Photoinduced Decarboxylative Radical Addition Reactions for Late Stage Functionalization of Peptide Substrates

Article abstract of DOI:10.1002/ejoc.202001178

Photoredox chemistry has greatly stimulated the application of radical based transformations in medicinal chemistry and early drug discovery in recent years. Carboxylate groups have been identified as traceless leaving groups that can be converted into radical intermediates capable of undergoing 1,4-conjugate addition reactions to Michael acceptors. Herein, we show the successful C-terminal derivatization of small peptide substrates by using this methodology in a parallel synthesis setting. Finally, we outline a general strategy for the γ-homologation of several drugs derived from α-amino acids in a late stage functionalization (LSF) approach.

Full text of DOI:10.1002/ejoc.202001178

European Journal of Organic Chemistry
Accepted Article
Accepted Article
Title: Photoinduced decarboxylative radical addition reactions for Late
Stage Functionalization of peptide substrates
Authors: Patricia Fernandez-Rodriguez, Fabien Legros, Thomas Maier,
Angelika Weber, María Méndez, Volker Derdau, Gerhard
Hessler, Michael Kurz, Ana Villar-Garea, and Sven Ruf
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.202001178
Link to VoR: https://doi.org/10.1002/ejoc.202001178
01/2020

10.1002/ejoc.202001178
European Journal of Organic Chemistry
COMMUNICATION
Photoinduced decarboxylative radical addition reactions for Late
Stage Functionalization of peptide substrates
Patricia Fernandez-Rodriguez,[a] Fabien Legros,[a] Thomas Maier,[a] Angelika Weber,[a] María Méndez,[a]
Volker Derdau,[a] Gerhard Hessler,[a] Michael Kurz,[a] Ana Villar-Garea,[a] and Sven Ruf*,[a]
[a]
Sanofi Aventis Deutschland GmbH
R&D, Integrated Drug Discovery
Industriepark Hoechst
65926 Frankfurt am Main, Germany
*E-mail: sven.ruf@sanofi.com
Abstract: Photoredox chemistry has greatly stimulated the
application of radical based transformations in medicinal chemistry
and early drug discovery in recent years. Carboxylate groups have
been identified as traceless leaving groups that can be converted to
radical intermediates capable of undergoing 1,4-conjugate addition
reactions to Michael acceptors. Here, we show the successful C-
terminal derivatization of small peptide substrates by using this
methodology in a parallel synthesis setting. Finally, we outline a
general strategy for the γ-homologation of several drugs derived from
α-amino acids in a late stage functionalization (LSF) approach.
Michael donor, avoiding the need for direct organometallic
activation or propagation. In 2016, Yoshimi further expanded the
scope of N-protected amino acid precursors and introduced N-
acryloyl amino acid esters and peptides as Michael acceptors.11
Similarly, in 2014, MacMillan et al. reported radical 1,4-conjugate
additions of readily available carboxylic acids and N-protected α-
amino acids to electrophilic alkenes, using an iridium-based
photoredox catalyst.12 In 2017, this photoinduced decarboxylative
1,4-addition was further expanded and applied to the
macrocyclization of peptides with the report of the synthesis of the
macrocyclic somatostatin analog COR00513 by an intramolecular
decarboxylative radical addition as the key step.14
Introduction
1,4-Conjugate addition of carbon nucleophiles to electron-
deficient alkenes is one of the most versatile methods for the
formation of C-C bonds. A classic example is the Michael reaction,
in which a CH-acidic Michael donor reacts with an α,β-
unsaturated carbonyl compound as acceptor under basic
1
conditions. Tremendous efforts have been invested over
decades into this reaction leading to the development of traceless
activation groups (TAGs) like cuprates, 2 boronic acids, 3 and
Grignard4 reagents as carbon nucleophiles.
While these developments enabled the use of a structurally
diverse set of carbon nucleophiles, acceptors or nucleophiles
containing sensitive chemical functionalities still remained largely
out of scope.5,6 This fact also hampered the use of Michael-type
additions in derivatizations of more complex molecules and
intermediates in drug discovery at an advanced stage (late stage
functionalization, LSF), an approach that was recently shown to
be versatile for efficient exploration of chemical and property
space.7 For instance, substrates like peptides do not tolerate
strongly basic reaction conditions required for the deprotonation
of the Michael donor.
