Solid-Phase Synthesis of Duocarmycin Analogues and the Effect of C-Terminal Substitution on Biological Activity

Article abstract of DOI:10.1021/acs.joc.5b01373

The duocarmycins are potent antitumor agents with potential for use in the development of antibody-drug conjugates (ADCs) as well as being clinical candidates in their own right. In this article, we describe the synthesis of a duocarmycin monomer (DSA) th

Full text of DOI:10.1021/acs.joc.5b01373

Subscriber access provided by FLORIDA STATE UNIV
Article
The solid phase synthesis of duocarmycin analogues and
the effect of C-terminal substitution on biological activity.
Michael J Stephenson, Lesley A. Howell, Maria A. O'Connell, Keith R Fox, Claire Adcock,
Jenny Kingston, Helen Sheldrake, Klaus Pors, Stephen P Collingwood, and Mark Searcey
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.5b01373 • Publication Date (Web): 10 Sep 2015
Downloaded from http://pubs.acs.org on September 11, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Organic Chemistry is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
The solid phase synthesis of duocarmycin analogues and the effect of C-terminal
substitution on biological activity.
1
1
1
3
Michael J. Stephenson, Lesley A. Howell, Maria A. O’Connell, Keith R. Fox, Claire
4
4
3
3
4
Adcock, Jenny Kingston, Helen Sheldrake, Klaus Pors, Stephen P. Collingwood
1,2*
and Mark Searcey
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
Schools of Pharmacy and Chemistry, University of East Anglia, Norwich Research
3
Park, Norwich, NR4 7TJ, UK, Centre for Biological Sciences, Life Sciences Building 85,
4
University of Southampton, Southampton SO17 1BJ, UK, Novartis Institutes for
Biomedical Research, Novartis Pharmaceuticals UK Limited, Horsham Research
Centre, Wimblehurst Road, Horsham, West Sussex RH12 5AB, UK, Institute for
Cancer Therapeutics, University of Bradford, Bradford, West Yorkshire BD7 1DP, UK.
m.searcey@uea.ac.uk
5
Abstract
The duocarmycins are potent antitumour agents with potential in the development
of antibody drug conjugates (ADCs) as well as being clinical candidates in their own
right. In this paper, we describe the synthesis of a duocarmycin monomer (DSA) that
is suitably protected for utilisation in solid phase synthesis. The synthesis was
performed on a large scale and the resulting racemic protected Fmoc-DSA subunit
was separated by supercritical fluid chromatography (SFC) into the single
enantiomers. Application to solid phase synthesis methodology gave a series of
monomeric and extended duocarmycin analogues with amino acid substituents. The
DNA sequence selectivity was similar to previous reports for both the monomeric
and extended compounds. The substitution at the C-terminus of the duocarmycin
caused a decrease in antiproliferative activity for all of the compounds studied. An
extended compound containing an alanine at the C-terminus was converted to the
primary amide or to an extended structure containing a terminal tertiary amine but
this had no beneficial effects on biological activity.
Introduction
The duocarmycin family of natural products incorporates the parent molecule
1
2
duocarmycin SA (1, Figure 1), several naturally occurring analogues and the
3
4
extended and sandwiched compounds CC-1065 (2) and yatakemycin (3). The mode
of action of these compounds, involving reversible alkylation of the N3 of adenine
through shape dependant activation on binding to the minor groove of DNA, has
5
been the subject of extensive investigation and has led to the design and synthesis
6-12
of numerous analogues.
Most recently, research has focussed on prodrugs that
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
3,14
15,16
are reductively
or oxidatively-activated
(4-6) or that carry glycosidic
17,18
linkages.
This desire to generate prodrug structures is due to the ultrapotent
activity of the drug molecules, which significantly narrows the therapeutic window
for this class of compound as alkylation and subsequent anti-proliferative effects
occur in all dividing cells, as with classical antitumour cytotoxics.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Duocarmycin SA, and biooxidative and bioreductive prodrugs
One route into the design of new molecular entities with therapeutic potential whilst
minimising cytotoxicity is via tumour cell targeting, rather than prodrug design.
Antibody-drug conjugates, in which the antibody targets a highly cytotoxic molecule
19
to the tumour site of action, have met with recent success in the clinic. The
enediyne antibiotic calicheamicin was conjugated to an anti-CD33 antibody to
generate the clinically utilised agent gemtuzumab ozogamicin for the treatment of
20
acute myelogenous leukemia (AML). Although this was subsequently withdrawn
2
1
due to toxicity problems, it paved the way for brentuximab vedotin and
2
2
trastuzumab emtansine, which are used in the treatment of Hodgkins lymphoma
2
ACS Paragon Plus Environment

Page 3 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
and breast cancer, respectively. Key to the design of antibody drug conjugates is the
linker between the cytotoxic drug and the antibody. In gemtuzumab ozogamicin, the
linker was cleaved under the acidic conditions that are formed in the endosomal
[11]
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
compartment when the antibody is internalised in the cell. Brentuximab vedotin
utilises a cathepsin cleavable linker to the antimitotic agent monomethylauristatin
23
E. Trastuzumab emtansine contains a non-cleavable linker, such that as the protein
is degraded in the environment of the endosome, the small molecule is released
with the linker and a lysine residue still attached, nevertheless still exerting a
24
cytotoxic effect.
These latter two approaches require the attachment of the potent drug molecule
through a peptidic linker. The ability to prepare multiple analogues via solid phase
peptide chemistry is thus particularly attractive for ADC design. The duocarmycin SA
alkylation subunit (termed (+)-DSA) is an amino acid ester and so is ideally placed to
be incorporated into peptide synthesis. To our knowledge, there have been no
descriptions to date of the incorporation of (+)-DSA directly into solid phase peptide
methodology and we describe here the synthesis of a monomer 7 that is suitably
protected for Fmoc-based solid phase synthesis. We also describe preliminary
studies of the incorporation of the subunit onto the solid phase with a peptide
moiety at the C-terminus for both a simple alkylating agent (8-13) and an extended
DSA analogue (14-18). The effects of the presence of the differing side chains on
DNA binding and biological activity are shown. Particularly striking was the discovery
that substitutions at this position have a profound effect on the antiproliferative
activity of the compounds. Hydrolysis of the simple ester 30 to give the free acid 31
completely eliminates the activity whereas the presence of a free amino acid
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
demonstrates considerable variability. The combination of a free amino acid with an
extended group on the N-terminus increases the antiproliferative activity but
incorporation of a terminal amine or amide has little effect on activity.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Compounds 7-18 made in this study.
Results and Discussion
Synthesis of the monomer for Fmoc-chemistry
2
5-28
Several synthetic routes to the (+)-DSA subunit have been described,
but key to
success in this project was the scalability, so that large amounts of a protected form
were available for application to the solid phase. There are also a smaller number of
4
d, 28
stereoselective routes to the target compound and analogs,
but in this instance,
it was felt a late stage separation of the enantiomers, on a large scale, would be
useful as it was felt that the seco form of DSA that would be generated from the
planned racemic route would serve as a suitable solid phase building block. This is
because the halide leaving group allows in vivo spiro-cyclisation, thus providing
compounds of biological interest in fewer synthetic steps post cleavage. Our
synthesis began with 50 g of commercially available 2-hydroxy-4-nitroaniline 19
4
ACS Paragon Plus Environment

