Synthesis and structure reassignment of malylglutamate, a recently discovered earthworm metabolite

Article abstract of DOI:10.1021/acs.jnatprod.8b01083

Malylglutamate, a newly identified metabolite in earthworms, was synthesized using a traditional peptide coupling approach for assembling the amide from protected malate and glutamate precursors. The proposed structure (1) and a diastereomer were synthesized, but their NMR spectra did not match the natural sample. Further analysis of the natural sample using HMBC spectroscopy suggested an alternative attachment of the malyl moiety, and β-malylglutamate (2) diastereomers were synthesized, L,L-2 and D,D-2. NMR spectra were an excellent match with the natural sample, and chiral-phase chromatography was employed to identify (a)-β-l-malyl-l-glutamate (2) as the isomer native to Eisenia fetida.

Full text of DOI:10.1021/acs.jnatprod.8b01083

Note
pubs.acs.org/jnp
Cite This: J. Nat. Prod. XXXX, XXX, XXX−XXX
Synthesis and Structure Reassignment of Malylglutamate, a
Recently Discovered Earthworm Metabolite
Corey M. Griffith,†,§ Abigail Feceu,‡,§ Cynthia K. Larive,‡ and David B. C. Martin*,‡
†Environmental Toxicology Graduate Program and ‡Department of Chemistry, University of California, Riverside, California 92521,
United States
S
* Supporting Information
ABSTRACT: Malylglutamate, a newly identified metabolite in earthworms, was
synthesized using a traditional peptide coupling approach for assembling the amide
from protected malate and glutamate precursors. The proposed structure (1) and a
diastereomer were synthesized, but their NMR spectra did not match the natural sample.
Further analysis of the natural sample using HMBC spectroscopy suggested an
alternative attachment of the malyl moiety, and β-malylglutamate (2) diastereomers
were synthesized, L,L-2 and D,D-2. NMR spectra were an excellent match with the
natural sample, and chiral-phase chromatography was employed to identify (−)-β-L-
malyl-L-glutamate (2) as the isomer native to Eisenia fetida.
alylglutamate (1) is a newly identified metabolite found
M
in earthworm species Eisenia fetida.1 It is a relatively
abundant metabolite, present at an average of 7 μg/mg in
whole earthworms, and similar in structure to metabolites such
as N-acetylaspartylglutamate (NAAG, 3) and β-citrylglutamate
(4). Malylglutamate is prominent in the NMR metabolite
profiles of whole earthworm, coelomic fluid, and coelomocyte
extracts.1−3 Coelomic fluid is a biofluid that fills the body
cavity of the worm and is critical for movement, excretion,
metabolism, and nutrient storage, while coelomocytes are free-
moving immune and liver-like cells in the coelomic fluid.4 The
metabolite profiles of coelomic fluid and coelomocytes are
primarily composed of amino acids, organic acids, and
polyamines, including N-trimethylated amino acids (i.e.,
betaine analogues).1 Malylglutamate is a suggested anionic
osmolyte and a charge-balance counterpart to betaine
analogues in earthworms, which are known cationic
osmolytes.1,5 Additionally, malylglutamate is postulated to be
a store for glutamic acid and malic acid, and a chelator, due to
its similarity in structure to β-citrylglutamate (4), a known
chelator.1,6,7
allow easy deprotection by hydrogenolysis in the final step. A
Fischer esterification of L-glutamic acid with benzyl alcohol
gave doubly protected intermediate L-6 as the tosylate salt
(Scheme 1); D-6 was also synthesized in the same manner.9
For the malic acid fragment, the two carboxylic acids were
differentiated using an acetonide protecting group, favoring the
five-membered ring.8 Benzylation of the free acid provided
acetonide intermediate L-8. Selective hydrolysis of the
acetonide ester group gave monobenzyl ester L-9. Coupling
of the two fragments was explored with a variety of common
amide coupling reagents. Acceptable results were obtained
using EDCI and HOBt in the presence of NEt3 to deprotonate
the tosylate salt L-6, giving amide L,L-10 in moderate yield.