Scheme 1. Overview on important published earlier work in the field
Even though peptide drugs play a major role in several
therapeutic approaches,15 these are to our knowledge the most
important examples applying photoredox catalyzed 1,4-additions
to Michael acceptors for the C-terminal derivatization of
peptides.16
We now envisioned the potential of this visible light-mediated
decarboxylative radical 1,4-addition as a late stage function-
alization strategy for peptidic substrates. Our main focus was
directed towards establishing and using this methodology in an
industrial parallel synthesis setup.
A breakthrough towards the use of peptide substrates in 1,4-
conjugated additions was accomplished by Yoshimi et al. in 2009,
taking advantage of photoredox catalysis as an easy entry into
radical chemistry.8 Initially, Yoshimi reported the decarboxylative
reduction of N-Boc-α-amino acids mediated by phenanthrene and
1,4-dicyanobenzene, in the presence of a thiol as the hydrogen
9
donor. Realizing the potential of the radical intermediate
generated via oxidative decarboxylation, the Yoshimi group
expanded this chemistry to radical 1,4-conjugate additions of N-
Boc-α-amino acids to a variety of electron-deficient alkenes.10 In
this case, the radical intermediate, formed after oxidative
decarboxylation of the starting material, readily reacts as a
1
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202001178
European Journal of Organic Chemistry
COMMUNICATION
Results and Discussion
We started our studies with the photoinduced radical 1,4-addition
of the model compound N-benzoyl glycine 1a to ethyl vinyl ketone
5a under reported reaction conditions,10 affording addition product
6aa with a 40% yield (Table 1, entry 1). In an attempt to improve
these results, we screened several photocatalysts and bases
(Table 1), leading to slightly higher yield with catalyst PC-2 (44%,
Table 1, entry 2). Therefore, we decided to use both PC-1 and
PC-2 during our further studies. The reactions were performed in
DMF to avoid any solubility problems with peptidic substrates.
Nevertheless, in our general experience ACN and DMF are the
most suitable solvents for performing photochemical reactions.
Table 1. Evaluation of test system I:
Figure 1. General structures of studied Michael donor precursors (R, R1, R2
side chains of biologically relevant amino acids).
1a
Entry
5a
6aa
E
E
Yield
(%)
Photocatalyst
(P*/P-)
(V)
(P/P-)
(V)
Base
1
2
(PC-1)
(PC-2)
+ 1.21
- 1.37
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
K2HPO4
K3PO4
CsF
40
44
18
21
0
+ 1.32
- 1.37
5a
5b R = Me
5c R = Et
5d
5h
5l
5e
3
(PC-3)
+ 1.65
- 0.82
4
(PC-4)
+ 1.35
- 1.04
5
Ru(bpz)3(PF6)2
+ 1.45
-0.80
6
(PC-1)
(PC-1)
-
-
-
-
-
-
-
-
-
-
25
34
32
16
10
7
5f
5g
5i
(PC-1)
(PC-1)
(PC-1)
8
9
Na2CO3
DBU
10
a Isolated yields, bpz: 2,2‘-bipyrazine
5j
5k
5m
The mechanism of the described transformation (shown in
Scheme 2) requires the oxidation of a carboxylate ion and
subsequent loss of CO2 to form the radical intermediate A in the
first SET step.10 Typically carboxylate ions have redox potentials
between +0.97V to +1.27V (vs SCE)20 thus only strongly oxidizing
photocatalysts in the excited state can be expected to be effective
for the desired conversion.
5n
5o
Figure 2. Scope of Michael acceptors.
In the second SET step the reduction of an -acyl radical
intermediate (EWG = acyl) is required; published redox potentials
of related derivatives are at -0.6 V vs. SCE.21 In our investigations
only photocatalysts with E(P/P-) ≤ -0.82 V were able to complete
the catalytic cycle (Table 1: entries 3 and 5). Best yields were
obtained when using inorganic salts as a base to deprotonate the
N-protected amino acid. On the contrary, the organic base DBU
only delivered 10% yield.