Page 5 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
2 3
which was protected as the benzyl ether 20 in two batches (BnBr, K CO , DMF, RT,
9
9%, Scheme 1) and then regioselectively iodinated, again in two batches, using N-
iodosuccinimide (NIS) and catalytic acid to give over 100 g of the iodo compound 21
in 92% yield overall. This set the stage for investigation of cross-coupling reactions to
introduce an alkyne substrate and potentially to induce cyclisation to the indole in
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
29
one pot as described previously. The Negishi coupling worked smoothly (methyl
propiolate, Pd(PPh ) Cl , ZnBr , N,N-Diisopropylethylamine (DIPEA), DMF, 66 °C,
3
2
2
2
7
7%) but we had elected on the large scale not to use the dimesylation/mesylation
2
9
approach as it increased the number of manipulations required. As such, in spite of
several attempts under differing conditions, we could not induce the cyclization of
2
2 to occur. As a consequence, the coupling was followed by cyclisation using
3
0
tetrabutylammonium fluoride, which generated the indole 23 and this was
immediately protected as the N1-Boc compound 24 (Boc O, 4-
2
dimethylaminopyridine (DMAP), CH
2 2
Cl ). The two steps were conducted on a 60 g
(
184 mmol) scale and gave a modest 39% yield of the pure protected indole 24 after
purification by flash column chromatography. We also investigated a Sonogashira
route to the same indole, which gave similar yields on a small scale but could not be
scaled up due to problems associated with the purification. The rest of the synthesis
3
1
to the di-Boc protected indole followed published methodology and utilized the 5-
3
2,33
exo-trig free radical cyclisation reaction
to introduce the dihydropyrrole ring
structure. Hydrolysis of the ester 28 was followed by Boc-removal under acidic
conditions to generate a substrate that was regioselectively Fmoc protected with
Fmoc-Cl to give the substrate for peptide synthesis. The final yield for the full
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
synthesis was 3% over 13 steps and generated over 8 g of the Fmoc protected
monomer 7 as a racemic mixture.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 1. i: BnBr, K
Pd(PPh Cl , DIPEA, Methyl propiolate, DMF, 66 °C, N
steps). vi: Zn, NH Cl, Boc O, DMAP, THF:Water. vii: NIS, H
1,3-dichloropropene, DMF, 62%. ix: AIBN, Tris(trimethylsilyl)silane (TTMSS), Toluene, 90 °C, N
THF:MeOH:Water, 100%. xi: 4M HCl in Dioxane. xii: Fmoc-Cl, NaHCO , THF:Water, 80% (2 Steps).
2
CO
3
, DMF, 98.5% (mean 2 batches). ii: NIS, H
, 77%. iv: TBAF, THF, 66 °C. v: Boc
SO , DMF, 59% (3 steps). viii: Potassium tert-butoxide,
, 70% x: LiOH,
2
SO
4
, DMF, 92% (mean 2 batches). iii: ZnBr
2
,
3
)
2
2
2
2
O, DMAP, CH Cl , 39% (2
2
2
4
2
2
4
2
3
Separation of enantiomers using supercritical fluid chromatography.
The application of supercritical fluid chromatography for the chiral resolution of
3
4,35
racemic mixtures is well established.
Supercritical fluids combine the density, and
dissolution character of a liquid, with a viscosity, and diffusion behaviour more
comparable to a gas. The low viscosity improves mass-transfer kinetics, and permits
the use of fast flow rates with high acuity columns. These properties make them
ideal mobile phases, and allow for highly efficient separations. As such this technique
is particularly attractive for preparative scale work, and was employed here for the
isolation of each enantiomer of 7.
6
ACS Paragon Plus Environment

