REVIEW 1559

Information Theory Description of Synthetic Strategies in the Polyquinane Series.

The Holosynthon Concept
M. Chanon,*a R. Barone,a C. Baralotto,c M. Julliard,a J.B. Hendricksonb
a
AM3 Lab. Case 561. ESA CNRS 6009. Fac. Sciences St Jérôme, 13397 Marseille Cedex 20, France
E-mail: michel.chanon@am3.u-3mrs.fr
b
Department of Chemistry. Brandeis University. Waltham, MA 02254–9110. USA; E-mail: jbh@jbh.chem.brandeis.edu
c
Provence Technologies. IMT-Technopôle de Château-Gombert, 13451 Marseille Cedex 20, France
Received 4 August 1998

Abstract: Information theory makes it possible to give a semi-
quantitative graphical representation of the various strategies used
to reach a given synthetic target. Skeletal complexity and similar-
ity of the precursors with respect to the target structure provide
figures which monitor the progress made from the starting mate-
rial en route toward the target. Examples selected from the
triquinane family are used to illustrate the benefits but also the
present limits of such an approach. Whereas for silphinene and
hirsutene various synthetic strategies appear in a clear graphical
form when treated within this framework, coriolin shows that the
counterpart of these reaction-centered approaches: the structural
entities which make them possible. Such structural entities (holo-
synthons) call attention to synthetic strategies where a global part
(holos: whole) of the target is looked at, this view complements
the more classical bond by bond, disconnection approach.
Key words: information theory, startegy, synthesis, triquinane,
silphinene, hirsutene, coriolin, holosynthon

Downloaded by: University of Texas at San Antonio. Copyrighted material.

skeleton-only approach provides graphics which can be mislead- 1 Molecular Complexity and Information Theory
ing. To improve this limitation, progress will have to be made in 2 Similarity, an Indispensable Complement of Molecular Com-
the treatment of functional complexity from a synthetic point of plexity
view. From a more general point of view, a practical treatment of 3 Present Limitation of Complexity and Similarity Quantification
stereochemistry within the information theory framework is still 4 Semi-Quantitative Description of Synthetic Strategies Reported
waited for. for Polyquinane Natural Products
The graphical treatment displays clearly the key step(s) in a given 4.1 Silphinene
strategy. Such steps are often characterized by a large change in 4.2 Hirsutene
complexity and/or similarity. This semiquantitative representation 4.3 Coriolin
converges with, on one hand, the interest of some rearrangements 5 Holosynthons
in shorter synthesis and, on the other hand, the interest and the 5.1 Definition and Further Development
limits of the class of reactions variously christened as cascades, 5.2 Good and Less Good Holosynthons
domino, tandem. The treatment shows also the indissociable Appendix

1 Molecular Complexity and Information Theory cation will also reveal its limits in the analysis to come. If
these limits are clearly identified, this quantitative ap-
Chemists have had intuitive feelings about molecular proach has the advantage of providing a semi-quantitative
complexity. In 1981 Bertz1 developed a quantitative ap- framework to express the beauty of building a complex
proach based on information theory2 and graph represen- molecule out of simple fragments.
tation of molecules.3 In this approach, the molecular com-
plexity is measured as a function of the number and nature
of its constitutive atoms and of the number and nature of
the constitutive bonds. The overall complexity of the mol-
ecule is calculated as being the sum of complexities asso-
ciated with connectivity factors and complexities associ-
ated with the presence of heteroatoms. Hendrickson and
Toczko4 have developed a simple algorithm for calculat-
ing this complexity for any organic compound. The results
converge with a chemist's intuition on many structural
features: a cyclic compound is considered more complex
than its acyclic counterpart; a ramified hydrocarbon is
more complex than its linear counterpart; a molecule with
several carbons replaced by heteroatoms is more complex.
The virtue of this approach is to provide figures allowing
easy graphical comparisons of synthetic strategies. We
have used it here as a convenient tool for such compari-
sons. Figure 1 shows on simple examples the convergence
of its results with chemical intuition. Its systematic appli- Figure 1. Complexity changes in some classical reactions
1560 M. Chanon, R. Barone, C. Baralotto, M. Julliard, J. B. Hendrickson SYNTHESIS

Biographical Sketches
Michel Chanon is Professor in the Faculté de Sciences de Saint-Jérôme. He has held
visiting appointments at the University of London, Indiana University, Kansas State Uni-
versity, Texas University, Brandeis University, Osaka University Hyderabad School of
Chemistry, Universitad Autonoma de Barcelona. Besides Computer-Aided Synthesis,
his research interests include experimental determination of mechanisms and reactivity
concepts, electron transfer and catalysis.

Christophe Baralotto was born in 1968 and studied chemistry at the University of
Marseille (France) where he received his Ph.D. degree under the supervision of Professor
Michel Chanon in 1997. His thesis subject was devoted to the study of the holosynthon
concept and its application in practical organic synthesis. In 1998 he left the academic
world to create his own company.

Downloaded by: University of Texas at San Antonio. Copyrighted material.

Michel Julliard is research director at the CNRS. In 1973 he received his Chemical
Engineer and Ph.D. degree at the University of Marseille (France). After post-doctoral
work at the Industrial Physical and Chemistry School (Paris) he moved to Professor
Jaques Metzger’s team where he studied the photochemical behaviour of diaryl triazenes.
Since 1983 he has worked on redox photochemistry, electrocatalytic processes, support-
ed photosensitization and photoactivation of oxygen. In 1993 he began to apply the pho-
toredox properties of phthalocyanines to the photodynamic therapy of cancer cells.

Rene Barone has been Directeur de Recherche at CNRS since 1986. He has worked in
the field of Computer-Aided Synthesis since 1970. In 1979 he did post-doctoral study in
M.L.H. Green’s group to develop the first program able to predict the products and by-
products of processes catalyzed by transition metal complexes. In 1981 he began to de-
sign programs working on personal computers with the aim of making available to me-
dium sized laboratories tools for use in synthesis design. His recent programs HOLO-
WIN, CONAN, SESAM are developed along this line.

