Professional Documents
Culture Documents
Strategies and Tactics in Organic Synthesis 1St Edition Michael Harmata Full Chapter PDF
Strategies and Tactics in Organic Synthesis 1St Edition Michael Harmata Full Chapter PDF
https://ebookmass.com/product/strategies-and-tactics-for-the-mbe-
strategies-tactics-for-the-mbe-book-1-ebook-pdf/
https://ebookmass.com/product/organic-synthesis-using-
biocatalysis-1st-edition-goswami/
https://ebookmass.com/product/firefighting-strategies-and-
tactics-3rd-edition-ebook-pdf-version/
https://ebookmass.com/product/strategies-tactics-for-the-mbe-bar-
review/
eTextbook 978-1305506381 Managerial Economics:
Applications, Strategies and Tactics
https://ebookmass.com/product/etextbook-978-1305506381-
managerial-economics-applications-strategies-and-tactics/
https://ebookmass.com/product/disaster-response-and-recovery-
strategies-and-tactics-for-resilience-2nd-edition-ebook-pdf/
https://ebookmass.com/product/carbon-monoxide-in-organic-
synthesis-carbonylation-chemistry-gabriele-bartolo-ed/
https://ebookmass.com/product/organic-corrosion-inhibitors-
synthesis-characterization-mechanism-and-applications-
chandrabhan-verma/
https://ebookmass.com/product/stereochemistry-and-organic-
reactions-conformation-configuration-stereoelectronic-effects-
and-asymmetric-synthesis-1st-edition-dipak-kumar-mandal/
EDITORIAL BOARD
Volume 15
Edited by
Michael Harmata
University of Missouri,
Columbia, MO, United States
Elsevier
Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
Copyright © 2021 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by
any means, electronic or mechanical, including photocopying, recording, or any
information storage and retrieval system, without permission in writing from the
publisher. Details on how to seek permission, further information about the Publisher’s
permissions policies and our arrangements with organizations such as the Copyright
Clearance Center and the Copyright Licensing Agency, can be found at our website:
www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under
copyright by the Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and
experience broaden our understanding, changes in research methods, professional
practices, or medical treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge
in evaluating and using any information, methods, compounds, or experiments
described herein. In using such information or methods they should be mindful of
their own safety and the safety of others, including parties for whom they have a
professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or
editors, assume any liability for any injury and/or damage to persons or property as a
matter of products liability, negligence or otherwise, or from any use or operation of
any methods, products, instructions, or ideas contained in the material herein.
Library of Congress Cataloging-in-Publication Data
A catalog record for this book is available from the Library of Congress
Wen-Ju Bai, Amgen Research, One Amgen Center Drive, Thousand Oaks, CA, United
States
Korey Bedard, Brock University, St. Catharines, ON, Canada
Tanner W. Bingham, Department of Chemistry, University of Illinois, Urbana, IL,
United States
Gaëlle Blond, Université de Strasbourg, CNRS Strasbourg, France
Reinhard Brückner, Institut für Organische Chemie, Albert-Ludwigs-Universität,
Freiburg, Germany
Eric M. Ferreira, Department of Chemistry, University of Georgia, Athens, GA,
United States
Saswata Gupta, Department of Chemistry, University of Illinois at Chicago, Chicago,
IL, United States
Lucas W. Hernandez, Department of Chemistry, University of Illinois, Urbana, IL,
United States
Tomas Hudlicky, Brock University, St. Catharines, ON, Canada
Stephen K. Jackson, OmegaChem, Lévis, QC, Canada
Phil C. Knutson, Department of Chemistry, University of North Carolina at Chapel
Hill, Chapel Hill, NC, United States
Julia Kopp, Institut für Organische Chemie, Albert-Ludwigs-Universität, Freiburg,
Germany
Daesung Lee, Department of Chemistry, University of Illinois at Chicago, Chicago, IL,
United States
Christopher C. McAtee, Department of Chemistry, University of Michigan, Ann
Arbor, MI, United States
Ryan P. Murelli, Department of Chemistry, Brooklyn College, The City University of
New York, Brooklyn, NY, United States; PhD Program in Chemistry, The Graduate
Center, The City University of New York, New York, NY, United States; PhD
Program in Biochemistry, The Graduate Center, The City University of New York,
New York, NY, United States
Eric T. Newcomb, Pfizer Boulder Research and Development, Boulder, CO, United
States
Thomas R.R. Pettus, Department of Chemistry and Biochemistry, University of
California, Santa Barbara, CA, United States
xv
xvi Contributors
xvii
xviii Preface
Michael Harmata
January, 2021
Chapter 1
Amaryllidaceae isocarbostyril
alkaloids
Tanner W. Bingham, Lucas W. Hernandez and David Sarlah*
Department of Chemistry, University of Illinois, Urbana, IL, United States
*Corresponding author at: sarlah@illinois.edu.
