Novel polyphenols that inhibit colon cancer cell growth affecting cancer cell metabolisms

Article abstract of DOI:10.1124/jpet.118.248278

New series of polyphenols with a hydrophilic galloyl-based head and a hydrophobic N-acyl tail, linked through a serinol moiety, have been synthesized and tested against colon cancer cell growth. Our structure activity relationship studies revealed that galloyl moieties are essential for growth inhibition. Moreover, the length of the N-acyl chain is crucial for the activity. Introduction of a (Z) double bond in the acyl chain increased the anticancer properties. Our findings demonstrate that 16, the most potent compound within this series, has inhibitory effects on colon cancer cell growth and metabolism (glycolysis and mitochondrial respiration) at the same time that it activates 59AMP-activated kinase (AMPK) and induces apoptotic cell death. Based on these results, we propose that 16 might reprogram colon cancer cell metabolism through AMPK activation. This might lead to alterations on cancer cell bioenergy compromising cancer cell viability. Importantly, these antiproliferative and proapoptotic effects are selective for cancer cells. Accordingly, these results indicate that 16, with an unsaturated C18 chain, might be a useful prototype for the development of novel colon cancer cell growth inhibitors affecting cell metabolism.

Full text of DOI:10.1124/jpet.118.248278

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
NOVEL POLYPHENOLS THAT INHIBIT COLON CANCER CELL GROWTH
AFFECTING CANCER CELL METABOLISM
Marta Gómez de Cedrón, Teodoro Vargas, Andrés Madrona, Aranza Jiménez, María-Jesús Pérez-Pérez, José-
Carlos Quintela, Guillermo Reglero, Ana San-Félix, Ana Ramírez de Molina
Molecular Oncology and Nutritional Genomics of Cancer, IMDEA-Food Institute, CEI UAM+CSIC, Madrid 28049, Spain (M.G.d.C.,
T.V., G.R., A.R.d.M.); Instituto de Química Médica (IQM-CSIC), Juan de la Cierva 3, Madrid 28006, Spain (A.M., A.J., M.J.P.-P., A.
S.-F.); Natac Biotech S.L., Parque Científico de Madrid, Campus de Cantoblanco, Madrid 28049, Spain (J.-C.Q); Actual address:
Novartis Farmacéutica S.A. Gran Vía de les Corts Catalanes, 764, Barcelona, 08013, Spain (A.M.); Actual address: Agencia
Española de Medicamentos y productos Sanitarios, Campezo 1, Edificio 8, Madrid 28022 (A.J.)
1

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
Running Title
Polyphenols targeting cancer cell metabolism
Corresponding author: Ana Ramírez de Molina. IMDEA Food Institute, Crta. Cantoblanco 8, 28049,
Madrid, Spain. Tel.: +34 912796957; fax: +34 911880756.
E-mail address: ana.ramirez@imdea.org
Corresponding author: Ana San-Félix. Instituto de Química Médica (IQM-CSIC), Juan de la Cierva 3, 28006,
Madrid, Spain. Tel.; +34 912587617; +34 915644853
E-mail address: anarosa@iqm.csic.es
Number of text pages: 36
Number of tables: 2
Number of figures: 9
Number of references: 58
Number of words in the Abstract: 178
Number of words in the Introduction: 682
Number of words in the Discussion: 641
Abbreviations: DIPEA, N,N-diisopropylethylamine; DMAP, dimethylaminopyridine; ECAR, extra-
cellular acidification rate; FCCP, cyanide-4-(trifluoromethoxy)phenylhydrazone; HATU, (1-
[bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxid hexafluorophosphate); MOM,
methoxymethyl; OCR, oxygen consumption rate; TEA, trimethylamine; TFH; tetrahydrofurane
Recommended section assignment: Drug Discovery and Traslational Medicine
Address correspondence to: Ana Ramírez de Molina. IMDEA Food Institute, Crta. Cantoblanco 8, 28049,
Madrid, Spain. Tel.: +34 912796957; fax: +34 911880756.
E-mail address: ana.ramirez@imdea.org
2

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
Abstract
New series of polyphenols with a hydrophilic galloyl based “head” and a hydrophobic N-acyl “tail”,
linked through a serinol moiety, have been synthesized and tested against colon cancer cell growth. Our
structure activity relationship studies revealed that galloyl moieties are essential for growth inhibition.
Moreover, the length of the N-acyl chain is crucial for the activity. Introduction of a (Z) double bond in the
acyl chain increased the anti-cancer properties. Our findings demonstrate that 16, the most potent compound
within this series, has inhibitory effects on colon cancer cell growth and metabolism (glycolysis and
mitochondrial respiration) at the same time that activates AMPK and induces apoptotic cell death. Based
on these results we propose that 16 might reprogram colon cancer cell metabolism through AMPK
activation. This might lead to alterations on cancer cell bioenergy compromising cancer cell viability.
Importantly, these anti-proliferative and pro-apoptotic effects are selective for cancer cells. Accordingly,
these results indicate that 16, with an unsaturated C18 chain, might be a useful prototype for the
development of novel colon cancer cell growth inhibitors affecting cell metabolism.
Visual Abstract
3

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
1. Introduction
In recent years, the understanding of differences in cell metabolism between normal and cancer cells
has been considered essential in the design and development of new anticancer drugs. In this respect,
proliferating cancer cells exhibit a higher dependency on aerobic glycolysis than normal cells (Lopez-
Lazaro, 2010; Marchetti et al., 2014), a phenomenon known as the “Warburg effect” (Warburg, 1956). This
metabolic difference has led to the hypothesis that inhibition of glycolysis may preferentially kill cancer
cells and may have significant therapeutic implications (Pelicano et al., 2006). Indeed, several glycolytic
inhibitors (2-deoxy-D-glucose, lonidamine, 3-bromopyruvate and dichloroacetate) have shown anticancer
activities both in vitro and in vivo, and some of them have entered clinical trials. However, they show
important drawbacks (low potency, instability…) that may limit their use as therapeutic drugs. Thus,
development of new generations of glycolytic inhibitors with high potency, chemical stability and good
safety profiles represents an important task in the anti-cancer field.
In addition, cancer cells and normal cells metabolize oxygen (O2) differently. Normal cells use oxygen
(O2) to generate energy in the form of ATP at mitochondria, meanwhile cancer cells use O2 to generate high
levels of reactive oxygen species (ROS), and aerobic glycolysis is used to provide energy and satisfy
structural demands. This difference in the metabolism of O2 can be also exploited to kill cancer cells
selectively (Marchetti et al., 2014).
Lately, natural and synthetic polyphenols have become a focus of interest on cancer chemoprevention
and chemotherapy. In fact, epidemiological studies suggest that the intake of polyphenols may protect
against tumor growth (Arts and Hollman, 2005; Neuhouser, 2004), particularly gastrointestinal cancers
(Pierini et al. 2008). The growth inhibition induced by polyphenols in tumor cells include several inhibitory
mechanisms such as the activation of caspases, cell cycle arrest and the inhibition of survival or proliferative
signaling pathways (NF-kB, JAK/STAT) among others (Gonzalez-Vallinas et al., 2013; Hadi et al., 2007;
Kubatka et al., 2016; Zhang et al., 2015).
Moreover, there are many examples of dietary polyphenols affecting different metabolic pathways of
cancer cells (Cerella et al., 2013). Flavonoids, quercetin, oleuropein, hydroxytyrosol and epigallocatechin-
3-gallate (ECGC) have been shown to affect glycolysis-related factors, meanwhile others like curcumin, or
resveratrol target mitochondrial metabolism inducing cell death and/or apoptosis in a variety of cancer
models (Boyer et al., 2012; Kim et al., 2009; Teiten et al., 2010). In many cases, the underlying mechanism
is associated to 5’AMP-activated kinase (AMPK) activation (Hung, et al., 2012; Shin et al., 2009).
4

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
AMPK is a central metabolic sensor that regulates anabolism or catabolism in response to energetic
demands and/or oxidative stress. When activated, AMPK inhibits the main anabolic pathways that promote
cell growth (DG., 2014; Li et al., 2015; Hardie, 2014). In cancer, where the energetic demands are elevated
to support rapid cell growth and division, AMPK activators can be suitable candidates for therapeutic
intervention. In fact, AMPK is emerging as a promising target for cancer prevention and therapy (Luo et
al., 2010).
As a part of an ongoing work in our laboratories directed to the synthesis and biological evaluation of
polyphenols as antiviral and antibacterial agents (Flores et al., 2014; Rivero-Buceta et al., 2015a; Rivero-
Buceta et al., 2015b), we prepared molecules of general formula I, with a hydrophilic “head”, composed of
two galloyl (3,4,5-trihydroxybenzoyl) units (R1), and a hydrophobic “tail” (R) composed of an N-acyl-
chain, linked through a serinol moiety (Figure 1). Systematic modifications have been performed in these
compounds. First, the effect of the N-acyl chain length of the hydrophobic tail was analyzed by introduction
of different aliphatic acyl residues with variable number of carbons (C2, C7, C11 and C17). The effect of
unsaturation (one or two double bonds) on the hydrophobic “tail” was also analyzed. Modifications on the
“head” were also performed by replacing the galloyl moieties, with 3 OHs, by other phenolic units with 2
OHs. Finally, ethanolamine, instead of serinol, was used as a linker.
For all these compounds inhibition of colon cancer cell proliferation and colon cancer energetic
metabolism (glycolysis and oxygen consumption), together with activation of AMPK and induction of
apoptotic cell death have been determined and are herein described.
5

