New insights into cyclobutenone rearrangements: A total synthesis of the natural ROS-generating anti-cancer agent cribrostatin 6

Article abstract of DOI:10.1002/chem.201102263

Aryl- and heteroarylcyclobutenone rearrangements proceed in excellent yield under continuous-flow conditions. The former shows a Hammett correlation with σ_I providing strong evidence that electrocyclisation is the rate-determining step and has a late transition state. The reaction has been modelled by using DFT and CCSD(T) methods, with the latter giving excellent correlation with the experimental rate constant. A short and efficient total synthesis of cribrostatin 6, an anti-neoplastic and anti-microbial agent, provides a topical demonstration of the value of this method.

Full text of DOI:10.1002/chem.201102263

DOI: 10.1002/chem.201102263
New Insights into Cyclobutenone Rearrangements: A Total Synthesis of the
Natural ROS-Generating Anti-Cancer Agent Cribrostatin 6**
Mubina Mohamed,[a] Thꢀo P. GonÅalves,[a] Richard J. Whitby,[a] Helen F. Sneddon,[b] and
David C. Harrowven*[a]
13698
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 13698 – 13705

FULL PAPER
Abstract: Aryl- and heteroarylcyclobu-
tenone rearrangements proceed in ex-
cellent yield under continuous-flow
conditions. The former shows a Ham-
mett correlation with sI providing
strong evidence that electrocyclisation
is the rate-determining step and has a
late transition state. The reaction has
been modelled by using DFT and
CCSD(T) methods, with the latter
giving excellent correlation with the ex-
perimental rate constant. A short and
efficient total synthesis of cribrosta-
tin 6, an anti-neoplastic and anti-micro-
bial agent, provides a topical demon-
stration of the value of this method.
Keywords: density functional calcu-
lations · flow chemistry · Hammett
correlation · reaction mechanisms ·
thermochemistry
Introduction
Thermal rearrangements of aryl- and heteroarylcyclobute-
nones have become established as useful methods for the de
novo synthesis of many polyaromatic and heteroaromatic
ring systems, especially those with dense substitution pat-
terns.[1–3] Though widely used, little is known about the fac-
tors that influence the course of these reactions, or indeed
the optimal conditions for effecting them. Herein we present
a detailed study of the rearrangement and show how it is
possible to achieve near-quantitative conversions under con-
tinuous flow. In turn, this has allowed us to establish a Ham-
mett relationship for the reaction which,[4] in conjunction
with an in silico study, provides new insights into the mecha-
nistic course. Extensions to heteroarylcyclobutenone rear-
rangements include a short and efficient total synthesis of
the marine natural product cribrostatin 6,[5–9] which displays
useful anti-microbial and anti-cancer activity through a reac-
tive-oxygen species (ROS)-generating mechanism.[5,7]
Scheme 1. Arylcyclobutenone rearrangements under continuouse flow
(with isolated yields quoted for the formation of 3 from 1).
ment of cyclobutenone 1a to benzohydroquinone 2a in
99% isolated yield in less than 1 h.
By reasoning that the marked improvement in efficiency
was due to a tight control of temperature across the narrow
tubing, we were pleased to observe similar results for the re-
arrangement of
a range of related substrates 1b–i
(Scheme 1). In each case, reactions took longer to proceed
to completion than the parent compound 1a, with electron-
withdrawing substituents (F, CF3, amide) and some electron
donors (OMe) slowing the reaction down significantly. To
delineate the nature of substituent effects, the progress of
each reaction was determined at various residence times by
Results and Discussion
Reaction optimization under continuous flow and establish-
ment of a Hammett relationship: Our investigation began
with a survey of the arylcyclobutenone rearrangement, 1!
2+3 (Scheme 1).[1] In batch, reactions of this type are usual-
ly conducted in xylenes at reflux (ca. 1358C) and typically
give yields of 70–85% after 2–10 h.[1–3] By contrast, under
continuous flow on a Vapourtec R4/R2+ instrument with
stainless-steel tubing of 1 mm diameter,[10] it was possible to
employ dioxane as the solvent at 1508C to induce rearrange-
1
using H NMR analysis to assess the extent of conversion.[11]
Though complicated by incomplete aerial oxidation of the
product hydroquinones 2 to the respective benzoquinones 3,
the method proved reliable in establishing that each rear-
rangement exhibited first-order kinetics (Figure 1).
With these data a Hammett relationship for the reaction
was sought. No correlation was evident with either the sm or
sp parameter sets, as would be expected if the electrocyclic
opening of the cyclobutenone were involved in the rate-de-
termining step (Figure 2, 1a!5). A reasonable correlation
was given with the sI (inductive) parameter set (R2 =0.8584,
Figure 3, grey line),[4b] with the parent compound 1a (X=
H) and those with large substituents (tBu, Me3Si, CF3) show-
ing greatest deviation from the line of best fit. This suggest-
ed a significant steric component to the reaction.[4,12] Indeed,
by introducing a small steric correction factor (sIꢀ6% Es)
the correlation was improved to R2 =0.989 (Figure 3, bold
line).[4,12]
[a] M. Mohamed, T. P. GonÅalves, Prof. R. J. Whitby,
Prof. D. C. Harrowven
Chemistry, University of Southampton
Highfield, Southampton, Hampshire, SO17 1BJ (UK)
Fax : (+44)23-8059-6805
E-mail: dch2@soton.ac.uk
[b] Dr. H. F. Sneddon
c/o GlaxoSmithKline Medicines Research Centre
Gunnels Wood Road, Stevenage, SG1 2NY (UK)
[**] ROS=reactive-oxygen species
Computational studies on the reaction mechanism: The
Hammett relationship observed suggests that the electrocyc-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201102263.