Chapter Outline
1. Introduction 1 4. Narciclasine and lycoricidine 33
2. Synthetic strategy 12 4.1 Optimization of Ni(II)
2.1 Retrosynthetic analysis 12 conditions 33
2.2 Development of the 4.2 Initial synthesis of
dearomative narciclasine 36
carboamination reaction 12 4.3 Scalable route to
2.3 Mechanistic hypothesis 17 (þ)-lycoricidine (3) and
2.4 Optimization of our new (þ)-narciclasine (4) 40
nickel-catalyzed 4.4 Synthesis and biological
conditions 19 evaluation of C-7 analogs 42
3. Pancratistatins 21 4.5 Metabolic studies 46
3.1 Initial investigations 21 5. Conclusion 46
3.2 Optimizing the route 24 Acknowledgments 49
3.3 Methods for lactam References 49
formation 27 Further reading 52
3.4 C-7 oxidation and final
route 31
1. Introduction
The anticancer properties of crude plant extracts from the Amaryllidaceae
family have been recorded in the literature since the time of the ancient Greeks,
with Hippocrates prescribing Narcissus oils for the treatment of uterine tumors.1
Recently, the therapeutic effects of these extracts have been attributed to the
isocarbostyril alkaloids (þ)-7-deoxypancratistatin (1), (þ)-pancratistatin (2),
(þ)-lycoricidine (3), and (þ)-narciclasine (4) (Fig. 1.1). The first of this family
Strategies and Tactics in Organic Synthesis. https://doi.org/10.1016/B978-0-12-822212-6.00004-7
Copyright © 2021 Elsevier Inc. All rights reserved. 1
2 Strategies and Tactics in Organic Synthesis
OH OH
2
HO OH OH
1 3
10
9 10a 4a 4 O
O OH O 10b OH
NH 6 NH 5
O NH
O O 8 6a
7
R O O R O
(+)-7-deoxypancratistatin (1, R=H) (+)-lycoricidine (3, R=H)
(+)-pancratistatin (2, R=OH) (+)-narciclasine (4, R=OH)
Narciclasine (4) 0.016 0.042 0.011 0.010 0.010 0.027 0.011 0.011
3
4 Strategies and Tactics in Organic Synthesis
of MX-1 breast carcinoma tumors,12 and invasive Hs683 and GL19 human
glioblastomas.13 Furthermore, narciclasine (4) inhibited mitosis in carcinoma,
glioma, and melanoma and was found to impair cancer cell migration.8 In
addition to their potent anticancer activity, 1e4 also showed significant anti-
viral activity against Japanese encephalitis, yellow fever, Rift Valley fever, and
dengue type 4 viruses,14 and narciclasine (4) has been found to attenuate diet-
induced obesity15 and possess antiinflammatory properties.16
Although these compounds have shown a myriad of promising activities,
their precise mode of action has yet to be uncovered. Studies suggest that
pancratistatin’s (2) anticancer activity proceeds through intrinsic apoptosis, as
evidenced by release of caspase-9 and caspase-3, exposure of phosphatidyl
serine, and destabilization of mitochondrial membrane potential.10,11,17
Interestingly, the cytotoxic activity of narciclasine (4) has been attributed to
extrinsic apoptosis as evidenced by the release of caspase-8 and the activation
of the Fas and death receptor 4 (DR4) or death-inducing signaling complex
(DISC).9 Yet, despite their structural homology, there have been no studies
directly comparing their mechanisms of action.9,18e22 Aside from its cytotoxic
activity, narciclasine has also displayed activity in various cytostatic pathways.
It has been shown to disrupt cytokinesis through two different pathways,
disruption of actin bundle formation through the binding of the translation
elongation factor eEF1A,25 and the formation of F-actin stress fibers through
the activation of GTPase RhoA.13 Additionally, narciclasine has been shown to
bind to the A-site of the 60S ribosome, thereby directly blocking peptide
synthesis.23,24 Notably, many of these promising activities of 4 have yet to be
shown for 2, thus necessitating further investigation and comparison of these
compounds and their structureeactivity relationships (SAR).
To fully elucidate their mechanism(s) of action and promote their pre-
clinical development, scalable access to isocarbostyrils 1e4 was needed.
