Enantio- And diastereoselective synthesis of duocarmycine-based prodrugs for a selective treatment of cancer by epoxide opening

Article abstract of DOI:10.1002/chem.200700988

For the enantio- und diastereoselective synthesis of the prodrug 2, the N-tert-butyloxycarbonyl-protected amine 7 was alkylated with the enantiopure epoxide 14 to give the amide 10. A regio- and facial-selective metal-mediated cyclisation by using a cuprate led to 17 with an inversion of configuration at C10. Subsequent transformation of the hydroxy group in 17 by using the Appel procedure afforded (1S, 10R)-9 with an unusual double inversion owing to neighbouring-group participation of the N-tert-butoxycarbonyl group. (1S, 10R)-9 is the key intermediate in the synthesis of the prodrug 2, which has been developed for a selective treatment of cancer based on the antibody-directed enzyme prodrug therapy as an analogue of the natural antibiotic duocarmycine SA (1).

Full text of DOI:10.1002/chem.200700988

FULL PAPER
DOI: 10.1002/chem.200700988
Enantio- and Diastereoselective Synthesis of Duocarmycine-Based Prodrugs
for a Selective Treatment of Cancer by Epoxide Opening
Lutz F. Tietze,*[a] Heiko J. Schuster,[a] Sonja M. Hampel,[a] Stephan Rühl,[b] and
Roland Pfoh[b]
Dedicated to Professor Manfred T. Reetz on the occasion of his 65th birthday
Abstract: For the enantio- und diaste-
reoselective synthesis of the prodrug 2,
the N-tert-butyloxycarbonyl-protected
amine 7 was alkylated with the enan-
tiopure epoxide 14 to give the amide
10. A regio- and facial-selective metal-
mediated cyclisation by using a cuprate
led to 17 with an inversion of configu-
ration at C10. Subsequent transforma-
tion of the hydroxy group in 17 by
using the Appel procedure afforded
(1S,10R)-9 with an unusual double in-
version owing to neighbouring-group
participation of the N-tert-butoxycar-
bonyl group. (1S,10R)-9 is the key in-
termediate in the synthesis of the pro-
drug 2, which has been developed for a
selective treatment of cancer based on
the antibody-directed enzyme prodrug
therapy as an analogue of the natural
antibiotic duocarmycine SA (1).
Keywords: Appel reactions · cup-
rates · epoxides · prodrugs · total
synthesis
Introduction
age with a conjugate of an appropriate enzyme and a mono-
clonal antibody that binds to tumour-associated antigens.
Prodrug 2 is one of the best prodrugs developed so far for
ADEPT: it contains a dimethylaminoethoxyindol-2-carbox-
ylic acid (DMAI) side chain[5] for binding to the minor
groove of DNA as well as improving the water solubility by
formation of a salt, and has b-galactose as a protecting
group. This prodrug has a rather low cytotoxicity of IC50 =
3600 nm, but a very high cytotoxicity in the presence of the
cleaving enzyme b-galactosidase to give the seco-drug 3,
which then leads to the final drug 6 with an IC50 value of
0.75 nm. It results in a QIC50 value of 4800,[6a] which together
with the high cytotoxicity of 3, makes 2 into a promising
candidate for a selective treatment of cancer. The excellent
results in vitro have been confirmed by experiments in
vivo.[7] The diastereomeric prodrug 4 cannot be used as the
cytotoxicity of the corresponding seco-drug 5 (enantiomer of
3) is too low.
The natural antibiotic duocarmycine SA[1](1) is a highly
potent cytostatic compound with an IC50 value of about
10 pm against different cancer cell lines and thus one of the
strongest anticancer agents known so far. The mode of
action is a specific alkylation of 3-N of adenine in DNA,
which occurs with opening of the cyclopropane moiety and
establishes an aromatic state of ring B in 1.[2] We have devel-
oped novel prodrugs such as 2 that are based on the natural
antibiotic duocarmycine (1) and tested them for their applic-
ability in antibody-directed enzyme prodrug therapy
(ADEPT; Scheme 1).[3,4] In this approach, the toxic effector
is liberated from a prodrug such as 2 by an enzymatic cleav-
[a] Prof. Dr. L. F. Tietze, Dipl.-Chem. H. J. Schuster,
Dipl.-Chem. S. M. Hampel
The tricyclic skeleton 9 of the prodrug 2 was obtained by
a radical cyclisation of 8, which was easily accessible by al-
kylation of 7 with 1,3-dichloro-2-butene (Scheme 2). Howev-
er, the cyclisation is rather unselective leading to the two
possible diastereomers as racemic mixtures 9a–d in an
almost 1:1:1:1 ratio.[6b] Although the diastereomers could be
separated by chromatography on silica gel and resolution of
both diastereomers was possible by chromatography on a
chiral support (Chiralpak IA), this synthesis does not meet
the requirements for efficiency and high yields to make the
Institut für Organische und Biomolekulare Chemie
der Georg-August-Universität Gçttingen
Tammannstrasse 2, 37077 Gçttingen (Germany)
Fax : (+49)551-399476
E-mail: ltietze@gwdg.de
[b] Dr. S. Rühl, Dipl.-Chem. R. Pfoh
Institut für Anorganische Chemie
der Georg-August-Universität Gçttingen
Tammannstrasse 4, 37077 Gçttingen (Germany)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2008, 14, 895 – 901
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
895

Herein we describe a new
synthetic approach towards
(1S,10R)-9a, which gives direct
access to the desired enantio-
and diastereopure compound.