Similarly, the coupling of protected D-glutamate gave amide
L,D-10. Removal of the benzyl protecting groups using
Following up on the recent identification of malylglutamate
by two of the authors,1 we sought to confirm the proposed
structure of malylglutamate and fully assign the relative and
absolute configuration. Due to complexity of 1H NMR spectra
of earthworm extracts and the difficulty in isolating a pure
sample for further studies, we turned our attention to synthesis
using a traditional peptide coupling approach. This strategy
allows the incorporation of either L- or D-isomers of both
glutamic and malic acids to determine both relative and
absolute configuration. As we describe below, these efforts
have led to a reassignment of the structure of natural
(−)-malylglutamate as the β-isomer L,L-2.
Our synthesis strategy takes advantage of previous reports
that describe the selective protection of either carboxylic acid
of malic acid.8 Benzyl ester protecting groups were chosen to
Received: December 21, 2018
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
DOI: 10.1021/acs.jnatprod.8b01083
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Note
a
Scheme 1. Synthesis of Both Diastereomers of the Originally Proposed Structure 1
a
a. BnOH, 1.22 equiv of p-TsOH, toluene, rt; b. 2,2-dimethoxypropane (DMP), 1 mol % p-TsOH, CH2Cl2; c. BnBr, Cs2CO3, DMF; d. AcOH,
H2O, 60 °C; e. 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDCI), N-hydroxybenzotriazole (HOBt), NEt3, DMF; f. H2, 5 mol
% Pd/C, EtOH.
hydrogenolysis gave the two diastereomeric products L,L-1
and L,D-1 in 73% and 71% yields, respectively.
1
Comparison of the H NMR data for synthetic L,L-1 and
L,D-1 in D2O with a naturally derived sample of
malylglutamate showed that the compounds were not identical
(Figure 1). In particular, the resonances for proton H-2 were
Figure 2. 1H−13C HMBC spectrum of an 80-worm pooled
coelomocyte sample. The peaks of malylglutamate are annotated,
and the H-3′a/C-4′ and H-3′b/C-4′ peaks suggest a reassignment of
the structure to β-malylglutamate (2).
1
3, and H2-4 protons nicely show the correct assignment.
Looking at the malyl moiety, equal intensities were observed
between the carbonyls and H-2′ proton, which did not aid with
the assignment, labeled as H-2′/C-1′ and H-2′/C-4′ peaks
(Figure 2); however, a very intense correlation between the
amide carbonyl (C-4′) and H2-3′ protons was observed, noted
as H-3′a/C-4′ and H-3′b/C-4′ (Figure 2 and Scheme 2). This
suggests that H2-3′ protons were positioned directly next to
the amide carbonyl and indicates that the malyl moiety was
attached to glutamate in the β position (i.e., β-malylglutamate
2).
Figure 1. H NMR spectrum of an earthworm coelomocyte sample
(top) compared to L,L-1 and L,D-1 (Scheme 1).
slightly shifted to the left in both synthetic molecules, and
proton H-2′ was significantly shifted to the left in both α-
malylglutamate (1) isomers (Scheme 1). To ascertain whether
the inconsistencies between the synthetic standards and native
sample were due to matrix or other effects, heteronuclear
multiple bond correlation spectroscopy (HMBC) was
employed to explore long-range correlations between protons
and carbons.10 Pooled earthworm coelomocyte samples were
used to conduct the analyses (Figure 2). Correlations between
both of glutamate’s carbonyls (C-1 and C-5) with the H-2, H2-
An alternative structure that would be consistent with the
available MS fragmentation data, HMBC analysis, and the
lower chemical shift of proton H-2′ was proposed. In the
B
DOI: 10.1021/acs.jnatprod.8b01083
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Note
a
Scheme 2. Synthesis of Both Diastereomers of the Revised Structure 2
a
a. TFAA, BnOH; b. EDCI, HOBt, NEt3, DMF; c. H2, 10 mol % Pd/C, EtOH.
revised proposal, the β-carboxylic acid of malic acid forms the
amide to glutamate rather than the α-carboxylic acid (i.e., 2).
For comparison, both α-malylarginine and β-malylarginine
isomers are found in apple and pear trees.11 This alternative
structure was therefore synthesized using a different protecting
group strategy.