Various isolation protocols for narciclasine (4) are reported in the literature,
mostly from the bulbs of the Narcissus plant, the yields of which can range
from 30 to 200 mg/kg depending on the species and time of year.19 When
harvested from the Hawaiian wilderness, Hymenocallis litoralis is the highest
yielding source of 2 (144 mg/kg) and 3 (222 mg/kg).5 However, when culti-
vated in fields and greenhouses in Arizona, the isolation yield dropped
significantly to 22 mg/kg for 2 and 15 mg/kg for 3 in the peak month of
October. This lack of availability has inhibited the investigations into their
bioactivities. Additionally, although the initial studies into these compounds
revealed their potential use as therapeutics, such as their notoriously low
aqueous solubility, there are drawbacks to overcome for these compounds to
become medicinally relevant.26
Owing to the lack of scalable access and the need for further biological
evaluations, as well as their structural complexity, 1e4 have attracted
considerable attention from the synthetic community. To date, there have been
11,27e38 13,30,39e50 14,27,38,51e62 and 843,57,62e67 syntheses of 1e4,
Amaryllidaceae isocarbostyril alkaloids Chapter j 1 5
respectively, as well as many formal, semi, epi-, and analog syntheses pub-
lished over the years.18e22 As such, there have been various strategies
developed for installing the correct stereocenters on the cyclitol core and
constructing the lactam ring. One notable approach, reported by the Hudlicky
group in 1992, involves the use of microbial arene oxidation to stereo-
selectively install the syn-diol present on the cyclitol core (Scheme 1.1),54
subjecting bromobenzene (5) to the bacterial dioxygenases of Pseudomonas
putida, followed by acetonide protection of the resulting diol-delivered diene
6. Subsequent stereoselective [4 þ 2] cycloaddition of diene 6 with an acyl
nitroso species provided bicycle 7a, which contains all the stereocenters
present in the cyclitol core of (þ)-lycoricidine (3). Reduction of the bromide
followed by cleavage of the NeO bond was accomplished with mercury
aluminum amalgam. Silyl protection followed by imide formation with
2-bromo-piperonyloyl chloride furnished key intermediate 9a. Closure of the
lactam was attempted with various methods including radical cyclization,
epoxide-opening, and the Pd-catalyzed Heck reaction. Ultimately, they found
that the modified Heck conditions, initially reported by Grigg and co-
workers,68 and optimized for a diastereomer of 9 by Chida and coworkers,53
were the only conditions that could provide the benzolactam. However, the
product was isolated as a mixture of fully protected, desilylated, and free
amide products. Fortunately, these could be isolated together and subjected to
global deprotection to furnish lycoricidine 3 in 9 steps and 25% yield overall.
Notably, 3 years later, Hudlicky was able to take advantage of this same
strategy of microbial arene oxidation for his synthesis of (þ)-pancratistatin.
Stereoselective aziridination followed by radical dehalogenation of diene 6
provided vinyl aziridine 7b. Installment of the necessary trans relationship
between the arene and amine then became possible by the nucleophilic ring
opening of the aziridine at the allylic position by cuprate 8b. Arylated inter-
mediate 9b was then further elaborated to yield (þ)-pancratistatin (1) in the
first asymmetric total synthesis in 14 steps and 2% overall yield. These
landmark total syntheses were the state-of-the-art before we started our work
on these molecules, and they demonstrated the power of dearomative pro-
cesses for the rapid introduction of complexity to feedstock petrochemical
starting materials.
Key steps in other notable approaches to this family of natural products are
shown in Fig. 1.2. In Trost’s 1995 synthesis of (þ)-pancratistatin, a desym-
metrization reaction was employed to define three existing stereocenters and
install a fourth in the product 12.41 In 1998, Magnus applied a beazidonation
strategy toward (þ)-pancratistatin, to yield allylic azide 14.42 Rigby reported
an interesting photocyclization approach in 2000.15 While the absolute ste-
reochemistry was dictated by the TBS protected alcohol in 15, the required
relative stereochemistry was imparted by the 6p-conrotatory photochemical
cyclization and subsequent suprafacial [1,5]-H shift to yield product 16. In
2009, Madsen utilized a Zn-mediated fragmentation to afford diene 19 that
6 Strategies and Tactics in Organic Synthesis
OH O
O Bu4NIO4 Br 1. PhI=NTs, O 8b
O AlH2(Hg) Br O Cu(acac)2 BF3·Et2O O O
CbzNHOH O
THF CH2Cl2 2. Bu3SnH, O 75% NHTs
O TsN O
91% O 74% O AIBN CONMe2
NHCbz N 42% OTBS
8a Cbz 7a 6 7b 9b
OSiMe2iPr OH 5 OH
O
O OH HO OH
1. Pd(OAc)2 7 steps
Br TlOAc, DIPHOS O O
O O O O OH
2. Pd/C, EtOH OH NHBoc
NCbz cyclohexene NH NH O CONMe2
O O O
OH
O 23% O OH O
9a (+)-lycoricidine (3) (+)-pancratistatin (1) 10
9 steps 10% overall yield 14 steps 2% overall yield
SCHEME 1.1 Hudlicky’s dearomative approach to (þ)-lycoricidine (3) and (þ)-pancratistatin (2).
