Studies towards the Total Synthesis of Drimentine C. Preparation of the AB and CDEF Ring Fragments

Article abstract of DOI:10.1002/ejoc.202000158

The drimentine family is a class of hybrid isoprenoids derived from actinomycete bacteria. Members of this family display weak antitumor and antibacterial activity. Herein we report our efforts toward the total synthesis of drimentine C using three distinct approaches incorporating palladium-catalyzed cyanoamidation, reductive cross-coupling, and photoredox-catalyzed α-alkylation of an aldehyde as key steps. Our synthetic efforts use a convergent synthesis to assemble the terpenoid and alkaloid portions of drimentine C from readily available l-tryptophan, l-proline, and (+)-sclareolide.

Full text of DOI:10.1002/ejoc.202000158

A Journal of
Accepted Article
Title: Studies towards the total synthesis of drimentine C. Preparation
of the AB and CDEF ring fragments
Authors: Sarah M. Pound, Steven J. Underwood, and Christopher J.
Douglas
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Org. Chem. 10.1002/ejoc.202000158
Link to VoR: http://dx.doi.org/10.1002/ejoc.202000158
Supported by

10.1002/ejoc.202000158
European Journal of Organic Chemistry
FULL PAPER
Studies towards the total synthesis of drimentine C. Preparation
of the AB and CDEF ring fragments
Sarah M. Pounda, Steven J. Underwooda, Christopher J. Douglasa*
Abstract: The drimentine family is a class of hybrid isoprenoids
derived from actinomycete bacteria. Members of this family display
weak antitumor and antibacterial activity. Herein we report our efforts
toward the total synthesis of drimentine C using three distinct
approaches incorporating palladium-catalyzed cyanoamidation,
reductive cross-coupling, and photoredox-catalyzed α-alkylation of
Figure 1. General scaffold of drimentines and structure of indotertine A.
an aldehyde as key steps. Our synthetic efforts use a convergent
synthesis to assemble the terpenoid and alkaloid portions of
drimentine C from readily available L-tryptophan, L-proline, and (+)-
Results and Discussion
sclareolide.
We initially envisioned the synthesis of drimentine C as a
showcase of the cut-and-sew alkene cyanoamidation
methodology studied by our lab and Takemoto.6–11
Introduction
Retrosynthetically (Scheme 1), we proposed accessing the
natural product through late-stage functionalization of lactam 6
via α-aminonitrile12 and reductive condensation to form the E
Drimentines are a family of indolosesquiterpenes isolated as
secondary metabolites from several strains of Streptomyces
bacteria.1,2 They and the structurally-related indotertines are
ring, hydrolysis of the nitrile to a carboxylic acid,cit and
condensation
with
L-proline.
Palladium-catalyzed
formed through the same biosynthetic pathway where
a
cyanoamidation would form lactam 6 from cyanoformamide 7,
prepared by cyanoformylating benzylaniline 8. Organometallic
insertion into Weinreb amide 9, followed by Wittig olefination of
the resulting ketone would grant access to benzylaniline 8.
Weinreb amide 9 can be prepared from (+)-sclareolide, (+)-10, in
two steps from a literature procedure.13
diketopiperazine links with farnesylpyrophosphate via the classic
mevalonate pathway.3 Biochemically, drimentines exhibit
anticancer, antibacterial, and antifungal properties; in particular,
drimentine C (1) exhibits modest cytoxicity against human tumor
cell lines HCT-8, Bel-7402, BGC-823, A549, and A2780.2 Being
secondary metabolites, isolation of the drimentines from natural
sources is difficult, offering a challenge that synthetic efforts can
solve.
Drimentines A (2), F (3), G (4), and indotertine A (5) were
previously synthesized by the Li lab,4 using a radical conjugate
addition to form the crucial C–C bond. This approach had
several benefits, notably the construction of a challenging all-
carbon quaternary stereocenter as well as late-stage
diversification of the diketopiperazine moiety, which allowed
several members of the family to be accessed. However, this
approach’s late-stage installation of the exocyclic alkene
ultimately required a low-yielding sequence of organocerium
addition followed by dehydration of the resulting tertiary alcohol.