Selective benzylation of the desired α-carboxylic acid was
achieved using a known procedure with trifluoroacetic
anhydride to give ester L-11 (Scheme 2).12 Amide coupling
of L-11 with L-glutamate derivative L-6 followed by reductive
removal of all three benzyl groups gave the new target amide
L,L-2. A similar sequence starting with D-malic acid gave
diastereomer D,L-2 and enantiomer D,D-2. Purification of the
protected intermediates was easier in these cases, and three of
the four stereoisomers of β-malylglutamate were efficiently
accessed for comparison to the natural product.
standards (Figure 4), indicating that it is an isomer of
264.071 m/z. To identify the enantiomer of malylglutamate
present in the natural sample, coelomic fluid, coelomocyte, and
whole earthworm extracts were injected on the LC-MS, and
the retention time and MS/MS spectra revealed L,L-2 as the
correct diastereomer (Figure 4A). To confirm the assignment,
L,L-2 and D,D-2 were spiked into separate coelomic fluid
samples where an increase in the L,L-2 peak intensity was
observed in the L,L-2-spiked sample (Figure 4B), while a new
peak was observed in the D,D-2-spiked sample and the L,L-2
peak intensity stayed consistent with the original sample
(Figure 4C). The specific rotation of synthetic L,L-2 was
determined to be [α]25 −13.6 (c 1.0, MeOH), thereby
D
establishing the natural product to be (−)-β-L-malyl-L-
glutamate.
In summary, we synthesized several diastereomers and
constitutional isomers of the originally proposed structure of
malylglutamate using a selective protecting group strategy. By
comparison to natural malylglutamate derived from earth-
worms, we identified the correct structure to be the β-isomer
L,L-2, resulting in the reassignment of the originally proposed
structure. We also determined the absolute configuration,
confirming that natural (−)-malylglutamate is derived from L-
malic acid and L-glutamic acid, providing support for the
hypothesis that this new metabolite can serve as a store of
these two essential α-hydroxy acid and α-amino acid
derivatives.
Access to synthetic β-malylglutamate allowed comparison of
1
the H NMR data in D2O to the natural sample. While the of
D,L-2 isomer differed in J-coupling of the H2-4 protons
(Figure 3), the L,L-2 isomer provided an excellent match in
chemical shift and J-coupling in the earthworm extract,
establishing the correct relative configuration.
EXPERIMENTAL SECTION
■
General Experimental Procedures. Optical rotation was
measured on a Rudolph Research Analytical Autopol IV automatic
polarimeter. IR spectra were recorded on a Bruker Alpha FT-IR
spectrometer. 1H and 13C NMR spectra of natural samples were
recorded on a Bruker Avance 600 MHz NMR spectrometer equipped
1
Figure 3. H NMR spectrum of an earthworm coelomocyte sample
1
with a SmartProbe. H and 13C NMR spectra of synthetic samples
(top) compared to L,L-2 and D,L-2 (Scheme 2).
were recorded on a Varian Inova 400 MHz spectrometer unless
otherwise indicated and were internally referenced to residual solvent
To finalize the configurational assignment of β-malylgluta-
mate, we sought to determine the absolute configuration. A
chiral-phase chromatography method was developed to
separate L,L-2 and D,D-2 using liquid chromatography−mass
spectrometry (LC-MS). Excellent separation was achieved,
yielding retention times of 1.61 min for L,L-2 and 8.28 min for
D,D-2 and identical MS/MS spectra (Figure 4). The D,D-2
standard contained a second peak at 3.14 min with identical
m/z as our standards; however, the MS/MS spectrum of the
unknown peak was inconsistent with L,L-2 and D,D-2
1
signal (note: D2O was referenced at 4.79 ppm for H and CDCl3
1
referenced at 7.26 ppm for H NMR and 77.16 ppm for 13C NMR,
1
respectively). Data for H NMR are reported as follows: chemical
shift (δ ppm), integration, multiplicity (s = singlet, d = doublet, t =
triplet, q = quartet, m = multiplet), and coupling constant (Hz). Data
for 13C NMR are reported in terms of chemical shift, and no special
nomenclature is used for equivalent carbons. High-resolution mass
spectrometry data were recorded on an Agilent LCTOF instrument
using direct injection of samples in dichloromethane into the
electrospray source (ESI) with positive ionization. Chiral-phase
C
DOI: 10.1021/acs.jnatprod.8b01083
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Note
Figure 4. Chiral-phase separation of L,L-2 and D,D-2 isomers using LC-MS at m/z 264.071. Chromatograms of (A) coelomic fluid sample; (B)
coelomic fluid sample spiked with L,L-2; and (C) coelomic fluid sample spiked with D,D-2, where *denotes an unknown isomer of the [M + H]+
ion. Positive-ion ESI-MS and annotated MS/MS of malylglutamate (calculated m/z 264.0719 for [M + H]+ ion) are shown for each isomer: (D)
MS1 of L,L-2 isomer and (E) its annotated MS/MS spectrum for the [M + H]+ ion; (F) MS1 of D,D-2 isomer and (G) its annotated MS/MS
spectrum for the [M + H]+ ion; and (H) MS1 of the unknown isomer and (I) its annotated MS/MS spectrum for the [M + H]+ ion.