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
2. Materials and Methods
2.1. Synthesis
2.1.1. General Methods
Commercial reagents and solvents were used as received from the suppliers without further purification
unless otherwise stated. Dichloromethane was dried prior to use by distillation from CaH2 and stored over
Linde type activated 4Å molecular sieves.
Analytical thin-layer chromatography (TLC) was performed on aluminum plates pre-coated with silica
gel 60 (F254, 0.25 mm). Products were visualized using an ultraviolet lamp (254 and 365 nm) or by heating
on a hot plate (approx. 200 ºC), directly or after treatment with a 5% solution of phosphomolybdic acid or
vanillin in ethanol.
The compounds were purified by: a) flash column chromatography on silica gel (60 Merck 230-400
mesh), b) trituration with cold diethyl ether and dichloromethane.
NMR spectra (1H, 13C NMR) were recorded on a Varian UNIT INOVA-300 (300 MHz), Bruker
AVANCE 300 (300 and 75 MHz), Varian INOVA-400 (400 and 100 MHz), Varian MERCURY-400 (400
and 100 MHz) and Varian-500 (500 and 125 MHz) spectrometers, using (CD3)2SO and CDCl3 as solvents.
Chemical shift (δ) values are reported in parts per million (ppm) relative to tetramethylsilane (TMS) in 1H
and CDCl3 (δ = 77.0) in 13C NMR. Coupling constant (J values) are reported in hertz (Hz) and multiplicities
of signals are indicated by the following symbol: s (singlet), d (doublet), t (triplet), q (quadruplet), m
(multiplet) and bs (broad singlet). Some two-dimensional spectra (COSY, HSQC and HMBC) were
performed to identify the structure.
Final compounds were lyophilized using a Telstar 6-80 system.
2.1.2. General procedure for the synthesis of serinol derivates 2-6
A stirred solution of 2-amino-1,3-propanediol (serinol) 1 (1 eq) and triethylamine (TEA) in MeOH (10
mL) was cooled at -20 ºC and treated drowpwise with a solution of the corresponding acyl chloride (1.1 eq)
in THF (5 mL). The reaction mixture was allowed to warm to room temperature and stirred overnight. Then
it was poored into brine and extracted with dichloromethane (3 x 10 mL). The combined organic phases
were washed with brine, dried over MgSO4 and concentrated under reduced pressure. The residue was
6

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
purified by flash column chromatography on silica gel using as eluent EtOAc/MeOH (10:1 to 7:1) to
provide the corresponding N-acyl serinol derivatives.
2.1.3. N-propanoylserinol (2)
According to the general procedure serinol (150 mg, 1.64 mmol) was treated with propanoyl chloride
(1.86 mmol, 0.16 mL) and triethylamine (0.4 mL) to give the title compound (202 mg, 81%) as an
amorphous white solid. 1H NMR (400 MHz, DMSO-d6) δ 7.40 (d, J = 8.1 Hz, 1H, NH), 4.55 (t, J = 5.5 Hz,
2H, OH), 3.74-3.64 (m, 1H, CH), 3.37 (t, J = 5.6 Hz, 4H, CH2OH), 2.06 (t, J = 7.4 Hz, 2H, CH2C=O), 1.51
– 1.42 (m, 2H, CH2), 1.30-1.15 (m, 8H, CH2), 0.86 (t, J = 6.8 Hz, 3H, CH3).
2.1.4. N-dodecanoyl serinol (3)
According to the general procedure serinol (200 mg, 2.19 mmol) was treated with dodecanoyl chloride
(625 mg, 2.85 mmol) and triethylamine (0.6 mL) to give the title compound (478 mg, 80 %) as an
amorphous white solid. Characterization of this compound is consistent with those found in the literature
(Boyer et al., 2012; Kim et al., 2009).
2.1.5. N-octadecanoyl serinol (4)
According to the general procedure serinol (150 mg, 1.64 mmol) was treated with octadecanoyl chloride
(545.24 mg, 1.80 mmol) and triethylamine (0.4 mL) to give the title compound (421 mg, 75%) as an
amorphous white solid. Characterization of this compound is consistent with those found in the literature
(Bieberich et al., 2002a; Merg et al., 2015; Shulz and Robins, 2004).
2.1.6. (N-oleoyl serinol) (5)
Serinol (55 mg, 0.60 mmol) was treated with 9-octadecenoyl chloride (oleoyl chloride) (198.6 mg, 0.66
mmol,) and triethylamine (0.1 mL) to give the title compound (149 mg, 70%) as an amorphous white solid.
Characterization of this compound is consistent with those found in the literature (Bieberich et al., 2002a;
Landau and Siegel, 2014; Osornio et al., 2012).
2.1.7. N-linoleoyl serinol (6)
7

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
Serinol (55 mg, 1.64 mmol) was treated with linoleoyl chloride (1.81 mmol, 0.6 mL) and triethylamine
(0.1 mL) to give the title compound (169 mg, 79 %) as an amorphous white solid. Characterization of this
compound in the literature was incomplete (Oette and Tschung, 1980). Missing data (NMR) was now
included. 1H NMR (400 MHz, DMSO-d6) δ 7.41 (d, J = 8.1 Hz, 1H, NH), 5.35 – 5.27 (m, 2H, CH=CH),
4.56 (t, J = 5.5 Hz, 2H, OH), 3.74 – 3.63 (m, 1H, CH, NH), 3.38 (t, J = 6.0 Hz, 4H, CH2OH), 2.05 (t, J =
7.5 Hz, 2H, CH2C=O), 2.01 – 1.92 (m, 4H, CH2HC= CHCH2), 1.50 – 1.42 (m, 2H, CH2), 1.34 – 1.14 (m,
22H, CH2), 0.85 (t, J = 6.8 Hz, 3H, CH3).
2.1.8. 3,4,5-tris(methoxymethoxy)benzoic acid (7)
To a cooled (0 ºC) mixture of methylgallate (500 mg, 2.71 mmol) and DIPEA (1.6 mL, 8.96 mmol) in
dichloromethane (20 mL) was added methoxymethyl chloride (MOM-Cl) (0.7 mL, 8.97 mmol). The
reaction was stirred at 0 ºC for 20 min and then at room temperature overnight. The solvent was evaporated
to dryness and the residue was dissolved in ethyl acetate (20 mL) and washed successively with a saturated
solution of ammonium chloride (3 x 20 mL) and water (3 x 20 mL). The organic layer was dried over
anhydrous Na2SO4, filtered, and concentrated under vacuum. The residue was then purified by flash column
chromatography using hexane/EtOAc (6:1 to 2:1) as the eluent to give 543 mg (63 %) of methyl 3,4,5-
tris(methoxymethoxy)benzoate as a white amorphous solid that was used for the next step. 1H-RMN (300
MHz, CD3Cl) δ 7.51 (s, 2H, H-Ph), 5.24 (s, 4H, 2 x CH2MOM), 5.13 (s, 2H, 1 x CH2MOM), 3.83 (s, 3H,
OCH3), 3.51 (s, 3H, 1 x CH3MOM), 3.41 (s, 6H, 2 x CH3MOM).
To a solution containing the above mentioned methyl ester derivative (540 mg, 1.71 mmol) in THF (15
mL) at 0 ºC (ice-bath), a solution of LiOH·H2O (220 mg, 5.12 mmol) in water (10 mL) was added, and the
mixture was stirred at room temperature overnight. Then 1 N hydrochloric acid aqueous solution was added
to reach pH = 2, and volatiles were evaporated to dryness. The residue was dissolved in ethyl acetate (20
mL) and washed with H2O (3 × 20 ml). The organic phase was dried over anhydrous Na2SO4, filtered and
evaporated to dryness to give 488.7 mg (95%) of the title compound that was directly used for the next step.
Characterization of this compound is consistent with those found in the literature (Gazak et al., 2011).
2.1.9. General coupling procedure for the synthesis of the MOM protected intermediates 8-12
A solution of the MOM protected gallic acid 7 (Gazak et al., 2011) (2.2 eq), HATU (2.2 eq) and DMAP
(1.25 eq) in dry CH2Cl2 (10 ml) was stirred at room temperature for 15 minutes and then added to a second
8