Chem. Eur. J. 2011, 17, 13698 – 13705
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13699

D. C. Harrowven et al.
Figure 3. Hammett plot for the arylcyclobutenone rearrangement 1!2+
&
3 with the sI parameter set ( ) and with a steric correction (sIꢀ6% Es,
[13]
^
).
bond, which is developing between the arene and the car-
bonyl in 6, is established; this explains the observed influ-
ence of inductive rather than resonance effects. To test this
hypothesis, the course of the reaction was modelled by DFT
Figure 1. Determination of rate constants for 1!2+3. X=CF3: k=4.4ꢁ
10ꢀ4 sꢀ1, R2 =0.96; X=F: k=4.6ꢁ10ꢀ4 sꢀ1, R2 =0.98; X=OMe: k=6.6ꢁ
10ꢀ4 sꢀ1, R2 =0.99; X=CONiPr2: k=8.0ꢁ10ꢀ4 sꢀ1, R2 =0.98; X=Ph: k=
8.2ꢁ10ꢀ4 sꢀ1, R2 =1.00; X=tBu: k=10.2ꢁ10ꢀ4 sꢀ1, R2 =0.99; X=SiMe3:
k=12.1ꢁ10ꢀ4 sꢀ1, R2 =0.98; X=Me: k=13.1ꢁ10ꢀ4 sꢀ1, R2 =0.97; X=H:
k=14.9ꢁ10ꢀ4 sꢀ1, R2 =1.00. [SM]t =concentration of starting material at
a given time.
calculations at the UB3LYP/6-311GACTHNUTRGNEUNG(d,p) level with the
Gaussian 09 program.[14,15] The estimated E+ZPVE values
(Figure 2, in green) were interesting in that they showed
little difference between the activation barrier for electrocy-
clic ring opening (21.6 kcalmolꢀ1) and ring closing (21.9 kcal
molꢀ1). However, when these were corrected to reflect free
energy, the calculated values for DG at 1508C (23.4 and
25.3 kcalmolꢀ1, respectively) supported our postulate that
electrocyclisation of vinylketene 5!7 is rate determining.
Importantly, they also showed the electrocyclic opening of
1a!5 to be reversible with the equilibrium favouring the
cyclobutenone rather than the ketene.
A limitation of the DFT method was exposed when we
sought to relate the predicted DG values to the reaction
rates observed experimentally for the rearrangement 1a!
2a. The calculated values implied a reaction rate substantial-
ly faster than the observed one, underestimating the energy
requirements by nearly 5 kcalmolꢀ1. Consequently, we re-
fined our analysis further by employing high ab initio single-
point energy calculations [RCCSD(T)/6-31G(d)] with the
GAMESS(US) package.[17,18] The results attained predicted
a
rate constant for the rearrangement of 1a!2a of
0.0016 sꢀ1 after correction for the free energy at 1508C; this
is in excellent agreement with the observed value
(0.00149 sꢀ1). In addition, these calculations reaffirmed that
the electrocyclic closure 5!7 is rate limiting (30.5 kcalmolꢀ1
compared with 28.9 kcalmolꢀ1 for 1a!5).
The calculated geometry for transition-state 6 (Figure 4)
is also instructive as it shows an angle of incidence of 40.08
between the developing s bond and the plane of the arene.
The angle is reduced to 22.18 as the reaction progresses to
intermediate 7. Thus, interaction between this nascent s
bond and the residual p system is limited as it develops to
become part of the s framework—an observation that is
consistent with a late transition state under the influence of
inductive rather than resonance effects.
Figure 2. Summary of the energies calculated for the rearrangement 1a!
2a in the gas phase at 1508C computed by DACTHNUTRGNEUNG(E+ZPE) UB3LYP/6-311G-
A
ACHTUNGERTN(NUNG d,p)) (in blue) and DG-
A
ACHTUNGTRENNUNG
lisation of ketene 5 to bicyclic ketone 7 is rate determining
and has a late transition state (i.e., 6 is more akin to inter-
mediate 7 than precursor ketene 5, Figure 2). The rate of re-
action is thus dictated by the ease with which the sigma
Further exemplifications of the method and a total synthesis
of cribrostatin 6: Our attention next turned to the rear-
13700
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 13698 – 13705

Cyclobutenone Rearrangements
FULL PAPER
the build-up of ROS.[20] As an approach to cancer chemo-
therapy, ROS-inducing therapies are still in their infancy.
Esclomol, for example, was recently advanced to phase III
clinical trials in combination with taxol, though these were
halted because of increased mortality.[20] To date, four total
syntheses of 19 have been reported.[6–9,20] The shortest, by
Knueppel and Martin, was reported in 2009 and featured an
alkynylcyclobutenone rearrangement as a key step, namely,
18!19.[9] Although this step gave a low yield (26%), it al-
lowed the total synthesis to be completed in just five linear
steps.