Page 7 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Separation was affected using a Chiralpak AD-H column (250 x 30 mm, 5 micron),
2
and an isocratic flow of 50% CO , and 50% IPA containing 0.1% TFA, at 45 mL per
min. The back pressure was regulated at 10 MPa, and column temperature
o
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
controlled at 40 C. A racemate of 7 (9.8 g) was dissolved in THF:MeOH 1:1 (100 mL),
and 1.25 mL (125 mg) injected every 9 min. Fractions were monitored by UV (220
nm), collected, combined and dried to afford 2.82 g of peak 1 (5.5 min), and 3.1 g of
peak 2 (7 min), both as cream solids (α = 1.27).
Analytical supercritical fluid chromatography of peak 1, showed a 7% impurity with
similar retention time to peak 2. The mass of this peak suggested the loss of Cl as
opposed to racemisation. NMR analysis showed no evidence of this impurity, with
1
the H NMR of both enantiomers being identical to that of the racemate, with the
exception of a small amount of residue IPA. We propose that the observed impurity
may be an artefact of supercritical fluid chromatography, and represent a transient
3
6
alternative spirocyclisation through the indole nitrogen, perhaps promoted by the
pressure and acidic buffer.
Peak 1 was assigned as the natural enantiomer based on the sign of specific rotation
3
7
25
matching that of the well characterised seco-Boc-DSA derivative. Peak 1 [α]
D
–20°
2
5
(
c 0.05, DMF), Peak 2 [α] +20° (c 0.05, DMF). This assignment was further
D
confirmed by the DNA alkylation sequence selectivity of the extended agent 14,
which matches that found for the natural products (i.e. alkylation of an A at the 3’-
10
end of an AT sequence).
Application of Fmoc-protected monomer to solid phase peptide synthesis.
The suitability of 7 to serve as a building block for Fmoc-based solid phase synthesis
was initially explored using a racemic batch of the compound accessed from a small
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
scale pilot synthesis. In this preliminary study (Scheme 2), a commercially available
2 2
Wang supported alanine resin was first prepared for coupling by swelling in CH Cl
for 30 min, followed by DMF for a further 30 min. Subsequent treatment with 40%
piperidine in DMF for 10 min, and 20% piperidine in DMF for 5 min twice, removed
the alanine’s Fmoc protection. A positive Kaiser Test confirmed the present of the
free amine. The resin was then treated with a modest excess of 7 (1.1 eq. based on
the manufacture’s resin loading) which had been pre-activated for 30 s prior to
addition, by treatment with an equimolar quantity of N,N,N′,N′-tetramethyl-O-(1H-
benzotriazol-1-yl)uronium hexafluorophosphate (HBTU), and a twofold excess of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
both hydroxybenzotriazole (HOBt.H O) and DIPEA in DMF. After two hours of
2
shaking, the Kaiser test was repeated. A negative result was observed, and suggested
complete coupling of 7 to the Wang supported alanine. The resin was treated again
with piperidine in DMF as previously described, affecting removal of Fmoc protection
from the indoline nitrogen. This allowed coupling of a further alanine residue, using
a fivefold excess of Fmoc-protected alanine for 45 min, which had been preactivated
for 30 s with equimolar quantities of HBTU and HOBT.H
DIPEA in DMF. The resin was then prepared for cleavage by removal of the final
Fmoc group and extensive washing with CH Cl , followed by drying under a stream of
2
O, and a twofold excess of
2
2
2
N .
8
ACS Paragon Plus Environment

Page 9 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Preliminary solid phase synthesis studies using racemic 7
3
8
The resin was cleaved using standard conditions for peptide synthesis. The dried
resin was treated with a solution of 95% TFA, 2.5% triisopropylsilane (TIPS), and 2.5%
H
2
O for 2 h. The cleavage mixture was filtered and concentrated, followed by
precipitation with the addition of cold Et O. HPLC analysis of the crude product
2
(
Figure 3a) revealed the formation of several significant side products. It was
suspected that these products may have resulted from degradation during cleavage
as opposed to representing problems during the synthesis. As a result the synthesis
was repeated to allow optimisation of the cleavage conditions.
The dried resin was divided into 10 mg portions, and subjected to different cleavage
conditions. The results of the HPLC analysis of the crude products are shown in
Figure 3. Reducing the concentration of TFA to 50% in CH
scavengers (TIPS, H O), led to a reduction in the number of side products (Figure 3b
versus 3a). However the product peak at 9 min was still accompanied by two
significant side products at 9.5 and 9.7 min. Cleavage with 95% TFA and 5% CH Cl in
2 2
Cl and omission of the
2
2
2
the absence of scavengers (Figure 3c) led to almost complete absence of the product
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
peak with the HPLC trace now being dominated by the side products at 9.5 and 9.7
min. Finally, reducing the concentration of TFA and including the scavengers (Figure
3
2 2 2
d, 47.5% TFA, 47.5% CH Cl , 2.5% TIPS, 2.5% H O) produced an HPLC trace
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
dominated by the product peak at 9 min. Mass spec analysis confirmed this peak to
be the desired product.
Interestingly closer inspection of the HPLC trace under the most destructive cleavage
conditions (Figure 3c) revealed that, not only did the product peak no longer
dominate, but there was also a large decrease in intensity compared to the other
conditions despite the crudes having been analysed in the same volume of
methanol. This suggested that the amount of material recovered from the resin had
been reduced. Repeating the cleavage did not lead to recovery of more material.
This led to the proposal that Wang resin may not be the most appropriate choice for
syntheses incorporating 7. Indeed, Wang resin has previously been suggested to be
3
9
problematic for peptides containing the indole side chain of tryptophan. The Wang
linker can be cleaved at more than one position leading to free benzyl cations that
can attack the nucleophilic positions of the indole. Furthermore, resin-bound cations
have been speculated to undergo an irreversible back alkylation with the tryptophan
4
0
side chain leading to reduced yields. Considering that in the indole structure of 7, it
is the more nucleophilic 3- position that is unsubstituted in comparison to the 2-
position of the tryptophan side chain, it may be particularly susceptible to such side
reactions.
10
ACS Paragon Plus Environment

Page 11 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3: HPLC analysis of crude HO-Ala-DSA-Ala-NH
varying conditions. a) 95% TFA, 2.5% TIPS, 2.5% H O. b) 50% TFA, 50% CH
c) 95% TFA, 5% CH Cl . d) 47.5% TFA, 47.5% CH Cl , 2.5% TIPS, 2.5% H O. 10
mg of dried resin was cleaved under either conditions a, b, c, or d with 5 mL
of the respective cleavage cocktail for 2 h. The cleavage mixture was filtered
and evaporated to dryness. The crude was dissolved in 1 mL of CH OH and
analysed by HPLC at 254 nm. Agilent Eclipse XDB-C18 column, 4.8 x 150 mm,
5 µM. Solvent A: [Water and 0.05% TFA], Solvent B: [ACN and 0.05% TFA].
Gradient: 0% [B] to 95% [B], from 0 min to 15 min, 95% [B] to 0% [B] from 15
to 20 min. Monitored UV 254 nM.
2
after cleavage under
2
2
Cl .
2
2
2
2
2
2
3
11
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
As a result of these concerns a qualitative analytical screen of available resins was
conducted. For these tests the C-terminal alanine was replaced with lysine and the
indoline nitrogen capped with 5 equiv. of AcCl and 10 equiv. of DIPEA in DMF for 45
min (Scheme 3). This synthesis was carried out on the same 0.038 mM scale with
identical reaction times, on four different resins: 2Cl-Trt resin, Rink amide resin,
Wang resin, and Novo Syn Wang resin. All were cleaved using 47.5% TFA, 47.5%
CH Cl , 2.5% TIPS, 2.5% H O at the same volume for 2 h. The crude cleavage mixtures
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
2
3
were filtered, evaporated to dryness, and dissolved in the same volume of CH OH (1
mL) for HPLC analysis. The area of the product peak for each of the crudes was
compared. As the same volume of CH OH was used to dissolve the crudes, the area
3
of the peak doses give a qualitative indication of the quantity of product recovered
from each resin. Both Wang resins appeared to perform the worst and gave the
same level of product as judged by peak area. The best performing was 2-Cl-Trt
resin, and Rink amide resin fell between the two extremes. 2-Cl-Trt resin can only be
cleaved at the desired position, and the cation formed is more sterically hindered by
the surrounding linker structure, making any potential back alkylation less likely. As
2
-Cl-Trt resin has the further advantage of allowing cleavage to be effective with as
little as 1% TFA, it was chosen as the resin of choice for further work.
12
ACS Paragon Plus Environment