Parallelization of the studied photochemical reactions was
enabled by the use of the PhotoRedoxBox® photoreactor. 22
Depending on the chosen scale up to 32 reactions can be carried
out simultaneously in this device. We usually chose a 0.25 to 0.5
mmol scale to perform the reactions. After completion of the
reaction, the samples were simply filtered on a robotic platform
and directly submitted to our automated purification and isolation
process using HPLC chromatography with autonomous peak
detection (via UV and MS), fraction collection and subsequent
lyophilization and reformatting (including aliquoting for analysis).
We only report isolated yields after these fully automated
Figure 3. Relevant Ir-based17 and organic photocatalysts18,19 Abbreviations:
ppy, 2-phenylpyridyl; dtbbpy, 4,4‘-di-tert-butyl-2,2‘-bipyridyl; bpy, 2,2'-bipyridyl;
Cz, carbazole.
2
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202001178
European Journal of Organic Chemistry
COMMUNICATION
purification and isolation steps; therefore, we consider any yields
> 40% satisfactory.23
When using α-substituted amino acids, such as N-benzoyl valine,
we found that the introduction of an alkyl side chain led to slightly
lower yields (6ba, 6bc and 6bg). As expected for reactions at a
radical center, in these cases the chiral center in the starting
material was racemized under the reaction conditions (see
supporting information for epimerization studies with dipeptide 2g
and ent-2g).10
We then proceeded to assess the applicability of this method to
the functionalization of dipeptides, using Boc-Leu-Gly-OH 2a as
radical precursor and our initial set of Michael acceptors (Scheme
4). In general, the reactions proceeded smoothly, with PC-2 as
catalyst delivering slightly better yields than PC-1 for the
substrates tested (see product 7ab in scheme 4).
Scheme 4. Functionalization of Boc-Leu-Gly-OH 2a: Michael acceptors scope
(I).
Scheme 2. Proposed mechanism for the discussed decarboxylate alkylation.
With a suitable set of conditions in hand we proceeded to
investigate the scope of this transformation. Functionalization of
N-benzoyl glycine 1a with a selected set of Michael acceptors
afforded the corresponding coupling products in acceptable yields
up to 54%. (see Scheme 3). The best yields were obtained with
acrylate derivatives 5c and 5d, which have the highest
electrophilic reactivities among the selected Michael acceptors.24
Less reactive Michael acceptors such as 5e or 5g finally led to
decreased reaction efficiency (23% and 14% respectively).
Unreacted starting material (amino acid, peptide) was often
observed together with the decarboxylated amino acid derivative
as side product. These compounds were only observed by LCMS
and not isolated.
2a
5
7a
7aa (39%)
7ab (65%)b
7ac (43%)b
7ad (47%)
Scheme 3. Functionalization of N-benzoyl-α-amino acid derivatives 1: Michael
acceptors scope (I).
7ae (28%)
7af (21%)
a
Reactions conditions: PC-1 (1 mol%), Boc-Leu-Gly-OH 2a (1.5 equiv.),
1
5
6
Cs2CO3 (1.5 equiv.), Michael acceptor 5 (1 equiv.), DMF, r.t., 24h, Hepatochem
b
photoreactor (λ = 450 nm); PC-2 (1 mol%) , 2a (1.0 equiv.), and Penn PhD
photoreactor25 (λ = 450 nm) were used for this reaction.
The results obtained (cf. scheme 4) prompted us to stay with the
highly reactive ethyl acrylate 5c as Michael acceptor for further
reactivity assessment of various di- and tripeptides 2 or 3. We
were pleased to see that some of the tested reactions proceeded
with improved yields > 40% for these larger substrates (Scheme
5).
6aa (40%)
6ad (54%)
6ac (38%)
6ae (23%)
Scheme 5. Functionalization of various di- and tripeptides 2 and 3 with ethyl
acrylate 5c.
6ag (14%)
6af (29%)
2 or 3
5c
7 or 8
6bc (41%)
6ba (25%)
6bg (21%)
7bc (28%)
7dc (49%)b
7cc (25%)
a Reactions conditions: PC-1 (1 mol%), N-benzoyl-α-amino acid 1 (1.5 equiv.),
Cs2CO3 (1.5 equiv.), Michael acceptor 5 (1 equiv.), DMF, r.t., 24 h, Hepatochem
photoreactor (λ = 450 nm).