Our plan was to achieve a synthesis of cribrostatin 6 in a
similar step count, but with greater efficiency. To that end,
nitrile 14 was reduced to the corresponding amine 15 with
alane (AlH3) (Scheme 3).[21] Acylation to 16 was followed by
Figure 4. Calculated structures for the transition state 6 and intermediate
7.[16]
rangement of heteroarylcyclobutenones. With a myriad of
options available, we limited our study to representative
electron-rich (thiophene) and electron-poor (pyridine) sys-
tems, and an exemplification through total synthesis. The
thermolysis of (2-pyridyl)cyclobutenones provides rapid
access to quinolizidones,[2a] for example, 9!10, which are
used widely in medicinal chemistry as isosteres for naphtha-
lenes. In spite of this, the reaction has found little favour,
perhaps due to modest yields (29–60%) and a need to pro-
tect the alcohol moiety formed on addition of a 2-lithiopyri-
dine to a cyclobutendione. Under continuous-flow condi-
tions, the thermal rearrangement of pyridylcyclobutenone 9
in dioxane at 1008C gave quinolizidone 10 in quantitative
yield after a residence time of just 10 min without the need
for alcohol protection (Scheme 2). Although the related syn-
Scheme 3. A short total synthesis of cribrostatin 6 (19). EDC=1-ethyl-3-
(3-dimethylaminopropyl)carbodiimide.
cyclisation with POCl3 to give imidazopyridine 17 in a near-
quantitative yield. Halogen–lithium exchange then facilitat-
ed the union of 17 and 2-ethoxy-3-methylcyclobutendione,
to give a separable 3:1 mixture of adduct 20 and a regioiso-
mer 20a derived from addition to the vinylogous ester car-
bonyl. Finally, thermolysis of 20 in dioxane for 1 h at 1108C
under continuous flow, followed by exposure to air for a fur-
ther 45 min, gave 19 in 90% yield after purification by
column chromatography.
Conclusion
Scheme 2. Representative
heteroarylcyclobutenone
rearrangements
under continuous-flow conditions.[17, 18]
In conclusion, we have shown that aryl- and heteroarylcyclo-
butenone rearrangements can be conveniently performed
under continuous flow in dioxane at 1508C and proceed
with excellent yields. The approach has allowed us to deter-
mine a Hammett relationship for the reaction. This, in con-
junction with DFT and ab initio modelling,[11] provides
strong evidence that the electrocyclisation of ketene 5 to bi-
cyclic ketone 7 is rate determining in arylcyclobutenone re-
arrangements and has a late transition state. From a compu-
tational perspective, the excellent performance of
thesis of quino[b]thiophene 13 from thiophene 11 required
a higher temperature (1508C) and an aerial oxidation, it too
proceeded efficiently, to give a 98% yield over the two steps
(Scheme 2).[19]
To demonstrate the value of the method, we chose to
tackle the synthesis of the marine natural product cribrosta-
tin 6 (19), a popular target since it was identified by Pettit
et al. in 2003, that exhibits useful anti-neoplastic and anti-
microbial activity.[5,6] In 2010, cribrostatin 6 was reported to
induce death in cancer cells by inducing oxidative stress and
RCCSD(T)/6-31G(d)//UB3LYP/6-311G
ACHTUNGTREN(NUGN d,p) in predicting
reaction kinetics is notable. The short and efficient total syn-
Chem. Eur. J. 2011, 17, 13698 – 13705
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13701

D. C. Harrowven et al.
2,3-Dimethoxy-6-methylnaphthalene-1,4-dione (3b):[1a] Compound 3b
could be formed in 94% yield by using the general procedure with a resi-
dence time of 2 h and stirring the resulting solution in air for 3 h. M.p.:
89–918C (MeOH; previously reported: 90–928C (aq MeOH));[1a]
1H NMR (400 MHz, CDCl3): d=7.96 (d, J=7.8 Hz, 1H), 7.87 (brs, 1H),
7.53–7.47 (ddq, J=7.8, 1.5, 0.8 Hz, 1H), 4.12 (s, 3H), 4.10 (s, 3H),
2.48 ppm (brs, 3H); 13C NMR (100 MHz, CDCl3): d=182.3 (C), 181.8
(C), 147.6 (C), 144.9 (C), 134.4 (C), 130.8 (CH), 128.6 (C), 126.7 (CH),
126.4 (CH), 119.7 (C), 61.4 (2ꢁCH3), 21.8 ppm (CH3); IR (CHCl3): n˜ =
2950, 1659, 1611, 1600, 1306, 1269, 1040 cmꢀ1; MS (ES+): m/z (%): 482
[2M+NH4]+ (60), 233 [M+H]+ (100). See Scheme 4, Figure 6 and
Table 2.
thesis of cribrostatin 6 (19), a useful anti-neoplastic and
anti-microbial agent, provides a topical demonstration of
the value of the method for the rapid construction of con-
densed quinones.
Experimental Section
General procedure for continuous-flow reactions: Aliquots of 1a–i
(2 mL) in dioxane were taken from bulk solutions (0.25 g in 25 mL) and
heated at 1508C under continuous flow in stainless-steel tubing (internal
diameter 1 mm, capacity 10 mL) for the stated residence time by using a
Vapourtec R4/R2+ device. The resulting solutions were concentrated in
1
vacuo then analyzed by H NMR to determine the composition (1, 2 and
3) by comparison of the respective integrals as indicated in the Support-
ing Information.