Trost - 1995 Magnus- 1998
OTIPS OTIPS
OCO2Me OCO2Me PhIO
(pC3H7PdCl)2 (0.5 mol%)
O O TMSN3
(R,R)-DACH-phenyl (0.75 mol%) Ar Ar
95%
O TMSN3 O N3
O
82% 13 Ar = 14
OCO2Me N3 O
11 12 OMe
Madsen - 2009
Rigby - 2000 O OMe
O O
TBSO hn TBSO I
OH
Zn
30% O TESO OBn
17 THF, O OBn
O NPMB O NPMB O Br H2O, MeOH O
O
OMOM
Padwa - 2007 Banwell - 2007 OMOM
OH
MeO2C SnBu3 OH
O Br OMOM
16 MeO2C O
O IO O B NH2
27
CuCl/Pd(0) O O O OH
LiCl, DMSO 1. Pd(PPh3)4, K2CO3, mwave
O NPMB O NPMB O NH
O 2. BSBr O
82%
O O 30% over
24 25 MOMO OMe OH O ent-4
26 two steps
7
FIGURE 1.2 Selected approaches toward the Amaryllidaceae isocarbostyril alkaloids.
8 Strategies and Tactics in Organic Synthesis
O Tf2O O O HCl O
OAc OAc OAc OAc
O HN OR O N O N O NH
O
OPG O OPG OPG OPG O
OR
28
R=OMe, OEt, OtBu 29 or 30 31
9
Arene derivatives Lactam derivatives
TMS O O O
OH OH OH OH OH
NH NH O O O NH O OH
TMS
O O O O OMe
32 33 34 35 36
Banwell - 200770 Hudlicky - 200671 Chapleur - 200472 Chapleur - 199373 Kornienko - 200674
inactive inactive inactive inactive inactive
Cyclitol derivatives Cyclitol skeleton derivatives
OH OH OH OH
OH HO HO OH
O OH O O OH O OH
O NH O NH O NH O NHBz
O O O
37 38 39 40
McNulty - 200575 Banwell - 200770 Kornienko - 200976 McNulty - 200877
inactive inactive inactive inactive
More potent derivatives
OH OBz OH OH OH
OH OH BzHN OH OH
O O O O N
OH OH OH OH
O NH O NH O NH O NH
OH O OH O OH O OH O
41 42 43 44
(+)-transdihydronarciclasine6 Hudlicky - 201278 Marion - 200979 Hudlicky - 201580
IC50 = 3.2 nM (P-388) IC50 = 50 nM (HCT116) IC50 = 4.7 nM (HCT116) IC50 = 193.7 nM (HCC1954)
3774 and 3869 show the beneficial role of each of the alcohols on the core,
while structural derivates 3975 and 4076 showcase the need for the polyol to be
cyclic. Together, this information suggests that much of the core scaffold of
the natural product is necessary for the high activity. However, the increased
activity of (þ)-trans-dihydronarciclasine (41) (IC50 ¼ 3.2 nM in P388 leuke-
mia cells) over (þ)-pancratistatin (2) (IC50 ¼ 52 nM in P388 leukemia cells)
suggests that the C-1 hydroxyl in 1 is detrimental for its activity.6 This led
Hudlicky, and Marion, to synthesize analogs 4277 and 43,78 respectively, which
have increased activity when compared with 1e4. Since the C-1 position is
spatially close to the C-10 position, Hudlicky and coworkers wanted to explore
this region as well. This led to analog 44,79 which was slightly less active in
HCC1954 human T-cell leukemia cells (193.7 nM) when compared with its
natural counterpart, narciclasine (4) (173.2 nM). While this analog did not
show increased activity, this position has been understudied as no other ana-
logs of the C-10 position have been synthesized, and its proximity to the C-1
position suggests that hydrophobic groups may lead to increased activity.
Lastly, little effort has gone into derivatizing the C-7 position. As previously
stated, the presence of a free phenol at that position increases activity. Other
than protected phenols, few other analogs of this position have been produced
despite its importance in the activity of these compounds.18e22 All this in-
formation together points at three distinct positions of the pharmacophore that
are underexplored: C-1, C-7, and C-10 (Fig. 1.4). As most previously reported
strategies required a de novo synthesis to access each analog, and hampered
efficient SAR studies, the development of a more facile route toward these
natural products as well as a means of late-stage functionalization is highly
desirable. With this in mind, we set out to develop more concise and scalable
syntheses of these compounds that would also expedite the synthesis of
analogs.
Hydrophobic
Groups
C-10 C-1 OH
OH
O OH
O NH
O
C-7
Hydrophilic
Groups
FIGURE 1.4 Potential sites of diversification.
12 Strategies and Tactics in Organic Synthesis
2. Synthetic strategy
2.1 Retrosynthetic analysis
In our retrosynthetic analysis of the Amaryllidaceae isocarbostyril alkaloids,
we were aware that the greatest synthetic challenge resided in the densely
functionalized cyclitol core encompassing six and four contiguous stereocenters.