Results and Discussion
The retrosynthesis of the target
molecule 9a leads via the cor-
responding hydroxy compound
(9a: OH instead of Cl) to the
epoxy derivative 10, which is
accessible by N-alkylation of
the naphthalene derivative 7
with
the
epoxide
12
(Scheme 3). The latter com-
pound can be obtained from
11. The epoxide 12 can exist as
four stereoisomers. However,
we have prepared and trans-
formed only the enantiomeric
epoxides (2R,3R)- and (2S,3S)-
12 as well as (3’R,4’R)-10 and
(3’S,4’S)-10 as the substrate for
the epoxidation 11 is available
mainly as an E isomer. We as-
sumed that the cyclisation re-
action of 10 with opening of
the epoxide would take place
in an anti fashion with inver-
sion of configuration at C3’.
However, we did not know at
that time, whether the Appel
reaction[8] would occur with in-
version, as usual, or with reten-
tion of configuration.
Scheme 1. Duocarmycine SA (1) and the prodrugs 2–5.
The known building block 7
was synthesised starting from
the corresponding acid through
a Curtius rearrangement and a
Koenigstein iodination.[9] For
Scheme 2. Former synthesis of 9: a) NaH, DMF, RT, 1 h, then (E/Z)-1,3-dichloro-2-butene, DMF, RT, 2 h,
98%; b) TTMSS, AIBN, toluene, 808C, 5 h. Yields: anti: 44%, syn: 42%, 9a/9b/9c/9dꢀ1:1:1:1. AIBN=azoi-
sobutyronitrile, Bn=benzyl, TTMSS=tris(trimethylsilyl)silane.
compound 2 available in the large quantities necessary for
clinical trials.
the alkylation of 7 to give the desired amides 10, 2,3-epoxy-
butanol derivatives 14 and 16a–e with R,R and S,S configu-
ration were synthesised starting from commercially available
Scheme 3. Retrosynthetic analysis of 9a.
896
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 895 – 901

Synthesis of Duocarmycine-Based Prodrugs
FULL PAPER
crotylalcohol (E/Z mixture 19:1; 11) by using the Sharpless–
Katsuki procedure with l-(+)-diisopropyl tartrate and d-
(À)-diethyl tartrate, respectively. The primarily formed
(2R,3R)- and (2S,3S)-2,3-epoxyalcohols 13 and 15 were ob-
tained with 71% ee and 74% ee, respectively, which corre-
sponds with the literature.[10] A separation of the mixture of
enantiomers to give the enantiopure compounds by chroma-
tography on a chiral support on a multigram scale at this
point proved to be unrealistic. The other two diastereomers
(resulting from small amounts of (Z)-crotylalcohol) were re-
moved by fractioned distillation.
After transformation of the enantioenriched 13 and 15
into the nosylates 14 and 16a, respectively, the two enantio-
mers could be obtained by fractioned crystallisation[11] in
greater than 99% ee for 14 and greater than 95% ee for
compounds 16 (Scheme 4). The following alkylation of 7
ment[14] proved to be false as the following metal-induced
cyclisation of the mixture led to more than 65% yield of the
desired product 17. The existence of separable rotamers
about the amide bond is less likely as these isomers are usu-
ally not stable at room temperature. Moreover, the addition-
al existence of rotameric forms in the two separated isomers
can be deduced from the NMR spectra. The two com-
pounds, both with very similar spectroscopic data, do not
convert into each other even at elevated temperatures of up
to 1008C. We assume that they are diastereomers owing to
À
the existence of a hindered rotation about the aryl N single
bond.
To support this assumption, we have prepared compound
10 without the iodine atom at the aryl moiety. This com-
pound, as expected, does not show any atropisomerism. The
atropisomerism of 10 and related compounds will be studied
in more detail in the future.[15]
For the metal-mediated cyclisation of 10 to give 17, we
had to explore several different conditions to give reasona-
ble results. Thus, many examples for metal-catalysed cyclisa-
tion reactions with terminal epoxides and alkyl-/arylhalides
through a metal-exchange reaction can be found in the liter-
ature; however, only a few cases for the reaction of non-ter-
minal epoxides have been reported so far.[16,17a] Common re-
agents for the necessary halogen–metal exchange are
RMgX,[17] BuLi,[18] CuI·PR3,[19] CuBr·Me2S,[18a] ZnX·nMe-
Li[20] and CuCN·nLiX.[17a,15] In the case of epoxide 10, there
are two possible reaction sites for the attack of the organo-
metallic centre leading to the 5-exo-trig and/or to the 6-
endo-trig products. However, in all the cyclisation reactions,
only the exo product was obtained. To achieve a stereoselec-
tive reaction, we concentrated on the corresponding lithium-
containing cuprates as it was known that lithium, as in the
Parham cyclisation, tends to coordinate to the oxygen of the
epoxide part to allow the formation of a sterically fixed
transition state.[21] Thus, for the epoxide opening of
(3’R,4’R)-10, we propose a lithium-coordinated transition
state TS-10/17 in which the attack of the metallated carbon
atoms C1 to C10 occurs in an anti fashion to give (1S,10R)-
17 with the S configuration at the newly formed stereogenic
centre (then C1; Scheme 5). As the second carbon (C11) of
the epoxide moiety does not take part in the reaction, its
original configuration should remain. It should be men-
tioned that the exact transition state using “cyano-Gilman
cuprates”, especially the location of the cyanide, has long
been a subject of controversial discussions.[22] The case of
the zincate with SCN ligand transition states have not been
Table 1. N-Alkylation of amide 7 with the epoxides 14 and 16a–e.
