Biomimetic synthesis of nudicaulins i and II, yellow pigments from the Iceland poppy: Papaver nudicaule

Article abstract of DOI:10.1039/c9cc07943b

Indole and the anthocyanin orientalin proceed through a unique cascade sequence that leads to nudicaulins I and II in 92% yield. This biomimetic synthesis confirms the biosynthesis proposal for these structurally unprecedented flavoalkaloids that play a key role in the colour range displayed by the Iceland poppy.

Full text of DOI:10.1039/c9cc07943b

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Biomimetic synthesis of nudicaulins I and II,
yellow pigments from the Iceland poppy Papaver
nudicaule†
Cite this: Chem. Commun., 2019,
55, 13594
Received 10th October 2019,
Accepted 18th October 2019
Rory Devlin and Jonathan Sperry
*
DOI: 10.1039/c9cc07943b
rsc.li/chemcomm
Indole and the anthocyanin orientalin proceed through a unique
cascade sequence that leads to nudicaulins I and II in 92% yield. This
biomimetic synthesis confirms the biosynthesis proposal for these
structurally unprecedented flavoalkaloids that play a key role in the
colour range displayed by the Iceland poppy.
The petals of the Iceland poppy (Papaver nudicaule) display
colours ranging from red and orange to dark and light yellow.1
The pigments responsible for the red colouration2 are the
anthocyanins pelargonidin 3-O-b-sophoroside-7-O-b-glucoside
(orientalin; 1), pelargonidin-3-O-b-sophoroside (2) and pelargo-
nidin 3-O-b-[(6-malonyl)sophoroside] (3) (Fig. 1A).3,4 The struc-
tural assignment of the pigment responsible for the orange
and yellow colours, termed nudicaulin, has been the subject of
multiple investigations over many years since its isolation was
first reported by Robert Robinson.5–7 In 2013, Bringmann,
Schneider and co-workers8 finally solved this decades-old
conundrum and reported that nudicaulin is a diglycosylated
pentacyclic dihydrobenzofurocyclopentaindole that exists as a
mixture of C3–C11 cis-fused diastereomers, termed nudicaulins
I and II (4a and 4b) (Fig. 1B).9
Fig. 1 Pigments responsible for the range of colours displayed by the
Iceland poppy (Papaver nudicaule).
From a biosynthetic perspective, the nudicaulins are fascinating.
The nudicaulins are flavoalkaloids, a rare class of natural
products,10 that possess a b-glucoside at C7-OH and
a
change is observed when red petals are exposed to an aqueous
b-sophoroside at C11-OH that correspond to the substitution solution of indole.13,14
pattern in orientalin (1). It is therefore logical to assume that
A plausible biosynthesis12 of the nudicaulins is shown in
the pelargonidin glycosides and the nudicaulins are biosynthe- Scheme 1A. Nucleophilic attack by indole onto the orientalin
tically related. Through a series of labelling studies,11,12 Schneider C2-site15 followed by retro-6p-electrocyclic ring opening would
and co-workers showed that the nudicaulin framework is generate the quinone methide 5. Diastereoselective bis-
assembled in vivo from indole and kaempferol-3-O-b-sophoroside- cyclisation would form the cis-fused 5,5-ring system16 that upon
7-O-b-glucoside, the biosynthetic flavonol precursor to orientalin dehydrogenation would generate nudicaulins 4a and 4b. This
(1). Furthermore, the level of anthocyanins (red) decreases as the cascade process was recently shown to be viable; the quercetin-
level of nudicaulin (yellow) rises in the flower and the same colour derived anthocyanidin 6 was converted to the racemic nudi-
caulin aglycons 7 upon reaction with several indoles under
acidic conditions (Scheme 1B).13 However, the yields for this
process are poor, an excess of the indole partner is required and
School of Chemical Sciences, University of Auckland, 23 Symonds Street, Auckland,
New Zealand. E-mail: j.sperry@auckland.ac.nz
the products do not contain the correct substitution pattern nor
the glycosides present in the nudicaulins.
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c9cc07943b
13594 | Chem. Commun., 2019, 55, 13594--13597
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Communication
ChemComm
Scheme 3 Regioselective b-glycosylation of phloroglucinaldehyde 8.