This challenging sequence and the desire for material for
additional bioactivity screenings motivated our lab to pursue
synthesis of members of the drimentine family. Herein we report5
our attempts to synthesize drimentine C using a set of
approaches allowing for early installation of the exocyclic alkene
and the potential for diversification of the diketopiperazine ring.
Scheme 1. Initial retrosynthesis of drimentine C through a cyanoamidation
strategy.
Critical to this strategy is the synthesis of cyanoamidation
precursor 15, which contains the cyanoformamide and alkene
functionality required for palladium-catalyzed cyanoamidation.
Lactone ring opening of (+)-sclareolide by the dimethylaluminum
amide derived from N-methoxy-N-methylamine yields the
desired Weinreb amide (Scheme 2).13 Dehydration of the
resulting tertiary alcohol 11 yields exocyclic alkene 9; performing
this reaction at low temperatures prevented the formation of the
internal alkene 12. Anticipating challenges with selectively
[a]
S. M. Pound, S. J. Underwood, C. J. Douglas
Department of Chemistry
University of Minnesota – Twin Cities
Smith Hall, 207 Pleasant St SE, Minneapolis, MN 55455
cdouglas@umn.edu
URL: http://www1.chem.umn.edu/groups/douglas/
Twitter: @CJDouglasGroup
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000158
European Journal of Organic Chemistry
FULL PAPER
installing the exocyclic alkene at the late stage, we prioritized
maintaining it throughout the synthesis; therefore, the exocyclic
olefin is retained in all conditions reported herein.
suggested that the resulting epimer (epi-20) would give the
undesired diastereomer of the natural product.
To maintain the stereochemistry at C11, we considered radical-
based methodologies, which have been successfully applied to
pyrroloindoline alkaloid synthesis.16 We looked specifically at
reductive cross-coupling, which is proposed to proceed via
radical intermediates without the need for organometallic
nucleophiles.17–19 While the initial work on this sp2-sp3 reductive
For the aniline coupling partner, the benzyl protecting group
was chosen to enhance diastereoselectivity for the later
cyanoamidation step,6 and was easily installed by reductive
amination of 2-iodoaniline with benzaldehyde. With 14 in hand,
we attempted to form the di-lithiate through deprotonation of the
aniline hydrogen and subsequent lithium-halogen exchange.
After addition of 9, only starting 9 and deiodinated 14 was
recovered. Attempts to use other protected anilines also proved
unsuccessful.
cross-coupling used primary or secondary alkyl halides19,20
,
more recent examples have been reported of successful
reductive coupling with tertiary halides16,21,22
Bromopyrrolidine 20 was synthesized according to a literature
procedure23 through protection of L-tryptophan as the N-Boc-N‘-
Boc methyl ester and bromocyclization using NBS with PPTS,
yielding the desired cyclization product in 74% yield. Primary
bromide 18 was synthesized in two steps from intermediate 9:
hydrolysis of the Weinreb amide to carboxylic acid 22 followed
by decarboxylation. Initial hydrolysis attempts showed significant
amounts of remaining starting material, as well as formation of
methyl amide 23 (Table 1). Longer reaction times ensured full
conversion of starting material (Entry 2), as did addition of 18-
crown-6. Large excess of potassium hydroxide, 18-crown-6 or
increased reaction time reduced selectivity for carboxylic acid 22,
suggesting 23 exists as a decomposition byproduct. Reactions
sometimes overpressurized at high temperature, leading to
solvent loss and decreased selectivity for 22. Performing the
reaction in a sealed pressure vessel (Entries 6–7) mitigated
these issues. Alternatively, carboxylic acid 22 could be prepared
in two steps by base-catalyzed ring opening of the lactone
followed by dehydration, however this sequence was not as
selective for the exocyclic alkene compared to the Weinreb
amide route. Carboxylic acid 22 was converted to the primary
Scheme 2. Synthesis of terpenoid fragment 9 and attempts at organolithium
coupling.