[a]
Epoxide
EquivAdditeisv
T/t
Yield of
[8C]/[h] 10 [%]
A
–
–
25/1.5
25/1.5
99
99
64
ACHTREUNG
16b (OMs)
1.5
[18]crown-6 ether (1.5), 25/240
TBAI (0.1)
16c (Cl)
16d (OTf)
16e (ONf)
1.3
1.0
1.5
[18]crown-6 ether (1.5) 70/2
69
14
0
–
–
0/0.5
25/96
[a] TBAI=tetrabutylammonium iodide.
was performed by using nosylates 14 and 16a to give 10 in
99% yield. Besides the nosylate 16a, we have also prepared
the chloride 16b, the mesylate 16c, the triflate 16d[12] and
the nonaflate 16e,[13] which could not be purified and more-
over gave much less satisfying results in the alkylation
(Table 1).
To our surprise, all alkylation reactions yielded a mixture
of two compounds in a ratio of 65:35 to 50:50, which could
be separated by chromatography on silica gel.
The primary assumption that regioisomers might have
been formed owing to an intramolecular Payne rearrange-
Scheme 4. Synthesis of derivatives of 2,3-epoxybutanol. a) NsCl, Et3N, toluene, RT, 30 min; 60% yield for 14 and 58% yield for 16a; b) CCl4, PPh3,
NaHCO3, reflux, 2 h, 92%; c) MsCl, Et3N, toluene, 08C, 1.5 h, 98%; d) nBuLi, À788C, 30 min, Tf2O, À788C for 1 h followed by RT for 1 h; e) NfF, Et3N,
DMF, RT, 15 h. Ms=mesylate, Ns= nosylate, Nf=nonafluorobutanesulfonate, Tf=triflate.
Chem. Eur. J. 2008, 14, 895 – 901
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
897

L. F. Tietze et al.
ing of the epoxide prior to the
ring closure can be widely ex-
cluded.
The obtained products could
be further purified by recrys-
tallisation to give, for example,
the desired enantiomer with
greater than 99% ee. The
structure of (1S,10R)-17 was
confirmed by X-ray analysis
(Figure 1).[25]
Scheme 5. Transition state with lithium coordination.
investigated so far, but a possible ligation of SCN cannot be
omitted owing to the tendency of zinc derivatives to form
complexes with a tetrahedral configuration.[18b,23] In any
case, a transition state of type TS-10/17 as well as a noncoor-
dinated anti attack should lead to (1S,10R)-17 starting from
(3’R,4’R)-10; this is indeed the case.
The best results for the cyclisation reaction (Table 2) were
achieved by using the copper organyl Li2Cu(CN)Me2 and
Table 2. Cyclisation reactions of (3’R,4’R)-10 (99% ee) to give (1S,10R)-
17.
Entry
Metal organyl
Yield [%]
Selectivity[a]
1
2
3
4
5
6
7
8
Li2Cu(CN)Me2
Li3Cu(CN)Me3
Li2Cu(CN)Me2
Li2ZnMe4
78
34
65
13
72
24
38
7
98:2
–
94:6
–
98:2
–
97:3
–
[b]
Figure 1. X-ray crystal structure of the alcohol (1S,10R)-17.
Li2Zn
A
LiZnMe3
iPrMgCl·LiCl
nBuLi
The final step towards the desired compound (1S,10R)-9
was the replacement of the hydroxy group in (1S,10R)-17 by
a chloride group by using an Appel reaction. Usually, the
Appel reaction proceeds with inversion of configuration.
However, in the transformation of (1S,10R)-17, we observed
a retention of configuration. The best results were obtained
by using a mixture of CH3CN/CCl4 (3:1) as the solvent and
2.2 equivalents of triphenylphosphine to give the desired
(1S,10R)-9 in 87% yield and 99.8% ee. We assume that this
stereochemical outcome is caused by a double inversion
under neighbouring-group participation of the N-tert-buty-
loxycarbonyl (Boc) protecting group (Scheme 6). From car-
bohydrate chemistry, many examples of neighbouring-group
participation are known with unhindered acyl derivatives.[26]
However, the involvement of a sterically highly demanding
Boc group is astounding. As part of the mechanism for the
Appel reaction of (1S,10R)-17, we can assume that primarily
the phosphane 19 is formed by nucleophilic attack of the
phosphonium salt 18. This is followed by an attack of the
carbonyl oxygen of the Boc protecting group with inversion
of the configuration and displacement of triphenylphosphine
oxide to give 20. The subsequent nucleophilic attack at C10
with chloride takes place with a second inversion of the con-
figuration in a SN2 fashion to give the desired compound
(1S,10R)-9.
[a] Determined by HPLC on Chiracel IA, see also Ref. [24]. [b] the reac-
tion was performed starting with (3’S,4’S)-10 with greater than 95% ee.
the zinc organyl Li2Zn(SCN)Me3, which led to the product
E
(1S,10R)-17 in 78% and 72% yield, respectively (Table 2,
entries 1 and 5); in contrast, employing Li3Cu(CN)Me,
Li2ZnMe4 and LiZnMe3 afforded (1S,10R)-17 in only 34%,
13% and 24% yield, respectively (Table 2, entries 2, 4 and
6). The use of the Grignard reagent iPrMgCl·LiCl gave 38%
of (1S,10R)-17, whereas employing nBuLi nearly exclusively
afforded the dehalogenated product with only 7% of the de-
sired product (Table 2, entries 7 and 8). This last reaction
clearly shows that the use of intermetal reagents is superior
for transformations with nonterminal epoxides when com-
pared with literature-known reactions with terminal epox-
ides in which the use of nBuLi gives good yields.[13] As ex-
pected, the transformation of the other enantiomer (3’S,4’S)-
10 with Li2Cu(CN)Me2 led to the alcohol (1R,10S)-17 in
65% yield (Table 2, entry 3).