Scheme 1 (A) Proposed biosynthesis of nudicaulin; (B) synthesis of the
nudicaulin aglycon.13
In view of our ongoing interest in biomimetic synthesis17
and flavoalkaloids,18 we were intrigued by the biosynthesis
proposal for the nudicaulins and instigated a programme to
validate this cascade process depicted in Scheme 1A. This study
required access to orientalin (1), a diglycosylated anthocyanin
not previously synthesised. It was assumed that orientalin
would be accessible by the condensation of the phloroglucinal-
dehyde 8 and the acetophenone 9 harbouring the requisite
b-glucoside and b-sophoroside units, respectively (Scheme 2).19,20
The simpler glycoside 8 was the initial target. The b-selective
glycosylation of the commercially available donor 2,3,4,6-tetra-
O-acetyl-a-D-glucopyranosyl bromide (10) with the known
phloroglucinaldehyde 1121 proceeded with the desired regio-
selectivity in the presence of silver(I) oxide and quinoline22 to
give b-glycoside 8 (Scheme 3).
Due to propensity for sophorose donors to undergo a-glyco-
sylation with unreactive acceptors in the absence of neighbouring
group effects,23 it was anticipated that the diastereoselective
synthesis of the b-sophoroside 9 would require more than one
donor to be screened (Scheme 4). Treatment of the 1,2-glucoside
1224 with acetic anhydride in the presence of catalytic quantities
of sulfuric acid gave octaacetylsophorose 13 that was readily
converted into the two sophorose donors, a-bromide 14 and the
a-trichloracetimidiate 1525 in good yield.
Scheme 4 Synthesis of sophorose donors 14 and 15.
Table 1 Glycosylation of sophorose donors 14 and 15 with acceptor 16
Temp.
Solvent (1C)
Time Yield 9
(h) (a : b)
Entry Donor Activator (eq.)
1
2
3
4
14
14
15
15
Ag2CO3 (2)
CH2Cl2 rt
43
—
—
BF3ÁEt2O (1)
BF3ÁEt2O (0.5)
TMSOTf (0.5)
CH2Cl2 À41 - rt 24
CH2Cl2 À41 - rt
CH2Cl2 À41 -
À20
2
24% 2 : 1
0.8 99% a
only
0.25 74% 1 : 2
0.5 28% 1 : 1
0.25 Trace ND
5
15
15
15
15
TMSOTf (0.5)
TMSOTf (0.5)
TMSOTf (0.5)
Pd(MeCN)4(BF4)2
(0.2)
MeCN À41
EtCN À41
MeCN À41
6
7a
8
CH2Cl2 À41 - rt 24
16% 1 : 2
9
15
Pd(MeCN)4(BF4)2 (1) CH2Cl2 rt - 40
Pd(MeCN)4(BF4)2 (1) CH2Cl2 rt
17
24
1
24% 1 : 2
Trace ND
60%
10a 15
11
15
AuCl3 (0.1)
CH2Cl2 À78 - rt
1 : 1.3
a
Reverse mode addition.
With the sophorose donors 14 and 15 in hand, b-glycosylation
with the hydroxyacetophenone 1626 as acceptor could be investi-
gated (Table 1). The sophorose donor 14 has been reported
to undergo b-selective glycosylations with reactive secondary
alcohols in the presence of silver carbonate,25 but applying
these conditions using the less reactive hydroxyacetophenone
16 failed (entry 1). The two donors 14 and 15 were subsequently
screened in the glycosylation using the standard promotor BF3Á
Et2O in an effort to determine which was most suitable. The
glycosylation failed using the donor 14 (entry 2), but the donor
Scheme 2 Proposed synthesis of orientalin (1).
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun., 2019, 55, 13594--13597 | 13595

View Article Online
ChemComm
Communication
to determine if this 2 : 3 ratio is in line with that observed in vivo.
The 92% yield for this reaction is striking, especially considering
the complexity of the biomimetic cascade and the yields obtained
on simpler model systems (6–17%)13 (Scheme 1B). The nudi-
caulins form yellow aqueous solutions, in line with their reported
role as pigments in Papaver nudicaule petals.
In conclusion, the total synthesis of the flavoalkaloid pigments
nudicaulins I and II has been accomplished, 80 years after their
first reported isolation. The fusion of indole and the anthocyanin
orientalin via an unprecedented biomimetic cascade affords
nudicaulins I and II in excellent yield. This biomimetic synthesis
verifies the biosynthesis of these pigments and provides material
for colour-scent association studies.