PG = Boc, 7ec (22%)
3
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202001178
European Journal of Organic Chemistry
COMMUNICATION
a
PG = Cbz, 7fc (78%))b
Reactions conditions: PC-2 (1 mol%), N-benzoyl glycine 1a or Boc-Leu-Gly-
OH 2a (1.5 equiv.), Cs2CO3 (1.5 equiv.), Michael acceptor 5 (1 equiv.), DMF,
r.t., 24h, Hepatochem photoreactor (λ = 450 nm).
Although the use of the Cbz protecting group (exemplified for the
Pro-Gly-OH dipeptide) led to more efficient reactions in
comparison with Boc (Scheme 5, entries 7ec vs 7fc), we preferred
Boc-derivatives for our studies due to their better commercial
availability.
7gc (32%)
7hc (52%)b
We then progressed to expand further the scope of Michael
acceptors with N-benzoyl glycine 1a and Boc-Leu-Gly-OH 2a
(Scheme 6) in combination with PC-2 as photocatalyst.
8ac (49%)b
8bc (19%)
In addition to classical Michael acceptor systems, our
methodology also allows for the C-terminal modification of
protected amino acids and dipeptides with vinyl substituted aryls.
Reasonable yields were obtained with 4-vinylpyridine (6al and
7aI), and even simple styrene was able to react with N-benzoyl
glycine 1a in low yield (6ao). As expected, the highest yields were
observed with the doubly activated Michael acceptor 5k. However,
using cyclopentenone 5h as Michael acceptor led to only
moderate yield (34%) 10
a Reactions conditions: PC-1 (1 mol%), dipeptide 2 or tripeptide 3 (1.5 equiv.),
Cs2CO3 (1.5 equiv.), ethyl acrylate 5c (1 equiv.), DMF, r.t., 24 h, Hepatochem
photoreactor (λ = 450 nm); b PC-2 (1mol%) , 2 or 3 (1.0 equiv.), and Penn PhD
photoreactor (λ = 450 nm) were used for this reaction.
Scheme 6. Functionalization of N-benzoyl glycine 1a and Boc-Leu-Gly-OH 2a:
Michael acceptors scope (II).
1a or 2a
5
6 or 7
Scheme 7. Application of methodology in LSF of established drugs Nateglinide
(4b), Alrestatin (4c) or drug-precursor to benazepril (4a).
6ah (23%)
6ai (9%)
4
5c
9
Entry
Drug
Producta (%)
6aj (28%)
6ak (53%)
6am (30%)
1
6al (46%)
4a
9ac (46%)b
2
6ao (11%)
6an (7%)
4b
9bc (44%)
3
7ah (34%)
7ai (14%)
4c
9cc (4%)
a Reactions conditions: PC-2(1 mol%), drug 4 (1.5 equiv.), Cs2CO3 (1.5 equiv.),
ethyl acrylate 5c (1 equiv.), DMF, r.t., 24 h, Penn PhD photoreactor (λ = 450
nm); b 1 equiv. of benazepril precursor 4a was used.
7aj (53%)
7ak (58%)
As late stage functionalization strategies are an emerging and
important topic for medicinal chemistry7, we looked for known
amino acid derived drugs, which could undergo a C3 carbon
homologation by the described protocol. We chose 4a, a
precursor of benazepril, Nateglinide 4b and Alrestatin 4c
(Scheme 7) to test our optimized reaction conditions. While 4a
and 4b reacted successfully with ethyl acrylate 5c, we observed
only minimal yield (4%) starting from 4c. We hypothesize that this
limitation is caused by the lower reactivity of the generated radical
7al (37%)
7an (20%)
7ao (6%)
4
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202001178
European Journal of Organic Chemistry
COMMUNICATION
due to the presence of the two carbonyl substituents in α-position
of the nitrogen.
Conclusions
We have demonstrated that decarboxylative 1,4-conjugate
additions are a valuable and broadly applicable synthetic strategy
for the C-terminal modification of small peptide substrates.