In all cases, the reactions proceed with little loss of stereo
integrity with a selectivity[24] of 97:3 to 98:2 starting with a
compound of 99% ee (Table 2, entries 2, 5 and 7). Thus, the
formation of a carbocation as an intermediate by ring open-
898
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 895 – 901

Synthesis of Duocarmycine-Based Prodrugs
FULL PAPER
Scheme 6. Proposed neighbouring-group effect in the Appel reaction of (1S,10R)-17 to give (1S,10R)-9.
mined on a Bruker Vektor 22, UV/Vis spectra were recorded on a
Perkin-Elmer Lambda 2 and mass spectra were recorded on a Varian
MAT 311 A, Varian MAT 731 for EI-HRMS and a Bioapex fourier
transformation ion cyclotron resonance mass spectrometer for ESI-
HRMS.
1H NMR spectra were recorded either on a Varian VXR-200 MHz or
Varian UNITY-300 MHz. 13C NMR spectra were recorded at 50 or
75 MHz. Spectra were recorded at room temperature (except when oth-
erwise stated) and in deuterated solvents as indicated with the solvent
peak as the internal standard.
We also performed the reaction in pure CCl4; under these
conditions a 1:1 mixture of syn- and anti-9 was obtained.
This validates the above described mechanism as the reac-
tion proceeds via the polar intermediate 20 and should have
a lower energy of activation in a polar solvent. In all cases,
an undesired b elimination took place to some extent as a
consequence of the instability of the intermediate carboxo-
nium ion 20. An increase in b elimination took place in the
order of decreasing polarity of the solvent used in the trans-
formation (CH3CN-CCl4 <CCl4-CH2Cl2 <CCl4).
Selected experimental data, including representatives of each of the syn-
thetic methods, are given below. Full experimental data for all other reac-
tions are given in the Supporting Information.
Synthesis of (2R,3R)- and (2S,3S)-2,3-epoxy-1-nitrobenzenesulfonyloxy-
butane (14) and (16a):[11] Nosylchloride (3.54 g, 16.0 mmol, 1.05 equiv)
was added portionwise to a stirred solution of (2R,3R)-13[10] (71% ee) or
(2S,3S)-15 (74% ee; 1.34 g, 15.2 mmol, 1.0 equiv) and triethylamine
(2.54 mL, 1.84 g, 18.2 mmol, 1.2 equiv) in toluene (30 mL) at 08C. Stirring
was continued for 30 min at 258C and then the orange-coloured solution
was filtered over Celite and washed with toluene (50 mL). The solvent
was evaporated and the crude material was placed on a short silica gel
column and eluted with diethyl ether (250 mL). After removal of 50% of
the solvent, the white precipitated solid (racemate) was filtered off and
the mother liquor evaporated to dryness. Repeated crystallisation (ap-
proximately two times) from diethyl ether (100 mL) afforded the nosy-
lates 14a (1.66 g, 40%) and 16a (1.50 g, 36%), respectively, as yellow
crystals with stable optimal rotation values; mp 65–65.58C; [a]2D5 (CHCl3,
c=0.8, 99.8% ee)=+39.68; 1H NMR (300 MHz, CDCl3): d=1.32 (d, J=
5.3 Hz, 3H; 4-H3), 2.87–2.99 (m, 2H; 2-H, 3-H), 4.04 (dd, J=11.5, 6.4 Hz,
1H; 1-Ha), 4.42 (dd, J=11.5, 3.1 Hz, 1H; 1-Hb), 8.10–8.17 (m, 2H; 2”
Ph-H), 8.39–8.45 ppm (m, 2H; 2”Ph-H); 13C NMR (75 MHz, CDCl3):
d=16.88 (C4), 52.60, 55.34 (C2, C3), 71.38 (C1), 124.5, 129.3 (Ph-Co, Ph-
Cm), 141.5, 150.8 ppm (Ph-Ci, Ph-Cp); IR (KBr) n˜ : 3112, 1524, 1365, 1196,
1097 cmÀ1; UV lmax (log e): 251 nm (0.55); MS (DCI): m/z (%): 291
(100%) [M+NH4]+; HRMS (ESI): m/z: calcd for C11H15N2O6S
[M+NH4]+: 291.06459; found: 291.06453;
Conclusion
We developed a method of enantioselective access to
(1S,10R)-9, which is the key intermediate in the synthesis of
the highly potent prodrug 2 used in ADEPT. After alkyla-
tion of amide 7 with the enantiopure epoxy-nosylate 14 to
give 10, a copper organyl mediated cyclisation was found to
best lead to 17. Astoundingly, 10 was obtained as a mixture
of atropisomers owing to a hindered rotation about the
À
aryl N single bond. In the last step, the Appel reaction for
the transformation of the hydroxy group into the desired
chloride (1S,10R)-9 proceeded with retention of configura-
tion owing to neighbouring-group participation of the Boc
protecting group. This is a rare example for an Appel reac-
tion with induced double inversion with very high selectivity.
In conclusion, the sequence allows the synthesis of the enan-
tiopure anti-methyl-seco-CBI (cyclopropabenzindole) unit
(1S,10R)-9 on a gram scale with an overall yield of 67% in
three steps from the known amide 7.
