Synthesis of the Integrastatin Nucleus Using the Ramberg - Baecklund Reaction

Article abstract of DOI:10.1021/ol035786v

(Equation presented) The first synthesis of the tetracyclic nucleus of the Integrastatins, natural products that have been shown to selectively inhibit HIV-1 integrase, is reported. Key steps of this synthesis involve a novel cis-selective Ramberg-Baecklund reaction and an unusual Lewis acid-promoted cyclization step.

Full text of DOI:10.1021/ol035786v

ORGANIC
LETTERS
2003
Vol. 5, No. 23
4441-4444
Synthesis of the Integrastatin Nucleus
Using the Ramberg−Ba1cklund Reaction
,†
Jonathan S. Foot,† Gerard M. P. Giblin,‡ and Richard J. K. Taylor*
Department of Chemistry, UniVersity of York, Heslington, York, YO10 5DD, UK, and
Department of Medicinal Chemistry, Neurology and GI CEDD, GlaxoSmithKline,
The Frythe, Welwyn, Herts AL6 9AR, UK
rjkt1@york.ac.uk
Received September 16, 2003
ABSTRACT
The first synthesis of the tetracyclic nucleus of the Integrastatins, natural products that have been shown to selectively inhibit HIV-1 integrase,
is reported. Key steps of this synthesis involve a novel cis-selective Ramberg−Ba1cklund reaction and an unusual Lewis acid-promoted cyclization
step.
Integrastatin A (1a) and B (1b) (Scheme 1) are two recently
discovered natural products isolated from both an unnamed
contain two chiral centers, they exist in nature in racemic
form ((R,R)-form shown).
As part of our ongoing interest in the synthesis of highly
oxygenated natural products,2 it was decided to investigate
the total synthesis of both Integrastatins. Given the novel
structure of the tetracyclic nucleus 2, we first explored
synthetic approaches to this simplified analogue by func-
tionalization of alkene 3, obtained by retrosynthetic analysis
(Scheme 1).
Scheme 1. Integrastatins A & B and the Integrastatin
Tetracyclic Nucleus, 2
Failure of Wittig, Grignard, Julia, and lithiation chemistries
to produce 3 led us to the Ramberg-Ba¨cklund reaction
(RBR).3
To this end, the two coupling partners 6 and 7 (Scheme
2) were synthesized from the commercially available 2-me-
thylacetophenone (4) and 2-hydroxyacetophenone (5). Cou-
pling and oxidation of the resultant thioether proceeded
smoothly to afford the sulfone 9 in 52% yield overall.
Sulfone 9 was then subjected to the in situ chlorination-
Ramberg-Ba¨cklund reaction using the conditions described
fungal source (ATCC74478) and from an endophytic Asco-
chtya species (ATCC74477), which have been found to
selectively inhibit the strand-transfer reaction of recombinant
HIV-1 integrase at micromolar concentrations.1 They are
based on a novel [6.6.6.6] tetracycle, and although they
(1) Singh, S. B.; Fink. D. L.; Quamina, D. S.; Pelaez, F.; Teran, A.;
Felock, P.; Hazuda, D. J. Tetrahedron Lett. 2002, 43, 2351.
(2) (a) Harvey, J. E.; Raw, S. A.; Taylor, R. J. K. Tetrahedron Lett.
2003, 44, 7209. (b) Lewis, A.; Stefanuti, I.; Swain, S. A.; Smith, S. A.;
Taylor, R. J. K. Org. Biomol. Chem. 2003, 1, 104. (c) Runcie, K. A.; Taylor,
R. J. K. Org. Lett. 2001, 3, 3237. (d) Ragot, J. P.; Taylor, R. J. K. Org.
Lett. 2000, 2, 1981. (e) Alcevaz, L.; MacDonald, G.; Ragot, J.; Lewis, N.
J.; Taylor, R. J. K. Tetrahedron 1999, 55, 3707 and references therein.
(3) (a) Ramberg, L.; Ba¨cklund, B. ArkiV Kemi. Mineral. Geol. 1940, 27,
13A. (b) Casy, G.; Taylor, R. J. K. Org. React. 2003, 62, 357.
† University of York.
‡ GlaxoSmithKline.
10.1021/ol035786v CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/11/2003

Scheme 2. Formation of the RBR Precursor
Scheme 3. Ramberg-Ba¨cklund Experiments
by Meyers et al. (Scheme 3),4 to give the corresponding
Attempting to transform 10 into 13 through Lewis acid-
promoted ketal removal,7 we discovered that the tetracycle
17 was formed as a minor byproduct, along with the desired
olefin 13 as a 2:1 mixture of E:Z isomers (Scheme 5).
olefin 10 in 83% yield as a 1:1 E:Z mixture.5 It was found,
however, that both the free acetophenone 11 and the
corresponding benzyl alcohol 12 also underwent the RBR,
under the same conditions, to afford predominantly the cis-
isomers of 13 (E:Z ) 1:8, 50%) and 14 (E:Z ) 1:16, 89%).
In the literature, methyl stilbenes are normally formed with
the trans-isomer predominating.4,6 We are currently carrying
out further studies to rationalize these unexpected results.
We next intended to dihydroxylate the newly formed
double bond, by using osmium, ruthenium, or permanganate
salts, but none gave the desired diol 15 (Scheme 4). Instead
we observed either oxidatively cleaved products or the enol
16, a result that led us to discontinue this approach.
Scheme 5. Formation of the Deoxy-nucleus, 17
Scheme 4. Attempted Dihydroxylation Experiments
As analysis of these results indicated that the (Z)-isomer
underwent preferential cyclization, we then subjected a
(4) Meyers, C. Y.; Malte, A. M.; Matthews, W. S. J. Am. Chem. Soc.
1969, 91, 7510.
4442
Org. Lett., Vol. 5, No. 23, 2003