We are indebted to the Royal Society of New Zealand for a
Rutherford Discovery Fellowship (J.S) and the University of
Auckland for a doctoral scholarship (R.D). We thank Professor
Bernd Schneider (Max Planck Institute for Chemical Ecology)
for insightful discussion and providing the 13C NMR spectra for
nudicaulins I and II.
Scheme 5 Biomimetic synthesis of nudicaulins I and II.
Conflicts of interest
15 did provide the glycoside product 9 (entry 3), albeit in poor
yield and as an inseparable mixture favouring the undesired
a-anomer. The sophorosyl trichloracetimidate donor 15 was
deemed the more suitable substrate and was thus employed in
an optimisation study. Switching the Lewis acid to TMSOTf led
to a marked increase in yield, but only the a-sophoroside was
formed (entry 4). Glycosylation of 15 via the acetonitrile derived
a-nitrilium-nitrile conjugate27 let to an anomeric mixture
favouring the b-sophoroside in good yield (entry 5). This ratio
decreased upon use of propionitrile (entry 6) and reverse mode
addition28 led to no reaction (entry 7). Cationic palladium(II)-
catalysed glycosylation29 gave a good ratio favouring the desired
b-anomer, but in poor yield (entry 8). Attempts to improve the
yield and selectivity by employing stoichiometric amounts of the
palladium(II)-salt were unsuccessful (entry 9), with reverse mode
addition failing to provide any improvement (entry 10). The
reported b-selectivity exhibited by gold(III)-salts in glycosylations30
partially translated onto this system, slightly favouring the for-
mation of the b-sophoroside (entry 11). It was at this point we
returned to the reaction conditions outlined in entry 5 and 9 was
used as the 1 : 2 mixture in the subsequent step.
There are no conflicts to declare.
Notes and references
1 B. Dudek, A. C. Warskulat and B. Schneider, Plants, 2016, 5, 28.
2 (a) G. M. Robinson and R. Robinson, Biochem. J., 1931, 25,
1687–1705; (b) G. M. Robinson and R. Robinson, Biochem. J., 1932,
26, 1647–1664.
3 J. B. Harborne, Phytochemistry, 1963, 2, 85–97.
4 G. Cornuz, H. Wyler and J. Lauterwein, Phytochemistry, 1981, 20,
1461–1462.
5 J. R. Price, R. Robinson and R. Scott-Moncrieff, J. Chem. Soc., 1939,
1465–1468.
6 J. B. Harborne, Phytochemistry, 1965, 4, 647–657.
7 W. Schliemann, B. Schneider, V. Wray, J. Schmidt, M. Nimtz,
¨
A. Porzel and H. Bohm, Phytochemistry, 2006, 67, 191–201.
¨
8 E. C. Tatsis, A. Schaumloffel, A. C. Warskulat, G. Massiot,
B. Schneider and G. Bringmann, Org. Lett., 2013, 15, 156–159.
9 For structural variants of the nudicaulins throughout other papa-
veraceous species, see: E. C. Tatsis, H. Bohm and B. Schneider,
Phytochemistry, 2013, 92, 105–112.
10 L. M. Blair, M. B. Calvert and J. Sperry, Flavoalkaloids-Isolation,
Biological Activity and Total Synthesis, in The Alkaloids, ed. H. J.
Knolker, Elsevier, Oxford, 2017, vol. 77, pp. 85–115.
¨
ˇ
11 E. C. Tatsis, E. Eylert, R. K. Maddula, E. Ostrozhenkova, A. Svatos,
W. Eisenreich and B. Schneider, ChemBioChem, 2014, 15, 1645–1650.
Hydrogen chloride gas was bubbled through a solution of 12 A. C. Warskulat, E. C. Tatsis, B. Dudek, M. Kai, S. Lorenz and
B. Schneider, ChemBioChem, 2016, 17, 318–327.
13 B. Dudek, F. Schnurrer, H.-M. Dahse, C. Paetz, A. C. Warskulat,
phloroglucinaldehyde 8 and the acetophenone 9, inducing
heteroannulation to give a peracetylated orientalin derivative
C. Wiegel, K. Voigt and B. Schneider, Molecules, 2018, 23, 3357.
that upon global deprotection gave orientalin (1) as a bright red 14 The nudicaulins play a key role in colour-scent association that
´
influences pollinator behaviour. J. Martınez-Harms, A. C. Warskulat,
solid (Scheme 5). With 1 in hand, the biomimetic synthesis of
the nudicaulins was attempted. A buffered solution of orientalin
B. Dudek, G. Kunert, S. Lorenz, B. S. Hansson and B. Schneider,
ChemBioChem, 2018, 19, 1553–1562.