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
solution, also stirred for 15 min, containing the corresponding serinol derivate (1eq) and DMAP (1.25 eq)
in dry CH2Cl2 (10 ml). The mixture was stirred at room temperature overnight and then evaporated to
dryness. The residue was dissolved in ethyl acetate (20 mL) and washed successively with saturated
solutions of citric acid (3 x 20 mL), NaHCO3 (3 x 20 mL) and NaCl (1 x 20 mL). The organic layer was
dried over anhydrous Na2SO4, filtered, and concentrated under vacuum. The residue was then purified by
flash column chromatography using hexane/EtOAc (10:1 to 1:1) as the eluent.
2.1.10. Bis-O-[3,4,5-tris(methoxymethoxy)galloy]-N-propanoyl serinol (8)
Following the general procedure, 2 (50 mg, 0.34 mmol) was treated with 7 (Gazak et al., 2011) (226.24
mg, 0.74 mmol), HATU (284.4 mg, 0.74 mmol) and DMAP (226.0 mg, 1.84 mmol) to afford 366 mg (75
1
%) of the title compound as an amorphous white solid. H-RMN (300 MHz, CD3Cl) δ 7.52 (s, 4H, Ph),
6.03 (d, J = 8.30 Hz, 1H, NH), 5.24 (s, 8H, 4 x CH2MOM), 5.22 (s, 4H, 2 x CH2MOM), 4.72 (m, 1H, CH),
4.46 (m, 4H, 2 x CH2), 3.61 (s, 6H, 2 x CH3MOM), 3.50 (s, 12H, 4 x CH3MOM), 2.23 (q, J = 7.56 Hz, 2H,
CH2), 1.13 (t, J = 7.43 Hz, 3H, CH3).
2.1.11. Bis-O-[3,4,5-tris(methoxymethoxy)galloy]-N-dodecanoyl serinol (9)
Following the general procedure, 3 (50 mg, 0.18 mmol) was treated with 7 (Gazak et al., 2011) (120
mg, 0.40 mmol), HATU (152 mg, 0.40 mmol) and DMAP (244.34 mg, 0.44 mmol) to afford 72 mg (47 %)
of the title compound as an amorphous white solid. 1H-RMN (300 MHz, CD3Cl) δ 7.57 (s, 4H, Ph), 6.01
(d, J = 8.30 Hz, 1H, NH), 5.24 (s, 8H, 4 x CH2MOM), 5.22 (s, 4H, 2 x CH2MOM), 4.73 (m, 1H, CH), 4.45
(m, 4H, 2 x CH2), 3.61 (s, 6H, 2 x CH3MOM), 3.51 (s, 12H, 4 x CH3MOM), 2.18 (t, J = 7.28 Hz, 2H, CH2),
1.59 (m, 2H, CH2), 1.25 (m, 16H, 8 x CH2), 0.87 (t, J = 6.92 Hz, 3H, CH3).
2.1.12. Bis-O-[3,4,5-tris(methoxymethoxy)galloy]-N-octadecanoyl serinol (10)
Following the general procedure, 4 (50 mg, 0.14 mmol) was treated with (Gazak et al., 2011) (90.68
mg, 0.30 mmol), HATU (114 mg, 0.30 mmol) and DMAP (42.8 mg, 0.34 mmol) to afford 100.48 mg (75
%) of the title compound as an amorphous white solid. 1H-RMN (300 MHz, CD3Cl) δ 7.56 (s, 4H, Ph), 6.02
(d, J = 8.30 Hz, 1H, NH), 5.24 (s, 8H, 4 x CH2MOM), 5.22 (s, 4H, 2 x CH2MOM), 4.72 (m, 1H, CH), 4.44
(m, 4H, 2 x CH2O), 3.60 (s, 6H, 2 x CH3 MOM), 3.50 (s, 12H, 4 x CH3 MOM), 2.05 (t, J = 7.2 Hz, 2H,
CH2CO), 1.44 (m, 2H, CH2), 1.17 (m, 28H, 14 x CH2), 0.81 (t, J = 6.92 Hz, 3H, CH3).
9

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
2.1.13. Bis-O-[3,4,5-tris(methoxymethoxy)galloy]-N-oleoyl serinol (11)
Following the general procedure, 5 (50 mg, 0.14 mmol) was treated with 7 (Gazak et al., 2011) (90.68
mg, 0.30 mmol), HATU (114 mg, 0.30 mmol) and DMAP (42.8 mg, 0.34 mmol) to afford 106 mg (81%)
of the title compound as an amorphous white solid. 1H-RMN (300 MHz, CD3Cl) δ 7.52 (s, 4H, Ph), 6.02
(d, J = 8.20 Hz, 1H, NH), 5.33 (m, 2H, CH=CH), 5.24 (s, 8H, 4 x CH2MOM), 5.22 (s, 4H, 2 x CH2MOM),
4.73 (m, 1H, CH), 4.46 (m, 4H, 2 x CH2), 3.61 (s, 6H, 2 x CH3MOM), 3.51 (s, 12H, 4 x CH3MOM), 2,19
(t, J = 8.55 Hz, 2H, CH2), 1.99 (m, 2H, CH2), 1.59 (m, 4H, 2 x CH2), 1.27 (m, 20H, 10 x CH2), 0.88 (t, J =
7.44 Hz, 3H, CH3).
2.1.14. Bis-O-[3,4,5-tris(methoxymethoxy)galloy]-N-linoleoyl serinol (12)
Following the general procedure, 6 (50 mg, 0.14 mmol) was treated with 7 (Gazak et al., 2011) (90.68
mg, 0.30 mmol), HATU (114 mg, 0.30 mmol) and DMAP (42.8 mg, 0.34 mmol) to afford 98.72 mg (76%)
of the title compound as an amorphous white solid. 1H-RMN (300 MHz, CD3Cl) δ 7.52 (s, 4H, Ph), 6.00
(d, J = 8.20 Hz, 1H, NH), 5.34 (m, 4H, 2 x CH=CH), 5.24 (s, 8H, 4 x CH2MOM), 5.22 (s, 4H, 2 x
CH2MOM), 4.73 (m, 1H, CH), 4.46 (m, 4H, 2 x CH2), 3.61 (s, 6H, 2 x CH3MOM), 3.50 (s, 12H, 4 x
CH3MOM), 2.76 (t, J = 7.70 Hz, 2H, CH2), 2.19 (m, 2H, CH2), 2.03 (m, 4H, 2 x CH2), 1.29 (m, 16H, 8 x
CH2), 0.88 (t, J = 8.01 Hz, 3H, CH3).
2.1.15. Synthesis of the deprotected serinol derivatives 13-17
A solution of the corresponding MOM protected derivative in methanol was treated with aqueous
hydrochloric acid (37%). The solution was stirred at room temperature overnight and then evaporated to
dryness. The residue was co-evaporated several times with dichloromethane and then purified by triturating
with cold diethyl ether. The precipitate that appeared was filtered and washed repeatedly with gently
quantities of cold dichloromethane and diethyl ether to afford the corresponding unprotected derivative as
a white solid.
2.1.16. Bis-O-galloy]-N-propanoyl serinol (13)
Following the general procedure, 8 (73 mg) was dissolved in methanol (1.5 mL) and treated with
aqueous HCl 37% (0.08 mL) to give 41.4 mg (90%) of the title compound as a white amorphous solid. 1H-
10