Synthesis of (2S,3S)-2,3-epoxy-1-methanesulfonyloxybutane (16b): A so-
lution of mesylchloride (650 mg, 5.70 mmol, 1.0 equiv) in toluene (2 mL)
was added dropwise to a stirred solution of (2S,3S)-15[10] (74% ee;
500 mg, 5.67 mmol, 1.0 equiv) and triethylamine (0.92 mL, 0.67 g,
6.67 mmol, 1.2 equiv) in toluene (10 mL) at 08C within 10 min. After stir-
ring at 08C for 1 h, the solution was filtered over Celite and the solvent
removed under slightly reduced pressure to yield the analytically pure
product 16b (920 mg, 98%) as a colourless liquid. 1H NMR (300 MHz,
CDCl3): d=1.33 (d, J=5.0 Hz, 3H; 4-H3), 2.91–3.04 (m, 2H; 2-H, 3-H),
3.05 (s, 3H; S-CH3), 4.08 (dd, J=12.0, 6.1 Hz, 1H; 1-Ha), 4.45 ppm (dd,
J=12.0, 2.8 Hz, 1H; 1-Hb); 13C NMR (50 MHz, CDCl3): d=17.00 (C4),
37.81 (S-CH3), 52.56, 55.70 (C2, C3), 69.79 ppm (C1); MS (DCI): m/z
(%): 184 (100%) [M+NH4]+; HRMS (ESI): m/z: calcd for C5H14NO4S
[M+NH4]+: 184.06375; found: 184.06381;
Experimental Section
General: All reactions were performed in flame-dried glassware under
an atmosphere of argon. Solvents were dried and purified according to
the method defined by Perrin and Armarego. Commercial reagents were
used without further purification. Crotylalcohol (E/Z 19:1) was pur-
chased at Fluka. TLC was carried out on precoated Alugram SIL G/
UV254 (0.25 mm) plates from Macherey-Nagel & Co. Column chroma-
tography was carried out on silica gel 60 from Merck with a particle size
of 0.063–0.200 mm for normal pressure and 0.020–0.063 mm for flash
chromatography (pentane was used as the solvent). Melting points were
recorded on a Mettler FP61 and are uncorrected. IR spectra were deter-
Synthesis of (2S,3S)-1-chloro-2,3-epoxybutane (16c): A mixture of CCl4
(7 mL), PPh3 (1.78 g, 6.80 mmol, 1.2 equiv), NaHCO3 (100 mg,
Chem. Eur. J. 2008, 14, 895 – 901
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
899

L. F. Tietze et al.
1.19 mmol, 0.2 equiv) and (2S,3S)-15[10] (74% ee) (500 mg, 5.67 mmol,
1.0 equiv) was refluxed for 2 h. Filtration over a short silica gel column,
washing with diethyl ether (50 mL) and removal of the solvent under
slightly reduced pressure afforded the target compound 16c (560 mg,
(0.15); MS (ESI): m/z (%): 442 (100) [M+Na]+; HRMS (ESI): m/z:
calcd for C26H29NO4Na [M+Na]+: 442.19861; found: 442.19888.
Chromatographic resolution of rac-(1S,10R)-17:
A solution of rac-
(1S,10R)-17 (500 mg, 1.19 mmol) in a 3:2 mixture of n-heptane and
CH2Cl2 (10 mL) was separated (injection volume: 0.2 mL) by semiprepar-
ative HPLC (Chiralpak IA, 250”20 mm, particle size: 5 mm, n-heptane/
92%) as
a
colourless liquid. [a]2D0 (C6H6, c=1.0, 74% ee)=À36.58;
1H NMR (300 MHz, CDCl3): d=1.36 (d, J=5.1 Hz, 3H; 4-H3), 2.93–3.01
(m, 2H; 2-H, 3-H), 3.53–3.58 ppm (m, 2H, 1-Ha,b), 13C NMR (50 MHz,
CDCl3): d=17.1 (C4), 44.6 (C1), 55.0, 58.0 ppm (C2, C3); IR (film) n˜ :
CH2Cl2 78:22, flow: 18 mLminÀ1
; UV-detector: l=250 nm, Jasco-
module) to provide (À)-(1R,10S)-17 (tR =13.7 min) and (+)-(1S,10R)-17
(tR =17.9 min). The optical purity was determined by analytical HPLC
(Chiralpak IA, 250”4.6 mm, particle size: 10 mm, n-hexane/iPrOH 7:3,
3444, 2971, 1439, 1380, 1265 cmÀ1
.
Synthesis of (3’R,4’R)- and (3’S,4’S)-2-amino-4-benzyloxy-N-(3,4-epoxy-
butyl)-1-iodo-N-(tert-butoxycarbonyl)-naphthalene (3’R,4’R)-10) and
(3’S,4’S)-10): Coupling with the nosylates 14/16a. A NaH (168 mg 60%
suspension, 4.28 mmol, 4.0 equiv) was added to a stirred solution of the
carbamate 7 (500 mg, 1.05 mmol, 1.0 equiv) in DMF (5 mL) at 258C and
stirring was continued for 30 min. After addition of the nosylate 14 or
16a (431 mg, 1.58 mmol, 1.5 equiv), the solution was stirred at 258C for
1.5 h. The reaction was quenched by the addition of a saturated NaHCO3
solution (100 mL) and the mixture was then extracted with CH2Cl2 (3”
50 mL). The combined organic extracts were washed with brine (50 mL)
and dried over Na2SO4. Column chromatography on silica (pentane/ethyl
acetate 10:1) yielded the title compound as a yellow solid (562 mg of
(3’R,4’R)-10 and (3’S,4’S)-10, respectively (99%)). 1H NMR (300 MHz,
C2D2Cl4, 1008C, both atropisomers, differing signals assigned with *): d=
1.20, 1.22 (2”d, 3H, J=5.1 Hz; 12-H, 5’-H*), 1.33–1.44 (m, 9H; C-
flow: 0.8 mLminÀ1
;
UV detector: l=250 nm, Kontron-module): (À)-
(1R,10S)-17: 91.4% ee (tR =5.2 min); [a]2D0 =À20.58(CHCl3, c=0.8); (+)-
(1S,10R)-17: 98.9% ee (tR =5.9 min); [a]2D0 =+22.58(CHCl3, c=0.5).