separation was performed on a Waters UPLC coupled to a G2-XS Q-
TOF.
L,L-2 and D,D-2 standards were dissolved in ultrapure H2O and
diluted with MeOH to 10 μM prior to a 1 μL injection. Coelomic
fluid, coelomocyte, and whole earthworm extracts were reconstituted
in 1 mL of 2:1.8 MeOH/H2O, centrifuged for 1 min at 16100g prior
to analysis, and divided into 200 μL aliquots. For the spiking
experiments, 2 μL of 1.7 mM L,L-2 standard was added to a coelomic
fluid aliquot and 2 μL of 0.4 mM D,D-2 standard was added to
separate coelomic fluid aliquot prior to LC-MS/MS analysis.
Chromatograms were processed using a mean smoothed function
with a window size of 3 and 2 smooths. Mass spectra were
deisotoped, lock mass corrected, and centered.
Synthesis of Authentic (−)-β-L-malyl-L-glutamate (L,L-2). To
a flame-dried flask equipped with a magnetic stir bar were sequentially
added acid L-11 (500 mg, 2.23 mmol, 1 equiv), L-6 (1.34 g, 2.67
mmol, 1.2 equiv), HOBt (409 mg, 2.67 mmol, 1.2 equiv), EDCI (868
mg, 4.53 mmol, 2.0 equiv), and dimethylformamide (DMF) (37.2
mL, 60 mM). The flask was sealed with a septum and degassed by
sparging with N2. The reaction mixture was stirred for 10 min
followed by addition of Et3N (1.24 mL, 8.92 mmol, 4.0 equiv)
dropwise. The reaction mixture was stirred for 24 h and diluted with
EtOAc. The organic layer was washed with water. The layers were
separated, and the aqueous layer was extracted with EtOAc (2×). The
combined organic layers were washed with brine and dried over
Na2SO4. The solvent was removed in vacuo, and the crude residue
was purified by column chromatography on silica gel using EtOAc to
afford (S)-dibenzyl 2-((S)-4-(benzyloxy)-3-hydroxy-4-
oxobutanamido)pentanedioate as a white solid (868 mg, 73%
yield): IR (film) 3360, 2915, 1745, 1715, 1170 cm−1; 1H NMR
(400 MHz, CDCl3) δ 1.94−2.06 (m, 1H), 2.16−2.27 (m, 1H), 2.33−
2.48 (m, 1H), 2.62 (dd, J = 7.16, 15.32 Hz, 1H), 2.72 (dd, J = 3.68,
15.32 Hz, 1H), 3.61 (bs, 1H), 4.50 (dd, J = 3.72, 7.12 Hz, 1H), 4.63−
4.70 (m, 1H), 5.09 (s, 2H), 5.15 (s, 2H), 5.20 (s, 2H), 6.56 (d, J =
7.84, 1H), 7.28−7.39 (m, 15H); 13C NMR (100 MHz, CDCl3) δ
173.3, 172.8, 171.6, 170.0, 135.8, 135.2, 128.8, 128.7, 128.7, 128.6,
128.5, 128.5, 77.5, 77.2, 67.9, 67.8, 67.8, 67.6, 66.7, 51.9, 40.0, 31.0,
30.3, 27.3; HRMS (ESI) m/z calcd for C30H32NO8 (M + H)+
534.2122, found 534.2119.