Page 13 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Qualitative analysis of the best resins and coupling reagents for addition of 7 to the solid phase.
A similar test was used to compare available coupling reagents (Scheme 3). The
same analogue was synthesised at a 0.016 mM scale on 2-Cl-Trt resin. The only
variation being the reagents used to couple
7 to the resin-bound lysine. Again, a
qualitative comparison of the area of the product peak on HPLC was carried out. -
[Bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium-3-oxid
13
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
hexafluorophosphate (HATU) appeared to be the best performing coupling reagent
followed
hexafluorophosphate
EDCI), and N,N′-Diisopropylcarbodiimide (DIC). As a result, HATU was used as the
by
HBTU,
(Benzotriazol-1-yloxy)tripyrrolidinophosphonium
(PyBop), 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
coupling reagent for further work.
Following these initial experiments, the natural enantiomer of 7, accessed from the
large scale synthesis, was used to generate the library of analogues shown in Figure
. Full details of their synthesis and isolated yields can be found in the experimental
section (yields were 69%, 81%, 10, 56%, 11 48%, 12 38%, 13 19%, 14 20%, 15
6%, 16 35%, 17 17%, 18 34% based upon resin-loading). All analogues where
2
8
9
1
synthesised on 2-Cl-Trt resin, with the exception of 18, where Rink amide was
employed to provide a neutral terminal amide after cleavage. When using 2-Cl-Trt
resin, the cleavage concentration of TFA was reduced further to 1% for analogues
not requiring t-butyl side chain deprotection, increasing to 10 or 20% and extended
reaction times for those that did. It was found when the TFA concentration was
2
reduced to this level, H O no longer served as an effective scavenger, presumably as
it was not miscible with the larger volumes of dichloromethane now being used. This
was resolved by increasing the concentration of TIPS to 10%. Interestingly, if 10%
2
H O was also included, t-butyl side chain deprotection did not occur. This might be
as the immiscible water layer was acting as a proton sink, reducing the acidity of the
CH Cl layer.
2
2
The benzyl protecting group was removed via transfer hydrogenation using Pd-
C/HCO NH on crude cleavage products, and the active compounds purified by either
2
4
preparative reverse phase HPLC, or silica gel Flash chromatography, depending on
14
ACS Paragon Plus Environment

Page 15 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
the polarity of the side chain. This reaction proceeded smoothly for all but one of the
molecules synthesised in this study. Compound 14, containing the lysine side chain
and the extended DNA binding structure, required careful monitoring as an
alternative product, which had undergone dehalogenation, began to form and
became the predominant species. The reason for the susceptibility of 14 to this side
reaction is unknown. It was observed that the reaction was unusually slow compared
to the other analogues and required greater quantities of Pd-C and ammonium
formate. Suspicion that the amine of the lysine side chain might be poisoning the
catalyst, or promoting dehalogenation, led to attempts to perform benzyl
deprotection prior to removal of lysine side chain protection. This was achieved by
cleaving in 1% TFA to give the protected lysine product. However this made no
difference and the side reaction was still encountered. The highest yield and
optimum purity of the target compound was obtained when the reaction was
carefully carried out and quenched after 1 h.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DNA alkylation studies
(
+)-Duocarmycin SA binds to the minor groove and alkylates the N3 of adenine at the
’-end of an AT rich sequence, with the unnatural enantiomer displaying a similar
sequence preference but with alkylation occurring at the 5’-end of the sequence,
3
3
7
suggesting productive binding is in the other orientation, as might be expected.
The alkylation subunit (+)-Boc-DSA displays a similar sequence selectivity to the
unnatural enantiomer, essentially alkylating any A within a truncated AT sequence,
4
1
down to two bases. This can be rationalised by models of both compounds in the
minor groove that allow for productive alkylation of the same site, regardless of the
binding orientation. It was expected that the C-terminal amino acid-linked seco-DSA
15
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
compounds
8
–
13 would display a similar alkylation profile to the simple alkylation
subunit 30 (figure 4), although it was also considered that the presence of the free
acid may affect DNA binding, so the compounds were also analysed alongside the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
hydrolysed derivative 31
.
Figure 4. Structures of 30 and 31
3
2
Compounds
8-13 were incubated at 37 °C for 16 h with the 3’- P-end-labeled MS1
DNA fragment (a fragment approximately 166 base pairs in length designed to
4
2
contain all possible tetranucleotide sequences). The resulting solutions were
1
heated to 100 °C for 3 min in order to generate cleavage and were then subjected
to denaturing polyacrylamide gel electrophoresis and developed on a
phosphoimager. Examples of the results are shown in the left hand panel of Figure 5.
It is clear that the sequence selectivity of compounds 8-13 closely matches that for
the controls 30 and 31 (which were analysed as the racemic mixtures) with little
evidence for sequence selectivity other than the alkylation of an A in short AT runs,
4
1
as previously suggested, and as shown in the open bars in the lower panel of Figure
. An indication of the relative reactivity of the compounds towards DNA alkylation
and thermal-induced cleavage is given by the intensity of the full length DNA band
at the top of the gel) at similar concentrations. The presence of significant cleavage
5
(
for compound 31 suggests that the presence of a carboxylic acid does not affect the
DNA alkylating ability of the subunit, as the reactivity of 31 closely matches that for
16
ACS Paragon Plus Environment