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
RMN (400 MHz, DMSO-d6) δ 8.08 (d, J = 8.10 Hz, NH), 6.96 (s, 4H, Ph), 4.39 (m, 1H, CH), 4.23 (m, 4H,
2 x CH2O), 2.09 (q, J = 7.60 Hz, 2H, CH2), 0.97 (t, J = 7.60 Hz, 3H, CH3). 13C-RMN (75 MHz, DMSO-
d6) δ 173.35, 165.75, 145.61, 138.66, 119.19, 108. 86, 62.98, 47.08, 38.31, 28.56, 10.01. MS (ES+): m/z
452.1 (M+H)+ Anal. C20H21NO11 (C, H, N, O).
2.1.17. Bis-O-galloy]-N-dodecanoyl serinol (14)
Following the general procedure, 9 (79 mg) was dissolved in methanol (1.5 mL) and treated with
1
aqueous HCl 37% (0.08 mL) to give 51.3 mg (95%) of the title compound as a white amorphous solid. H-
RMN (400 MHz, DMSO-d6) δ 9.25 (s, 6H, OH), 8.08 (d, J = 8.10 Hz, NH), 6.96 (s, 4H, Ph), 4.39 (m, 1H,
CH), 4.23 (m, 4H, 2 x CH2O), 2.09 (q, J = 7.60 Hz, 2H, CH2), 1.50 (m, 2H, CH2), 1,17 (m, 16H, 8 x CH2),
0.97 (t, J = 7.60 Hz, 3H, CH3). 13C-RMN (75 MHz, DMSO-d6) δ 165.05, 144.08, 137.99, 108.12, 62.33,
30.66, 28.39, 28.35, 28.24, 28.15, 28.09, 21.47, 13.33. MS (ES+): m/z 578.2 (M+H)+. Anal. C29H39NO11
(C, H, N, O).
2.1.18. Bis-O-galloy]-N-octadecanoyl serinol (15)
Following the general procedure, 10 (70 mg) was dissolved in methanol (1.5 mL) and treated with
aqueous HCl 37% (0.08 mL) to give 39 mg (92%) of the title compound as a white amorphous solid. 1H-
RMN (400 MHz, DMSO-d6) δ 9.13 (s, 6H, OH), 8.07 (d, J = 6.94 Hz, NH), 6.93 (s, 4H, Ph), 4.38 (m, 1H,
CH), 4.21 (m, 4H, 2 x CH2O), 2.05 (t, J = 7.20 Hz, 2H, CH2CO), 1.43 (m, 2H, CH2), 1.17 (m, 28H, 14 x
CH2), 0.81 (t, J = 7.08 Hz, 3H, CH3). 13C-RMN (75 MHz, DMSO-d6) δ 172.5, 165.7, 145.5, 138.6, 119.1,
108.7, 63.0, 46.9, 38.3, 35.4, 31.3, 29.1, 28.9, 28.8, 28.7, 28.6, 25.3, 22.1, 14.0. MS (ES+): m/z 662.3
(M+H)+. Anal. C35H51NO11 (C, H, N, O).
2.1.19. Bis-O-galloy]-N-oleoyl serinol (16)
Following the general procedure, 11 (71.4 mg) was dissolved in methanol (1.5 mL) and treated with
aqueous HCl 37% (0.08 mL) to give 45.9 mg (90%) of the title compound as a white amorphous solid. 1H-
RMN (400 MHz, DMSO-d6) δ 9.23 (s, 4H, OH), 8.96 (s, 2H, OH), 8.07 (d, J = 6.94 Hz, NH), 6.95 (s, 4H,
Ph), 5.28 (m, 2H, CH=CH), 4.40 (m, 1H, CH), 4.22 (m, 4H, 2 x CH2), 2.06 (t, J = 7.20 Hz, 2H, CH2), 1.92
(m, 4H, 2CH2), 1.44 (m, 2H, CH2), 1.20 (m, 20H, 10 x CH2), 0.82 (t, J = 7.08 Hz, 3H, CH3). 13C-RMN (75
11

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
MHz, DMSO-d6) δ 172.4. 165.7, 145.5, 138.6, 129.7, 129.6, 119.1, 108.7, 63.0, 46.9, 35.4, 31.3, 29.1, 28.8,
28.7, 28.6, 28.5, 26.6, 25.3, 22.1, 14.0. MS (ES+): m/z 660.3 (M+H)+. Anal. C35H49NO11 (C, H, N, O).
2.1.20. Bis-O-galloy]-N-linoleoyl serinol (17)
Following the general procedure, 12 (61.5 mg) was dissolved in methanol (1.5 mL) and treated with
HCl 37% (0.08 mL) to give 42.5 mg (97%) of the title compound as a white amorphous solid. 1H-RMN
(400 MHz, DMSO- d6) δ 8.07 (d, J = 6.94 Hz, NH), 6.93 (s, 4H, Ph), 5.26 (m, 4H, 2 x CH=CH), 4.38 (m,
1H, CH), 4.20 (m, 4H, 2 x CH2), 2.71 (m, 2H, CH2), 2.05 (t, J = 7.20 Hz, 2H, CH2), 1.90 (m, 2H, CH2),
1.43 (m, 2H, CH2), 1.19 (m, 18H, 9 x CH2), 0.80 (t, J = 7.08 Hz, 3H, CH3). 13C-RMN (75 MHz, DMSO-
d6) δ 172.7, 165.8, 145.6, 138.7, 129.9, 129.8, 127.9, 127.8, 119.2, 108.9, 63.1, 47.1, 35.5, 31.0, 29.2, 28.8,
28.7, 28.6, 26.7, 25.5, 25.3, 22.1, 14.1. MS (ES+): m/z 658.3 (M+H)+. Anal. C35H47NO11 (C, H, N, O).
2.1.21. Bis-O-[2,3-bis(benzyloxy)benzoyl]-N-oleoyl serinol (19)
A solution of the Bn protected benzoic acid 18 (Belin et al, 2003) (166 mg, 0.49 mmol), HATU (214
mg, 0.56 mmol) and DMAP (41.5, 0.33 mmol) in dry CH2Cl2 (7.5 ml) was stirred at room temperature for
15 minutes and then added to a second solution, also stirred for 15 min, containing the N-oleoyl serinol
derivate 5 (80 mg, 0.22 mmol) and DMAP (41.5, 0.33 mmol) in dry CH2Cl2 (7.5 ml). The mixture was
stirred at room temperature overnight and then evaporated to dryness. The residue was dissolved in ethyl
acetate (20 mL) and washed successively with saturated solutions of citric acid (3 x 20 mL), NaHCO3 (3 x
20 mL) and NaCl (1 x 20 mL). The organic layer was dried over anhydrous Na2SO4, filtered, and
concentrated under vacuum. The residue was then purified by flash column chromatography using
hexane/EtOAc (10:1 to 1:1) as the eluent to afford 174 mg (77 %) of the title compound as a white
1
amorphous solid. H-RMN (400 MHz, CDCl3) δ 7.42-7.06 (m, 26H), 5.77 (d, 1H, NH), 5.33 (m, 2H,
CH=CH), 5.13 (m, 8H, 4 x CH2), 4.56 (m, 1H, CH), 4.39 (m, 2H, CH2), 4.28 (m, 2H, CH2), 4.26 13C-RMN
(75 MHz, CDCl3) δ 173.1, 166.0, 152.8, 148.5, 137.4, 136.6, 130.1, 129.9, 128.7, 128.6, 128.5, 128.3,
128.2, 127.7, 126.3, 124.2, 123.1, 118.3, 75.8, 71.4, 63.7, 47.4, 36.5, 32.0, 29.9, 29.7, 29.5, 29.4, 29.3, 27.4,
25.6, 22.8, 14.3.
2.1.22. Bis-O-(2,3-dihydroxybenzoyl)-N-octadecanoyl serinol (20)
12

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
A solution of 19 (150 mg, 0.152 mmol) in THF/MeOH (1:1) containing 30% wt% of Pd/C (10%) was
hydrogenated at 30 ºC overnight under atmospheric pressure using a reaction balloon filled with hydrogen
gas and a glass flask as the reaction vessel. The Pd/C was filtered through Whastman ® filter paper 42 and
the solvent was removed under reduced pressure to give a crude product which was then purified by
triturating with cold diethyl ether and dichloromethane to afford 84 mg (87 %) of the title compound as a
white amorphous solid.1H-RMN (400 MHz, DMSO-d6) δ 10.27 (s, 2H, OH), 9.38 (s, 2H, OH), 8.12 (d, J
= 8.20 Hz, NH), 7.26 (dd, J = 8.10 Hz, 2H, J = 0.60 Hz, 2H, Ph), 7.00 (dd, J = 7.80 Hz, J = 0.60 Hz, 2H,
Ph), 6.70 (t, J = 7.90 Hz, 2H, Ph), 4,54 (m, 1H, CH), 4.38 (m, 4H, 2 x CH2), 2.06 (t, J = 7.20 Hz, 2H, CH2),
1.43 (m, 2H, CH2), 1.16 (m, 28H, 14 x CH2), 0.82 (t, J = 6.60 Hz, 3H, CH3). 13C-RMN (75 MHz, DMSO-
d6) δ 172.6, 169.2, 149.5, 146.1, 120.8, 119.9, 118.8, 112.9, 64.0, 46.4, 35.4, 31.3, 30.4, 29.1, 29.0, 28.9,
28.8, 28.7, 28.5, 25.3, 22.1, 14.0. MS (ES+): m/z 630.4 (M+H)+. Anal. C35H51NO9 (C, H, N, O).
2.1.23. N,N-bis[3,4,5-tris(benzyloxy)benzoyloxyethyl]oleoylamide (23)
A solution of the benzyl protected gallic acid 21 (Rivero-Buceta et al., 2015b) (263 mg, 0.60 mmol),
HATU (258 mg, 0.68 mmol) and DMAP (52.5 mg, 0.40 mmol) in dry CH2Cl2 (10 ml) was stirred at room
temperature for 15 minutes and then added to a second solution, also stirred for 15 min, containing 22 (100
mg, 0.27 mmol) and DMAP (52.5 mg, 0.40 mmol) in dry CH2Cl2 (10 ml). The mixture was stirred at room
temperature overnight and then evaporated to dryness. The residue was dissolved in ethyl acetate (20 mL)
and washed successively with saturated solutions of citric acid (3 x 20 mL), NaHCO3 (3 x 20 mL) and NaCl
(1 x 20 mL). The organic layer was dried over anhydrous Na2SO4, filtered, and concentrated under vacuum.
The residue was then purified by flash column chromatography using hexane/EtOAc (4:1 to 1:1) as the
1
eluent to afford 216 mg (66%) of the title compound as an amorphous white solid. H-RMN (400 MHz,
CDCl3) δ 7.27 (m, 34H, Ph), 5.24 (m, 2H, CH=CH), 5.01 (m, 12H, CH2), 4.36 (m, 4H, CH2), 4.36 (m, 4H,
CH2), 3.66 (m, 4H, CH2), 2.29 (m, 2H, CH2),1.90 (m, 4H, 2 x CH2),1.50 (m, H, CH2), 1.17 (m, 20H, 10 x
CH2), 0.80 (t, J = 6.60 Hz, 3H, CH3). 13C-RMN (75 MHz, CDCl3) δ 173.3, 166.1, 165.9, 152.8, 152.7,
136.8, 136.6, 130.1, 129.9, 128.7, 128.6, 128.3, 128.2, 128.1, 127.7, 127.6, 125.0, 124.5, 109.2, 100.5, 75.3,
71.4, 71.3, 63.0, 62.6, 45.7, 33.4, 32.1, 29.9, 29.7, 29.6, 29.5, 29.4, 27.3, 25.5, 22.8, 14.3.
2.1.24. N,N-bis(3,4,5-trihydroxybenzoyloxyethyl)octadecanoylamide (24)
13