Earthworm Sample Preparation and Metabolite Extraction.
Earthworm (Eisenia fetida) samples were collected by removing
worms from a laboratory culture, the soil was rinsed off, and worms
were patted dry.1−3,13 Whole earthworms were immediately flash-
frozen, lyophilized, and homogenized by bead beating prior to
extraction. Coelomic fluid and coelomocyte samples were extruded by
placing the worm in 0.1% NaCl solution, and a voltage was applied 10
times for <1 s each using a 9 V battery. The sample was transferred to
an Eppendorf tube and centrifuged at 400g, 4 °C for 10 min. The
supernatant (coelomic fluid) was transferred to a fresh tube, dried via
Speedvac overnight, and stored at −80 °C until analysis. The pellet
(coelomocytes) was immediately flash-frozen in liquid nitrogen to
quench metabolism and kept at −80 °C until metabolite extraction.14
Coelomocyte metabolites were extracted using ice-cold 2:2:1.8
CHCl3/MeOH/H2O.15 The aqueous layer was transferred to a
fresh tube, dried by Speedvac, and kept at −80 °C until analysis.
Prior to NMR analysis, samples were reconstituted in 200 μL of
100 mM phosphate buffer (pD 7.45) in D2O (D, 99.9%) containing
0.2 mM ethylenediaminetetraacetic acid-d16 (EDTA-d16) and 0.25
mM sodium 2,2-dimethyl-2-silapentane-5-sulfonic acid-d6 (DSS-d6)
(Cambridge Isotope Laboratories) and transferred to a 3 mm NMR
Tube (New Era Enterprises). A Bruker Avance NMR spectrometer
operating at 599.88 MHz and equipped with a SmartProbe was used
to acquire one-dimensional and HMBC spectra.10 Further details on
NMR acquisition and processing parameters for earthworm samples
can be found in the Supporting Information.
Chiral-Phase Separation and Detection of L,L-2 and D,D-2.
Separation was performed on an Astec Chirobiotic R 150 × 2.1 mm, 5
μm column (Sigma-Aldrich) using isocratic conditions of 98%
ultrapure water containing 0.01% acetic acid (Fisher Scientific) and
2% LiChrosolv ethanol (EMD Millipore Corporation) at a flow rate
of 0.55 mL/min with the column maintained at 30 °C. Electrospray
ionization was used, and β-malylglutamate was detected in positive-
ion mode at m/z 264.071 using a spray voltage of 35 V and source
temperature of 150 °C. MS/MS spectra were acquired for the m/z of
interest.
D
DOI: 10.1021/acs.jnatprod.8b01083
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Note
To an oven-dried vial equipped with a magnetic stir bar were
sequentially added (S)-dibenzyl 2-((S)-4-(benzyloxy)-3-hydroxy-4-
oxobutanamido)pentanedioate (20.0 mg, 0.038 mmol) and 10% Pd/
C (4.0 mg, 0.004 mmol, 10 mol %). The vial was sealed with a septum
and pumped and backfilled three times with nitrogen. This was
followed by addition of dry ethanol (723 μL, 0.052 M) and careful
addition of H2 via balloon. The reaction mixture was stirred for 6 h
and then filtered through Celite, washing with dichloromethane.
Solvent was removed in vacuo to afford L,L-2 as a white solid (9.0 mg,
82% yield) without further purification: IR (film) 3352, 3034, 1733,
1536, 1386, cm−1; 1H NMR (400 MHz, D2O) δ 4.60 (dd, J = 4.5, 7.8
Hz, 1H), 4.46 (dd, J = 4.7, 8.2 Hz, 1H), 3.20−3.33 (m, 1H), 2.82 (dd,
J = 4.5, 15.2, 1H), 2.70−2.77 (m, 1H), 2.35−2.55 (m, 3H), 2.15−
2.34 (m, 1H), 1.90−2.05 (m, 1H); 13C NMR (100 MHz, D2O) δ
177.2, 175.3, 174.9, 174.5, 67.9, 66.9, 57.5, 51.9, 48.9, 38.6, 29.9, 16.8;
HRMS (ESI) m/z calcd for C9H14NO8 (M + H)+ 264.0714, found
(3) Griffith, C. M.; Morgan, M. A.; Dinges, M. M.; Mathon, C.;
Larive, C. K. J. Proteome Res. 2018, 17, 2611−2622.