Professor Hendrickson took his doctoral degree from Harvard under Professor R.B.
Woodward and was on the staff at UCLA before coming to Brandeis, developing re-
search programs both in synthetic chemistry and in application of chemistry to the com-
puter. Some 40 of his 150 research publications have been concerned with the latter area.
November 1998 Information Theory Description of Synthetic Strategies in the Polyquinane Series
In this review, we will analyze the different total synthe-
ses which have been reported for some natural products of
the triquinane family: silphinene; hirsutene and coriolin.
Fortunately, Mehta’s recent review provides an in-depth
coverage of the recent synthetic achievements in the field
of polyquinane natural products.5 This review will allow
us to lighten the weight of chemical considerations in or-
der to concentrate more on the field as a kind of bench-
mark to judge the strengths and weaknesses of informa-
tion theory as applied to the comparison of different build-
ing strategies.
2 Similarity, an Indispensable Complement
of Molecular Complexity
In 1981, when Bertz1 applied information theory to syn-
thetic strategies, his approach was a breakthrough. It per-
mitted the description, in a two dimensional way, of the
overall progression of a synthesis. Previously, this pro-
gression was only monitored in terms of steps or yields.
Bertz's graph plots the number of steps versus the increase
in complexity of the target. We propose, now, that a more
realistic way of looking at synthetic strategies has to be at
least tridimensional; Figure 2 illustrates this point. It com-
pares an overall synthetic approach to mountain climbing.
Starting from the valley (small sized starting material) the
chemist follows footpaths to reach a given summit (the
target). Looking at Figure 2 clearly shows the necessity of
adding something more to the complexity only descrip-
tion. Indeed, starting from the valley, one may progress
towards either summit A or summit B. In both cases the
complexity is increasing. The climber having only a com-
plexity measurement of his progress is “blind” with re-
1561
spect to the question “do I aim for A or for B?”. A conve-
nient way of circumventing this blindness is to define, at
every step, the distance6 between the partially synthesized
fragment of the target and the target itself. This distance is
directly connected to the similarity between this fragment
and the desired target. Actually, the chemist knows which
structural target he is aiming at; the other advantage of
completing complexity by similarity results from an al-
most material visualisation of the different strategic ap-
proaches.
We have established a rough method of evaluating the
similarity of two skeletons containing only carbon and hy-
drogen atoms (Appendix). For most of the strategies that
we will examine the progression of the synthesis will be
represented in terms of both complexity and similarity
changes.
Before leaving this mountain metaphor for strategies, let
us examine its virtues and shortcomings. Virtues first.
Downloaded by: University of Texas at San Antonio. Copyrighted material.
This representation clearly shows the crucial importance
of planning the strategy looking in all the directions from
the summit. In other terms the selection of starting mate-
rials is crucial. Indeed if the climber has missed the fact
that, not far from the summit, a large plateau exists where
a helicopter could easily land, he may waste his efforts
starting from a far deeper valley. The large plateau obvi-
ously corresponds to another natural product starting ma-
terial available in suitable quantity. A recent and spectac-
ular example has been provided by the synthesis of taxol
from 10-deacetyl baccatin III.7 Wipke's SST algorithm
has addressed this problem.8 What would be even more
useful is not yet available, i.e., a network-map of natural
products displaying the synthetic similarity of natural
compounds.
Figure 2.
Synthetic strategies and mountain climbing
1562 M. Chanon, R. Barone, C. Baralotto, M. Julliard, J. B. Hendrickson
Another advantage of the representation in Figure 2 is that
it demonstrates the difficulty of ranking strategies. In-
deed, depending upon the walker, the selected path may
drastically change. The security-minded walker (industri-
al approach) will rather follow a succession of smooth
valleys to reach, with a minimum of bad surprises, the se-
lected summit. The highly competitive sportsman will
certainly select a way which challenges his technical and
physical abilities. Another one would just focus on the
pleasure of exploring a given aspect of the available path-
ways (e.g. exploration of a radical reaction in a new struc-
tural context).
Despite its pedagogical attractiveness the shortcomings of
this representation follow from its roughness. In the next
section we will examine some of the limits associated
with the present treatment of complexity and similarity.
Even if these two concepts were totally satisfactory in
terms of quantitative treatment one would still have to an-
swer the question “Is a three dimensional space sufficient
to represent synthetic strategies?”
In previous work9 we applied the multivariate statistical
approach to explore the space of representation of a large
set of solvents. Comparing the present approach to the one
used for solvent classification shows that much remains to
be done to cleanly delineate the “space of synthetic strat-
egies”. Other efforts are in progress in this direction.10
3 Present Limitation of Complexity and
Similarity Quantification
As the biological applications of information theory have
already shown,11 this approach may be rather loose de-
spite its mathematical appearance. One of the reasons for
this looseness is that the data which are dealt with are of-
ten conceptual rather than material quantities. As such
they depend critically on the context.
Let us take the example of complexity. The Bertz ap-
proach could convey the misleading message “yes there is
a unique way of measuring complexity.” In actuality there
are several; the content of complexity in a natural product
is not the same for the pharmacologist, the material sci-
ence chemist and the synthetic chemist. Even for the syn-
thetic chemist the complexity defined by Bertz that we ex-
tensively use in this report is rather biased. It stresses the
skeletal aspects of structural complexity. As such it con-
verges with Hendrickson's description of synthetic prob-
lems,12 since adopted by several groups.13 It diverges
however from the actual overall knowledge accumulated
for decades in synthesis.
Indeed the synthetic aspects of complexity should be di-
vided into three components. The first one, skeletal com-
plexity, is predominant in polycyclic natural compounds.
The second one should be “functional complexity” con-
sidered from a synthetic point of view (chemo- and regio-
selectivities and protecting group strategies14). There is
presently no satisfactory quantitative treatment of this
SYNTHESIS
kind of complexity and we will see, with the coriolin case,
that such considerations can play a determining role in the
clever synthesis of some families of targets. The third
component is “stereochemical complexity”. No actual at-
tempts has been made to evaluate this complexity.15 These
three components are also reflected in computer-aided
synthesis programs which are either only skeleton-orient-
ed or deal also with functionality16 and stereochemistry.17
An equivalent state of imperfection characterizes “molec-
ular similarity”. As the blossoming of recent reviews and
monographs18 shows, this concept is gaining more and
more popularity in molecular sciences. The main incen-
tive for this growth is molecular recognition19 and its
pharmacological applications.20a,b
The scale of similarities for a series of natural compounds
as perceived by the active site of a given enzyme would be
quite inappropriate for ranking the similarities of the same
series from the point of view of synthetic strategy. Some
Downloaded by: University of Texas at San Antonio. Copyrighted material.
of the structural features of the studied series could be rel-
evant for both approaches but others would completely
differ. The reason is again a question of context. In a pop-
ulation of athletes the similarity perceived by an Ameri-
can football coach and a polo team coach obviously differ.
It is expected, however, that the American football coach
would do a better job than an average spectator would at
selecting the most efficient team. It is within this rather
pragmatic and limited spirit that we have built the similar-
ity scale described in Appendix 1.
4 Semi-Quantitative Description of Synthetic
Strategies Reported for Polyquinane Natural Products
In this section the synthetic achievements associated with
the targets silphinene, hirsutene and coriolin will be ex-
amined within the semi-quantitative framework described
in Sections 1 and 2. To keep this report within a reason-
able size, the publications after 1980 have been selected
and some synthetic attempts towards these natural prod-
ucts were not discussed if the intermediate finally ob-
tained was still too far from the actual target. Because of
the skeleton-centered approach the syntheses designed to
reach a given enantiomer are not examined here since
comparing the strategies to racemic products with those
for just one enantiomer of the target would be quite unfair.
A more exhaustive treatment may be found in Mehta's re-
cent review.5 All the figures describing complexity
changes versus the progression of the synthesis use a skel-
eton-only calculation of complexity. This means that all
the functions (multiple bonds, heteroatoms) have been ne-
glected in terms of complexity change. The perspective is
therefore a direct application of Hendrickson's simplify-
ing approach.12 The calculation of the overall yield is just
the multiplication of all the yields reported for each ele-
mentary step in the published work.
Every synthesis examined is also chemically represented
according to certain conventions. Only the steps associat-
November 1998 Information Theory Description of Synthetic Strategies in the Polyquinane Series 1563