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
A solution of 23 (200 mg, 0.164mmol) in THF/MeOH (1:1) containing 30% wt% of Pd/C (10%) was
hydrogenated at 30 ºC overnight under atmospheric pressure using a reaction balloon filled with hydrogen
gas and a glass flask as the reaction vessel. The Pd/C was filtered through Whastman ® filter paper 42 and
the solvent was removed under reduced pressure to give the crude product which was then purified by
triturating with cold diethyl ether and dichloromethane to afford 58 mg (53 %) of the title compound as a
white amorphous solid.1H-RMN (300 MHz, DMSO-d6) δ 9.14 (s, 3H, OH), 6.91 (s, 4H, Ph), 4.27 (m, 4H,
2 x CH2), 3.66 (m, 4H, 2 x CH2), 2.31 (t, J = 7.20 Hz, 2H, CH2), 1.41 (m, 2H, CH2), 1.29 (m, 28H, 14 x
CH2), 0.81 (t, J = 6.60 Hz, 3H, CH3). 13C-RMN (75 MHz, CDCl3) δ 172.6, 165.8, 165.7, 145.6, 145.5,
138.7, 138.5, 119.2, 118.9, 108.6, 61.9, 61.7, 46.8, 44.3, 32.1, 31.3, 29.1, 29.0, 28.9, 28.7, 24.9, 22.1, 14.0.
MS (ES+): m/z 676.4 (M+H)+. Anal. C36H53NO11 (C, H, N, O).
2.2. Biological methods
Stock solutions for each compound were prepared by dissolving the corresponding compound in
dimethyl sulfoxide (DMSO) and were stored at -20ºC in dark.
2.2.1. Cell culture
SW-480 and SW-620 colorectal cancer cell lines and CCD18-Co colonic fibroblastic colon cell line
were obtained from the American Type Culture Collection (ATCC) and were maintained under standard
conditions of temperature (37°C), humidity (95%) and carbon dioxide (5%). SW-620 and SW-480 colon
cancer cell lines were cultured in Dulbecco's Modified Eagle Medium (DMEM, Gibco-Invitrogen, Grand
Island, NY, USA) supplemented with 10% Fetal Bovine Serum (Gibco-Invitrogen), 2mM glutamine
(BioWhittaker, Lonza Group, Basel, Switzerland) and 1% antibiotics-antifungal (containing 10,000
units/mL of penicillin base, 10,000 µg/mL of streptomycin base and 25,000 ng/mL of amphotericin B;
Gibco-Invitrogen). CCD18-Co human colon fibroblasts (ATCC, CRL-1459) were grown in Minimum
Essential Medium (MEM) supplemented with 10% fetal bovine serum, 1% antibiotic/antimycotic, 2 mM
GlutaMAX, 0,1 mM Non-essential amino acids (NEAA) (Invitrogen).
2.2.2. Cell Viability Assay
The anti-proliferative and cytotoxicity properties of the compounds in SW-620 human colon cancer cell
line were measured by MTT assay. SW-620 cells were seeded in 24-well plates in exponential growth phase
14

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
using a density of 25 x 10³ cells per well. After 24 h, culture media were replaced with 500 µl of fresh
culture media containing serial concentrations of each compound (dissolved in DMSO) and the number of
viable cells was determined at time zero (control growth wells) using the MTT assay, as described below
for treated cells. After 48 hours of treatment, treated cells were incubated during 3 h at 37ºC with 3-(4,5-
dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (50µL/well at 5 mg/mL in PBS) (Sigma-
Aldrich, St. Louis, MO, USA). Over that time, MTT containing media were removed and the MTT reduced
to purple formazan by living cells was solubilized by adding 200 µL of DMSO per well. After 1 h with
mixing and dark, reduced formazan product, proportional to the number of viable cells, was measured at
560 nm using a scanning spectrophotometer microplate reader (Biochrom Asys UVM 340 Microplate
Reader, ISOGEN, De Meern, The Netherlands). The parameters IC50 (µM) (concentration needed for 50%
inhibition of cell proliferation), GI50 (µM) (concentration needed for 50% growth inhibition), TGI (µM)
(concentration needed for total growth inhibition) and LC50 (µM) (concentration needed for 50% cell death)
were calculated according to the NIH definitions using a logistic regression.
For all compounds, at least three independent experiments with three replicates per concentration were
performed.
2.2.3. Extracellular flux analysis of the oxygen consumption rate (OCR) and extracellular acidification rate
(ECAR)
Mitochondrial respiration and glycolytic function were analyzed with the XFe96 Cell Bionalyzer and
Cell Mito Stress Test and XF Glycolysis Stress kits respectively (Seahorse Biosciences, XFe96). Prior to
the assay, optimal cell density and oligomycin tritation were determined.
Glycolysis and oxidative phosphorylation are the two main energy-producing pathways in the cell. Most
cells possess the ability to switch between these two pathways, thereby adapting to changes in their
environment. Glucose in the cell is converted to pyruvate (referred to as glycolysis), and then converted to
lactate in the cytoplasm, or CO2 and water in the mitochondria. The conversion of glucose to pyruvate, and
subsequently lactate, results in a net production and extrusion of protons into the extracellular medium.
This, results in the acidification of the medium which is directly measured by XF6 analyzer and reported
as the acidification rate (ECAR).
The assay scheme is as follows: First, cells are incubated in the glycolysis stress test medium without
glucose or pyruvate and basal ECAR is monitored. The first injection is a saturating concentration of
15

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
glucose (10 mM). The cells catabolize this glucose to pyruvate through the glycolytic pathway, producing
ATP, NADH, water, and protons. The extrusion of protons into the surrounding medium leads to an increase
in basal ECAR. The difference between ECAR before (basal situation) and after the addition of glucose is
a measurement of the rate of glycolysis (glycolysis rate).
The second injection is oligomycin, which inhibits ATP synthase. Oligomycin inhibits mitochondrial
ATP production, and shifts the energy production to glycolysis, with the subsequent increase in ECAR
which is monitored as the cellular maximum glycolytic capacity. The difference between glycolytic
capacity and glycolysis rate defines glycolytic reserve.
The final injection is 2-deoxy-glucose (2-DG), a glucose competitor, which inhibits glycolysis through
competitive binding to hexokinase, the first enzyme in the glycolytic pathway. The resulting decrease in
ECAR confirms that the ECAR produced and monitored in the assay is due to glycolysis.
For glycoStress assay, 20000 cells were plated in a XFe-96 well plate and kept overnight (o/n) in
complete media. The next day, culture medium was changed to 0 mM glucose, 2 Mm glutamine in non-
buffered media to starve the cells for 1h. Basal ECAR was monitored (1 to 3 measurements). Next, 10 mM
glucose was injected to determine the capacity of the cells to upregulate glycolysis from the basal situation
(4 to 6 measurements) and difference between basal ECAR and ECAR in the presence of glucose was
determined to calculate Glycolysis value. Next, oligomycin was added to block ATP production from
mitochondrial respiration to determine the maximal glycolytic capacity. Finally, 50 mM DG was injected
to block completely glycolytic pathway.
The Cell Mito Stress Test measures key parameters of mitochondrial respiration by measuring the
oxygen consumption rate (OCR) of cells. Cell Mito Stress Test uses sequential drug injections to target
different electron transport chain (ETC) complexes in the mitochondria. In this way different parameters
such as basal respiration, ATP production, proton leak, maximal respiration, spare respiratory capacity, and
nonmitochondrial respiration were determined. Injections are as follow: oligomycin injection to inhibit
ATP synthase (complex V), FCCP injection to uncouple oxygen consumption from ATP production, and
rotenone and antimycin A injection to inhibit complexes I and III, respectively.
As indicated, oligomycin inhibits ATP synthase (complex V), and the decrease in OCR following
injection of oligomycin correlates to the mitochondrial respiration associated with cellular ATP production.
Carbonyl cyanide-4 (trifluoromethoxy) phenylhydrazone (FCCP) is an uncoupling agent which disrupts
the mitochondrial membrane potential and thus, electron flow through the electron transport chain (ETC)
16