2,3-Dimethoxynaphthalene-1,4-dione (3a):[1a, f] Compound 3a could be
formed in a near-quantitative yield by using the general procedure with a
residence time of 30 min and stirring the resulting solution in air for 1 h.
M.p.: 116–1188C (Et2O/petroleum ether; previously reported: 115–1178C
(Et2O));[1a,f] 1H NMR (300 MHz, CDCl3): d=7.71 (m, 2H), 8.07 (m, 2H),
4.13 ppm (s, 6H); 13C NMR (75 MHz, CDCl3): d=182.0 (2ꢁC), 147.5
(2ꢁC), 133.8 (2ꢁCH), 130.8 (2ꢁC), 126.3 (2ꢁCH), 61.5 ppm (2ꢁCH3);
IR (CHCl3): n˜ =3385, 2954, 1772, 1601, 1468, 1337, 1047, 994, 858 cmꢀ1
;
Scheme 4. Synthesis of 3b. CAN=ceric ammonium nitrate.
MS (ES+): m/z (%): 241 [M+Na]+ (46), 219 [M+H]+ (16). See Figure 5
and Table 1.
Figure 6. NMR spectra used to extract data presented in Table 2.
Table 2. Concentration of 1b after thermolysis in dioxane at 1508C for
the stated residence time, as determined by NMR integration; the reac-
tions were run in duplicate.
Figure 5. NMR spectra used to extract data presented in Table 1.
Time [s]
0
1b [%]
ln[SM]
100
100
4.61
4.61
4.41
4.36
4.28
4.21
4.14
4.04
3.84
3.78
Table 1. Concentration of 1a after thermolysis in dioxane at 1508C for
the stated residence time, as determined by NMR integration.
210
300
450
600
81.9
78.4
72.4
67.5
62.7
56.9
46.5
43.7
Time [s]
1a [%]
ln[SM]
0
100
4.61
4.15
3.98
3.68
3.22
2.84
300
450
600
900
1200
63.2
53.5
39.6
25.1
17.1
13702
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 13698 – 13705

Cyclobutenone Rearrangements
FULL PAPER
As an alternative to the general method, the crude reaction mixture
could be concentrated in vacuo, dissolved in CH2Cl2 then exposed to
23% CAN on silica. The procedure facilitates the quantitative oxidation
of the benzohydroquinone 2b to benzoquinone 3b with remaining start-
ing-material 1b converted to 3-methoxy-4-(4-methylphenyl)cyclobuten-
1,2-dione (21). 1H NMR (400 MHz, CDCl3): d=7.94 (d, J=8.1 Hz, 2H),
7.32 (d, J=8.1, Hz, 2H), 4.60 (3H, s), 2.44 ppm (s, 3H); 13C NMR
(100 MHz, CDCl3): d=193.8 (C), 184.7 (C), 174.9 (C), 143.9 (C), 129.9
(2ꢁCH), 127.9 (2ꢁCH), 125.0 (C), 116.3 (C), 61.6 (CH3), 22.0 ppm
(CH3); IR (CHCl3): n˜ =1784, 1595, 1369 cmꢀ1; MS (ES+): m/z (%): 266
[M+MeCN+Na]+ (100), 203 [M+H]+ (2).[22]
ture was allowed to warm to RT then extracted with CH2Cl2 (20 mLꢁ3).
The combined organic layers were washed with brine (20 mLꢁ2), dried
(MgSO4), filtered, concentrated in vacuo and purified by flash column
chromatography (5%!40% EtOAc/petroleum ether with 2% NEt3) to
afford the title-compound 11 as a white solid (0.257 g, 1.14 mmol, 32%).
M.p.: 66–688C (Et2O/petroleum ether; previously reported: 68–698C
(Et2O));[1a] 1H NMR (400 MHz, CDCl3): d=7.33 (dd, J=5.0, 1.3 Hz,
1H), 7.11 (dd, J=3.5, 1.3 Hz, 1H), 7.02 (dd, J=5.0, 3.5 Hz, 1H), 4.11 (s,
3H), 4.02 (s, 3H), 3.56 ppm (s, 1H); 13C NMR (100 MHz, CDCl3): d=
183.1 (C), 166.0 (C), 140.7 (C), 135.2 (C), 127.2 (CH), 126.2 (CH), 125.0
(CH), 85.7 (C), 60.3 (CH3), 58.8 ppm (CH3); IR (CHCl3): n˜ =3400, 3008,
2956, 1781, 1644, 1635, 1470, 1348, 1040 cmꢀ1; MS (ES+): m/z (%): 227
[M+H]+ (100).
