.
Angewandte
Communications
considering the ability of p-acid catalysts to activate alkynes,
thus facilitating the intramolecular nucleophilic addition of
alcohols and aldehydes,[7] we hypothesized that ideally
a unique metal complex could promote the cycloisomeriza-
tion of alkyndiol 2 to give the enol ether 5 and also the
cycloisomerization of aldehyde 3 to give 6 (Scheme 1b). The
formal cycloaddition reaction between the intermediates 5
and 6 would result in the formation of the core structure of
berkelic acid 4 in an apparently very simple way.
Although the proposed reaction was risky in terms of
stereoselectivity, because just one chiral center (that of
alkyndiol 2) would induce the selective formation of the
other three newly formed ones in 4 (including a desymmet-
rization in 2), the process was very attractive from the
synthetic point of view as it avoided the synthesis of starting
materials having multiple stereocenters and tedious protec-
tion/deprotection steps. Once the compound 4 was obtained
we were confident about the possibilities of completion the
total synthesis of (À)-berkelic acid (1), because only the
installation of the lateral chain and a few conventional
functional group manipulations would remain.
Preparation of the key building blocks, 2 and 3, was
achieved in a concise and scalable way from commercially
available starting materials. Specifically, chiral fragment 2 was
synthesized from (S)-(À)-3-butyn-2-ol (7; 98% ee) in just
three conventional chemical transformations (Scheme 2).
Thus, initial mesylation of alcohol 7 led to the corresponding
derivative 8. Displacement of the mesylate by diethylmalo-
nate in the presence of cesium fluoride through an SN2
reaction furnished the diester 9. Finally, reduction with
lithium aluminium hydride provided the desired diol 2 in
70% overall yield and 96% ee (2.8 g scale).[8] Synthesis of
building block 3 was also achieved in three steps (Scheme 2).
Namely, triflic anhydride reacted selectively with the hydroxy
group in para-position of commercially available ester methyl
2,4,6-trihydroxybenzoate (10) to provide compound 11.[3b]
Reaction of this intermediate through a Sonogashira-type
cross-coupling reaction employing the potassium trifluorobo-
rate salt derived from 1-heptyne was successful, thus allowing
the synthesis of alkyne 12 in high yield.[9] Finally, introduction
of the aldehyde functionality was achieved using a sequence
involving an initial hydroxymethylation with formaldehyde in
the presence of calcium chloride and subsequent oxidation
with manganese dioxide. Every reaction was performed on
a gram scale, thus providing the fragment 3 in 60% overall
yield from 10 (4.8 g scale).
At this stage, with substantial quantities of fragments 2
and 3 in hand, we turned our attention to the key reaction of
our synthesis, that is, the catalytic cycloisomerization of these
compounds and subsequent formal cycloaddition reaction
between the in situ formed intermediates 5 and 6 to give
product 4 (Scheme 1). Our experience in cycloisomerization
reactions indicated that our desired transformation could be
catalyzed by a metallic complex derived from typical carbo-
philic Lewis acids.[10] In fact, we found that the reaction
proceeded in the presence of 5 mol% of AgOTf. To avoid
possible decomposition of compound 4 as a result of the
presence of the new reactive pyran ring, we directly per-
formed the hydrogenation of the carbon–carbon double bond
of this pyran ring under conventional conditions (Scheme 3).
Scheme 3. Synthesis of the central core of 1. Reagents and conditions:
a) AgOTf (5 mol%), THF, 08C and then H2, Pd/C (5 mol%), MeOH,
À58C. THF=tetrahydrofuran.
It should be remarked that in this global transformation four
new chiral centers are formed and, surprisingly, only two
diastereoisomers were observed in the crude reaction mixture
(d.r. = 2:1). Even more pleasant was the confirmation that the
structure of the major diastereoisomer corresponded to that
of the desired product 13 with all chiral centers and
functionalities resembling those of the natural product. The
minor isomer could be easily separated at an advanced stage,
as it will be shown. Notably, this reaction could be performed
on a gram scale with 83% yield (2.4 g of this material were
synthesized). This new silver-catalyzed reaction allowed the
assembly of the central core of the natural product, which
contains four rings and five stereocenters, in just one step.
At this point we had achieved a formal synthesis of
berkelic acid because intermediate 13 was an advanced
intermediate in the total synthesis reported by Fꢀrstner and
co-workers.[3c] It should be also remarked that the introduc-
tion of the lateral chain at an advanced stage is beneficial and
highly desirable because it makes the synthesis much more
Scheme 2. Synthesis of building blocks 2 and 3. Reagents and con-
ditions: a) MsCl, Et3N, CH2Cl2, 08C; b) diethyl malonate, CsF, THF,
458C; c) LAH, THF, 758C; d) Tf2O, 2,6-lutidine, CH2Cl2, RT; e) potas-
sium trifluoro(hept-1-yn-1-yl)borate, DIPEA, [PdCl2(dppf)] (5 mol%),
MeOH, 658C; f) formaldehyde, CaCl2·2H2O, KOH, MeOH, RT and
then MnO2, CH2Cl2, RT. DIPEA=N-ethyl-N,N-diisopropylamine,
dppf=1,1’-bis(diphenylphosphino)ferrocene, LAH=lithium aluminium
hydride, Ms=methanesulfonyl, Tf=trifluoromethylsulfonyl.
2
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!