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
is inhibited, and oxygen is maximally consumed by complex IV. The FCCP-stimulated OCR is used to
calculate spare respiratory capacity, defined as the difference between maximal respiration and basal
respiration. Spare respiratory capacity is a measure of the ability of the cell to respond to increased energy
demand.
The third injection is a mix of rotenone, a complex I inhibitor, and antimycin A, a complex III inhibitor.
This combination shuts down mitochondrial respiration and enables the calculation of nonmitochondrial
respiration driven by processes outside the mitochondria. Proton leak which is the remaining basal
respiration not coupled to ATP production, can be a sign of mitochondrial damage or can be used as a
mechanism to regulate the mitochondrial ATP production.
For mitostress assay, 40000 cells were plated and kept o/n in complete media. The next day, culture
medium was changed to 10 mM glucose, 2 mM glutamine, 1 mM pyruvate non-buffered media and cells
were kept for 1h at 37 ºC in an incubator without CO2 Basal OCR was monitored (1 to 3 measurements).
Next, 1µM oligomycin was injected to determine the amount of oxygen dedicated to ATP production at
mitochondria (3 to 6 measurements). Then, carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone
(FCCP) was injected at 0.4 µM to free H+ gradient from the intermembrane space to the mitochondrial
matrix, and so to determine the maximal respiration rate (MRR) or Spare Respiratory Capacity. Finally,
antimycin A and rotenone (0.5µM) were added to completely inhibit mitochondrial respiration (7 to 9
measurements).
2.2.4. Western blot
Cells were lysed in Laemmli buffer, proteins were separated by SDS–polyacrylamide gel
electrophoresis and transferred onto a nitrocellulose membrane (Bio-Rad Laboratories, Hercules, CA,
USA). The membranes were blocked using 5% nonfat dry milk in TBS 0.05% Tween-20. Primary
antibodies were incubated overnight at 4ºC and upon 1h incubation with secondary antibodies signal
detection was performed using the Clarity Western ECL Substrate (Bio-Rad). β-actin determination or
Ponceau stain were used as loading controls.
2.2.5. Caspase activation assay
SW-620 cells were plated in 96-multiwell and treated for 48 hours with 1.66 µM concentration of
compounds 16 and 14. Staurosporin (1.5 µM) was used as a positive control. Activity of caspase 3 and
17

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
caspase 7 was quantified using the Caspase-Glo 3/7 assay kit (Promega), following manufacturer’s
instructions.
2.2.6. Measurement of cellular ATP content
Relative cellular ATP content was measured by the ATP based assay CellTiter-Glo Luminescent Cell
Viability kit (Promega, Madison, WI, USA; Cat # G7571) with modifications from the manufacturer's
protocol. Briefly, after 48h of treatment with compound 16 at IC50 (0.83 M) and 2 x IC50 (1.66 M), SW-
620 cells were plated in 96-well clear bottom black polystyrene plates. 10000 cells were plated per well.
Non-treated cells were plated in a similar way. Cells were then maintained for 6 h in complete media,
without treatments, to allow cells to attach to the plate. Then, an equal volume of the single-one-step reagent
provided by the kit was added to each well and rocked for 15 minutes at room temperature. Cellular ATP
content was measured by a luminescent plate reader.
2.2.7. Statistical analysis
Dose-response curves for cell viability assays were analyzed by analysis of variance (ANOVA) with
Bonferroni and Tukey as post hoc tests. Data were presented as mean ± SEM of at least three independent
experiments each performed in triplicate. Statistical significance was defined as *p<0.05, **p<0.01;
***p<0.001. The statistical analyses were performed by use of the R statistical software version 2.15
(www.r-project.org).
18

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
3. Results
3.1. Chemical results
For the synthesis of the proposed final compounds, the key N-acyl serinol derivatives (2-6) were
prepared first (Scheme 1). Selective N-acylation of 2-amino-1,3-propanediol (serinol, 1) was undertaken
by reaction of the commercially available serinol with the corresponding acyl chlorides in the presence of
triethylamine at -20 ºC to afford compounds 2-6 in good yields (70-81 %). It should be mentioned that the
synthesis of these compounds were previously described in the literature following different procedures
(Bieberich, 2002a; Bieberich, 2002b; Landau, 2014; Merg et al., 2015; Schulz and Robins, 2004; Oette and
Tschung, 1980; Osornio et al., 2012). In our case, a unique procedure that implies the use of acyl chlorides
and triethylamine has been followed for all of them.
Subsequently, condensation of the N-acyl serinol derivatives (2-6) with the MOM-protected galloyl acid
(3,4,5-tris(methoxymethoxy)benzoic acid) (7) in the presence of HATU, as coupling reagent, and DMAP
as base, followed by deprotection of the MOM ether intermediates 8-12 with aqueous hydrochloric acid
(37%), gave the deprotected final compounds 13-17 in good yields (Scheme 2). It should be mentioned that
the MOM-protected galloyl intermediate 7 was previously synthesized by other authors (Gazak et al.,
2011). In our case, a modification of the previously described procedure, that implies the use of methyl
gallate as starting material, gives this compound with good overall yield (60% for the two steps) (Scheme
2).
Next, compound 20, with two 2,3-dihydroxybenzoyl residues, instead of galloyl, as phenolic
substituents, was prepared (Scheme 3). In this case, the N-oleynol serinol derivative 5 was treated with the
benzyl-protected benzoic acid 18 (Belin et al. 2003) in the presence of HATU, as coupling reagent, and
DMAP as base (Scheme 3) to give 19 (77%). Catalytic hydrogenation of 19, at atmospheric pressure in the
presence of 10% Pd/C, afforded the fully deprotected derivative 20 in 87% yield (Scheme 3).
The same synthetic methodology was applied for the synthesis of compound 24, in which
diethanolamine, instead of serinol, has been used as linker (Scheme 4). In this case, the 3,4,5-benzyl
protected galloyl 21 (Belin et al., 2003) was used as the phenolic synthon. Condensation of 21 with the
corresponding N-oleoyl diethanolamine derivative 22 (Bieberich et al., 2002c; Sagnella et al., 2011)
afforded 23 that was submitted to catalytic hydrogenation, using H2/Pd-C, to give the deprotected derivative
24 (53% yield).
19

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
3.2. Inhibition of colon cancer cell growth
Firstly, the anti-proliferative properties of the new synthetic phenolic derivatives (13-17, 20 and 24)
were tested in SW-620 human colon cancer cell line. Compound SF6, containing two galloyl moieties at
both ends of a C12 N-acyl chain, from our collection of phenolic derivatives, was included for comparative
purposes (Rivero-Buceta et al., 2015a) and 4,4´-Di-O-methylellagic acid, an effective phenolic compound
against colon cancer cells (Ramirez de Molina et al., 2015) was used as the reference drug. The results are
shown in Table 1.
Compound SF-6 did not show any growth-inhibitory effect at the concentration tested (up to 100 µM)
(Table 1) while compound 14, in which the two galloyl units are at the same side of the C12 N-acyl chain
affected human colon cell viability in the µM range. This data points to the importance for the growth-
inhibitory effect of a polar “head”, with two galloyl subunits.
Compound 13, with one short N-acyl chain (C3) resulted inactive at the concentration tested. Further
elongation of the length of the N-acyl chain led to an increase in the inhibitory activity. Thus, among the
galloyl derivatives with a saturated N-acyl chain (13-15) the best inhibitory effect was observed for
compound 15, containing a C18 alkyl substituent, followed by 14, containing a C12 alkyl chain.
On the other hand, the growth-inhibitory effect observed for compound 16, with an unsaturated C18
chain, is better than that obtained for 15, with a saturated C18 chain, and also better than that of 17, with
two double bonds in the acyl chain.
Compound 20 with a C-18 alkyl chain and where the phenolic substituents contain only 2 hydroxyl
groups, was 22-fold less active than its corresponding C18 analogue (15), with two gallic acid residues (3
OHs on the aromatic ring).
On the basis of all of these results, it can be concluded that the cell growth inhibition depends
significantly on both the size and unsaturation of the alkyl chain on the “tail” and the presence of galloyl
groups on the “head”.
The importance of serinol as the linker is supported by the loss of activity of the compound 24, with an
ethylene amine as linker, with respect to those observed for 15, with serinol as linker. This result indicates
that simple changes in structure can result in differences in growth inhibition, being serinol the best linker.
Finally, for comparative purposes we considered that it could be of interest to determine the growth-
inhibitory effect of compound 16, the most potent, in the cancer cell line SW480 (Table 2). SW480 and
20