(1) and indole in an open flask went from bright red to yellow over 15 L. Jurd, Tetrahedron, 1975, 31, 2884–2888.
16 Due to the level of ring strain, trans-fused 5,5-ring systems are quite
a period of seven hours, from which the nudicaulins 4a and 4b
were isolated in 92% yield. The product existed as a 2 : 3 mixture
of C3–C11 cis-diastereomers favouring nudicaulin II, suggesting
the sugar units are imparting modest diastereoselectivity during
the biomimetic cascade process.31 The diastereomeric ratio of
4a :4b was not disclosed in the isolation report, so we are unable
rare. For some examples, see: (a) L. A. Paquette, A. T. Hamme,
L. H. Kuo, J. Doyon and R. Kreuzholz, J. Am. Chem. Soc., 1997, 119,
1242–1253; (b) G. A. Molander, P. J. Nichols and B. C. Noll, J. Org.
Chem., 1998, 63, 2292–2306; (c) T. F. Jamison, S. Shambayati,
W. E. Crowe and S. L. Schreiber, J. Am. Chem. Soc., 1997, 119,
4353–4363; (d) M. Kock, A. Grube, I. B. Seiple and P. S. Baran, Angew.
Chem., Int. Ed., 2007, 46, 6586–6594.
13596 | Chem. Commun., 2019, 55, 13594--13597
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Communication
ChemComm
¨
17 (a) E. M. Boyd and J. Sperry, Org. Lett., 2015, 17, 1344–1346; (b) E. K. 25 N. Hoffmann, J. Pietruszka and C. Soffing, Adv. Synth. Catal., 2012,
Davison, P. A. Hume and J. Sperry, Org. Lett., 2018, 20, 3545–3548.
18 M. B. Calvert and J. Sperry, Chem. Commun., 2015, 51, 6202–6205.
19 A. Siegel, Ambix, 2008, 55, 62–82.
354, 959–963.
26 J. Huang, J. Li, J. Zheng, W. Wu, W. Hu, L. Ouyang and H. Jiang, Org.
Lett., 2017, 19, 3354–3357.
20 For recent experimental procedures: (a) O. Dangles and H. Elhajji, 27 R. R. Schmidt, M. Behrendt and A. Toepfer, Synlett, 1990, 694–696.
Helv. Chim. Acta, 1994, 77, 1595–1610; (b) Q. Zhang, N. P. Botting 28 For reverse mode additions benefiting glycosylations, see: (a) K. B.
and C. Kay, Chem. Commun., 2011, 47, 10596–10598.
21 M. Kraus, E. Biskup, E. Richling and P. Schreier, J. Labelled Compd.
Radiopharm., 2006, 49, 1151–1162.
22 M. L. Belyanin, E. V. Stepanova and V. D. Ogorodnikov, Carbohydr.
Res., 2012, 363, 66–72.
Mishra, A. K. Singh and J. Kandasamy, J. Org. Chem., 2018, 83,
4204–4212; (b) P. Peng and R. R. Schmidt, Acc. Chem. Res., 2017, 50,
1171–1183; (c) P. Peng and R. R. Schmidt, J. Am. Chem. Soc., 2015,
137, 12653–12659.
29 J. Yang, C. Cooper-Vonosdell, E. A. Mensah and H. M. Nguyen,
J. Org. Chem., 2008, 73, 794–800.
23 (a) H. P. Wessel and G. Englert, J. Carbohydr. Chem., 1994, 13, 1145–1158;
(b) W. Seebacher, E. Haslinger, K. Rauchensteiner, J. Jurenitsch, 30 P. Peng and R. R. Schmidt, J. Am. Chem. Soc., 2015, 137, 12653–12659.
A. Presser and R. Weis, Monatsh. Chem., 1999, 130, 887–897.
24 B. Coxon and H. G. Fletcher Jr, J. Org. Chem., 1961, 26, 2892–2894.
31 The 1H and 13C NMR data for the natural and synthetic samples are
in excellent agreement. See ESI† for full details.
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun., 2019, 55, 13594--13597 | 13597