At this point, we chose to change strategies, favoring a
convergent route with the key disconnection between the
alkaloid and terpenoid fragments. We believed that a late-stage
condensation could be used to form the diketopiperazine ring,
thereby allowing access to other drimentines depending on the
amino acid used. Key to this route was determining a method of
forming the C9–C19 bond from coupling partners 18 and 20,
derived from 10 and L-tryptophan (21), respectively.
bromide by
a
thermally-initiated Barton decarboxylative
bromination.24,25
Table 1. Optimization of the synthesis of carboxylic acid 22.
Entry
KOH
(equiv)
18-crown-6
(equiv)
Temp
(°C)
Time
(h)
9 : 22 : 23
Yield
(%)a
1b
2c
4.0
–
140
130
130
130
130
130
130
0.66
45
1 : 7 : 5
0 : 5 : 1
1 : 0 : 0
0 : 3 : 1
0 : 5 : 1
0 : 4 : 1
0 : 7 : 1
–
70d
–
4.0
–
3e,f
4c
10.0
10.0
10.0
10.0
10.0
–
43
1.0
1.0
1.0
1.0
25
37
53g
67
50
5c
27
Scheme 3. Revised retrosynthesis via a convergent approach.
6h
7h
27
55
The first approach we considered with this convergent route
involved primary organocuprate 19. Using cyclopropanation
conditions reported by Rainier and Espejo,14 we believed that
the key carbon-carbon bond formation could be achieved with
release of ring strain. However, this ring opening would give the
wrong stereochemistry at C11, and work by Reisman et al.15
a. Isolated yield of 22. b. Reaction run on 0.04 g scale in a microwave reactor.
c. Reaction run on 0.1 g scale in a sealed vial with a PTFE-lined cap using an
aluminum block. d. Yield given as an average of two trials. e. Reaction run on
a 0.3 g scale in a sealed vial with a PTFE-lined cap using an aluminum block. f.
1,4-dioxane used instead of MeOH. g. Yield given as an average of three trials.
h. Reaction run in a sealed pressure vessel in an oil bath.
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000158
European Journal of Organic Chemistry
FULL PAPER
Mixing 22 and Barton salt 24 with triethylamine in
bromotrichloromethane at 90 °C for twenty minutes, followed by
quench with hydroquinone and acidic workup yielded bromide
18 in a 43% yield after purification (Figure 2). We did not further
optimize this reaction.
deformylation of a fused isoprenoid alkaloid like intermediate 35.
Formation of the key C–C bond could occur from α-alkylation
under photoredox conditions of two intermediates: the fully
formed diketopiperazine core from 17 and 21 and a masked
aldehyde of (+)-10, such as an enamine30–32 or silyl enol ether.33
Figure 2. Synthesis of primary bromide 18.
To test the viability of reductive cross-coupling in the synthesis
of drimentine C, we studied the reaction on a model system.
Beginning with tryptamine 25, double Boc protection with
subsequent bromocyclization23 yielded bromopyrrolidine (±)-27
(Scheme 4a). We reacted (±)-27 with 2-cyclohexylbromide and
iodide using a variety of nickel catalysts, ligands, and terminal
reductants (Scheme 4b). These reactions consistently formed
four major alkaloid products, confirmed by spectroscopic
analysis and independent synthesis (see supplementary
information for individual trials and details of independent
synthesis). Dimerization of the pyrroloindoline occurred, with
both meso and chiral dimers formed. Reduction of the C–Br
bond occurred to yield 33. Mass spectrometry of the fourth
alkaloid product suggested coupling of (±)-27 and the
(cyclohexyl)methyl moiety, but spectroscopic analysis after
purification revealed the structure of 31, where coupling occured
at the 5 position with opening of the pyrrolidine ring. Other
Scheme 5. Revised retrosynthesis using a photoredox-catalyzed α-alkylation
approach.