Compared to previous reports we have expanded the scope of
Michael acceptors to acyclic unsaturated ketones and in particular
less reactive acceptor systems - such as cinnamic acid ester and
vinyl-substituted (hetero)aromatic systems.
The use of the Hepatochem PhotoRedOxBox® device and the
robustness of the reaction conditions identified allowed us to
establish this reaction type in a parallel setup and made the use
of our automated purification and isolation protocols possible.
This enables fast derivatizations of a wider range of substrates
and allows the methodology to be employed in late stage
functionalization of amino acid containing drugs and drug
precursors, which was demonstrated for a precursor of benazepril,
Nateglinide and Alrestatin.
Acknowledgements
The authors thank Mrs. Anette Manss and Mr. Holger Theils for
execution of experiments and Mrs. Ute Messinger for analytical
support.
Keywords: 1,4-conjugate addition • photoredox • peptides • late
stage functionalization • parallel syntheses
KEY TOPIC:
Photochemistry, Late Stage Functionalization
TABLE OF CONTENT TEXT:
In our contribution we showcase an application of photochemistry
for Late Stage Functionalization (LSF) of amino acids and small
peptides in an industrial medicinal chemistry environment. The
radical intermediates generated under photochemical conditions
from the C-terminal carboxylates undergo rapid 1,4-conjugate
additions to a variety of Michael acceptors. The studied
methodology is applicable to the functionalization of established
drugs and drug precursors.
5
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202001178
European Journal of Organic Chemistry
COMMUNICATION
1
a) A. Michael, J. Prakt. Chem. 1887, 35, 349; b) A. Michael, J. Prakt. Chem. 1894, 49, 20.
2
a) M. S Kharasch, P. O. Tawney, J. Am. Chem. Soc 1941, 63, 2308, b) B. L. Feringa, R. Badorrey, D.Peña, S. R. Harutyunyan, A. J. Minnaard
PNAS, 2004, 101, 5834; c) A. C. Spivey, L. Shukla, J. F. Hyler, Org. Lett. 2007, 9, 891; d) T. Hellenbrand, K. T. Wanner, Eur. J. Org. Chem. 2014, 20,
4398.
3
a) C. A. Malapit, I. K. Luvaga, D. R. Caldwell, N. K. Schipper, A. R. Howell, Org. Lett. 2017, 19, 4460; b) F. Albrecht, O. Sowada, M. Fistikci, M.M.K.
Boysen, Org. Lett. 2014, 16, 5212; c) X. Lu, S. Lin, J. Org. Chem. 2005, 70, 9651.
4
5
G. Cardillo, L. Gentilucci, A. Tolomelli, C. Tomasini, Tetrahedron 1999, 55, 6231.
Selected reviews on 1,4-conjugate additions: P. Perlmutter, in Conjugate Addition Reactions in Organic Synthesis; Pergamon, Oxford, 1992. H.-G. Schmalz
in Comprehensive Organic Synthesis, (Eds. B.M. Trost, I. Flemming), Pergamon Press, Oxford, 1991, Vol4. Chapter 1.5.
6
Selected reviews on nucleophilic conjugate additions: a) N. Krause, A. Gerold, Angew. Chem,. Int. Ed. 1997, 36, 186; b) B. E. Rossiter, N. M. Swingle, Chem.
Rev. 1992, 92, 771; c) J. Leonard, E. Diez-Barra, S. Merino, Eur. J. Org. Chem. 1998, 2051; d) M. P. Sibi, S. Manyem, Tetrahedron 2000, 56, 8033, e) J. L.
Vicario, D. Badia, L. Carillo, Synthesis 2007, 2065; f) A. Alexakis, J. E. Bäckvall, N. Krause, O. Pàmies, M. Diéguez, Chem. Rev. 2008, 108, 2796; g) S. R.
Harutyunyan, T. den Hartog, K. Geurts, A. J. Minnaard, B. L. Feringa, Chem. Rev. 2008, 108, 2824.