4-Hydroxy-2,3-dimethoxy-4-(pyridin-2-yl)cyclobut-2-enone (9): nBuLi
(3.52 mL, 2.40m solution in hexane, 8.45 mmol) was added to a solution
of 2-bromopyridine (0.82 mL, 8.45 mmol) in THF (30 mL) at ꢀ788C over
5 min. After 15 min the resulting solution was added, through a cannula,
to a solution of dimethyl squarate (1.00 g, 7.04 mmol) in THF (10 mL) at
ꢀ788C, giving a red solution. After 30 min, saturated NH4Cl (20 mL) was
added. The reaction mixture was allowed to warm to RT then extracted
with CH2Cl2 (50 mLꢁ3). The combined organic layers were washed with
brine (50 mLꢁ2), dried (MgSO4), filtered, concentrated in vacuo and pu-
rified by flash column chromatography (5%!50% EtOAc/petroleum
ether with 2% NEt3) to afford the title-compound 9 as a pale brown oil
(0.97 g, 4.37 mmol, 62%). 1H NMR (400 MHz, CDCl3): d=8.62 (d+fine
splitting, J=4.5 Hz, 1H), 7.77 (td, J=7.5, 1.5 Hz, 1H), 7.48 (d, J=7.5 Hz,
1H), 7.30 (ddd, J=7.5, 4.5, 1.0 Hz, 1H), 6.07 (s, 1H), 4.09 (s, 3H),
3.98 ppm (s, 3H); 13C NMR (75 MHz, CDCl3): d=182.6 (C), 164.4 (C),
154.6 (CH), 148.3 (C), 137.6 (C), 137.4 (CH), 123.2 (CH), 120.0 (CH),
80.0 (C), 59.9 (CH3), 58.8 ppm (CH3); IR (CHCl3): n˜ =3463, 2951, 1776,
1632, 1468, 1337, 1061 cmꢀ1; MS (ES+): m/z (%): 222 [M+H]+ (100);
HRMS (ES+): m/z: calcd for C11H12NO4: 222.0775 [M+H]+; found:
222.0766.
5,6-Dimethoxybenzo[b]thiophene-4,7-dione (13):[1a]
A solution of 11
(5.0 mg, 0.022 mmol) in dioxane (2 mL) was heated at 1508C in stainless-
steel tubing for a residence time of 30 min by using a Vapourtec R4/R2+
device. The resulting solution was stirred in air for 1 h then concentrated
in vacuo to give the title-compound 13 as an orange solid (4.9 mg,
0.022 mmol, 98%). M.p.: 169–1728C (Et2O/petroleum ether; previously
reported:
171.5–1738C
(CH2Cl2/petroleum
ether));[1a]
1H NMR
(400 MHz, CDCl3): d=7.63 (d, J=5.0 Hz, 1H), 7.50 (d, J=5.0 Hz, 1H),
4.09 ppm (apparent s, 6H); 13C NMR (100 MHz, CDCl3): d=178.0 (C),
176.6 (C), 147.2 (C), 146.7 (C), 141.4 (C), 139.5 (C), 133.4 (CH), 125.9
(CH), 61.5 (CH3), 61.6 ppm (CH3); IR (CHCl3): n˜ =3033, 3008, 2929,
1651, 1340, 1292, 858 cmꢀ1; MS (ES+): m/z (%): 225 [M+H]+ (100).
(3-Bromopyridin-2-yl)methanamine (15):[21] To a cooled (08C) solution of
3-bromo-2-cyanopyridine (0.100 g, 0.546 mmol) in toluene (20 mL) was
added alane·Me2NEt complex (0.5m solution in toluene, 2.2 mL,
1.093 mmol) over 4 min. The resulting mixture was warmed to RT and
then, after 16 h, was re-cooled to 08C. Methanol (10 mL) and saturated
sodium potassium tartare (50 mL) were cautiously added, then the aque-
ous phase was separated and extracted with CHCl3 (20 mLꢁ3). The com-
bined organic phases were then washed with brine (20 mLꢁ2), dried
(MgSO4), filtered and concentrated in vacuo to afford the title-compound
15 as an yellow oil (60.5 mg, mmol, 60%). 1H NMR (300 MHz, CDCl3):
d=8.52 (dd, J=5.0, 1.5 Hz, 1H), 7.85 (dd, J=7.9, 1.5 Hz, 1H), 7.11 (dd,
J=7.9, 5.0 Hz, 1H), 4.14 (s, 2H), 2.77 ppm (s, 2H); 13C NMR (75 MHz,
CDCl3): d=162.4 (C), 150.2 (CH), 145.2 (CH), 123.1 (C), 115.9 (CH),
1-Hydroxy-2,3-dimethoxy-4H-quinolizin-4-one (10):
A solution of 9
(5.0 mg, 0.023 mmol) in dioxane (2 mL) was heated at 1008C in stainless-
steel tubing for a residence time of 10 min with a Vapourtec R4/R2+
device. The resulting solution was stirred in air for 1 h and then concen-
trated in vacuo to give 10 as a pale brown solid (4.9 mg, 0.022 mmol,
99%). M.p.: 110–1128C (EtOAc/petroleum ether); 1H NMR (400 MHz,
[D6]DMSO): d=8.88 (s, 1H), 8.70 (d, J=7.6 Hz, 1H), 7.73 (d, J=9.0 Hz,
1H), 7.16 (dd, J=8.0, 6.6 Hz, 1H), 6.86–6.97 (m, 1H), 4.03 (s, 3H),
3.87 ppm (s, 3H); 13C NMR (100 MHz, [D6]DMSO): d=152.1 (C), 150.2
(C), 131.7 (C), 127.2 (C), 126.7 (C), 125.4 (CH), 125.2 (CH), 120.0 (CH),
114.0 (CH), 60.9 (CH3), 59.3 ppm (CH3); IR (CHCl3): n˜ =2955, 2925,
1733, 1634, 1595, 1457, 1288, 1122 cmꢀ1; MS (ES+): m/z (%): 222 [MH]+
(100); HRMS (ES+): m/z: calcd for C11H12NO4: 222.0775 [M+H]+;
found: 222.0766.