Page 17 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
30. For most of the amino acid compounds, the reactivity towards DNA is decreased
relative to 30 and 31, with lower levels of cleavage at 5 µM concentration. The
exception is the serine derivative 11, which maintains similar reactivity to the
controls. This suggests that the presence of the amino acid group has a negative
effect on DNA binding, though this effect is unlikely to be a consequence of the
presence of the terminal carboxylic acid.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The lysine functionality in 13 does not enhance the binding of the monomeric
subunit, which was unexpected as the amino acid was incorporated so that it would
enhance interactions with the sugar-phosphate backbone. This appears not be the
case and might suggest that the lysine and other side chains lie along the floor of the
groove on binding, rather than extending out into the solvent environment close to
the alkylation site.
17
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. DNA cleavage of DNA fragment MS1 by duocarmycin derivatives. Left hand panel compounds 9-13 and
control compounds 30 and 31. Right hand panel 14, compared with compounds 30 and 31. Tracks labelled GA are
sequence markers for purines; control is DNA in the absence of added ligand. Ligand concentrations (µM) are
indicated at the top of each gel lane. The lower panel shows the relative cleavage (compared with control lane)
at each position for 14 (filled bars) and 31 (open bars) for a section of the MS1 fragment that incorporates the
site of best cleavage for 14. The asterisk indicates the location of the best cleavage with 14.
The extended agent 14, which is more similar in structure to the full natural product
duocarmycin SA than to the simple alkylation subunit, displayed significantly
enhanced DNA alkylation activity compared with the other compounds (Figure 5,
right hand panel) producing cleavage products at much lower concentrations. As
expected, 14 reacted predominantly with As at the 3’-end of AT sequences, with
fewer cleavage sites than the other compounds. The major cleavage on this
fragment is observed at AATTA, (asterisk in Figure 5) demonstrating a sequence
18
ACS Paragon Plus Environment

Page 19 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
selectivity that is closer to that expected for the natural product. The enhancement
of the alkylation seen with 14 compared with 13 was significant with clear alkylation
at 10 nM compared with 5 ꢀM for the latter.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Antiproliferative activity
The antiproliferative activity of the compounds was measured using the human
leukaemia cell line HL60 and an MTS assay (Table1). An initial, striking, observation
with the control compounds 30 and 31 is the difference in biological activity. While
racemic 30 maintains an anti-proliferative effect that is close to that seen with L1210
mouse leukaemia cells in the literature ((+)-N-Boc-DSA, 6 nM, (-)-N-Boc-DSA, 60 nM,
37
this study (±)-N-Ac-DSA 30, 25 nM), hydrolysis of the ester to give racemic 31
effectively removes all antitumour effects. The DNA cleavage studies clearly showed
that there is no difference in alkylation ability between these two compounds,
strongly suggesting that the presence of the carboxylic acid does not affect DNA
binding. The lack of activity is likely to derive from the free carboxylic acid, even with
the benzoic acid-type of group in the DSA subunit, being substantially ionised at
physiological pH and that cellular uptake with this compound is not achieved.
Bearing in mind that the pKa of the amino acid carboxylic acid groups appended to
the duocarmycin subunit might be expected to be lower than that of 31 and that,
with the exception of 11, all compounds displayed a lower affinity for DNA under the
alkylation experimental conditions, we expected the new agents to similarly lack
antiproliferative activity. Surprisingly, this was not the case with only the Lys and β-
alanine compounds showing a complete lack of activity against HL60, although all
compounds were between approximately 1000-10000 fold less active than 30.
19
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 1. Antiproliferative activity of the duocarmycin compounds. Activity was assessed in HL60 human
leukaemia cell lines and was measured using the MTS assay. For details, see the experimental section.
It seems unlikely that there is a simple explanation for the increased activity of the
simple amino acid derivatives. The activity of 8 and 10 may be due to increased
hydrophobic nature of the side chains increasing passive diffusion into cells in spite
of the presence of the free acid group or indicate a role for transport mediated
uptake. The Ser analogue 11 may display better uptake but also benefit from its
more efficient DNA alkylation ability once in the cell. The lysine derivative 13 will
form a zwitterion in solution and this may explain its complete inactivity, due to the
lack of cellular uptake, although this is then countered by the more hydrophobic
extended structure of 14, which has reasonable activity. Compound 12, with two
acid groups has low activity and it is unclear why the presence of the linear
would lead to complete inactivity.
β-ala in 9
The introduction of extended hydrophobic structure of the 5-methoxyindole
derivatives significantly enhanced the biological activity of the compounds with the
lysine derivative 14 now having an activity of 374 nM and the serine and alanine
20
ACS Paragon Plus Environment

Page 21 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
derivatives 15 and 16 displaying a 260 and 2200 fold increase in activity, respectively.
As the lysine derivative 14 displayed enhanced activity on binding to DNA compared
with 30, the decreased cytotoxicity is likely to derive from what is still a limited
uptake for these compounds, with only the alanine compound 16 having activity that
approaches that of the simple analogue 30. In other studies, albeit in different cell
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
7
lines, neutral duocarmycin analogues have low pM antiproliferative activity.
In order to further investigate the effect of the carboxylic acid and it derivatives on
cytotoxicity, we also synthesised the -dimethylamino- compound 17 and the
N,N
neutral compound 18. Minor groove binders based upon distamycin often carry a
positively charged end group similar to this, which enhances solubility and DNA
binding. The inclusion of the amino-end group for 17 is to the detriment of its
biological activity, with a 2-fold decrease in activity compared with the free acid
compound 16, although this remains the relatively potent. Finally, the uncharged
analogue 18 was synthesised using Rink amide resin to directly generate the amide.
This compound was found to have similar, albeit slightly enhanced activity compared
with the other compounds.
Conclusions
The duocarmycins have great potential as antitumour agents but have yet to
progress to clinic use. Their utility in ADCs has been widely recognised, with recent
4
3,44
disclosures focussed on the design of conjugates.
In this paper, we show that the
duocarmycin alkylation subunit is suitable for solid phase synthesis methods and can
be added directly to a peptide-based linker for incorporation into a targeting
construct. It would be possible to block the reactivity of the duocarmycin linker
through phenol protection, either of a bioreductive, biooxidative or simple
21
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
enzymatic or chemical cleavage nature, but with the exquisite targeting potential of
antibodies, it remains to be seen if this is really necessary, as the duocarmycins tend
to be activated only in the presence of DNA. This methodology also has potential in
the design of, for example, covalent analogues of polyamide minor groove binders or
other DNA binding peptides.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The work here also clearly demonstrates that small changes in the duocarmycin
structure, even of the extended analogues, can have profound effects on the
antitumour activity of the compound and that this is more likely related to uptake
than to DNA alkylation ability. Secondary amides of the
C-terminus of the
duocarmycin structure have been described that maintain their potent antitumour
2
8
activity, whereas in this study, we demonstrated that the presence of an OH or NH
on C-terminus, either with a carboxylic acid or a amide, leads to a decrease in activity
such that the extended compounds only just match the antiproliferative activity of
the truncated compound 30. That this is likely to be due to uptake is demonstrated
by the difference in 31, where DNA alkylation ability is similar to that for 30, and for
the lysine analogue 14, which alkylates DNA with far more efficiency than 30, but still
has lower antitumour activity.
Work is ongoing to explore the attachment of duocarmycins to antibodies and other
proteins through their incorporation into peptidic linker structures and will be
reported in due course.
Experimental Section.
All chemicals were reagent grade. HPLC mobile phases were prepared using HPLC
grade solvents. THF and DMF where specified as dry were bought as such and
22
ACS Paragon Plus Environment