As discussed, several methods were already in place for the installation of the
benzolactam, i.e., the BischlerNapieralski reaction. Furthermore, in many of
the previous total syntheses, significant effort was required for the installation of
the requisite functionalities for the annulation, that led to a dramatic increase in
sequential operations (protection, lactam formation, and deprotection). Based on
these observations, we knew there were safer methods to install the lactam,
giving us the freedom to initially focus on the cyclitol core; particularly the
rapid and controlled formation of six contiguous stereocenters, enabling rapid
downstream installment of the lactam ring of the natural products (Scheme 1.3).
With this in mind, our first simplifying disconnect was lactam formation,
tracing back to aminocyclitol 45 and epoxy acetonide 46. We realized that
installation of the oxygen functionalities would be most straightforward at this
point, but we were not sure what selectivity would be inherent to the system to
install the correct relative stereochemistry and were resigned to an empirical
approach. By design, both disconnects revealed a convergent intermediate,
diene 47, which we designated as the key motif that would enable our syn-
thesis, that being: concise, scalable, and highly modular for the synthesis of
analogs. However, this meant that our total syntheses would rely on the
scalable preparation of diene 47. Additionally, while not of immediate priority,
but of overall importance, the asymmetric synthesis of these natural products
was crucial to allow us to test their biological activity, as the enantiomers were
shown to be inactive.69 Based on these requirements, we believed deriving this
product from benzene through a dearomative carboamination reaction would
be the most straightforward approach.
13
14 Strategies and Tactics in Organic Synthesis
R O
N
concurrent development [TM]
N NMe
O
HO O O R [M]
N N
HO OsO4 [TM]
N NMe N NMe dearomative carboamination R
O O NUR
R [M]
benzene-MTAD
cycloadduct (51)
N
N O
[TM]n+2
NMe
O
fraught with challenges. For example, the selective induction of trans ste-
reochemistry as well as formation of the desired 1,2-product rather than the
1,4-product would be difficult. Furthermore, we knew that the benzene-MTAD
cycloadduct (51) was unstable to temperatures above 40 C. Unfortunately,
most conditions for allylic substitution with softer nucleophiles required
elevated temperatures.
With these concerns in mind, we chose to simplify our problem signifi-
cantly. Based on our limited studies at the time, we knew that the generation
and isolation of the naphthalene-MTAD cycloadduct (52) was possible in a
51% yield (Fig. 1.5A). Furthermore, we knew that 52 was stable at higher
temperatures (i.e., 0 C) for a short period of time and believed this would
allow us to greatly expand the number of screening experiments we could run
at a time. With 52 in hand, we immediately tested traditional conditions with
softer nucleophiles to no avail.81e84 We then turned to harder nucleophiles and
found a report from Pineschi and coworkers that described a copper-catalyzed
allylic substitution of diazabicyclic systems employing Grignard reagents.85
We were particularly attracted to these conditions as not only could they be run
cryogenically, but they also employed rac-BINAP as the ligand which gave
hope to our long-term desire for asymmetric induction.
To our delight, when we tested Pineschi’s conditions, albeit with a higher
catalyst loading, our desired dearomative trans-carboaminated product 53 was
afforded with exclusive selectivity in 60% isolated yield. After a short ligand
screen, we eventually found that dppf gave us the highest yield of 72%. Yet, a
model system will always be a model system. When we moved our conditions
to begin mirroring our actual desired transformation, i.e., in situ generation of
the cycloadduct, we found a significant decrease in yield of ca. 42%
(Fig. 1.5B). Despite this reduction in efficiency, we pushed forward and tested
our conditions with benzene (49) as the substrate to find that no reaction was
observed (Fig. 1.5C). While many hypotheses could be put forth to explain this
dearth in reactivity and potentially allow for optimization to a working system,
we knew we needed to consider alternative options. As is often the case when
working with a model system in total synthesis, we had excellent conditions
that failed to perform on the actual substrate.
Undeterred, we decided to go for the home-run approach and merely swap
the transition metal and keep the remaining conditions consistent (Fig. 1.5D).