AHCTREUNG
a
2.38 mmol, 1.0 equiv) in CCl4 (10 mL) and CH3CN (30 mL) at À188C.
After 30 min, the mixture was allowed to warm to 08C and stirring was
continued for another 3 h. The reaction was quenched by the addition of
silica gel (5 g) and evaporation of the solvent. Column chromatography
on silica gel (pentane/ethyl acetate 10:1) yielded
a colourless solid
(905 mg, 87%). The optical purity was determined as described in the lit-
erature;[6b] [a]D20 =+28.08 (CHCl3, c=0.8, 99.8% ee);[6b] [a]2D0 =+28.08
(CHCl3, c 0.8, 99.9% ee); 1H NMR (300 MHz, CDCl3): d=1.48–1.56 (m,
(CH3)3), 2.65 (dq, J=5.1, 2.2 Hz, 0.5H; 4’-H*), 2.78 (dq, J=5.1, 2.2 Hz,
12H; C(CH3)3, 11-H3), 3.78 (mc, 1H; 1-H), 4.00 (mc, 1H; 2-Ha), 4.26 (mc,
G
1H; 2-Hb), 4.53 (mc, 1H; 10-H), 5.19 (s, 2H; OCH2Ph), 7.09–7.63 (m,
8H; 7-H, 8-H, 9-H, 5”Ph-H), 7.78 (br s, 1H; 4-H), 8.23 ppm (d, J=
8.2 Hz, 1H; 6-H); 13C NMR (150 MHz, CDCl3): d=23.74 (C11), 28.50
0.5H; 4’-H), 2.81 (dt, J=5.3, 2.2 Hz, 0.5H; 3’-H), 3.16 (dddc, J=6.5, 4.8,
*
2.2 Hz, 0.5H; 3’-H*), 3.30 (dd, J=14.6, 6.4 Hz, 0.5H; 2’-Ha ), 3.41 (dd, J=
14.6, 5.3 Hz, 0.5H; 2’-Ha), 3.90 (dd, J=14.6, 4.7 Hz, 0.5H; 2’-Hb*), 4.11
(dd, J=14.6, 5.0 Hz, 0.5H; 2’-Hb), 5.26, 5.24 (2”s, 2H; OCH2Ph), 6.83,
6.90 (2”s, 1H; 3-H), 7.27–7.60 (m, 7H; 6-H, 7-H, 5”Ph-H), 8.18,
8.18 ppm (2”mc, 2H; 5-H, 8-H); 13C NMR (300 MHz, C2D2Cl4, 908C,
both atropisomers, differing signals are assigned with *): d=16.96 (C5’),
(C
A
AHCTREUNG
C9), 127.6, 127.9, 128.5 (5”Ph-CH), 127.2 (C8), 130.4 (C9a), 136.9 (Ph-
Ci), 142.1 (C3a), 152.4 (C=O), 155.8 ppm (C5); m/z found [ESI]: 460
(100%, [M+Na]+). C26H28NO3Na requires 460.166.
28.1, 28.2 (C
ACHTREUNG
56.46 (C3’*), 56.84 (C3’), 70.59*, 70.61
A
ACHTREUNG
(C1), 107.8, 108.0* (C3), 122.4 (C5), 125.5 (C4a), 126.0, 126.1*, 127.1,
127.2* (C6, C7), 127.9, 128.2, 128.3, 128.4, 128.5 (5”Ph-CH), 132.5 (C8),
135.2*, 135.3 (C8a), 136.3*, 136.4 (Ph-Ci), 143.1, 143.7* (C2), 153.4, 153.69*
(C=O), 155.5 ppm (C-4); IR (KBr): n˜ =2986, 1691, 1592, 1379, 1334,
1150, 762 cmÀ1; UV/Vis (MeCN): lmax (log e)=216 (1.14), 244 (0.65),
305 nm (0.21); HRMS (ESI): m/z: calcd for C26H29INO4 [M+H]+:
546.11334; found: 546.11358; m/z: calcd for C26H28INO4Na [M+Na]+:
568.09530; found: 568.09552.
Acknowledgements
This research was supported by the Deutsche Forschungsgemeinschaft
and the Fonds der chemischen Industrie. H.J.S. would like to thank the
Friedrich-Ebert-Stiftung for a PhD scholarship.
Synthesis of (1S,10R)-5-benzyloxy-3-(tert-butoxycarbonyl-1-(10-hydroxy-
ACHTREUNG
[1] a) L. J. Hanka, A. Dietz, S. A. Gerpheide, S. L. Kuentzel, D. G.
Martin, J. Antibiot. 1978, 31, 1211; b) D. G. Martin, C. Biles, S. A.
Gerpheide, L. J. Hanka, W. C. Krueger, J. P. McGovren, S. A.
Mizsak, G. L. Neil, J. C. Stewart, J. Visser, J. Antibiot. 1981, 34,
1119; c) M. Ichimura, T. Ogawa, K. Takahashi, E. Kobayashi, I. Ka-
wamoto, T. Yasuzawa, I. Takahashi, H. Nakano, J. Antibiot. 1990,
43, 1037; d) B. K. Bhuyan, K. A. Newell, S. L. Crampton, D. D.
von Hoff, Cancer Res. 1982, 42, 3532; e) D. G. Martin, C. G. Chidest-
er, D. J. Duchamp, S. A. Mizsak, J. Antibiot. 1980, 33, 902.
[2] a) D. L. Boger, D. L. Hertzog, B. Bollinger, D. S. Johnson, H. Cai, J.
Goldberg, P. Turnbull, J. Am. Chem. Soc. 1997, 119, 4977; b) D. L.