As with the reductive cross-coupling strategy, we studied α-
cyclohexyl-containing
products
observed
were
alkylation
on
a
model
system
using
simplified
dicyclohexylethane 33 and ketone 34.
bromopyrroloindole 20 and the enamine 39 formed from
cyclohexylacetaldehyde and pyrrolidine (
Table 2). Initially, the reaction was irradiated with a 15 W
fluorescent lamp in the presence of Ru(bpy)3Cl2 and common
additives (triethylamine, 2,6-lutidine, and Hantzsch ester) using
dimethylformamide as a solvent, but only reduced product 33
was observed (entries 1-3). Cyclic voltammetry of 20 gave a first
reduction
potential
of
–0.800 eV, suggesting that the pyrroloindoline radical forms
through a reductive quench of the excited photocatalyst. Since
many of the additives screened can act as both electron and
hydrogen donors, their presence likely led to formation of 33. To
remove the need for said additives, we moved to the Ir3+ family
of photocatalysts. With fac-Ir(ppy)3 used as photocatalyst (entry
4) a new product was observed; the formate ester of 20,
indicating that the solvent acted in a non-innocent manner (see
supplementary information for additional details).34 In addition,
the formate ester of 20 experienced epimerization at the
stereogenic center bearing the methyl ester.
To minimize these background reactions, we optimized the
solvent. Fac-Ir(ppy)3 was insoluble in ethereal solvents like THF
and 1,4-dioxane, as well as aromatic solvents like toluene. While
only sparingly soluble in acetonitrile (entry 5), new products
were observed: coupling product 40 and epi-40. Changing the
solvent to dichloromethane increased the solubility of the
photocatalyst, and changing the photocatalyst to the more
strongly reducing fac-Ir(tbppy)3 and light source to a blue LED
increased conversion of 20, forming 40 in 74% yield after acidic
workup. Alternate radical acceptors were tested, specifically the
morpholine-derived enamine and the TMS, TIPS, and TBS silyl
Scheme 4. Reductive cross-coupling a) substrate synthesis and b) model
study
Because no desired coupling was observed in our model
study, we considered other strategies involving single-electron
transfer. In this context, photoredox catalysis was enticing, as
the stereocenter at C11 should not be affected, and photoredox
catalysis has previously been used for related alkaloid
synthesis.4,29 Retrosynthetically (Scheme 5. Revised retrosynthesis
using
a photoredox-catalyzed α-alkylation approach.Scheme 5), we
envisioned drimentine
C
stemming from
a
late-stage
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000158
European Journal of Organic Chemistry
FULL PAPER
ethers, but under the reaction conditions only 33 was observed
as a product.
Two diastereomers are possible in the bromocyclization step,
assuming no epimerization in the diketopiperazine ring; 36 and
dia-36. Because the relative stereochemistry of 36 could not be
conclusively determined by NOESY, 36 and dia-36 were
examined computationally. Single point NMR calculations
(B3LYP/6-311+G(2d,p)/M06-2X/6-31+G(d,p))36 of 36 and dia-36
showed closer agreement for the synthesized sample being
compound 36.
Table 2. Optimization of photoredox-catalyzed α-alkylation.
Initial reactions between bromopyrollidine 20 and enamine 41
were promising—a new product was observed by high resolution
mass spectrometry with the chemical formula C38H54N2O7,
consistent with productive coupling between 20 and 41. This
product could not be isolated by column chromatography,
decomposing after several attempts at purification. Attempts to
perform the photoredox-catalyzed coupling with 36 and 41 were
also attempted (Scheme 7). After 28 h, 36 was fully consumed,
an ion consistent with iminium ion 44 was observed by high
resolution mass spectrometry (see supplementary information).
Attempts to hydrolyze the product to an aldehyde were
unsuccessful, and this product could not be isolated.
Entry
Deviation from above
Yield
40 (%)
1
none
74%
2
Ru(bpy)3Cl2 (2.5 mol%), NEt3 (2 equiv), 15W
fluorescent lamp, DMF
0
3
4
Ru(bpy)3Cl2 (2.5 mol%), Hünig’s base (2 equiv), 15W
0
0
fluorescent lamp, DMF
Ru(bpy)3Cl2 (2.5 mol%), Hantzsch ester (2 equiv),
15W fluorescent lamp, DMF
Scheme 7. Proposed intermediate formed after reaction of 36 and 42 under
optimized photoredox-catalyzed conditions
5
6
Fac-Ir(ppy)3 (5 mol%), 15W fluorescent lamp, DMF
0a
Fac-Ir(ppy)3 (2.5 mol%), 15W fluorescent lamp,
8%
MeCN
Y = morpholinyl, -OTMS, -OTIPS, -OTBS
Conclusions
7
0
Yields determined by 1H NMR using 1,3,5-trimethoxybenzene, isolated yields
represented in parentheses. a. Formate ester of 20 isolated as major product.