45.2 ppm (CH2); MS (ES+): m/z (%): 189 [M
[M
(79Br)+H]+ (100).
N-((3-Bromopyridin-2-yl)methyl)acetamide
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
(16):[21]
(0.850 mL, 14.83 mmol) and triethylamine (4.13 mL, 29.7 mmol) were
added sequentially to a solution of N-(3-dimethylaminopropyl)-N’-ethyl-
carbodiimide hydrochloride (2.84 g, 14.83 mmol) in dichloromethane
(30 mL) at RT. After 1 h, a solution of 15 (1.422 g, 5.93 mmol) in DMF
(15 mL) and CH2Cl2 (10 mL) was added, followed by 2m sodium carbon-
ate (30 mL) after an additional 1 h. The aqueous phase was separated
and extracted with dichloromethane (3ꢁ20 mL) and then the organic
phases were dried (MgSO4), concentrated in vacuo and purified by
column chromatography (EtOAc) to give the title-compound 16 as a
white solid (0.730 g, 3.187 mmol, 52%). M.p.: 62–648C (EtOAc);
1H NMR (400 MHz, CDCl3): d=8.50 (d, J=4.4, 1.3 Hz, 1H), 7.89 (d, J=
7.9, 1.3 Hz, 1H), 7.18 (brs, 1H), 7.16 (dd, J=7.9, 4.4 Hz, 1H), 4.62 (d,
J=4.3, 2H), 2.13 ppm (s, 3H); 13C NMR (100 MHz, CDCl3): d=170.0
(C), 153.7 (C), 146.9 (CH), 140.4 (CH), 123.5 (C), 120.3 (CH), 44.1
(CH2), 23.3 ppm (CH3); IR (CHCl3): n˜ =3314, 1642, 1560, 812 cmꢀ1; MS
(ES+): m/z (%): 231 [M
8-Bromo-3-methylimidazoAHCTNUGTRENNNUG
N
(79Br)+H]+ (100).
ACHTUNGTRENNUNG
(1.07 mL, 11.5 mmol) was added to a solution of acetamide 16 (0.730 g,
3.19 mmol) in toluene (10 mL) at RT over 5 min. The reaction mixture
was heated at reflux for 4 h then cooled to 08C and saturated sodium bi-
carbonate (30 mL) was added. The aqueous phase was separated and ex-
tracted with ethyl acetate (10 mLꢁ3), then the combined organic phases
were washed with water (10 mLꢁ2), dried (MgSO4) and concentrated to
give the title-compound 17 as a brown oil (0.680 g, 3.222 mmol, 99%, ca.
98% purity). 1H NMR (400 MHz, CDCl3): d=7.68 (d, J=7.3 Hz, 1H),
7.46 (s, 1H), 6.92 (d, J=6.8 Hz, 1H), 6.47 (apparent t, J=7.1 Hz, 1H),
2.69 ppm (s, 3H); 13C NMR (100 MHz, CDCl3): d=179.0 (C), 136.6 (C),
4-Hydroxy-2,3-dimethoxy-4-(thiophen-2-yl)cyclobut-2-en-1-one (11):[1a] 2-
Bromothiopene (0.376 mL, 3.87 mmol) was added to a solution of nBuLi
(2.42 mL, 1.6m solution in hexane, 3.87 mmol) in THF (15 mL) over
5 min at ꢀ788C. After 15 min, a solution of dimethyl squarate (0.50 g,
3.52 mmol) in THF (10 mL) was added over 5 min, giving an orange solu-
tion. After 1 h, saturated NH4Cl (20 mL) was added. The reaction mix-
Chem. Eur. J. 2011, 17, 13698 – 13705
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13703

D. C. Harrowven et al.
130.0 (C), 120.8 (CH), 119.8 (CH), 119.5 (CH), 112.7 (CH), 12.6 ppm
[3] For an excellent review see a) R. L. Danheiser, G. B. Dudley, W. F.
Austin in Science in Synthesis, Vol. 23, (Ed.: R. Danheiser), Thieme
Chemistry, Stuttgart, 2006, pp. 493–568; see also b) D. C. Harrowv-
en, D. D. Pascoe, D. Demurtas, H. O. Bourne, Angew. Chem. 2005,
117, 1247–1248; Angew. Chem. Int. Ed. 2005, 44, 1221–1222;
c) D. C. Harrowven, D. D. Pascoe, I. L. Guy, Angew. Chem. 2007,
119, 429–432; Angew. Chem. Int. Ed. 2007, 46, 425–428.
[4] a) L. P. Hammett, Physical Organic Chemistry, 2nd ed., McGraw-
Hill, New York, 1970; b) O. Exner, Correlation Analysis of Chemical
Data, Plenum Press, New York, 1988.
(CH3); IR (CHCl3): n˜ =3314, 1642, 1560, 812 cmꢀ1; MS (ES+): m/z (%):
213 [M
3-Ethoxy-4-hydroxy-2-methyl-4-(3-methylimidazo
clobut-2-enone (20): nBuLi (1.6m in hexane, 0.79 mL, 1.26 mmol) was
added to a solution of imidazo[1,5-a]pyridine 17 (242 mg, 1.15 mmol) in
ACHTUNGTRENNUNG ACHTUNGTRENNUGN
(81Br)+H]+ (100), 211 [M(79Br)+H]+ (100).