ed with a change in the carbon skeleton are shown in the
schemes. When a transformation has several steps the first
is usually the one inducing a skeletal change. The symbols
“+” and “–” are respectively associated with steps which
are holosynthetic and steps with a yield lower than 60%.
The symbol “+–” designs steps possessing both character-
istics (strategically interesting but impeded by low yields).
The bold numbers placed under each structural formula are
the skeletal complexities calculated by Hendrickson's pro-
gram.4 The italic number stands for a measure of similarity
of the given structure with respect to the target. For the sake
of clarity we have gathered the syntheses which, at first
sight, seems to be the best. We stress the point on the arbi-
trariness of this choice. In Section 2 we have specifically
explained that reasons of context may completely upset
ranking of syntheses based on simplistic considerations
such as number of steps or overall apparent yields. Con-
cerning the question of uncertainties attached to yields the
reader is referred to Hudlicky's lucid discussion.21
Figures 3 and 5 gather the descriptions of eight reported
synthetic strategies (Schemes 1–8). In Figure 3 the select-
ed strategies share a common feature: the skeletal com-
plexity of silphinene (290) is reached before the seventh
step and as soon as the fourth step a value of 250 is
reached. Of the three displayed complexity curves two
show that a precursor on the way to the target is of higher
complexity than the target itself. This corresponds to a
negative situation in terms of atom economy23 if the at-
oms bringing the excedent of complexity have to be elim-
inated in a further step toward the target. In both cases the
intermediate is obtained by a holosynthetic transforma-
tion (see definition in Section 5) so that nothing can be
done to improve further the strategy concerning this point.
Whereas all the strategies gathered in Figure 3 include at
least one step in which ∆C (increase in complexity) is
higher than 100, Figure 5 describes only two strategies
(Itô and Yamamura) which contain such a step. These

Downloaded by: University of Texas at San Antonio. Copyrighted material.

4.1 Silphinene (Schemes 1–8)
Silphinene synthesis has attracted considerable attention
because its angular triquinane skeleton makes it a lead
into the series laurenene, coriolin, retigeranic acid.22
steps with a large change of complexity often correspond
to key steps. Before jumping to the conclusion that the
three other strategies depicted in Figure 5 contain no actu-
al key step one has to add two other considerations. The
first is that only skeletal complexities have been displayed
here. If a fair semi-quantitative treatment of functional

a) 58%. b) HClO4. c) NaOMe. d) Me2CuLi. e) LiCl, H2O, Me2SO; 89% (4 Steps). f) EtCO2TMS, TBAF; 89%. g) hν; 86%. h) TMSI; 89%. i)
(n-Bu)3SnH; 98%. j) LDA, (EtO)2POCl; 64%. k) Li, CH3NH2 liq.; 99%. (Overall yield : 22%).
Scheme 1. Crimmins’ strategy for silphinene synthesis24

a) Na+Cp-; 75%. b) Li(Me)C=CH2. 71%. c) 160˚C; d) H2O, PyH, OTs; 68% (2 Steps). e) O3. f) KOH; 75% (2 Steps). g) Jones’ reaction; 95%.
h) Pb(OAc)4, Cu(OAc)2; 68%. i) Me2CuLi; 89%. j) N2H4, K2CO3; 93%. (Overall yield : 15%).
Scheme 2. Sternbach’s strategy for silphinene synthesis25
1564 M. Chanon, R. Barone, C. Baralotto, M. Julliard, J. B. Hendrickson SYNTHESIS

a) 1. Li. 2. NH3. 3. NH4Cl; 87%. b) hν; 35%. c) Li, CH3NH2 liq.; 66%. (Overall yield : 20%).
Scheme 3. Wender’s strategy for silphinene synthesis26

Downloaded by: University of Texas at San Antonio. Copyrighted material.

a) CuBr, Me2S. b) HCl. c) MsCl. d) DBU. e) MeLi. f) PCC; 70% (5 Steps). g) CuBr, Me2S. h) HCl; 100% (2 Steps). i) p-MPTC; 100%. j) 200˚C;
84%. k) CH3Li. l) TsOH. m) CH3CO3H. n) BF3-Et2O. o) N2H4, K2CO3; 17% (5 Steps). (Overall yield : 10%).
Scheme 4. Paquette’s strategy for silphinene synthesis27

a) AlCl3. b) HO(CH2)2OH; 80% (2 Steps). c) NaIO4, OsO4. d) NaBH4. e) HCl, MeOH; 73% (3 Steps). f) Jones’ reaction. g) CH2N2; 76% (2
Steps). h) LDA, MeI; 86%. i) LiAlH4. j) PCC. k) N2H4, KOH; 72% (3 Steps). l) TMS, NaI; 99%. m) DBU; 87%. n) CuI; 70%. o) HCl; 98%.
p) POCl3; 81%. q) MeLi. r) SOCl2. s) m-CPBA. t) BF3-Et2O; 33% (4 Steps). u) N2H4, KOH; 90%. (Overall yield : 4%).
Scheme 5. Itô’s strategy for silphinene synthesis28

a) LDA, TMSCl. b) HC≡C-CO2Et, ZrCl4; 90% (2 Steps). c) CF3SO3TMS; 94%. d) 85%. e) hν; 85%. f) TBAF; 100%. g) H2SO4; 100%. h)
CF3SO3SiMe2Thexyl. i) SeO2; 76% (2 Steps). j) C6H5CH2OCNHCCl3; 80%. k) DIBAH. l) MnO2; 73% (2 Steps). m) 80%. n) MnO2; 85%. o)
BF3-Et2O. p) TBAF; 50%(2 Steps). q) (COCl)2, DMSO; 89%. r) LDA., MeI; 63%. s) Me2CuLi; 78%. t) N2H4, KOH. u) N2H4, KOH; 46% (2
Steps). (Overall yield : 2%).
Scheme 6. Franck-Neumann’s strategy for silphinene synthesis29
November 1998 Information Theory Description of Synthetic Strategies in the Polyquinane Series 1565

Downloaded by: University of Texas at San Antonio. Copyrighted material.

a) AlCl3; 80%. b) Anodic oxidation; 54%. c) DIBAH. d) Ac2O, Pyr.; 82% (2 Steps). e) H2O; 76%. f) MeMgBr; 72%. g) LiAlH4; 98%. h)
Pb(OAc)4; 100%. i) NaClO2; 99%. j) MOMCl; 97%. k) PDC; 35%. l) HCl; 96%. m) NaOEt/EtOH; 100%. n) Me2Cu(CN)Li2; 92%. o, p) See
Crimmins; 63% (2 Steps). (Overall yield : 6%).
Scheme 7. Yamamura’s strategy for silphinene synthesis30

a) NaH; 88%. b) LDA; 90%. c) 1. Li, NH3 liq. 2. NH4Cl; 60%. d) RuCl3, NaIO4; 100%. e) CH2N2; 100%. f) HO(CH2)2OH; 60%. g)
CH3P(O)(OMe)2, n-BuLi; 100%. h) HCl; 100%. i) (n-Bu)4NOH; 70%. j) NaBH4; 72%. k) p-MPTC; 70%. l) (n-Bu)3SnH; 70%. m, n) See
Crimmins; 63% (2 Steps). (Overall yield : 7%).
Scheme 8. Nagarajan’s strategy for silphinene synthesis31