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
SW620 are colon carcinoma cell lines from a single patient derived either from the primary tumor (SW480)
or from a lymph node metastasis (SW620) of the same tumor. SW620 cells are more highly tumorigenic and
metastatic than SW480 (Hewitt et al., 2000).
As it was shown in Table 2 the human colon cancer cell line SW480 is also growth inhibited by 16,
although to different extent than SW620 cell line. In fact, a better inhibitory effect of 16 in SW620 cells than
in SW480 cells (IC50 = 0.83 µM vs IC50 = 4.03 µM) was observed. This result can be of great therapeutic
interest due to the fact that SW620 is much more aggressive than SW480.
Differences in sensitivity between those two cell lines towards a particular drug have also been
described in the literature (R. Rashmi et al., 2003). Those differences might be explained by proteomic and
metabolic differences between those cell lines that required further studies.
3.3. Toxicity in non-cancer cells
Next, toxicity in non-cancers cells for compound 16, the one with the highest growth inhibitory effect,
was determined. With this purpose, CCD18-Co colonic fibroblastic colon cell line was used. The IC50 for
compound 16 in these primary epithelial colonic cells was 30 µM vs 0.83 µM in SW-620 colon cancer cells.
This result indicates that there exists a therapeutic window for the use of compound 16 in colon cancer.
3.4. Cell bioenergetic analysis
Very recently a rapid and easy experimental methodology has been developed to analyze the energetic
metabolism of a cell (glycolysis and oxidative phosphorylation) in real time (Pike Winer and Wu, 2014;
TeSlaa and Teitell, 2014). When applied to the anticancer-field, this novel methodology facilitates a better
understanding of the cancer cell metabolism and may be useful to discover agents that target specific
metabolic pathways.
Glycolysis, conversion of glucose to lactate with the production of two molecules of ATP, is monitored
through the measurement of the Extra-Cellular Acidification Rate (ECAR) at the surrounding media while
oxidative phosphorylation through the measurement of the Oxygen Consumption Rate (OCR). High levels
in basal ECAR may be indicative of reliance on glycolysis to support cell proliferation. The response to an
input of glucose after starvation, -this is, the increase in ECAR compared to the basal situation-, determines
the preferential use of the glycolytic pathway vs oxidative mitochondrial pathway. On the other hand,
21

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
monitorization of OCR allows to determine the dependency of the oxidative metabolism by the cell. Thus,
high basal OCR indicates a highly active oxidative metabolism.
Taking all of this into consideration we investigated the role of compound 16 (IC50: 0.83 µM), the one
with the highest growth inhibitory effect, on cell bioenergy as a possible cause of its anticancer activity.
With this aim we monitored ECAR and OCR as readouts of glycolysis and mitochondrial respiration
respectively using an extracellular flux bioanalyzer (Seahorse Biosciences XF96). Compound 14 (IC50:
29.5 µM), with a moderate effect on cell viability, was also evaluated for comparative purposes.
We first tested the effect of compound 16 on the glycolytic activity using a glycostress standard assay
(Fig. 2). SW-620 cancer cells were pretreated for 48 h with 16 at two concentrations (0.83 µM = IC50 and
1.66 µM = 2 x IC50). Similarly, SW-620 cancer cells were pretreated with 14 at 0.83 µM and 16.66 µM (1
fold and 20 fold of the IC50 showed for compound 16). Then cells were re-suspended and plated overnight.
The next day, media was changed and cells were glucose starved for 1h. Basal ECAR was monitored (1 to
3 measurements). Next, 10 mM glucose was injected to determine the capacity of the cells to increase
glycolysis from the basal situation (4 to 6 measurements). The difference between ECAR before (basal
situation) and after the addition of glucose is a measurement of the glycolytic rate (glycolysis rate). As it
can be observed in Fig. 2, compound 16 at a concentration of 1.66 µM significantly reduced glycolysis rate
compared to control situation.
When oligomycin was added to inhibit mitochondrial ATP synthase and to block ATP production from
mitochondrial respiration, SW-620 cells increased their glycolytic flux to maintain cell bioenergetic
homeostasis. This increase in the glycolytic flux in response to a deficiency in mitochondrial ATP
production is known as glycolytic capacity. As shown in Figure 2, the glycolytic capacity was significantly
diminished in the presence of a 2 x IC50 dose of 16. On the contrary, cells treated with 14 behaved similarly
to control cells even at the highest concentration tested (16.66 µM), reinforcing the idea that compound 14
is less bioactive than 16.
The difference between glycolytic capacity and glycolysis rate defines glycolytic reserve. This
parameter was also significantly diminished at 2 x IC50 dose of 16.
From these assays, it can be concluded that in the presence of 16, glycolysis of SW620 cancer cells as
energetic pathway is diminished (reduced glycolysis rate, glycolytic capacity and glycolytic reserve).
We next analyzed mitochondrial respiration by monitoring the oxygen consumption rate (OCR) using
a mitostress standard assay (Figure 3). For this purpose, SW-620 cells were treated for 48h with compound
22

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
16 at 0.83 µM = IC50 and 1.66 µM = 2 x IC50, and with compound 14 at 0.83 µM and 16.66 µM. Non-
treated cells were kept as controls.
Basal respiration, that is, the energetic demand of the cells under baseline conditions, was measured
first. As shown in Fig.3A, cells treated with compound 16 at both concentrations displayed a reduced basal
respiration compared to 14-treated cells and control cells.
All other parameters of mitochondrial function -ATP production, spare respiratory capacity and proton
leak - were determined following sequential injections of specific inhibitors: oligomycin (inhibitor of ATP
synthase), carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP), a disrupter of mitochondrial
membrane potential, and rotenone/antimycin A (inhibitors of complex I and III of electron transport chain)
(Pike Winer and Wu, 2014).
OCR before oligomycin injection minus OCR after oligomycin injection determines the portion of
oxygen dedicated to ATP production by mitochondria. As shown in Fig. 3A this parameter is significantly
reduced in 16 treated cells at 2 x IC50 concentration compared to 14 treated cells and control non treated
cells.
Next, FCCP, which mimics an increased physiological energy demand, was injected to determine the
maximal respiration rate (MRR) or spare respiratory capacity. This parameter determines the amount of
extra ATP that can be produced by the cell in case of a sudden increase of energy demand as well as how
closely the cell is breathing to its theoretical maximum. As it can be observed in Fig. 3A, after adding
FCCP, cells treated with compound 16 at 2 x IC50 dose, displayed a significant reduction on the reserve
respiratory capacity parameter compared to 14 treated cells and control non treated cells.
Finally, antimycin A and rotenone were added to completely inhibit mitochondrial respiration. The
difference between OCR after oligomycin injection and OCR after rotenone and antimycin A injection was
used to determine the proton leak function. This parameter determines the remaining basal respiration not
devoted to ATP production and can be a sign of mitochondrial damage. As it can be observed in Fig. 3A,
cells displayed a significant reduction on proton leak at all tested concentrations of 16 compared to
compound 14 treated cells and control non-treated cells.
Next, we run a similar assay to compare mitochondrial respiration of 16 and 14 treated cells at 2 x IC50
dose of each compound, this is 1.66 µM for 16 and 60 µM for 14. As it is shown in Fig. 3B, 14 at 60 µM
diminished mitochondrial function in a similar way than 16 at 1.66 µM. These results indicate that 14 is
23