Scheme 6. Bromide and Displacement Trials: Formation of the Integrastatin nucleus
sample of 13 obtained through RBR of 12 (E:Z ) 1:16) to
appears to occur in the case of the cis-isomer, we deduce
that the double bond is left intact until quite late in the
mechanistic pathway. Further experiments are underway to
evaluate this hypothesis.
Initial attempts at the benzylic oxidation of 17 (SeO2, IBX,
PCC, PDC, etc) were unsuccessful. Radical bromination,
which gave 18, was successful (see Scheme 6) but subsequent
oxygen displacement reactions also failed. In the case of
SN2-type conditions, no reaction at all was observed, while
with SN1 reactions (using silver salts), the molecule under-
went rapid intramolecular rearrangement to afford the novel
benzofuran product 19.
excess tin(II) chloride dihydrate, affording the tetracycle 17
in an excellent 94% yield (Scheme 5) as colorless crystals.
The structure of 17 was confirmed by X-ray crystallography
(Figure 1).8
It was deduced that the bromide 18 was too hindered for
displacement reactions. We therefore sought a direct radical-
based oxidation process. Numerous reagents were investi-
gated, but success was achieved using TBHP and PDC
supported on Celite (Scheme 6).10,11 We discovered that it
was crucial to keep the temperature below 10 °C, and this
produced a slow conversion. However, using this procedure,
we observed the racemic integrastatin nucleus 2 in a 77%
Table 1. Comparison of Reported (Integrastatin A)1 and
Observed 13C NMR Shifts
Figure 1. ORTEP drawing of tetracycle 17 (50% probability
thermal ellipsoids).
We believe that the mechanism involves the initial loss
of the benzyl protection group (as occasionally seen with
Lewis acids9), followed by internal hemiacetal formation and
cyclization onto the double bond. Since the reaction only
(5) Representative RBR Procedure. To a stirred solution of sulfone in
t
CCl4 (5 mL/mmol), BuOH (5 mL/mmol), and water (1 mL/mmol) was
carbon
literature1 δC (1a )/ppma
observed δC (2)/ppma
added powdered KOH (20 equiv standard, 40 equiv if free alcohol was
present). The reaction was then heated at 80 °C until complete consumption
of sulfone was observed by TLC analysis. The solvent was removed in
vacuo and the residue extracted with EtOAc, washing with water and
saturated NaCl solution, before drying over magnesium sulfate. Filtration
and removal of solvent in vacuo, followed by purification (flash chroma-
tography, eluting in petroleum ether/EtOAc) afforded the desired olefin.
(6) Chan, T.-L.; Fong, S.; Li, Y.; Man, T.-O.; Poon, C. D. J. Chem.
Soc., Chem. Commun. 1994, 1771.
2
9
10
18
19
97.8
193.7
77.1
26.5
25.8
96.3
192.9
77.1
25.9
19.6
a Carried out in 1:1 CDCl3/CD3CN, 100 MHz.
(7) Ford, K. L.; Roskamp, E. J. Tetrahedron Lett. 1992, 33, 1135.
(8) CCDC no. 219622.
Org. Lett., Vol. 5, No. 23, 2003
4443

yield (based on recovered starting material; actual yield 41%,
with 47% unreacted starting tetracycle).
Full product analysis was in accordance with the proposed
structure, while comparison of the carbon-13 NMR shifts
(Table 1) provided convincing evidence that 2 is indeed the
Integrastatin core.
In summary, we have successfully synthesized the racemic
core of the Integrastatins in 11 steps, with an overall yield
of 33%, through application of a cis-selective Ramberg-
Ba¨cklund reaction and a novel Lewis acid-promoted cycliza-
tion. We are currently investigating the extension of our route
to the synthesis of the Integrastatin natural products.
(9) (a) Kadam, S. M.; Nayak, S. K.; Banerji, A. Tetraedron Lett. 1992,
33, 5129. (b) Park. M. H.; Takeda, R.; Nakanishi, K. Tetrahedron Lett.
1987, 28, 3823.
Acknowledgment. We thank Dr. A. C. Whitwood for
his help with X-ray crystallography and the E.P.S.R.C. and
GSK for their financial support (J.S.F).
(10) Chidambaram, N.; Chandrasekaran, S. J. Org. Chem. 1987, 52, 5048
(11) Procedure for Oxidation of 17. To a stirred solution of 17 (0.03
g, 0.12 mmol), PDC (0.26 g, 0.7 mmol), and Celite (0.2 g) in benzene (3.5
mL) at 6-10 °C under nitrogen was added TBHP (5.5 M in decane, 0.13
mL, 0.7 mmol). The reaction mixture was stirred below 10 °C for 4 days,
with further addition of TBHP (2 × 0.13 mL) after 36 and 72 h. Filtration
of the reaction through a pad of Celite, washing with EtOAc, and removal
of solvent in vacuo, followed by purification (flash chromatography, eluting
in 15:1 petroleum ether/EtOAc), afforded the integrastatin nucleus, 2 (0.013
g, 41%), and unreacted 17 (0.014 g, 47% recovered).
Supporting Information Available: Experimental pro-
cedures and full data for 14, 17, and 2. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL035786V
4444
Org. Lett., Vol. 5, No. 23, 2003