Page 23 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
assumed to conform to the manufacturer’s standards. All water used was distilled.
All DMF for solid phase synthesis was purchased as peptide grade.
1
13
H and C-NMR spectra were recorded in Fourier Transform mode operating at a
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
nominal H NMR frequency of 400 MHz, using the specified deuterated solvent. The
1
13
chemical shifts for both H and C-spectra were recorded in ppm and were
referenced to the residual solvent peak. Multiplicities in the NMR spectra are
described as: s = singlet, d = doublet, dd = doublet of doublets, t = triplet, q =
quartet, m = multiplet, br = broad, appt = apparent; coupling constants are reported
in Hz. High resolution mass spectra were obtained using a quadrupole MS. Ionisation
by nano-electrospray. Infrared spectra were recorded as neat samples. Thin layer
chromatography was performed on aluminium plates coated with 0.2 mm silica gel-
60 F254. After elution, the TLC plates were visualised under UV light. Flash
chromatographic separations were performed on silica gel for column
chromatography (particle size 60 µm). Unless otherwise stated analytical RP-HPLC
was performed using a C18 column, 4.6 x 150mm, 5µM and a flow rate of 1 mL/min.
Solvent A = H
5% B over 15 min, 95% B
Unless otherwise stated preparative RP-HPLC was performed using a C18 column,
1.2 x 150 mm, 5 µM and a flow rate of 20 mL/min and a flow rate of 20 mL/min.
Solvent A = 95% H O 5% MeOH + 0.05% TFA and Solvent B = 95% MeOH 5% H O +
.05% TFA. Gradient 0 % B 100% B over 15 min, 100% B 0% B over 5 min.
2
O + 0.05% TFA and Solvent B = MeOH + 0.05% TFA. Gradient 5% B
ꢀ
9
ꢀ
5 % B over 5 min. Detection wavelength 254 nm.
2
2
2
0
ꢀ
ꢀ
Detection wavelength 254 nm. Unless otherwise stated RP-Flash chromatography
was performed using a pre-packed 12 g C18 column, and a flow rate of 20 mL/min.
Solvent A = 95% H O 5% MeOH + 0.05% TFA and Solvent B = 95% MeOH 5% H O +
2
2
23
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0.05% TFA. Gradient 0% B
Detection wavelength 254 nm. Samples were dry loaded by adsorption on Celite.
-Nitro-2-(phenylmethoxy)benzenamine (20)
BnBr (21 mL, 178 mmol) was added dropwise to a stirring suspension of 2-amino-5-
nitrophenol (25 g, 162 mmol) and K CO (49.3 g, 357 mmol) in DMF (250 mL) at room
ꢀ 100% B over 15 min, 100% B ꢀ 0% B over 5 min.
4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3
temperature. After 20 h, the reaction mixture was poured over crushed ice. The
precipitate was collected by filtration and triturated with cold water prior to drying
O
at 40 C under vacuum overnight. The reaction was repeated and the two batches
combined to afford 78.05 g of 20 as yellow/brown amorphous solid (98.5% average
o
29
yield over the 2 batches). R 0.17 (20 % EtOAc in hexane); mp 147-149 C (Lit 151-
f
o
1
1
53 C). H NMR (CDCl
3
, 400 MHz) δ 7.83 ( 1H, dd,
=8.7), 5.15 (2H, s), 4.60 (2H, brs). C NMR (CDCl
00 MHz) δ 144.6, 143.6, 138.7, 135.9, 128.9, 128.7, 128.0, 119.5, 112.1, 107.4,
J=2.4, 8.7), 7.77 (1H, d, J=2.4),
1
3
7.37-7.46 (5H, m), 6.66 (1H, d,
J
3
,
1
7
1
7
1.0. IR (neat) νmax 3483, 3359, 3225, 3188, 3075, 2939, 2876, 1622, 1579, 1519,
480, 1455, 1386, 1282, 1222, 1176, 1091, 1007, 950, 914, 870, 853, 818, 797, 755,
-1
+
13 2 3
44, 727, 697, 643, 623 cm . HRMS (ES+) calculated for C13H N O (M+H) 245.0921
found 245.0923.
-Iodo-4-nitro-6-(phenylmethoxy)benzenamine (21)
SO (800 µL, 15.15 mmol) was added to a stirring solution of 20 (37 g, 151 mmol)
2
H
2
4
in DMF (555 mL), followed by portionwise addition of NIS (51.1 g, 227 mmol) at
room temperature. After 4 h, the reaction mixture was poured over crushed ice. The
precipitate was collected by filtration and triturated with cold water, followed by
O
cold hexane, prior to drying at 40 C under vacuum overnight. The reaction was
repeated with 39.1 g of 21 and the two batches combined to afford 105.65 g as a
24
ACS Paragon Plus Environment

