbourn, C. C. Biochim. Biophys. Acta 2014, 1840, 730. 20. Waki, A.; Fujibayashi, Y.; Magata, Y.; Yokohama, A.; Sadato, N.; Tsuchida, T.; Ishii,; Yonekura, Y. J. Nucl. Med. 1998, 2, 245. 21. Ahn, Y. H.; Hwang, Y.; Liu, H.; Wang, X. J.; Zhang, Y.; Stephenson, K. K.; Boronina, T. N.; Cole, R. N.; Dinkova-Kostova, A. T.; Talalay, P.; Cole, P. A. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 9590. 22. NMR spectra of A8, C8, H8, and K8 are shown in Figures S4–S7. The structures are shown in Figure 3. The anomeric stereochemistry was assigned by examining the coupling constant between anomeric proton and adjacent C– H: A8 (4.80 ppm, J = 7.9 Hz), C8 (4.73 ppm, J = 8.2 Hz), H8 (4.70 ppm, J = 8.0 Hz), K8 (4.67 ppm, J = 8.2 Hz). The structure of C8, H8, and K8 are shown as pyranoses, given that arabinose prefers pyranose over furanose in water. 23. Sahra, B. I.; Laurent, K.; Giuliano, S.; Larbret, F.; Ponzio, G.; Gounon, P.; Marchand-Brustel, L. Y.; Giorgetti-Peraldi, S.; Cormont, M.; Bertolotto, C.; Deckert, M.; Auberger, P.; Tanti, J.; Bost, F. Cancer Res. 2010, 70, 2465. 24. To prepare the N-aryl glycosides, a reaction mixture of 1 equiv aniline derivative (1–8) in 20% dimethyl sulfoxide (DMSO) was added to 16 equiv of preheated carbohydrate (A–L) in potassium phosphate buffer (pH 6.5) in a 4 mL glass vial and incubated at 70 °C for 16 h. The reactant mixture was filtered twice using cotton wool to remove any solid particles and a HPLC injection was made to purify the product. Products were purified by a semipreparative C-18 reversed-phase column HPLC with a gradient elution method (a mobile phase composition of water and acetonitrile, a flow rate of 10 mL/ min, injection volume of 3–4 mL, run time of 75 min and detection wavelengths of 214 nm and 254 nm). The yields after HPLC purification are in the range of 5–15%. Products were eluted at retention times between 30– 35 min and collected, flash-frozen in liquid nitrogen, and lyophilized under vacuum. Identity of individual product was confirmed by LC–MS ESI (Table S1). The products were then dissolved in DMSO and stored in 20 °C freezer for assays. 25. To prepare K8A, D-xylose (500 mg, 3.33 mmol) and 1-amino-napthalene (525 mg, 3.66 mmol) was refluxed in ethanol for 8 h to obtain brown sparkly crystalline solid (730 mg). The crystalline solid (500 mg) was reacted with acetic anhydride (8.17 mmol) in pyridine (0.79 mL, 9.81 mmol) in an ice bath for 3 h and in the cold room for another 13 h. The reactant mixture was then quenched by pouring into ice/water, extracted, washed with aq sodium bicarbonate, water and brine, and purified by column chromatography (1:4 ethyl acetate/hexane solvent conditions). After purification, a yellow solid K8A (622 mg, 1.55 mmol) was obtained and it was further recrystallized from ethanol to provide the white K8A crystals. 1H NMR (400 MHz, CDCl3) d 7.80 (t, J = 4.8 Hz, 1H), 7.72 (t, J = 4.8 Hz, 1H), 7.46 (t, J = 4.4 Hz, 2H), 7.37 (q, J = 8 Hz, 2H), 6.79 (d, J = 6.8 Hz, 1H), 5.44 (t, J = 9.6 Hz, 2H), 5.17 (t, J = 9.2 Hz, 1H), 5.13– 5.03 (m, 1H), 4.84 (d, J = 9.2 Hz, 1H), 4.16 (q, J = 5.6 Hz, 1H), 3.51 (t, J = 10.8 Hz, 1H), 2.1 (s, 3H), 2.07 (s, 3H), 2.06 (s, 3H). 26. Lee, A. S. Methods 2005, 35, 373.