Boger, B. Bollinger, D. L. Hertzog, D. S. Johnson, H. Cai, P. MØsini,
R. M. Garbaccio, Q. Jin, P. A. Kitos, J. Am. Chem. Soc. 1997, 119,
4987.
[3] Reviews: a) W. A. Denny, Cancer Invest. 2004, 22, 604; b) L. F.
Tietze, T. Feuerstein, Curr. Pharm. Des. 2003, 9, 2155; c) L. F.
Tietze, T. Feuerstein, Aust. J. Chem. 2003, 56, 841; d) W. A. Denny,
Eur. J. Med. Chem. 2001, 36, 577; e) M. Jung, Mini-Rev. Med. Chem.
2001, 1, 399; f) G. Xu, H. L. McLeod, Clin. Cancer Res. 2001, 7,
3314; g) K. N. Syrigos, A. A. Epenetos, Anticancer Res. 1999, 19,
605; h) G. M. Dubowchik, M. A. Walker, Pharmacol. Ther. 1999, 83,
67; i) C. J. Springer, I. Niculescu-Duvaz, Adv. Drug Delivery Rev.
1997, 26, 151; j) L. N. Jungheim, T. A. Shepherd, Chem. Rev. 1994,
94, 1553; first publication of the ADEPT concept: k) K. D. Bag-
shawe, Br. J. Cancer 1987, 56, 531.
a
2.92 mmol, 3.0 equiv) was added dropwise to a stirred suspension of
CuCN (131 mg, 1.46 mmol, 1.5 equiv) in THF (5 mL) at À788C and stir-
ring was continued for 30 min at À408C. After cooling again to À788C, a
solution of (3’R,4’R)-10 (530 mg, 0.972 mmol, 1.0 equiv) in THF (5 mL)
was added dropwise and stirring was continued for 1 h at À788C. The
mixture was then allowed to warm to 258C (30 min) and stirring was con-
tinued for and additional 45 min. Afterwards, the solvent was removed
under high vacuum. The residue was dissolved in ice water (200 mL), the
mixture extracted with CH2Cl2 (3”100 mL), the combined organic ex-
tracts washed with brine (100 mL), dried over Na2SO4 and the solvent fi-
nally evaporated. Purification by column chromatography on silica gel
(pentane/ethyl acetate 5:1) yielded (1S,10R)-17 as a colourless foam
(317 mg, 78%); 1H NMR (600 MHz, CDCl3): d=1.30 (d, J=6.5 Hz, 3H;
11-H3), 1.60 (s, 9H; C(CH3)3), 3.72 (mc, 1H; 1-H), 4.03 (mc, 1H; 2-Ha),
G
4.27 (mc, 1H; 2-Hb), 4.43 (mc, 1H; 10-H), 5.29 (s, 2H; OCH2Ph), 7.30–
7.60 (m, 7H; 7-H, 8-H, 5”Ph-H), 7.71 (d, J=8.3 Hz; 9-H), 7.93 (br s,
1H; 4-H), 8.31 ppm (d, J=8.6 Hz, 1H; 6-H); 13C NMR (125 MHz,
CDCl3): d=20.22 (11-CH3), 28.51 (C
68.33 (C10), 70.38 (OCH2Ph), 80.80 (C
ACHTREUNG
AHCTREUNG
122.4 (C5a), 122.5, 122.9, 123.5 (C6, C7, C9), 127.2 (C8), 127.6, 127.9,
128.5 (5”Ph-CH), 130.5 (C9a), 136.9 (Ph-Ci), 142.3 (C3a), 152.5 (C=O),
155.7 (C5); IR (KBr): n˜ =3442, 2925, 2854, 1698, 1626, 1583, 1458 cmÀ1
;
UV/Vis (MeCN): lmax (log e)=208 (0.48), 256 (0.75), 304 (0.14), 315 nm
900
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 895 – 901

Synthesis of Duocarmycine-Based Prodrugs
FULL PAPER
[4] a) L. F. Tietze, T. Herzig, A. Fecher, F. Haunert, I. Schuberth, Chem-
BioChem 2001, 2, 758; b) L. F. Tietze, T. Feuerstein, A. Fecher, F.
Haunert, O. Panknin, U. Borchers, I. Schuberth, F. Alves, Angew.
Chem. 2002, 114, 785; Angew. Chem. Int. Ed. 2002, 41, 759; c) L. F.
Tietze, B. Krewer, H. Frauendorf, F. Major, Angew. Chem. 2006,
118, 6720; Angew. Chem. Int. Ed. 2006, 45, 6570; d) L. F. Tietze, F.
Major, Eur. J. Org. Chem. 2006, 2314.
[5] J. B. J. Milbank, M. Tercel, G. Atwell, W. R. Wilson, A. Hogg, W. A.
Denny, J. Med. Chem. 1999, 42, 649.
[6] a) L. F. Tietze, F. Major, I. Schuberth, Angew. Chem. 2006, 118,
6724; Angew. Chem. Int. Ed. 2006, 45, 6574; b) L. F. Tietze, F.
Major, I. Schuberth, D. A. Spiegl, B. Krewer, K. Maksimenka, G.
Bringmann, J. Magull, Chem. Eur. J. 2007, 13, 4396.
[18] a) M. P. Cooke, I. N. Houpis, Tetrahedron Lett. 1985, 26, 3643;
b) J. H. Babler, W. E. Bauta, Tetrahedron Lett. 1984, 25, 4323.
[19] a) R. D. Rieke, D. E. Stack, B. T. Dawson, T.-C. Wu, J. Org. Chem.
1993, 58, 2483; b) M. S. Tichenor, J. D. Trzupek, D. B. Kastrinsky, F.