Page 25 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
f
bright yellow amorphoussolid (91.5% average yield over the 2 batches). R 0.31 (20 %
o
29
o
1
EtOAc in hexene); mp 103-106 C (Lit 105-106 C). H NMR (CDCl
1H, d, =2.3), 7.74 (1H, d, =2.3), 7.38-7.44 (5H, m) 5.16 (2H, s), 5.02 (2H brs). C
NMR (CDCl , 100 MHz) δ 144.1, 143.3, 138.9, 135.4, 129.0, 128.9, 128.3, 128.1,
3
, 400 MHz) δ 8.29
1
3
(
J
J
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
1
06.7, 178.5, 71.5. IR (neat) νmax 3476, 3379, 3359, 3091, 3056, 3030, 2357, 2333,
1602, 1568, 1497, 1451, 1425, 1386, 1282, 1237, 1099, 1037, 1025, 869, 849, 819,
-1
+
7
3
40, 726, 692 cm . HRMS (ES+) calculated for C H IN O (M+H) 370.9887 found
13 12 2 3
70.9890.
Methyl 3-[2-Amino-3-(phenylmethoxy)-5-nitrophenyl]-2-propynoic acid (22)
1 (40.8 g, 110 mmol) was dissolved in anhydrous DMF (1225 mL). The resulting
solution was degassed with a stream of N for 30 min prior to addition of methyl
propiolate (37.1 mL, 441 mmol), Pd(PPh Cl (3.87 g, 5.51 mmol), ZnBr (99 g, 441
2
2
3
)
2
2
2
mmol), and DIPEA (77 ml, 441 mmol) at room temperature . The reaction mixture
o
was then heated to 66 C and stirred overnight under N
2
. After cooling to room
temperature the reaction was poured over crushed ice, and the resulting chocolate
colour precipitate collected by filtration. The reaction was repeated with 51 g of 21,
and the precipitates were combined prior to adsorption on to 250 g of silica. Elution
through a 1 Kg silica plug with 50% ethyl acetate and hexane afforded 62 g of 22 as
an orange amorphous solid (77% yield). R
f
0.16 (20 % EtOAc in hexane); mp 136-139
o
1
C. H NMR (CDCl
3
, 400 MHz) δ 8.06 (1H, d,
J
=2.4), 7.76 (1H, d,
J
=2.4), 7.38-7.45 (5H,
1
3
m), 5.32 (2H, brs), 5.17 (2H, s), 3.86 (3H, s). C NMR (CDCl , 100 MHz) δ 154.1, 146.7,
3
1
44.5, 137.8, 135.2, 129.1, 129.0, 128.1, 123.0, 108.3, 101.0, 87.1, 81.1, 71.5, 53.1. IR
(neat) νmax 3499, 3391, 3351, 3087, 3063, 3030, 2951, 2204, 1698, 1611, 1455, 1430,
1
393, 1325, 1299, 1237, 1215, 1148, 1093, 1040, 1028, 1001, 886, 859, 755, 740,
25
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
-
1
+
731, 694, 657, 612 cm . HRMS (ES+) calculated for C17
found 327.0979.
-(1,1-Dimethylethyl)-2-methyl 5-nitro-6-(phenylmethoxy)indole-1,2-dicarboxylate
24)
15 2 5
H N O (M+H) 327.0975
1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
2
2 (60 g, 184 mmol) in anhydrous THF (858 mL) was treated with 1M TBAF in THF
o
solution (368 mL, 368 mmol) and refluxed at 66 C for 1 h. After cooling to room
temperature the THF was removed by rotary evaporation under reduced pressure.
The residue was dissolved in ethyl acetate (1000 mL) and washed 3 times with water
(1000 mL). Concentration of the ethyl acetate followed by co-evaporation of the
residue with CH Cl afforded crude product as a dark purple foam. The foam was
2
2
dissolved in CH
2
Cl
2
(1000 mL) and treated with Boc
2
O (80 g, 368 mmol), and DMAP
Cl gave a
(22.46 g, 184 mmol) at room temperature for 1.5 h. Removal of the CH
2
2
dark foam which was purified by silica gel chromatography using an automated flash
chromatography system. The crude was dry loaded on to a 1.5 kg pre-packed silica
column adsorbed on to 200 g of silica. A linear gradient of 0 to 30% ethyl acetate in
hexane was run over 23 column volumes and then held at 30% ethyl acetate until
complete elution of the product. Removal of the solvent afford 31 g of 24 as an
orange amorphous solid (39% yield of 2 steps). R
f
0.33 (20 % EtOAc in hexane); mp
=1.9), 7.67 (1H, d, =1.9), 7.49-
o
1
1
68-171 C. H NMR (CDCl
3
, 400 MHz) δ 8.26 (1H, d,
J
J
1
3
7.45 (2H, m), 7.41-7.34 (3H, m), 7.33 (1H, s) 5.33 (2H, s), 3.94 (3H, s), 1.47 (9H, s). C
NMR (CDCl , 100 MHz) δ 160.5, 149.3, 145.6, 143.6, 135.2, 130.2, 128.9, 128.7,
3
1
28.2, 126.4, 112.7, 112.5, 102.2, 86.5, 71.3, 52.5, 27.9, 27.3. IR (neat) νmax 3127,
3099, 3050, 2981, 2949, 1765, 1722, 1586, 1512, 1437, 1388, 1372, 1325, 1252,
26
ACS Paragon Plus Environment