and stereochemical complexities were available it could
well reveal strong complexity increases for these three
strategies. The second consideration follows from Figures
4 and 6. Within a given strategy, similarity is often corre-
lated to complexity, as might be expected. In Figure 2,
when one progresses from one valley towards the target,
the normal evolution is an increase in the complexity of
the intermediates and an increase in their similarity with
respect to the target. This trend reveals itself in Figure 4.
This usual trend led Bertz to state that complexity and
similarity are correlated.32 Figure 6, however, shows that
this trend is not always observed.
In the plot describing Franck-Neumann's strategy there is
indeed a step in which complexity decreases while simi-
larity increases. This step corresponds to g, h, i, j, k, l in
Scheme 6. The complexity value indeed decreases from
216 to 157 while the similarity with respect to silphinene
increases from 45% to 66%. One does not necessarily
need information theory to pick out this fact. The com-
plexity decreases on going from the first intermediate to
the second one because there is one ring less in the second
intermediate. A quick glance at the two intermediates and
at the structure of the target convinces any chemist that the
second intermediate is more similar to the target. This sit-
1566 M. Chanon, R. Barone, C. Baralotto, M. Julliard, J. B. Hendrickson SYNTHESIS

Complexity (Continuous line) Similarity vs Complexity

silphinene 300
(Dotted line)
250
300 200
100
150
100
250 Similarity vs
80 50 silphinene
0
200 0 20 40 60 80 100 120
60 Yamamura (16 Steps, 6%) Nagarajan (14 Steps, 7%)
150 Paquette (15 Steps, 10%) Franck-Neumann (21 Steps, 2%)
Itô (21 Steps, 4%)
40
100 Figure 6. Complexities versus similarity (with respect of the target) in
silphinene synthesis (B)
20
50
Synthetic uation, in which the main change brought by the transfor-
Intermediates
0 0 mation is on the similarity coordinate rather than on the
1 2 3 4 5 6 7 8 9 10 11 12 complexity one, is not exceptional. It will be met again
Silphinene Wender (3 Steps, 20%) later in this report and is widely illustrated in the
Crimmins (11 Steps, 22%) Sternbach (10 Steps, 15%)
literature33 (Equations 1–3).

Downloaded by: University of Texas at San Antonio. Copyrighted material.

Figure 3. Evolution of skeleton complexities and similarities (with re-
spect of the target) in silphinene synthesis (A)

Complexity

300
250
200
150
100
Similarity vs
Equation 133a
50
silphinene
0
0 20 40 60 80 100 120

Crimmins (11 Steps, 22%)

Wender (3 Steps, 20%)
Sternbach (10 Steps, 15%)

Figure 4. Complexities versus similarity (with respect of the target)

in silphinene synthesis (A)

Equation 233b
Complexity (Continuous line) Similarity vs
silphinene
(Dotted line)
300
100

250
80

200
60
Equation 333c
150

40 Returning to Figure 2, this would correspond to a path

100 which, for a while, seems to lead the climber further away
from the summit. Then this path crosses a valley which
50 20
goes more directly towards the selected summit. This pat-
Synthetic
Intermediates tern will help us in proposing a definition of holosynthon
0 0 (Section 5).
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Silphinene Itô (21 Steps, 4%) In this treatment of silphinene we have omitted Fraser-
Yamamura (16 Steps, 6%) Nagarajan (14 Steps, 7%) Reid's approach.34 The reason is that the selected strategy
Paquette (15 Steps, 10%) Franck-Neumann (21 Steps, 2%)
involves several protection-deprotection steps. We will
Figure 5. Evolution of skeleton complexities and similarities (with see with coriolin synthesis (Section 4.3) the limitations of
respect of the target) in silphinene synthesis (B) our approach for this kind of strategy. Taylor’s transition
November 1998 Information Theory Description of Synthetic Strategies in the Polyquinane Series
metal centered strategy35 was skipped for sake of conci-
sion.
4.2 Hirsutene (Schemes 9–26)
This sesquiterpene has been taken as a model for synthetic
strategies leading to the hirsutanes.36 The number of re-
Scheme 9. Biosynthesis of hirsutene37
a) 1. LiC(SPh)3. 2. s-BuLi. 3. FeCl3; 64%. b) Ni Raney; 93%. c)
HO(CH2)2OH, TsOH; 86%. d) Li, NH3 liq.; 92%. e) 1. BuLi. 2.
Cl2P(O)NMe2. 3. Me2NH. 4. Li, MeNH2 liq. 5. H2O, Me2CO, TsOH;
82%. f) CH2=PPh3, KOt-Bu; 96%. (Overall yield : 37%)
Scheme 10. Cohen’s strategy for hirsutene synthesis38
a) OsO4, NaIO4. b) HOAc; 62% (2 Steps). c) CH2=CHMgBr; 91%. d) CH3C(OEt)3, Hg(OAc)2, EtCO2H. e) KOH, H2O; 82% (2 Steps). f)
(COCl)2. g) CH3CHN2; 89% (2 Steps). h) Cu(acac)2; 94%. i) PbCO3, 580°C; 66%. j) H2/PtO2; 90%. k) See Cohen; 96% (1 Steps). (Overall
yield: 22 %).
Scheme 11. Hudlicky’s strategy for hirsutene synthesis39
1567
a) hν; 80%. b) TMSI; 95%. c) 1. Li, NH3 liq. 2. MeI; 47%. d) See
Cohen; 96% (1 Steps). (Overall yield: 33%).
Scheme 12. Iyoda’s strategy for hirsutene synthesis40
ported syntheses led us to treat them in four groups. The
criterion selected to classify these groups is overall yield.
Downloaded by: University of Texas at San Antonio. Copyrighted material.
In Figure 7 the complexity and similarity profiles of the
biosynthetic pathway37 and the four most efficient strate-
gies (by overall yield) are displayed. A rapid, early growth
of complexity and similarity characterizes the Cohen38
and Iyoda40 strategies. The target complexity value is
practically obtained in three steps. Cohen's synthesis is
particularly spectacular from this point of view. The
triquinane skeleton is built in a highly convergent way by
a bis-enolate oxidative coupling step (Scheme 10). This
one-pot procedure involves four elementary synthetic
steps. A characteristic shared by the three other
strategies37,39,41 described in Figure 7 is that the increase
in complexity precedes the increase in similarity. The
same pattern is associated with the biosynthetic approach.
Iyoda's synthesis (Scheme 12) displays the strategic pro-
file closest to the biosynthetic route. Hudlicky's
approach39 (Scheme 11) illustrates again the point men-
tioned above with the Franck-Neumann silphinene syn-
thesis. The comparative evolution of complexity and sim-
ilarity (steps i,j) shows that a large increase in similarity
occurs (58 to 95%) without any change in complexity.
In Figure 8 three types of strategies appear. The Funk43
and Franck-Neumann45 strategies contain a step with a
strong change in complexity about half way to the target.
1568 M. Chanon, R. Barone, C. Baralotto, M. Julliard, J. B. Hendrickson SYNTHESIS

a) PhSeSO2Ph, hν. b) H2O2; 70% (2 Steps). c) 160°C. d) Jones’ reaction; 72% (2 Steps). e) hν, CH2=CH2. f) NaN(TMS)2, TBSCl. g) m-CPBA;
68% (3 Steps). h) Al(Hg)x; 91%. i) Ph3P=CH2, (i-Pr)2O; 86%. j) I2; 91%. k) See Cohen; 96% (1 Steps). (Overall yield: 26%).
Scheme 13. Paquette’s strategy for hirsutene synthesis41