While many of the metals we tested either underperformed or did not react, we
found that in the case of Ni(cod)2, there was an increase in yield when
naphthalene (54) was used as substrate from 42% to 66%. However, there was
still one more reaction that needed to succeed before our proof of concept
could be validated, and that was the dearomative carboamination of benzene
(49, Fig. 1.5E). We knew that if this transformation did not work with the
reaction conditions as developed, we would either need to redo the screen with
benzene (49) or reevaluate our approach toward the Amaryllidaceae iso-
carbostyril alkaloids. With much anticipation, we set up the transformation and
a) Me [Cu] (10 mol%)
O N O O
[4+2] N Ligand (20 mol%)
+
PPh2
PPh2 Cu(OAc)2 Cu(OTf)2
O Fe
PPh2 dppb (n=4): 64% dppb (n=4): 64% PPh2
Ph2P PPh2 PPh2 PPh2
( )n
dpppent (n=5): 67% dpppent (n=5): 81%
DPEPhos: 57% dppf
dpphex (n=6): 57% dpphex (n=6): 65%
(rac)-BINAP
Cu(OAc)2: 19% Cu(OAc)2: 65%
Cu(OAc)2: 53%
Cu(OTf)2: 46% Cu(OTf)2: 72%
Cu(OTf)2: 60%
b) c)
Me visible light Me visible light
O N O O N O
+ then + then
N N Cu(OTf)2 (10 mol%) N N Cu(OTf)2 (10 mol%)
NUR NUR
naphthalene (54) MTAD (50) dppf (20 mol%) benzene (49) MTAD (50) dppf (20 mol%)
PhMgBr, CH2Cl2 53 PhMgBr, CH2Cl2 55
d) e)
Me visible light Me visible light
N O N O
O O then then
+ +
Ni(cod)2 (10 mol%) N N Ni(cod)2 (10 mol%)
N N NUR
dppf (20 mol%) NUR dppf (20 mol%)
naphthalene (54) MTAD (50) benzene (49) MTAD (50)
PhMgBr, CH2Cl2 53 PhMgBr, CH2Cl2 55
FIGURE 1.5 (A) Development of the copper-catalyzed dearomative carboamination. (B) Testing of the copper conditions on in situegenerated 52. (C) Testing of
the copper conditions with benzene as substrate. (D) Discovery of the nickel-catalyzed dearomative carboamination conditions. (E) Application of the nickel
conditions with benzene as substrate.
Amaryllidaceae isocarbostyril alkaloids Chapter j 1 17
were delighted to find whether the reaction was now proceeding in a syn-
thetically useful 72% isolated yield of 55.
d)
R'Li/R''X V R'Li/L*/R''X Me
O N O O
R [4+2] N [Ni]n
+
N N N NMe
visible PhMgBr
benzene (49) MTAD (50)
light O NUR
Cr(CO)2L*
III IV Cr(CO)3 benzene-MTAD 55
Chromium coordinated Chiral nucleophile cycloadduct (51)
chiral ligand
p-decomplexation p-complexation
[Ni]0 [Ni]0
[Ni]0
Mg O
N N
N O N NMe
Ph NMe
O
IX VI O
plane of symmetry Ph
FIGURE 1.6 (A) Methods for enantioselective synthesis of 1,2-trans products utilizing stoichiometric chromium complexes. (B) Generation and reactivity of
stoichiometric iron cyclohexadienyl cation complexes. (C) Transition metals capable of forming cyclohexadienyl complexes and their typical nucleophiles. (D)
Proposed method for the nickel-catalyzed dearomative trans-carboamination proceeding through a cyclohexadienyl complex.
Amaryllidaceae isocarbostyril alkaloids Chapter j 1 19
achiral ligands
dppe (n=2): 55%
dppp (n=3): 24% PPh2
Ph2P PPh2 Fe
O O ( )n dppb (n=4): 14% PPh2
PPh2 PPh2 PPh2 PPh2 dpppent (n=5): 17%
dpphex (n=6): 45%
xantphos: 21% DPEPhos: 57% dppf: 74%
chiral ligands
PPh2
PPh2 PPh2 N
O R PPh2
R = Bn (59%, 95:5 er)
Me Me Ph2P N Fe O PPh2
R = iPr (75%, 98:2 er)
(S,S)-BDPP tBu R = tBu (26%, 85:15 er)
27% (90:10 er) (S)-tBu-PHOX
30% (66:34 er) (R,Rp)-Phosferrox-type ligands (S)-BINAP
O 11% (63:37 er)
H
O N
PPh2 MeO PPh2
O PPh2 P PPh2 PPh2
P H
O OMe
(S,S)-NORPHOS
(S)-SEGPHOS (S,S)-DIPAMP (S)-QUINAP
65% (68:32 er)
27% (62:38 er) no reaction no reaction
FIGURE 1.7 Ligand screen for the nickel-catalyzed dearomative carboamination reaction.
Amaryllidaceae isocarbostyril alkaloids Chapter j 1 21
3. Pancratistatins
3.1 Initial investigations
With access to our key diene 47 from our retrosynthetic analysis, we were now
able to study the reactivity of the system (Scheme 1.5). While we had hy-
potheses as to how the olefins would react, we could still not be certain about
the stereo- and chemoselectivity. The urazole moiety is inductively with-
drawing, making the adjacent olefin more electron deficient. Therefore, we
believed that the distal olefin would be the first to undergo electrophilic attack.