In summary, we have developed three distinct approaches
toward the total synthesis of drimentine C. With each of these
approaches, we identified the quaternary stereocenter at C9 as a
key disconnection and prioritized early installation of the
exocyclic alkene. Of these approaches, the photoredox-
catalyzed coupling of enamine 41 and bromopyrrolidine 36
seems the most promising, but is hampered by challenging
isolation of the potential iminium ion product. Further studies are
underway in our lab, focusing on the key coupling reaction, as
well as completing the synthesis of the natural product.
With this promising result in hand, the coupling partners for α-
alkylation were synthesized. From Weinreb amide 9, a simple
DIBAL-H reduction yields aldehyde 37 in a 91% yield with high
enough purity to be carried forward without further purification.
From protected amino acids 17 and 21-OMe·HCl, the
diketopiperazine moiety can be formed by peptide coupling, Boc
cleavage, and condensation-cyclization in 54–70% yield over
three steps after recrystallization from methanol.35 Selective Boc
protection of the indole nitrogen occurs in good yield, with
bromocyclization forming 36 in a modest 51% yield as a single
diastereomer.
Acknowledgements
We thank the NIH for funding this work (R01 GM095559). We
thank Prof. Tom Hoye (UMN) for use of an LCMS.
Keywords: natural products • alkaloids • photoredox • cross
coupling
--
(1)
(2)
(3)
(4)
(5)
E. Lacey, M. Power, Z. Wu, R. W. Rickards. WO1998009968 A1,
1997.
Q. Che, T. Zhu, R. A. Keyzers, X. Liu, J. Li, Q. Gu, D. Li. J. Nat.
Prod. 2013, 76 (4), 759–763.
Q. Che, T. Zhu, X. Qi, A. Mándi, T. Kurtán, X. Mo, J. Li, Q. Gu, D. Li.
Org. Lett. 2012, 14 (13), 3438–3441.
Y. Sun, R. Li, W. Zhang, A. Li. Angew. Chemie Int. Ed. 2013, 52
(35), 9201–9204.
S. M. Pound. Development of an Intramolecular Alkene
Aminocyanation Reaction and Progress Toward the Total Synthesis
of Drimentine C, PhD Thesis, University of Minnesota–Twin Cities,
Scheme 6. Synthesis of coupling partners for α-alkylation.
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000158
European Journal of Organic Chemistry
FULL PAPER
2016.
(24)
(25)
(26)
(27)
(28)
(29)
(30)
(31)
(32)
(33)
(34)
(35)
D. H. R. Barton, D. Crich, W. B. Motherwell. Tetrahedron Lett. 1983,
24 (45), 4979–4982.
D. H. R. Barton, D. Crich, W. B. Motherwell. Tetrahedron 1985, 41
(19), 3901–3924.
M. Movassaghi, M. A. Schmidt. Angew. Chem. Int. Ed. 2007, 46
(20), 3725–3728.
M. Movassaghi, O. K. Ahmad, S. P. Lathrop. J. Am. Chem. Soc.
2011, 133 (33), 13002–13005.
S. P. Lathrop, J. Kim, M. Movassaghi. Chim. Int. J. Chem. 2012, 66
(6), 389–393.
L. Furst, J. M. R. Narayanam, C. R. J. Stephenson. Angew. Chemie
Int. Ed. 2011, 50 (41), 9655–9659.
D. A. Nicewicz, D. W. C. MacMillan. Science 2008, 322 (5898), 77–
80.
D. A. Nagib, M. E. Scott, D. W. C. MacMillan. J. Am. Chem. Soc.
2009, 131 (31), 10875–10877.
H.-W. Shih, M. N. Vander Wal, R. L. Grange, D. W. C. MacMillan. J.
Am. Chem. Soc. 2010, 132 (39), 13600–13603.