Downloaded by: University of Texas at San Antonio. Copyrighted material.

a) CH(OEt)3, TsOH; 95%. b) O3, Ph3P; 100%. c) 81%. d) LiAlH4; 99%. e) HCl; 100%. f) Na2SO4, Pyrrolidine; 99%. g) LiAlH4; 93%. h)
Mesitylene; 70%. i) Jones’ reaction; 97%. j) O3, Me2S; 99%. k) Ac2O, Pyr.; 90%. l) Jones’ reaction; 95%. m) 1. (COCl)2. 2. 2-Mercaptopyridin
N-oxide, t-BuSH; 82%. n) MeOH, Et3N; 95%. o) Ph3PCH3Br, KHMDS; 86%. p) 1. NaH. 2. (COCl)2. 3. Me2CuLi; 97%. q) O3, Me2S; 80%. r)
KOH; 97%. s) H2/PtO2; 100%. t) See Cohen; 96% (1 Steps). (Overall yield: 20%).
Scheme 14. Sternbach’s strategy for hirsutene synthesis4

Complexity (Continuous line) Similarity vs Complexity (Continuous line) Similarity vs

hirsutene hirsutene
(Dotted line) (Dotted line)
300 300
100 100
250 250
80 80
200 200

60 60
150 150

40 40
100 100

20 50 20
50
Synthetic Synthetic
Intermediates Intermediates
0 0 0 0
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Hirsutene Paquette (11 Steps, 26%) Hirsutene Curran (13 Stepas, 11%)
Iyoda (4 Steps, 33%) Cohen (6 Steps, 37%) Franck-Neumann (10 Steps, 13%) Funk (12 Steps, 13%)
Biosynthesis (5 Steps) Hudlicky (11 Steps, 22%) Sternbach (20 Steps, 20%)

Figure 7. Evolution of skeleton complexities and similarities (with Figure 8. Evolution of skeleton complexities and similarities (with
respect of the target) in hirsutene synthesis (A) respect of the target) in hirsutene synthesis (B)
November 1998 Information Theory Description of Synthetic Strategies in the Polyquinane Series 1569

In contrast, Curran’s approach44 saves the key step for the Complexity
very end of the synthesis. Sternbach's smooth approach42 300
provides the best efficiency by overall yield. This strate- 250
gy, however, is the longest one in terms of steps and the 200
complexity of the target is almost reached after only seven 150
steps (out of 17 overall), but to reach 100% similarity to 100
Similarity vs
hirsutene requires ten more steps. This part of the strategy 50
hirsutene
has been recently improved by Rawal46 using their ap- 0
0 20 40 60 80 100 120
proach of forming a latent triquinane ((±) endo-hirsutene)
out of a norbornane via a Paterno-Büchi reaction. Franck- Cossy (9 Steps, 5%) Wender (7 Steps, 4%)
Weedon (7 Steps, 9%) Ley (8 Steps, 8%)
Neumann’s strategy stands out in Figure 9 with similarity
plotted against complexity. This corresponds to the same Figure 11. Complexities versus similarity (with respect of the target)
type of approach for two similar targets (compare Figures in hirsutene synthesis (C)
6 and 9). For both targets the key step involves a decrease
in complexity coupled with an increase in similarity.
terms of number of steps. Indeed, here, the shortest strat-
Figures 10 and 11 reveal some new trends. Ley's egy (number of steps) is also the least efficient (overall
approach48 resembles Funk’s discussed above. This re- yield). In the same spirit of skepticism one may note that
semblance is partly fortuitous because our treatment ne- Wender's strategy,47 which was a winner with silphinene

Downloaded by: University of Texas at San Antonio. Copyrighted material.

glects functionality changes. The comparison of the two or isocomene is not among the most competitive when ap-
strategies calls attention to another limit of semi-quantita- plied to hirsutene. In Weedon’s strategy,50 the intramolec-
tive approaches ranking the “quality” of strategies only in ular McMurry coupling reaction displays its efficiency
more clearly when monitored in terms of similarity than
Complexity when monitored in terms of complexity. Figure 11 points
out one artifact caused by our treatment of similarity. In
300
Cossy's strategy49 the introduction of the gem-dimethyl
250
200
group by a cyclopropane opening appears as a discontinu-
150
ity in the complexity-similarity diagram. This discontinu-
100 ity has no strategic significance but just follows from our
Similarity vs choice (Appendix I) to retain the number of rings in struc-
50
hirsutene
0 tures as a key factor to establish their similarity.
0 20 40 60 80 100 120

Sternbach (20 Steps, 20%) Funk (12 Steps, 13%) Figure 12 gathers the syntheses with the lowest overall
Franck-Neumann (10 Steps, 13%) Curran (13 Steps, 11%) yields. One must stress again that classification by this cri-
Figure 9. Complexities versus similarity (with respect of the target) terion is just a matter of convenience. The gathering of all
in hirsutene synthesis (B) the strategies on a single graph would have been undeci-

Complexity (Continuous line) Similarity vs Complexity (Continuous line) Similarity vs

hirsutene hirsutene
(Dotted line) (Dotted line)
300 300
100 100

250 250

80 80
200
200
60
60
150
150
40
40 100
100

50 20
50 20
Synthetic
Synthetic Intermediates
Intermediates 0 0
0 0 2 4 6 8 10 12 14 16 18 20 22 24 26
1 2 3 4 5 6 7 8 9 10
Hirsutene Wender (7 Steps, 4%) Hirsutene Fukumoto (25 Steps, 3%)
Cossy (9 Steps, 5%) Ley (8 Steps, 8%) Hewson (17 Steps, 3%) Magnus (16 Steps, 3%)
Weedon (7 Steps, 9%) Mehta (10 Steps, 3%) Little (15 Steps, 2%)

Figure 10. Evolution of skeleton complexities and similarities (with Figure 12. Evolution of skeleton complexities and similarities (with
respect of the target) in hirsutene synthesis (C) respect of the target) in hirsutene synthesis (D)
1570 M. Chanon, R. Barone, C. Baralotto, M. Julliard, J. B. Hendrickson SYNTHESIS