Page 27 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
-
1
1223, 1151, 1115, 1073, 982, 875, 840, 822, 801, 778, 766, 742, 729, 697, 606 cm .
+
HRMS (ES+) calculated for C22
H
23
N
2
O
7
(M+H) 427.1500 found 427.1499.
1
-(1,1-Dimethylethyl)-2-methyl 5-[[(1,1-dimethylethoxy)carbonyl]amino]-4-iodo-7-
phenylmethoxy)indole-1,2-dicarboxylate (26)
4 (15 g, 35.2 mmol) was dissolved in THF (293 ml) and treated with zinc powder
(34.5 g, 528 mmol), NH Cl (18.82 g, 352 mmol), Boc O (23.03 g, 106 mmol), DMAP
430 mg, 3.52 mmol), and water (58.6 mL). The resulting suspension was stirred
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
2
4
2
(
vigorously at room temperature overnight. After removal of the zinc by filtration,
the THF was evaporated and the residue taken up in ether (500 mL). The ether was
washed 3 times with water (250 mL) and died over MgSO . Co-evaporation with
4
CH
2
Cl
2
gave crude 25 as a light brown foam. The reaction was repeated on the same
SO (0.375 mL,
.04 mmol) was added followed by portionwise addition of NIS (23.75 g, 106 mmol)
scale and the crudes combined and dissolved in DMF (352 mL). H
2
4
7
2
at room temperature. After 3 h the reaction was diluted with Et O (1000 mL), and
washed once with 50% saturated brine in water (1000 mL), twice with water (1000
mL), and once with saturated brine (1000 mL). The first wash was back extracted 3
times with Et
2
O (500 mL), which was subsequently combined and washed twice with
O was combined and concentrated to give a
saturated brine (1000 mL). All the Et
2
dark red foam which was purified by silica gel chromatography using an automated
flash chromatography system. The crude was dry loaded on to a 750 g pre-packed
silica column adsorbed on to 170 g of Celite. A linear gradient of 0 to 20% ethyl
acetate in hexane was run over 16 column volumes. Removal of the solvent afforded
26 g of 26 as an off white foam which dried to an amorphous solid (59% yield over 3
o
1
steps). R 0.37 (20 % EtOAc in hexane); mp 158-161 C. H NMR (CDCl , 400 MHz) δ
f
3
27
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
7.79 (1H, brs), 7.49-7.46 (2H, m), 7.30-7.38 (3H, m), 7.09 (1H, s), 6.77 (1H, brs), 5.24
13
(
3
2H, s), 3.91 (3H, s), 1.54 (9H, s), 1.41 (9H, s). C NMR (CDCl , 100 MHz) δ 160.9,
1
1
1
53.1, 149.9, 146.6, 136.0, 134.7, 131.5, 128.7, 128.6, 128.4, 127.8, 123.7, 114.6,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
02.6, 85.6, 81.0, 71.1, 52.3, 28.5, 27.3. IR (neat) νmax 3355, 2984, 2933, 1763, 1725,
716, 1615, 1575, 1541, 1505, 1449, 1393, 1361, 1338, 1310, 1256, 1221, 1152,
-1
1080, 980, 908, 878, 843, 817, 758, 723, 693 cm . HRMS (ES+) calculated for
+
C H O N I (M+H) 623.1249 found 623.1246.
2
7 32 7 2
1
-(1-,1-Dimethylethyl)-2-methyl
5-[(3-chloro-2-propenyl)[(1,1-
dimethylethoxy)carbonylamino]-4-iodo-7-(phenylmethoxy)indole-1,2-
dicarboxylate (27)
2
6 (26 g, 41.8 mmol) was dissolved in DMF (418 mL) and treated with t-BuOK (9.37 g,
84 mmol) and technical grade (90%) 1,3-dichloropropene as a mixture of cis and
trans isomers (12.90 mL, 125 mmol). After stirring for 1.5 h with the vessel
o
submerged in a room temperature water bath, the reaction was cooled to 0 C and
quenched with saturated aqueous NH
1000 mL), and washed twice with of 50% saturated brine in water (1000 mL), and
once with saturated brine (1000 mL). The Et O was dried over MgSO4, concentrated,
and co-evaporated with CH Cl 6 times to afford a brown foam which was purified by
4 2
Cl (20 mL). The mixture was diluted with Et O
(
2
2
2
silica gel chromatography using an automated flash chromatography system. The
crude was dry loaded on to a 220 g pre-packed silica column adsorbed on silica. A
linear gradient of 0 to 10% ethyl acetate in hexane was run over 16 column volumes.
Removal of the solvent afforded 18 g of 27 as a light brown foam which dried to an
f
amorphous solid (62% yield - mixture of E/Z isomers). R 0.31 (20 % EtOAc in
o
1
hexane); mp 104-106 C. H NMR (CDCl , 400 MHz, mixture of E/Z isomers) δ 7.28-
3
28
ACS Paragon Plus Environment

Page 29 of 47
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
7.44 (5H, m) 7.18 (1H, s), 6.65-6.47 (1H, m), 5.80-6.00 (2H, m]), 5.17-5.28 (2H, m),
4
.46 & 4.18 (1H, m), 4.33 & 3.73 (1H, m), 3.93 (3H, s), 1.53 (9H, s), 1.29 & 1.27 (9H,
1
3
s). C NMR (CDCl
3
, 100 MHz) δ 160.9, 154.2, 150.0, 145.7, 138.7, 135.9, 132.2,
28.9, 128.4, 128.0, 127.5, 125.4, 121.8, 120.7, 115.3, 109.7 86.0, 83.9, 80.6, 70.7,
2.4, 49.5, 46.2, 28.4, 27.3. IR (neat) νmax 2976, 2921, 1775, 1731, 1702, 1694, 1571,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
5
1535, 1467, 1454, 1435, 1391, 1372, 1299, 1251, 1227, 1150, 1118, 1077, 978, 932,
-1
8
85, 842, 829, 782, 764, 739, 731, 699 cm . HRMS (ES+) calculated for C H O N ClI
30 35 7 2
+
(
M+H) 697.1172 found 697.1174.
3.6-Bis(1,1-dimethylethyl)-2-methyl
8-(chloromethyl)-7,8-dihydro-4-
: 4,3-b’]dipyrrole-2,3,6-tricarboxylate (28)
7 (9 g, 12.9 mmol) was dissolved in anhydrous toluene and degassed with a stream
for 45 min prior to addition of AIBN (0.530 g, 3.23 mmol) and
(
phenylmethoxy)benzo[1,2-b
2
of N
2
Tris(trimethylsilyl)silane (TTMSS) (4.38 mL, 14.20 mmol). The resulting solution was
o
refluxed a 90 C under N
2
. After 1 h the reaction was allowed to cool to room
temperature before being concentrated and subjected directly to silica gel column
chromatography using an automated flash chromatography system. A 120 g pre-
packed silica column was used, and 0% ethyl acetate in hexane run for 5 column
volumes rising to 10% linearly over the subsequent 5 column volumes, holding at
1
0% until complete elution of the product. The reaction was repeated on the same
scale and the products combined, affording 10.28 g of 28 as white foam which dried
o
1
to an amorphous solid (70% yield). R 0.30 (20 % EtOAc in hexane); mp 115-118 C. H
f
6
NMR (DMSO-d , 400 MHz) δ 7.69 (1H brs), 7.47-7.29 (6H, m), 5.27 (2H, s), 4.13 (1H, t,
1
3
J
6
=9.7), 4.06-3.89 (4H, m), 3.87 (3H, s), 1.48 (9H, s), 1.39 (9H, s). C NMR (DMSO- d ,
1
00 MHz) δ 160.4, 151.4, 149.5, 145.1, 136.2, 128.4, 128.0, 127.9, 123.5, 113.2,
29
ACS Paragon Plus Environment