Furthermore, the aryl ring is sterically demanding, so we believed that the
reagent would approach from the opposite face. Finally, FürstePlattner
analysis revealed that regardless of the diastereomer formed in the epoxidation
reaction, we could still access the correct relative stereochemistry of the
cyclitol core.
Based on this analysis, we subjected diene 47 to epoxidation conditions
with mCPBA and found an intriguing bicyclic product 59, which we verified
by X-ray crystallography. This outcome confirmed our hypothesis about the
urazole moiety: the hydrazyl proton was acidic enough to direct epoxidation
with mCPBA in a manner like that of an allylic amide NeH. We believed that
this product could have formed in one of two ways. After directed epoxidation,
the epoxide could either be opened by TsOH, followed by an SN2 reaction, or
the formation of the allyl cation and ring closure. Either way, we knew this
product would not lead to the desired natural product, and we opted to mask
this acidic proton with a methyl group.
Our conditions for methylation turned out to be crucial as the diene 47 was
unstable to basic conditions, and if our base was fully soluble, we only ob-
tained an undesired biaryl that resulted from elimination of the urazole. We
found that the use of CH2Cl2 in combination with potassium carbonate and
methyl iodide led to quantitative formation of methylated diene product 60. At
this time, we were confident that if we performed the halohydrin reaction on
the system, we could not only form the desired bromohydrin but also
brominate the aromatic ring. This would give us a key functional handle
further in the synthesis that we could exploit to form the lactam. Upon
exposure of 60 to traditional halohydrin conditions, we formed dibromide 61
in 90% yield. Interestingly, we found that the bromide in fact approached from
the top face of the diene, leading to the unexpected diastereomer.
We first attempted to close the epoxide from bromohydrin 61 but observed
no reaction. Using more forcing conditions led only to decomposition, and we
were left with no choice but to move forward and hope to close the epoxide
later. When bromohydrin 61 was subjected to traditional Upjohn dihydrox-
ylation conditions, we observed partial conversion to bromotriol 62 (Scheme
1.6). This was mainly due to the insolubility of 61 in the reaction conditions.
After some optimization, we found that with the addition of acetone for
22 Strategies and Tactics in Organic Synthesis
O
O H
OH O O
mCPBA, pTSOH MeI, K2CO3
N CH2Cl2 O N O CH2Cl2 O N O
N O HN MeN
37% NMe quant. NMe
NMe
59 O 47 O 60 O
OTs
NBS
O 79%
OH THF, H2O
O N O
HN
OH
NMe
Br
O Br
O N
O O O MeN O
OH
O NMe
O N O O N O
HN HN O
NMe NMe 61
O O
SCHEME 1.5 First empirical investigation of diene reactivity involving the formation of 59 through directed epoxidation/ring opening. Then observation of the
facial selectivity in employment of traditional halohydrin reaction conditions.
OH OH OH
Br OsO4 (5 mol%), Br OH HO OH
Br Br Br
NBS citric acid, NMO NaOBz
OH OH
H2O, THF N O
H2O, tBuOH N O H2 N O
O MeN acetone O MeN O MeN
79% O O 45% O
NMe NMe NMe
91%
O N O
MeN
NMe
60 O
OH OH
O
HO HO OH
OsO4 (5 mol%),
mCPBA O NaOBz O NMO O OH
NaHCO3, CH2Cl2 O N O H2 O N H2O, tBuOH O N O
MeN MeN O MeN
68% NMe 93% 91% NMe
NMe
O O O
64 65 66
23
24 Strategies and Tactics in Organic Synthesis
SCHEME 1.7 Investigation of the urazole cleavage in tetraol 66, leading to the discovery of cyclic hydrazine hemiaminal 70.
25
26 Strategies and Tactics in Organic Synthesis
a) OH OH
HO HO
AgOBz, I2 O mCPBA, TsOH
O O
O N O
O N MeN CH2Cl2, HFIP, H2O O N
MeN O MeN O
NMe 74%
NMe O NMe
O O
72 60 65
b) c)
H
O H
CF3 O
H H H
F3C O H O
O H O HO O
O O O
H O O H
R O HO O
H H H
H H O H
73 H O 74
H
FIGURE 1.8 (A) One-pot dihydroxylation attempt using Prévost reaction, followed by success with HFIP. (B) Explanation for HFIP-mediated rate acceleration
for olefin epoxidation. (C) Jamison’s explanation of how HFIP can stabilize hydrogen bonding interactions of water molecules with the substrate. HFIP,
1,1,1,3,3,3-hexafluoro-2-propanol.