P. V. Pham, D. A. Nagib, D. W. C. MacMillan. Angew. Chemie Int.
Ed. 2011, 50 (27), 6119–6122.
Y.-Q. Zou, W. Guo, F.-L. Liu, L.-Q. Lu, J.-R. Chen, W.-J. Xiao.
Green Chem. 2014, 16 (8), 3787–3795.
Our synthesis of 36 is analagous to that used for other
diketopiperazine derived natural products, see: M. E. Kieffer, K. V.
Chuang, S. E. Reisman. J. Am. Chem. Soc. 2013, 135 (15), 5557–
5560.
(6)
A. M. Dreis, S. C. Otte, M. S. Eastwood, E. R. Alonzi, J. T.Brethorst,
C. J. Douglas. European J. Org. Chem. 2017, 2017 (1), 45–48.
Y. Kobayashi, H. Kamisaki, R. Yanada, Y. Takemoto. Org. Lett.
2006, 8 (13), 2711–2713.
Y. Kobayashi, H. Kamisaki, H. Takeda, Y. Yasui, R. Yanada, Y.
Takemoto. Tetrahedron 2007, 63 (14), 2978–2989.
Y. Yasui, H. Kamisaki, Y. Takemoto. Org. Lett. 2008, 10 (15), 3303–
3306.
Y. Yasui, H. Kamisaki, T. Ishida, Y. Takemoto. Tetrahedron 2010,
66 (11), 1980–1989.
G. B. Frost, N. A. Serratore, J. M. Ogilvie, C. J. Douglas. J. Org.
Chem. 2017, 82 (7), 3721–3726.
G. Moinet, T. Imbert, T. United States Patent: 4942221, 1990.
J. Boukouvalas, J.-X. Wang, O. Marion, B. Ndzi. J. Org. Chem.
2006, 71 (17), 6670–6673.
V. R. Espejo, X.-B. Li, J. D. Rainier. J. Am. Chem. Soc. 2010, 132
(24), 8282–8284.
L. M. Repka, J. Ni, S. E. Reisman. J. Am. Chem. Soc. 2010, 132
(41), 14418–14420.
C. Pꢀrez-Balado, A. R. de Lera. Org. Lett. 2008, 10 (17), 3701–
3704.
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
I. Colon, D. R. Kelsey. J. Org. Chem. 1986, 51 (14), 2627–2637.
C. Amatore, A. Jutand. Organometallics 1988, 7 (10), 2203–2214.
S. Biswas, D. J. Weix, J. Am. Chem. Soc. 2013, 135 (43), 16192–
16197.
(20)
(21)
(22)
(23)
D. A. Everson, R. Shrestha, D. J. Weix. J. Am. Chem. Soc. 2010,
132 (3), 920–921.
(36)
P. H. Willoughby, M. J. Jansma, T. R. Hoye. Nat. Protoc. 2014, 9 (3),
643–660.
X. Lu, Y. Wang, B. Zhang, J.-J. Pi, X.-X. Wang, T.-J. Gong, B. Xiao,
Y. Fu. J. Am. Chem. Soc. 2017, 139 (36), 12632–12637.
X. Wang, G. Ma, Y. Peng, C. E. Pitsch, B. J. Moll, T. D. Ly, X. Wang,
H. Gong. J. Am. Chem. Soc. 2018, 140 (43) 14490–14497.
V. R. Espejo, J. D. Rainier. J. Am. Chem. Soc. 2008, 130 (39),
12894–12895.
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000158
European Journal of Organic Chemistry
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
FULL PAPER
Total Synthesis*
Sarah M. Pound, Steven J. Underwood,
Christopher J. Douglas*
Page No. – Page No.
Studies towards the total synthesis of
drimentine C. Preparation of the AB
and CDEF ring fragments
Synthetic efforts towards the total synthesis of the natural product drimentine C are
described. Three distinct approaches are discussed incorporating palladium-
catalyzed cyanoamidation, reductive cross-coupling and photoredox-catalyzed α-
alkylation of an aldehyde as key steps to join the alkaloid and terpenoid fragments.
*alkaloid synthesis
This article is protected by copyright. All rights reserved.