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
indeed effective on targeting mitochondrial respiration, but it requires a higher concentration than 16 (60
µM for 14 vs 1.66 µM for 16) in accordance to the previously determined IC50 value.
The discovery that compound 16 affects cell bioenergetics in colon cancer cells allow us to propose that
its anticancer activity might be due, at least in part, to this property. It should be highlighted that, different
to other scenarios where the energetic metabolism is rewired from glycolysis to mitochondrial respiration
(Dai et al., 2015) or the other way around (Tesori et al., 2015), treatment with 16 diminishes both.
3.5. AMPK activation
As it was previously mentioned, 5’-AMP-activated kinase (AMPK) is an energy sensor that regulates
energy metabolism. This is a heterotrimeric Ser/Thr kinase complex characterized by a catalytic α subunit
and two regulatory subunits (β,γ) (Zadra et al., 2015). For the full activity of the kinase, a phosphorylation
at the residue Thr172 in the catalytic loop is required. The active phosphorylated form of AMPK (P-AMPK)
inhibits essentially all anabolic pathways that promote cell growth (Faubert et al., 2013). Thus, in cancer
where the energy demands of the cell are elevated, AMPK activators may suppress tumor growth. The role
of AMPK as a tumor suppressor is supported by the observation that patients on treatment with metformin,
an activator of AMPK, have a lower cancer incidence (Evans et al., 2005; Vazquez-Martin et al., 2009).
Due to the critical role of AMPK in the regulation of energy and redox homeostasis, we wanted to check
if compound 16 can contribute to its activation. As shown in Fig. 4A, treatment of colon cancer cells with
compound 16, even at the lowest concentration (0.83 µM), significantly increased the amount of the
phosphorylated active form of AMPK (P-AMPK (Thr172)). Consequently, the ratio P-AMPK(Thr172)/total
AMPK (figure 4B) is higher for 16 compared to control cells. Importantly, compound 14 was also able to
activate AMPK but it required a higher concentration than 16 (29.5 µM for 14 vs 1.66 µM for 16 (Figs 4A
and B). These results indicate that the two structurally related polyphenols 16 and 14, are able to activate
AMPK, but 16 was 18 times more potent than 14.
As SW620 cells treated with 16 showed defective mitochondrial ATP production, together with
decreased aerobic glycolysis, leading to AMPK activation, we analyzed the effect of this compound on the
cellular ATP content. For this purpose, colorectal cancer SW620 cells were treated for 48 h with compound
16 at two concentrations (0.83 µM = IC50 and 1.66 µM = 2 x IC50). Then, 10000 viable cells of non-treated
and treated cells were re-plated. After 6 h in complete media, without treatments, the ATP content was
measured by mean of the ATP based assay CellTiter-Glo Luminescent Cell Viability kit (Promega). As
24

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
shown in Figure 4C the cellular ATP content was significantly reduced with respect to the control (non-
treated cells) in the presence of the highest concentration of 16 (1.66 µM) (p<0.038). From this study we
concluded that compound 16 diminishes cell bioenergetics leading to AMPK activation as a sensor of ATP
depletion.
3.6. Induction of apoptosis
Finally, as compound 16 inhibited cell viability, diminishing cell bioenergetics and activating AMPK,
we wanted to determine if it might induce apoptosis.
Apoptosis is a vital biological process of multicellular organisms by which damaged, mutant and aged
cells are eliminated in a programmed way. As previously mentioned, a variety of natural polyphenols, like
curcumin or resveratrol, induces apoptotic cell death in a variety of cancer models. This process is mediated
by the activation of some specific proteases named caspases (Oliver and Vallette, 2005; Reed, 2000).
Apoptosis, mediated by caspase activation may result in induction of death in cancerous cells and therefore
represent a promising strategy to develop cancer chemotherapeutics.
There are more than 13 known caspases (procaspases or active cysteine caspases) that can be detected
using various types of caspase activity assays (Bayascas et al., 2002; García-Calvo et al., 1999; Le et al.,
2002; Mooney et al., 2002; Nicholson and Thornberry, 1997; Thornberry et al., 1997; Thornberry and
Lazebnik, 1998). Among them, caspase-3 and -7 are especially important because they play a key role in
the apoptotic pathway. In fact, their activation is being considered as a typical hallmark of apoptosis.
In the present study, we wanted to check if induction of apoptosis could be also implicated in the
observed cell growth inhibition promoted by 16. With this purpose the activity of caspase 3 and 7 was
monitored using a luminescent assay. For this assay a time period of 48 h has been chosen because in a
previous experiment carried out to determine the effect of the tested compounds on cell viability at different
times (24 h, 48 h and 72 h) it was found that 24 h is a period of time too short to induce cellular lethality
while 72 h is too long, ending with the cell death of most cells (data not shown).
Thus, SW-620 cells were treated for 48 h with compounds 14 and 16, both at 1.66 µM (2 fold of the
IC50 showed for compound 16). Staurosporine (1.5 µM) treated cells were used as positive control of
caspase 3/7 activation. Staurosporine is an alkaloid isolated from Streptomyces staurosporeus known to
activate apoptosis in several cancer cells (Yadav et al., 2015).
25

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
As it is shown in Figure 5, activity of caspase 3/7 after 48 h was significantly increased in SW-620
treated cells with compound 16 compared to control and 14. Remarkably, higher effect on caspase 3/7
activation was also observed with 16 compared to the positive control, staurosporine.
Our results showed that compound 16 is able to stimulate significantly caspase activation in SW-620
cancer cells.
4. Discussion
Although metabolic alteration is one of the oldest mechanisms associated with cancer development,
only recently its targeting became a promising anticancer strategy. In fact, altered metabolism is now
considered an emerging hallmark of cancer and researches in the field hope to find drugs that may lead to
a new class of anticancer drugs (Jang et al., 2013). However, this field is in a very preliminary stage and
the effective number of compounds under investigation as potential modulators of cell metabolism is still
very poor. For this reason, the identification of metabolically active agents can be of great interest for
anticancer therapies.
In this report we provide evidence of the inhibitory effects on colon cancer growth and metabolism of
a new family of synthetic polyphenols. These compounds have a hydrophilic “head”, composed of two
galloyl (3,4,5-trihydroxybenzoyl) units, and a hydrophobic “tail” linked through a serinol moiety. They
were very efficiently obtained (high yields and easy purifications) in three steps by N-acylation of serinol
followed by coupling with methoxy methyl (MOM) protected gallic acid and deprotection with aqueous
hydrochloric acid.
On the basis of the Structure-Activity-Relationship (SAR) studies performed on this family it appears
that the presence of the two galloyl moieties at the same side of the molecules is essential for inhibitory
growth, and that the length of the N-acyl chain is also crucial for activity, being compounds with C12 and
C18 chain length the best. A decrease of the acyl chain length, reduction of the number of OHs on the
aromatic ring or substitution of serinol by ethanolamine as linker lowers or eliminates their anticancer
properties. However, the introduction of a double bond on the hydrophobic tail increases the inhibitory
growth properties. Compound 16, with an unsaturated C18 chain, showed the highest anti-proliferative
effect in SW-620 colon cancer cells.
Our results indicate that compound 16 reduces the energetic metabolism (glycolysis and mitochondrial
respiration) in SW-620 colon cancer cells at the same time that activates AMPK and caspase 3 and 7
26

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
activity. All of these findings lead us to suggest that, through AMPK activation, compound 16 might lead
to alterations in colon cancer cell bioenergy compromising cancer cell viability. Importantly, these anti-
proliferative and pro-apoptotic effects have been shown to be selective for cancer cells, indicating a
therapeutic window for the use of this compound.
Of particular interest is the simultaneous inhibition of the two main energy-producing pathways
(glycolysis and oxidative phosphorylation) showed by 16 in colon cancer cells. In this respect, it has been
recently postulated that the pharmacological inhibition of several complementary metabolic pathways
(antimetabolic cooperativity) can be better to eliminate cancer cells than the inhibition of a single one
(Marchetti et al., 2015).
5. Conclusion
As recently emphasized by Jang and coworkers (Jang et al., 2013) cancer cell metabolism has become
one of the most exciting and promising fields for the development of new anticancer agents. Targeting
cancer metabolism opens an opportunity to develop broadly applicable drugs that can treat multiple cancer
types and hence may lead to a new class of anticancer drugs.
We concluded that the novel molecules here described, and in particular 16, the most potent within this
series, show antiproliferative properties in two different colon carcinoma cell lines (SW480 and SW620),
reduces the energetic metabolism (glycolysis and mitochondrial respiration) in SW-620 colon cancer cells
at the same time that activates AMPK and caspase 3 and 7 activity. Based on that, 16 and its analogues
could serve as hits to develop a new class of therapeutic agents that target tumor metabolism. Of particular
interest is the simultaneous inhibition of the two main energy-producing pathways (glycolysis and oxidative
phosphorylation) showed by 16 in colon cancer cells. Further research is needed to fully elucidate the
possible mechanism(s) of action to suppress tumor growth. This is a non-trivial task because many different
metabolic enzymes, oncogenic signalling pathways or/and regulatory network can be implied.
27

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
6. Acknowledgments
We acknowledge MA Bonache for help with the processing of the Supporting Information.
28

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
7. Authorship Contributions
Authors declare no potential conflict of interest. All co-author participated sufficiently in the work to take
responsibility for the content and all co-authors approved the final version.
Participated in research design: San-Félix, Pérez-Pérez, Quintela, Gómez de Cedrón, Ramirez de Molina,
Reglero
Conducted experiments: Madrona, Jiménez, Vargas, Gómez de Cedrón
Performed data analysis: San-Félix, Gómez de Cedrón
Wrote or contributed to the writing of the manuscript: San-Félix, Gómez de Cedrón
29

JPET Fast Forward. Published on June 5, 2018 as DOI: 10.1124/jpet.118.248278
This article has not been copyedited and formatted. The final version may differ from this version.
JPET#248278
8. Supplemental data
Copies of 1H and 13C NMR spectra are included.
30