Complexity pherable. The reader must keep in mind that: (1) a syn-
300 thesis of 10 steps with an average 80% yield amounts to
250 an overall yield of 10% whereas if the average yield is
200 75% this value drops to 5.6% (2) our description does
150 not account for the evolution of functional and stere-
100 ochemical complexities. Of the five strategies displayed
Similarity vs
50
0
hirsutene in Figure 12, two (Little52 and Mehta54) involve a holo-
0 20 40 60 80 100 120 synthetic step. Mehta's starts with a double jump in com-
plexity obtained by a sequence of [4+2] and [2+2] cy-
Hewson (17 Steps, 3%) Magnus (16 Steps, 3%)
Mehta (10 Steps, 3%) Little (15 Steps, 2%) cloadditions. This double jump yields a precursor whose
Fukumoto (25 Steps, 3%) skeletal complexity exceeds hirsutene complexity by 60
Figure 13. Complexities versus similarity (with respect of the target) units. This excess has to be compensated in the next step
in hirsutene synthesis (D) in which a 118 loss in complexity is compensated by a
22% gain in similarity.
Returning to Figure 2, this strategy illustrates the situation
wherein the climber follows a valley which leads him fur-
ther off from the selected summit; he then has to go down
into another one to find his original direction. One may

Downloaded by: University of Texas at San Antonio. Copyrighted material.

a) HO(CH2)2OH, TsOH. b) NaBH4. c) TsCl, Pyr. d) ∆; 50% (4 Steps).
e) TiCl4; 71%; f) MeNHOH; 75%. g) MeI. h) H2/Pd; 89% (2 Steps).
i) m-CPBA; 90%. j) PCC. k) H2/Pd; 65%. l) See Cohen; 96% (1 Step).
(Overall yield : 13%).
Scheme 15. Funk’s strategy for hirsutene synthesis43
note in Figure 12 that Hewson's strategy51 also contains a
holosynthetic step which “overdoes” what it has been
planned for. One may speculate about the advantages and
disadvantages of such types of strategy; the interesting
fact about the graphical representation is that it reveals
this characteristic at first glance. Fukumoto's strategy53
again illustrates the point that, in terms of overall yield, a
safe strategy may be as efficient as one containing half as
many steps. If one had, however, to extrapolate to larger
scale preparations the shortest strategy would be, for a
given overall yield, the least costly.
The intramolecular allylsilane addition key step used by
Majetich56 to assemble the polycyclic skeleton proved to
be efficient but the synthesis was not carried to the final
target. Thus this strategy was not discussed here. Two
other approaches of (±) hirsutene were also not dis-
cussed here because they reach precursors on the way to
the target.57,58 For reasons explained in the introduction
the asymetric synthesis of hirsutene were not dealt with
here.59 The “metallo ene reaction–carboxylation” cascade
strategy60 was left aside for saving space.

a) NaBH4, CeCl3. b) Ac2O, Pyr. c) CH2=CH(OTBS)2. d) ∆. e) PhSeCl. f) H2O2; 62% (6 Steps). g) 75%. h) DIBAH; 83%. i) (CF3CO)2O; Pyr.;
81%. j) (n-Bu)4NI; 75%. k) LiC≡C-TMS. l) TBAF; 75% (2 Steps). m) (n-Bu)3SnH; 63%. (Overall yield : 11%).
Scheme 16. Curran’s strategy for hirsutene synthesis44
November 1998 Information Theory Description of Synthetic Strategies in the Polyquinane Series 1571

a) 1. CuBr, Me2S. 2. HCl; 84%. b) PhCOCl, Pyr; 90%. c) Morpholine, TsOH; 91%. d) 57%. e) HI; 85%. f) H2/Pd; 91%. g) NaOMe; 95%. h)
Electrochmical reduction: 45%. i) PCC; 100%. j) See Cohen; 96% (1 Steps). (Overall yield: 13%).
Scheme 17. Franck-Neumann’s strategy for hirsutene synthesis45

Downloaded by: University of Texas at San Antonio. Copyrighted material.

a) 1. Mg. 2. Ac2O; 79%. b) hν; 23%. c) 10-camphorsulfonic acid;
71%. d) PhSH; 78%. e) H2, [Ir(cod)(pyr)(PCy3)]PF6; 80%. f) NaIO4;
86%. g) P(MeO)3, 170°C; 60%. (Overall yield : 4%).
Scheme 18. Wender’s strategy for hirsutene synthesis47
a) 1. KDA. 2. LiAlH4; 86%. b) Me2N=CHBr+ Br-; 90%. c) 1. Mg. 2.
CuI; 58%. d) NPSP, TiCl4; 45%. e) LDA, LiAlH4; 66%. f) NPSP, (n-
Bu)3P. g) Ni Raney; 63% (2 Steps). h) See Cohen; 96% (1 Step).
(Overall yield : 8%).
Scheme 19. Ley’s strategy for hirsutene synthesis48

4.3 Coriolin (Schemes 27–35)
The main structural characteristic of this triquinane ses-
quiterpenoid is its considerable functionality: the carbon
skeleton is substituted by two hydroxy and one keto
group, as well as by two epoxides. We have purposely se-
lected this target to show the limits of a “skeleton-only
complexity” approach.
Let us first consider the specific problem of the double
epoxidation needed to obtain the target. Tatsuta64 per-
formed this transformation by a one-pot reaction (62%).
The same year, Danishefsky68 proposed a four-step trans-
formation, far better in terms of stereochemical control,
and of comparable overall yield (58%). For saving space
in Schemes 27–35 and Figure 14–16 we have retained the
“shorter” Tatsuta's transformation.
In Figure 14 the graphical evolutions of skeletal complex-
ities and similarities of the three “best” approaches (com-
bination of yields and number of steps) are displayed.
Only Curran's synthesis63 (Scheme 29) presents a holo-
synthetic character in its last phase; the strategy used in
this phase had already met with success for hirsutene syn-
thesis (Section 4.2). Ikegami61 and Ito62 classically built
the triquinane skeleton starting from a precursor where
two out of the three rings constituting the triquinanic skel-
eton are already present (Schemes 27 and 28).
Figures 15 and 16 gather the “average” synthesis. In Fig-
ure 15, the graph describing Tatsuta’s approach64 unduly
suggests a highly holosynthetic step in the first phase of the
synthesis. Actually, the great increase in complexity is the
result of addition of the complexity of the reagent to the one
of the substrate (steps f,g,h,i,j in Scheme 30). The same
1572 M. Chanon, R. Barone, C. Baralotto, M. Julliard, J. B. Hendrickson
a) Pd(OAc)2, (i-PrO)3P; 60%. b) CH2N2, Pd(OAc)2; 99%. c) H2/PtO2;
82%. d) KH; 71%. e) LiAlH4; 90%. f) KAPA; 86%. g) PCC; 96%. h)
hν; 58%. i) 1. NiCl2(PPh3)2. 2. MeMgBr; 35%. (Overall yield : 5%).
Scheme 20. Cossy’s strategy for hirsutene synthesis49
a) hν; 29%. b) TBSCl, Pyr.; 89%. c) TiCl3, K; 55%. d) TBAF. e) H2/
PtO2. f) Jones’ reaction; 62% (3 Steps). g) See Cohen; 96% (1 Step).
(Overall yield : 9%).
Scheme 21. Weedon’s strategy for hirsutene synthesis50
a) NaH, TsSMe. b) HO(CH2)2OH, TsOH. c) NaIO4. d) CaCO3; 64% (4 Steps). e) EtNO2, Tetramethylguanidine; 93%. f) TiCl4; 77%. g) TFA;
79%. h) NaH; 83%. i) LDA, MeI. j) ∆. k) N2H4, K2CO3. l) m-CPBA; 67% (4 Steps). m) NaH; 72%. n) NaHPO3; 87%. o) HCO2H; 56%. p)
RuO2; 42%. q) See Cohen; 96% (1 Step). (Overall yield : 3%).
Scheme 22. Hewson’s strategy for hirsutene synthesis51
SYNTHESIS
holds true in Schuda’s synthesis66 (Scheme 32, steps a,b,c).
In Tatsuta’s approach64 the key step is better displayed in
terms of similarity rather than in complexity change (steps
k, l, m, n). This synthesis clearly illustrates the shortcoming
of the skeleton-only complexity approach when applied to
highly functionalized targets. Indeed, the graph (Figure 15)
would suggest that not much has happened in the synthetic
approach between steps 9 and 19 (steps j to q in Scheme
32). Actually this is not true in terms of functionality build-
ing. Tatsuta’s approach similarly shows a deceptively inac-
tive phase of 8 steps (Scheme 30, steps k to r).
The cyclopropanation technique used in Trost’s approach65
(step j, Scheme 31) introduces the hectic appearance of the
similarity evolution; this has more to do with our similarity
definition than with any significant strategic feature.
In Figure 16, Mehta’s54 (Scheme 35) and Wender’s67
(Scheme 33) strategies both stand out with an early holo-
synthetic step which leads to precursors whose complexi-
Downloaded by: University of Texas at San Antonio. Copyrighted material.
ty exceeds that of the target. Danishefsky’s strategy68 also
allows a very efficient building of the coriolin skeleton.
Nevertheless these three strategies which could be consid-
ered as holosynthetic from a skeletal perspective are the
ones with the lower overall yields. These results, obtained
in three excellent synthetic groups, suggest that the use of
a skeletal-only holosynthetic strategy is not necessarily a
trump card for any kind of target.
The compared scattering of the graphs displayed in Fig-
ures 17 and 18 which plot the sum of the complexities of
precursors versus the overall yield of the different synthe-
ses shows in a different way that the skeleton-only ap-
proach gives better results for targets less loaded with
functionalities.
The recent Kuwajima’s group69 approach presents a very
good overall yield of 10.5%. The skeleton obtained in this
work differs, however slightly from the actual skeleton of
coriolin. V.Singh’s70 novel strategy for the construction
November 1998 Information Theory Description of Synthetic Strategies in the Polyquinane Series 1573