Shiga, I. Hwang, D. L. Boger, J. Am. Chem. Soc. 2006, 128, 15683.
[20] a) Y. Kondo, T. Matsudaira, J. Sato, N. Murata, T. Sakamoto,
Angew. Chem. 1996, 108, 818; Angew. Chem. Int. Ed. Engl. 1996, 35,
736; b) M. Uchiyama, M. Kameda, O. Mishima, N. Yokoyama, M.
Koike, Y. Kondo, T. Sakamoto, J. Am. Chem. Soc. 1998, 120, 4934.
[21] a) I. R. Hardcastle, P. Quayle and E. L. M. Ward, Tetrahedron Lett.,
1994, 35, 1747; b) S. Mori, E. Nakamura, K. Morokuma, J. Am.
Chem. Soc. 2000, 122, 7294.
[22] a) N. Krause, Angew. Chem. 1999, 111, 83; Angew. Chem. Int. Ed.
1999, 38, 79; b) J. P. Snyder, D. P. Spangler, J. R. Behling, B. E. Ros-
siter, J. Org. Chem. 1994, 59, 2665; c) S. H. Bertz, G. Miao, M. Eriks-
son, Chem. Commun. 1996, 815; d) T. L. Stemmler, T. M. Barnhart,
J. E. Penner-Hahn, C. E. Tucker, P. Knochel, M. Behme, G. Frenk-
ing, J. Am. Chem. Soc. 1995, 117, 12489; e) E. Nakamura, S. Mori,
Angew. Chem. 2000, 112, 3902; Angew. Chem. Int. Ed. 2000, 39,
3750.
[7] F. Alves, M. Zientkowska, F. Major, B. Krewer, H. J. Schuster, I.
Schuberth, L. F. Tietze, unpublished work.
[8] R. Appel, Angew. Chem. 1975, 87, 863.
[9] D. L. Boger, J. A. McKie, J. Org. Chem. 1995, 60, 1271.
[10] (2S,3S)-13 and -15: a) J. D. White, P. Theramongkol, C. Kuroda, J. R.
R
Engebrecht, J. Org. Chem. 1988, 53, 5909; b) H.-C. Guo, X.-Y. Shi,
X. Wang, S.-Z. Liu, M. Wang, J. Org. Chem. 2004, 69, 2042;
(2R,3R)-13 and -15: B. E. Rossiter, T. Katsuki, K. B. Sharpless, J.
Am. Chem. Soc. 1981, 103, 464; (2S,3S)- and (2R,3R)-13 and -15: D.
Pons, M. Savignac, J.-P. Genet, Tetrahedron Lett. 1990, 31, 5023.
[23] E. Weiss, Angew. Chem. 1993, 105, 1565; Angew. Chem. Int. Ed.
Engl. 1993, 32, 1501.
[24] In contrast with the chloride anti-9, the separation was realised on a
Chirapak IA on a Kontron 322 system, a DAD 440 Kontron detec-
tor and with KromaSystem2000 as software. Despite an extensive
search, no conditions for a baseline separation were found.
[25] CCDC-652396 contains the supplementary crystallographic data for
this paper. The structure can be found in the ESI and was solved by
direct methods by using SHELXS-97[27] and was refined against F2
on all data by full-matrix least squares with SHELXL-97.[28] These
data can also be obtained free of charge from The Cambridge Crys-
tallographic Data Centre via ;www.ccddc.cam.ac.uk/data_request/cif.
[26] H. Paulsen, Angew. Chem. 1982, 94, 184; Angew. Chem. Int. Ed.
Engl. 1982, 21, 155.
[11] a) S. T. Pickard, H. E. Smith, P. L. Polavarapu, T. M. Black, A.
Rauk, D. Yang, J. Am. Chem. Soc. 1981, 103, 464; b) S. T. Pickard,
H. E. Smith, P. L. Polavarapu, T. M. Black, A. Rauk, D. Yang, J.
Am. Chem. Soc. 1992, 114, 6850; c) R. M. Hanson, Chem. Rev. 1991,
91, 437.
[12] S. Muto, K. Mori, Biosci. Biotechnol. Biochem. 2003, 67, 1559.
[13] P. N. M. Botmann, J. Fraanje, K. Boubutz, R. Peschar, J. M. Verho-
even, J. H. van Maarseven, H. Hiemstra, Adv. Synth. Catal. 2004,
346, 743.
[14] a) G. B. Payne, J. Org. Chem. 1962, 27, 3819; b) P. C. Bulman Page,
C. M. Rayner, I. O. Sutherland, J. Chem. Soc. Perkin Trans. 1 1990,
1375.
[15] The determination of the rotational energy barrier of 10 and deriva-
tives is currently in progress.
[27] G. M. Sheldrick, SHELXS-97 Program for Crystal Structure Solu-
tion 1997, Gçttingen, Germany.
[28] G. M. Sheldrick, SHELXL-97 Program for Crystal Structure Refine-
ment 1997, Gçttingen, Germany.
[16] a) K. Shankaran and V. Snieckus, J. Org. Chem. 1984, 49, 5022;
b) K. L. Dhawan, B. D. Gowland, T. Durst, J. Org. Chem. 1980, 45,
924; c) T.-C. Wu, R. D. Rieke, Tetrahedron Lett. 1988, 29, 6753.
[17] a) D. Corbel, J. M. Descesare, T. Durst, Can. J. Chem. 1978, 56, 505;
b) D. Corbel, T. Durst, J. Org. Chem. 1976, 41, 3648.
Received: June 28, 2007
Revised: October 10, 2007
Published online: November 21, 2007
Chem. Eur. J. 2008, 14, 895 – 901
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
901