Amaryllidaceae isocarbostyril alkaloids Chapter j 1 27
OH OH
Pinnick-Lindgren oxidation was HO OH HO OH
Failed formylation conditions:
irreproducible NaClO2, NaH2PO4 - Zn(CN)2/HCl - CH(OMe)3
O OH O OH
2-methyl-2-butene - Cl2CHOMe/ZnBr2
Other oxidation conditions tested: NH N - MeNO2/PPA
- KI/TBHP - H2O2 - MnO2 O THF, H2O O
- KHSO5 - AgNO3 - TEMPO O 57%
7-deoxypancratistatin (1) 76
b)
OH
OH OAc
HO OH
HO OH AcO OAc
Boc2O, Et3N Ph3P(O), Tf2O
O OH
O OH 1,4-dioxane, H2O; O OAc BF3 2O, CH2Cl2;
then Ac2O then NaOMe, MeOH O NH
O NH2 O NHBoc
76% 75% O
45 75 7-deoxypancratistatin (1)
SCHEME 1.8 (A) Completion of the first-generation synthesis of 7-deoxypancratistatin (1) through the use of the Duff formylation and the PinnickeLindgren
oxidation. (B) Fail-safe strategy involving the BischlereNapieralski reaction to close the lactam.
Amaryllidaceae isocarbostyril alkaloids Chapter j 1 29
expected reaction
OH OH
OH OH
HO OH HO OH
HO OH HO OH
Br2 Co2(CO)8, n-Bu4NBr CO2H
O OH
O AcOH O OH
OH OH 5N NaOH, PhH
O NH NH2
NH2 90% NH2 CO, hn (365 nm)
O O Br O
O
O
45 77 7-deoxypancratistatin (1) 78
25% 60%
NaCo(CO)4, n-Bu4NBr
dioxane:NaHCO3 (aq.)
CO, hn
80%
CuCN; Pd/Zn(CN)2
SCHEME 1.9 Development of alternative and more scalable conditions for lactam formation to provide 7-deoxypancratistatin (1).
Amaryllidaceae isocarbostyril alkaloids Chapter j 1 31
reaction system with 5N NaOH and benzene as solvents. The role of the base
was to activate the cobalt carbonyl, as well as maintain alkaline conditions
during the reaction as HBr was the major by-product. Fortuitously for them,
they found that the lactam would form in the aqueous phase and, upon closure,
would be sequestered in the organic phase. When we employed these condi-
tions in our system, we were gratified to find we could isolate
7-deoxypancratistatin (1) in 25% yield. Interestingly, we found that the
remainder of the mass balance resided in the amino acid 78 which likely
resulted by hydrolysis of the lactam of the natural product.
When considering the issue of lactam hydrolysis, we knew the major factor
was the use of 5N NaOH. Furthermore, as our UV lamps became warm during
the reaction, we noticed that the solution would also heat from their proximity,
potentially increasing the rate of hydrolysis. Therefore, we knew we needed to
obviate the need for such strongly alkaline conditions. Based on Caubere’s
report, the main purpose of the base was to convert Co2(CO)8 to the active
catalyst NaCo(CO)4. We hypothesized that if we could prepare and employ
NaCo(CO)4, then perhaps we could remove the strong base and utilize sodium
bicarbonate instead, potentially reducing the amount of hydrolyzed lactam.
After some optimization, we found this to be the case and obtained
7-deoxypancratistatin (1) in 80% yield and did not observe the amino acid by-
product.
FIGURE 1.9 Conditions tested to metalate/oxidize the C-7 position of 7-deoxypancratistatin (1).
Amaryllidaceae isocarbostyril alkaloids Chapter j 1 33
precursors for the transformation as they would be inexpensive and easier to use
due to their air stability (Table 1.2).99 One of our first goals was to reduce the
ratio of nickel to ligand, as well as lower the catalyst loading (entries 1e3). We
found that lowering the nickel loading to 5.0 mol%, as well as the ligand to
Me
O N O
N N
MTAD (50)
visible light
[Ni] (x mol%)
NUR
benzene (49) (R,Rp)-iPr-phosferrox (y mol%)
55
7.0 mol%, had minimal impact on yield and enantiomeric ratio. We found that
concentrating the reaction (entry 4) gave a lower yield, but higher er. We then
tested to see if 3.0 equivalents of our Grignard reagent were necessary (entries
4e6). Interestingly, lowering the Grignard equivalents impacted the yield and
enantioselectivity, but we found this to still be in the acceptable range. We then
tested a series of Ni(II) sources (entries 7e11) and found that many sources
could be employed in this transformation. However, Ni(acac)2 (entry 9) gave us
the highest yield and enantioselectivity out of the series tested. We then began
to lower the nickel loading, as well as Ni:ligand ratio (entries 12e17). Notably,
we found that as we lowered the catalyst loading, as well as Ni:ligand, a marked
increase in yield and er was observed. We eventually decided on the conditions
in entry 15, which we found to be the most practical and reproducible.