Downloaded by: University of Texas at San Antonio. Copyrighted material.

a) NaBH4; 74%. b) DIBAH; 76%. c) Ph3PCH=CH2; 88%. d) PCC; 88%. e) Et2NH; 91%. f) . g) KO2CN=NCO2K; 87% (2 Steps). h) 1.
Electrochemical reduction ; 55%. i) ∆; 63%. j) 1. BH3. 2. PCC, CeCl3; 56%. k) NaOMe, EtOCHO; 96%. l) n-BuSH, TsOH, MgSO4; 82%. m)
KOt-Bu, MeI; 62%. n) KOH, HO(CH2)2OH; 51%. o) See Cohen; 96% (1 Step). (Overall yield : 2%).
Scheme 23. Little’s strategy for hirsutene synthesis52

a) I2, KI, NaHCO3. b) DBU. c) LiAlH4; 58% (3 Steps). d) TBSCl. e) Ac2O, Pyr. f) TBAF; 82% (3 Steps). g) O3, Me2S; 93%. h) Ph3P=CHCOMe.
i) PPTS, MeOH; 69% (2 Steps). j) TsOH; 86%. k) LiOH. l) ortho-NO2PhSeCN, (n-Bu)3P. m) H2O2; 66% (3 Steps). n) 1. LDA. 2. TMSCl. 3.
Pd(OAc)2; 99%. o) H2/Pd. p) Ph3P=CH2. q) CH2I2, Et2Zn; 50% (3 Steps). r) HClO4. s) Ph3P=CH2; 61% (2 Steps). t) PCC. u) PdCl2, CuCl, O2.
v) (n-Bu)4NOH, KOH. w) H2/PtO2; 84%; (4 Steps). x) PCC; 74%. y) See Cohen; 96% (1 Step). (Overall yield : 3%).
Scheme 24. Fukumoto’s strategy for hirsutene synthesis53
1574 M. Chanon, R. Barone, C. Baralotto, M. Julliard, J. B. Hendrickson SYNTHESIS

of linearly fused cis:anti:cis tricyclopentanoids could pro-

vide a rapid access to coriolin in the future.

5 Holosynthons

5.1 Definition and Further Development

a) 80˚C; 90%. b) hν; 85%. c) 500˚C; 100%. d) benzyl benzoate,
317˚C; 53%. e) H2/Pd; 95%. f) NaH, MeI; 65%. g) Ph3P=CH2; 84%.
The preceding sections have shown that several strategies
were characterized by a step involving a large change in
complexity and/or similarity. In such steps, the transfor-
mation is “spread” over several bonds. This “spreading”
corresponds, at the conception stage, to a global view of
the structure as opposed to a more classical disconnection
of one bond at a time. We coined the word holosynthon
(holos = global in Greek) to characterize this type of strat-
egy.71 A holosynthon is a structural entity specifically
designed to make a major change of complexity and/or
similarity possible in a one-pot reaction.

Downloaded by: University of Texas at San Antonio. Copyrighted material.

h) LiAlH4; 90%. i) NaH, Imidazole, CS2, MeI; 88%. j) (n-Bu)3SnH; This definition focuses on the kind of starting structure
19%. (Overall yield : 3%). needed to design a synthesis with a holosynthetic key
Scheme 25. Mehta’s strategy for hirsutene synthesis54 step. By extension, one may define a holosynthetic mix-

a) K2CO3, TsN3. b) hν; 82% (2 Steps). c) NaBH4. d) SOCl2, Pyr.; 72% (2 Steps). e) DBU; 94%. f) NaOH; 96%. g) (COCl)2; 68%. h) AgBF4;
38%. i) 1. MeLi. 2. HgCl2, HgO; 93%. j) 1. BuLi. 2. CuI. 3. BF3-OEt2; 60%. k) BzEt3N+Cl-, KF; 84%. l) TsCl, Pyr.; 94%. m) LiN(TMS)2; 75%.
n) NaBH4. o) CS2, NaH, MeI. p) (n-Bu)3SnH, AIBN; 70% (3 Steps). (Overall yield : 3%).
Scheme 26. Magnus’ strategy for hirsutene synthesis55

a) H2O2. b) 40˚C; 66% (2 Steps). c) NaH, BzBr; 91%. d) NBS. e) H2/Pd; 80% (2 Steps). f) PCC. g) TBSCl. h) DBU; 80% (3 Steps). i) Me2CuLi;
95%. j) KOt-Bu, MeI; 77%. k) Li, NH3; 63%. l) DHP. m) TBAF. n) PCC; 90% (3 Steps). o) NaH; 94%. p) PdCl2, CuCl; 77%. q) KOt-Bu; 83%.
r) LDA, MeI; 78%. s) 1. LDA, PhSeBr. 2. H2O2. t) HOAc; 50% (2 Steps). u) KOt-Bu. v) m-CPBA. w) DBU; 48% (3 Steps). x) See Danishefsky;
58% (1 Step). (Overall yield : 1.1%).
Scheme 27. Ikegami’s strategy for coriolin synthesis61
November 1998 Information Theory Description of Synthetic Strategies in the Polyquinane Series 1575