AHCTUNGERTG[NNUN 1,5-a]pyridin-8-yl)cy-
AHCTUNGTRENNUNG
THF (5 mL) at ꢀ788C. After 30 min a solution of 3-ethoxy-4-methyl-3-
cyclobutene-1,2-dione (0.161 g, 1.147 mmol)[1,9] in THF (5 mL) was added
over 4 min, followed by saturated NH4Cl (20 mL) after an additional 1 h.
After warming to RT the aqueous phase was separated and extracted
with dichloromethane (20 mLꢁ3). The combined organic phases were
washed with brine (20 mLꢁ2), dried (MgSO4), concentrated in vacuo
and purified by flash column chromatography (0%!5% methanol/di-
chloromethane with 1% NEt3) gave firstly 20a (59 mg, 0.204 mmol,
[5] G. R. Pettit, J. C. Collins, J. C. Knight, D. L. Herald, R. A. Nieman,
M. D. Williams, R. K. Pettit, J. Nat. Prod. 2003, 66, 544–547; R. K.
Pettit, B. R. Fakoury, J. C. Knight, C. A. Weber, G. R. Pettit, G. D.
Cage, S. Pon, J. Med. Microbiol. 2004, 53, 61–65.
[6] S. Nakahara, A. Kubo, Heterocycles 2004, 63, 2355–2362.
[7] S. Nakahara, A. Kubo, Y. Mikami, J. Ito, Heterocycles 2006, 68, 515–
520.
[8] M. D. Markey, T. R. Kelly, J. Org. Chem. 2008, 73, 7441–7443.
[9] D. Knueppel, S. F. Martin, Angew. Chem. 2009, 121, 2607–2609;
Angew. Chem. Int. Ed. 2009, 48, 2569–2571.
[10] The flow system employed is shown in Figure 2 of S. V. Ley, I. R.
Baxendale, Chimia 2008, 62, 162–168. Further details may be found
on the manufacturerꢂs web site http://www.vapourtec.co.uk/products/
rseriessystem accessed 21/7/2011.
[11] a) A. Odedra, P. H. Seebergera, Angew. Chem. 2009, 121, 2737–
2740; Angew. Chem. Int. Ed. 2009, 48, 2699–2702; b) F. E. Valera,
M. Quaranta, A. Moran, J. Blacker, A. Armstrong, J. Cabral, D. G.
Blackmond, Angew. Chem. 2010, 122, 2530–2537; Angew. Chem.
Int. Ed. 2010, 49, 2478–2485.
18%) then the title-compound 20 as
a pale orange oil (175 mg,
0.642 mmol, 56%). 1H NMR (400 MHz, CDCl3): d=7.64 (d, J=7.0 Hz,
1H), 7.39 (s, 1H), 6.97 (d, J=6.8 Hz, 1H), 6.60 (t, J=6.9 Hz, 1H), 4.45
(dq, J=9.8, 7.1 Hz, 1H), 4.26 (dq, J=9.9, 7.1 Hz, 1H), 2.66 (s, 3H), 1.85
(s, 3H), 1.36 ppm (t, J=7.1 Hz, 3H); 13C NMR (100 MHz, CDCl3): d=
189.5 (C), 182.2 (C), 135.5 (C), 127.8 (C), 127.6 (C), 125.2 (C), 120.7
(CH), 118.5 (CH), 116.5 (CH), 112.0 (CH), 91.6 (C), 69.2 (CH2), 15.0
(CH3), 12.6 (CH3), 6.9 ppm (CH3); IR (CHCl3): n˜ =2928, 2861, 1715,
1731, 1617, 1332, 1135, 1078 cmꢀ1; MS (ES+): m/z (%): 273 [M+H]+
(100); HRMS (ES+): m/z: calcd for C15H17N2O3: 273.1161 [M+H]+;
found: 273.1232.
9-Ethoxy-3,8-dimethylimidazoACHTNUTRGNEUNG[5,1-a]isoquinoline-7,10-dione (19, Cribros-
tatin 6): Cyclobutenone 20 (59 mg, 0.217 mmol) in dioxane (2 mL) was
heated at 1108C in stainless-steel tubing for a residence time of 1 h by
using a Vapourtec R4/R2+ device. The resulting solution was stirred in
air for 30 min then concentrated in vacuo and purified by chromatogra-
phy (2% MeOH in CH2Cl2) to give 19 as a light-blue solid (52 mg,
0.193 mmol, 90%). M.p.: 167–1698C (acetone at ꢀ48C; previously re-
[12] a) V. P. Andreev, Chem. Heterocycl. Compd. 2010, 46, 184–195;
b) E. Kutter, C. Hansch, J. Med. Chem. 1969, 12, 647–651; c) T.
Fujita, C. Takayama, M. Nakajima, J. Org. Chem. 1973, 38, 1623–
1630; d) G. Ananchenko, E. Beaudoin, D. Bertin, D. Gigmes, P. La-
garde, S. R. A. Marque, E. Revalor. P. Tordo, J. Phys. Org. Chem.