7
a) T. Cernak, K. D. Dykstra, S. Tyagarajan, P. Vachal, S. W. Krska, Chem. Soc. Rev. 2016, 45, 546; b) M. Moir, J. J. Danon, T. A. Reekie, M. Kassiou,
Expert Opinion on Drug Discovery 2019, 14, 11379.
8
9
For a general review on photoredox catalysis see: C.-S. Wang, P. H. Dixneuf, J.-F. Soulé, Chem. Rev. 2018, 118, 7532.
Y. Yoshimi, T. Itou, M. Hatanaka, Chem. Commun. 2007, 5244.
10
11
12
13
Y. Yoshimi, M. Masuda, T. Mizunashi, K. Nishikawa, K. Maeda, N. Koshida, T. Itou, T. Morita, M. Hatanaka, Org. Lett. 2009, 11, 4652.
K. Osaka, M. Sugie, M. Yamawaki, T. Morita, Y. Yoshimi, Journal of Photochemistry and Photobiology A: Chemistry 2016, 317, 50.
C. Lingling, C. Ohta, Z. Zuo, D.W.C. McMillan, J. Am. Chem. Soc. 2014, 136, 10886.
a) V. Hornik, M. Afgaran, G. Gellerman, WO 99/65508, December 23, 1999; b) U. Plöckinger, U. Hoffman, M. Geese, A. Lupp, M. Buchfelder, J. Flitsch, W.
Vajoczy, W. Saeger, S. Schulz, C. Dohrmann, Eur. J. Endocrinol. 2012, 166, 223.
14
S. J. McCarver, J. X. Qiao, J. Carpenter, R. M. Borzilleri, M. A. Poss, M. D. Eastgate, M. M. Miller, D. W. C. MacMillan, Angew. Chem. Int. Ed. 2017, 56, 728.
J. L. Lau, M. K. Dunn, Bioorganic & Medicinal Chemistry, 2018, 26, 2700.
15
16
a) J-Q. Liu, A. Shatskiy, B. S. Matsuura, M. D. Kärkäs, Synthesis 2019, 51, 2759. b) F. J. A. Troyano, K. Merkens, K. Anwar, A. Gomez-Suarez, Angew.
Chem. Int. Ed. 2020, DOI: 10.1002/anie.202010157. c) L. Raynal. N. C. Rose, J. R. Donald, C. D. Spicer, Chem. Eur. J., 2020, DOI:
10.1002/chem202003779
17
18
19
T. Koike, M. Akita, Inorg. Chem. Front. 2014, 1, 562.
A. Joshi-Pangu, F. Lévesque, H. G. Roth, S. F. Oliver, L-C. Campeau, D. Nicewicz, D. A. DiRocco, J. Org. Chem. 2016, 81, 7244.
T-Y. Shang, L-H Lu, Z. Cao, Y. Liu, W-M, He, B. Yu, Chem. Commun. 2019, 55, 5408.
20
a) L. Capaldo, L. Buzzetti, D. Merli, M. Fagnoni, D. Ravelli, J. Org. Chem. 2016, 81, 7102; b) M. Galicia, F. J. Gonzalez, J. Electrochem. Soc. 2002, 149,
D46.
21
22
N. Bortolamei, A. A. Isse, A. Gennaro, Electrochim. Acta 2010, 55, 8312
PhotoRedoxBox® photoreactor is commercially available from Hepatochem, see https://www.hepatochem.com/photoreactors-leds-
accessories/photoredox-box/
23
a) P. Fricero, L. Bialy, W. Czechtizky, M. Méndez, J. P.A. Harrity, ChemMedChem, 2020, doi.org/10.1002/cmdc.202000187 b) T. C. Maier, W. Czechtizky
in Small Molecule Medicinal Chemistry: Strategies and Technologies, Wiley, Hoboken, 2015, pp. 323–351.
24
25
D. S. Allgäuer, H. Jangra, H. Asahara, Z. Li, Q. Chen, H. Zipse, A. R. Ofial, H. Mayr. J. Am. Chem. Soc. 2017, 139, 13318.
Penn PhD Photoreactor M2 is commercially available from Aldrich, see https://www.sigmaaldrich.com/catalog/product/sial/z744035?lang=de&region=DE
6
This article is protected by copyright. All rights reserved.