(4) Edwards, C. A.; Bohlen, P. J. Biology and Ecology of Earthworms,
3rd ed.; Chapman & Hall: London, UK, 1996; pp 11−12.
(5) Liebeke, M.; Bundy, J. G. Biochem. Biophys. Res. Commun. 2013,
430, 1306−1311.
(6) Hamada-Kanazawa, M.; Kouda, M.; Odani, A.; Matsuyama, K.;
Kanazawa, K.; Hasegawa, T.; Narahara, M.; Miyake, M. Biol. Pharm.
Bull. 2010, 33, 729−737.
(7) Narahara, M.; Hamada-Kanazawa, M.; Kouda, M.; Odani, A.;
Miyake, M. Biol. Pharm. Bull. 2010, 33, 1938−1943.
(8) Chan, T. M.; Kong, J.; McNamara, P.; Wong, J. K. Synth.
Commun. 2008, 38, 2252−2260.
(9) Du, Z.; Lu, Y.; Dai, X.; Zhang-Negrerie, D.; Gao, Q. J. Chem. Res.
2013, 37, 177−180.
(10) Bax, A.; Subramanian, S. J. Magn. Reson. 1986, 67, 565−569.
(11) Kasai, T.; Sakamura, S. Phytochemistry 1984, 23, 19−22.
(12) Schobert, R.; Jagusch, C. Synthesis 2005, 2005, 2421−2425.
(13) Bundy, J. G.; Osborn, D.; Weeks, J. M.; Lindon, J. C.;
Nicholson, J. K. FEBS Lett. 2001, 500, 31−35.
264.0721. The specific rotation is [α]24 = −13.6 (c 0.1, MeOH).
D
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jnat-
prod.8b01083.
■
S
(14) Lorenz, M. A.; Burant, C. F.; Kennedy, R. T. Anal. Chem. 2011,
83, 3406−3414.
(15) Bligh, E. G.; Dyer, W. J. Can. J. Biochem. Physiol. 1959, 37,
911−917.
Experimental procedures, NMR acquisition parameters,
1
characterization data, and H and 13C NMR spectra of
all new compounds (PDF)
AUTHOR INFORMATION
Corresponding Author
*Tel: 951-827-4429. E-mail: dave.martin@ucr.edu.
■
ORCID
Corey M. Griffith: 0000-0001-9104-9501
Abigail Feceu: 0000-0003-1316-9414
Cynthia K. Larive: 0000-0003-3458-0771
David B. C. Martin: 0000-0002-8084-8225
Author Contributions
§C. M. Griffith and A. Feceu contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
C.M.G. acknowledges support from the National Institute of
Environmental Health Sciences training grant (T32
ES018827), UC Riverside Graduate Research Mentorship
Program, and UC Riverside Dissertation Year Program. A.F.
acknowledges support from the UC Riverside Graduate
Research Mentorship Program. The authors acknowledge
support from the National Science Foundation (CHE-
1626673) and the U.S. Army (W911NF-16-1-0523) for
funding the NMR instrumentation used for this research at
the UC Riverside Analytical Chemistry Instrumentation
Facility (ACIF). Mass spectrometry instrumentation for this
research was supported by funding from the NSF (CHE-
0541848). D.B.C.M. is a member of the UC Riverside Center
for Catalysis. The authors thank Dr. J. Kirkwood at the UC
Riverside Metabolomics Core for his helpful discussions and
use of their LC-MS.
REFERENCES
■
(1) Griffith, C. M.; Williams, P. B.; Tinoco, L. W.; Dinges, M. M.;
Wang, Y.; Larive, C. K. J. Proteome Res. 2017, 16, 3407−3417.
(2) Yuk, J.; Simpson, M. J.; Simpson, A. J. Ecotoxicology 2012, 21,
1301−1313.
E
DOI: 10.1021/acs.jnatprod.8b01083
J. Nat. Prod. XXXX, XXX, XXX−XXX