2006, 19, 269–275.
ported: 165–1678C (acetone)[8,9]
, , 171–1728C
169–1718C (acetone)[5]
(acetone)[6,7]); 1H NMR (400 MHz, CDCl3): d=8.52 (s, 1H); 8.10 (d, J=
7.6 Hz, 1H), 7.62 (d, J=7.6 Hz, 1H), 4.49 (q, J=6.8 Hz, 2H), 3.09 (brs,
3H), 2.13 (s, 3H), 1.45 ppm (t, J=7 Hz, 3H); 13C NMR (100 MHz,
CDCl3): d=184.9 (C), 180.7 (C), 156.2 (C), 137. 7 (C), 130.1 (C), 125.9
(C), 125.0 (C), 124.7 (C), 123.9 (CH), 123.5 (CH), 107.6 (C), 69.6 (CH2),
16.0 (CH3), 12.6 (CH3), 9.2 ppm (CH3); IR (CHCl3): n˜ =2925, 1662, 1626,
1611, 1527, 1172 cmꢀ1; MS (ES+): m/z (%): 271 [M+H]+ (100).
[13] As the value of sI for CONiPr2 is not known, we used as an esti-
mate, namely, the known value for CONH2.
[14] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.
Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X.
Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J.
Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, N. J. Millam, M. Klene, J. E. Knox,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A.
Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, ꢃ.
Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian,
Inc., Wallingford CT, 2009.
CCDC-808547 contains the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Acknowledgements
We gratefully acknowledge GSK, ERDF (IS:CE-Chem & InterReg IVa
program 4061) and the EPSRC for funding, and for the support of the
Iridis 3 cluster. Rob Wheeler and Dr. Andy Craven are thanked for val-
uable advice and Dr. M. E. Light for the X-ray analysis.
[15] a) P. W. Musch, C. Remenyi, H. Helten, B. Engels, J. Am. Chem.
Soc. 2002, 124, 1823–1828; b) P. R. Schreiner, B. H. Bui, Eur. J. Org.
Chem. 2006, 1162–1165; c) P. W. Musch, B. Engels, J. Am. Chem.
Soc. 2001, 123, 5557–5562.
[1] a) H. W. Moore, S. T. Perri, J. Org. Chem. 1988, 53, 996–1003;
b) S. T. Perri, H. J. Dyke, H. W. Moore, J. Org. Chem. 1989, 54,
2032–2034; c) A. Enhsen, K. Karabelas, J. M. Heerding, H. W.
Moore, J. Org. Chem. 1990, 55, 1177–1185; d) J. M. Heerding, H. W.
Moore, J. Org. Chem. 1991, 56, 4048–4050; e) M. P. Winters, M.
Stranberg, H. W. Moore, J. Org. Chem. 1994, 59, 7572–7574;
f) H. W. Moore, O. H. W. Decker, Chem. Rev. 1986, 86, 821–830;
g) M. W. Reed, D. J. Pollart, S. T. Perri, L. D. Foland, H. W. Moore,
J. Org. Chem. 1988, 53, 2477–2482.
[2] a) A. G. Birchler, F. Liu, L. S. Liebeskind, J. Org. Chem. 1994, 59,
7737–7745; b) D. Zhang, I. Llorente, L. S. Liebeskind, J. Org.
Chem. 1997, 62, 4330–4338; c) S. Zhang, L. S. Liebeskind, J. Org.
Chem. 1999, 64, 4042–4049.
[16] All calculations were performed at the B3LYP/6–311GACHTUNGTRENNUNG(d,p) level
with Gaussian 09,[14] with GAMESS(US)[17] used for all RCCSD(T)/
6–31G(d) calculations.[18] For each optimized structure an analytical
Hessian was calculated to obtain vibrational frequencies, zero-point
energies and thermodynamic corrections at 423 K, additionally con-
firming each as a minimum or transition-state structure. Post-pro-
AHCTUNGTREGcNNUN essing visualization was carried out with the ChemCraft (G. A.
Zhurko, ChemCraft version 1.6; http://www.chemcraftprog.com) and
Gabedit programs (A. R. Allouche, J. Comput. Chem 2011, 32, 174–
182).
[17] Programme used: GAMESS(US), ver. 1-Oct-2010 (R1): M. W.
Schmidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert, M. S. Gordon,
13704
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 13698 – 13705

Cyclobutenone Rearrangements
FULL PAPER
J. H. Jensen, S. Koseki, N. Matsunaga, K. A. Nguyen, S. Su, T. L.
Windus, M. Dupuis, J. A. Montgomery, J. Comput. Chem. 1993, 14,
1347–1363.
[21] This reduction is reported to proceed in >80% yield with the bora-
ne·THF complex (a) Z. Jiajie Chinese Pat. CN101239978(A); b) S.
Ramsbeck, U. Heiser, M. Buchholz, A. J. Niestroj, US Pat. US2008/
234313 A1, 2008). In our hands, isolation of amine 15 from that re-
action proved difficult, prompting a switch to alane.
[22] A. H. Schmidt, G. Kircher, S. Maus, H. Bach, J. Org. Chem. 1996,
61, 2085–2094.
[18] P. Piecuch, S. A. Kucharski, K. Kowalski, M. Musial, Comp. Phys.
Commun. 2002, 149, 71–96.
[19] S. T. Perri, L. D. Foland, O. H. Decker, H. W. Moore, J. Org. Chem.
1986, 51, 3067–3068.
[20] M. T. Hoyt, R. Palchaudhuri, P. J. Hergenrother, Invest. New Drugs
2010, DOI: 10.1007 s10637–010–9390-x, accessed 1/12/2010.
Received: July 22, 2011
Published online: November 14, 2011
Chem. Eur. J. 2011, 17, 13698 – 13705
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13705