Downloaded by: University of Texas at San Antonio. Copyrighted material.

a) AlCl3; 90%. b) NaBH4. c) Jones’ reaction. d) KOt-Bu, MeI; 65% (3 Steps). e) LiAl(OMe)3H; 98%. f) O3, Me2S. g) Jones’ reaction; 76% (2
Steps). h) ∆; 84%. i) 1. NaOAm. 2. H2O. 3. CH3I; 100%. j) 1. AlH3. 2. MsCl, LiCl. 3. LiAlH4. k) MsCl, Pyr.; 80% (2 Steps). l) KO2. m) NaBH4;
82% (2 Steps). n) BzCl, Pyr. o) BH3, H2O2. p) Jones’ reaction. q) KOt-Bu; 54% (4 Steps). r) Hg(OAc)2; 100%. s) KOH; 60%. t) Isoprenyl
acetate, TsOH; 90%. u) 1. m-CPBA. 2. NaHCO3, LiOH; 75%. v) DHP. w) LDA, MeI. x) LDA, PhSeBr. y) H2O2. z) HOAc; 65% (5 Steps). aa)
See Danishefsky; 58% (1 Step). (Overall yield : 2%).
Scheme 28. Ito’s strategy for coriolin synthesis62

a) NaBH4, CeCl3. b) Ac2O, Pyr; 95% (2 Steps). c) CH2=C(OTBS)2. d) ∆. e) PhSeCl. f) H2O2; 62% (4 Steps). g) Li, CuBr. h) LiAlH4; 90% (2
Steps). i) PCC. j) LiC≡C__TMS. k) PCC. l) HO(CH2)2OH. m) TBAF. n) PCC; 30% (6 Steps). o) SmI2. p) TsOH; 60% (2 Steps). q) LDA,
TMSCl; 89%. r) DDQ; 72%. s, t, u) See Ikegami; 48% (3 Steps). v) See Danishefsky; 58% (1 Step). (Overall yield : 1.7%).
Scheme 29. Curran’s strategy for coriolin synthesis63
1576 M. Chanon, R. Barone, C. Baralotto, M. Julliard, J. B. Hendrickson SYNTHESIS

Complexity (Continuous line) Similarity vs

350 coriolin
(Dotted line)

300
100
250
80
200
60
150

40
100

20
50
Synthetic
Intermediates
0 0
2 4 6 8 10 12 14 16 18 20 22 24 26 28

Coriolin Curran (22 Steps, 1,7%)

Ito (27 Steps, 2%) Ikegami (24 Steps, 1,1%)

Downloaded by: University of Texas at San Antonio. Copyrighted material.

Figure 14. Evolution of skeleton complexities and similarities (with
respect of the target) in coriolin synthesis (A)

ture using the above definition where “structural entity” is

a) Br2. b) TsOH. c) KOt-Bu. d) NBS. e) AcOAg; 54% (5 Steps). f) hν;
35%. g) NaBH4. h) MOMCl. i) NaOMe; 74% (3 Steps). j) TsCl. k)
KHCO3, 85˚C; 90% (2 Steps). l) NaI; 81%. m) H2SO4; 94%. n) OsO4;
86%. o) CH3C(OMe)2CH3, TsOH; 98%. p) PCC; 89%. q) NaH, Me-
thyl-2-nitro-phenyl disulfide; 80%. r) Th(NO3)3; 68%. s) MeLi; 60%.
t) Li, NH3; 75%. u) TFA; 93%. v) Ac2O, Pyr.; 78%. w) MeSO2Cl,
Pyr. x) LiOH; 46% (2 Steps). y) See Danishefsky; 58% (1 Step).
(Overall yield : 0.3%).
Scheme 30. Tatsuta’s strategy for coriolin synthesis64
replaced by “set of reagents”. This extension has applica-
tions in terms of combinatorial chemistry.72 The transfor-
mations to be applied on the starting holosynthon have
been the center of much attention during the last decade.
The first review to gather data related to this idea was pro-
vided by Posner.73 Since this time several reviews have
been published to cover tandem reactions,74 cascade reac-
tions,75 domino reactions;76 a special issue of Chemical
Reviews77 edited by Wender (see also ref. 13e) superbly

a) Pd(P(Ph)3)4, DBU; 76%. b) NBS; 89%. c) 1.P(Ph)3. 2. K2CO3. 3. 40˚C; 79%. d) MeSH. e) HO(CH2)2OH; 92% (2 Steps). f) KH, (MeS)2;
79%. g) KH; 72%. h) m-CPBA; 63%. i) TBAF; 87%. j) CH2I2, ZnEt2; 82%. k) H2/PtO2; 94%. l) SOCl2; 92%. m) m-CPBA; 87%. n) HClO4. o)
DBU; 91% (2 Steps). p) Sodium naphthalenide. q) DBU; 52% (2 Steps). r) Li, NH3 liq. s) m-CPBA; 63% (2 Steps). t) CF3C(O)N(TMS)2. u)
LDA, TMSCl. v) Me3N+,I-. w) MeI. x) DBU; 46%. y) HF; 85%. z) See Danishefsky; 58% (1 Step). (Overall yield : 0.7%).
Scheme 31. Trost’s strategy for coriolin synthesis65
November 1998 Information Theory Description of Synthetic Strategies in the Polyquinane Series 1577

Complexity (Continuous line) Similarity vs Complexity (Continuous line) Similarity vs

coriolin coriolin
350 (Dotted line) (Dotted line)
350
300
100 300
100
250
250
80
80
200
200
60
150 60
150
40
100 40
100

50 20
Synthetic 20
50
Intermediates Synthetic
0 0 Intermediates
2 4 6 8 10 12 14 16 18 20 22 24 26 0 0
2 4 6 8 10 12 14 16 18 20 22
Coriolin Trost (26 Steps, 0,7%)
Schuda (25 Step, 0,7%) Tatsuta (25 Steps, 03%) Coriolin Wender (14 Steps, 0,4%)
Metha (19 Steps, 0,8%) Danishefsky (22 Steps, 0,2%)
Figure 15. Evolution of skeleton complexities and similarities (with

Downloaded by: University of Texas at San Antonio. Copyrighted material.

Figure 16. Evolution of skeleton complexities and similarities (with
respect of the target) in coriolin synthesis (B)
respect of the target) in coriolin synthesis (C)

a) HO(CH2)2OH; 70%. b) 1. Br2. 2. NaOMe. 3. HCl; 59%. c) LiAl(OMe)3H, CuBr; 84%. d) KOt-Bu, MeI; 85%. e) Li, NH3 liq.; 70%. f) KH,
BzBr; 89%. g) OsO4; 96%. h) NaIO4. i) NaBH4; 85% (2 Steps). j) t-BuCOCl, Pyr; 48%. k) (CF3SO2)2O, Pyr. l) TBAI. m) Zn; 81% (3 Steps).
n) KOH; 100%. o) RuO2, NaIO4; 93%. p) HCl; 85%. q) CH(OMe)3, TsOH. r) 160˚C; 82% (2 Steps). s) Hg(OAc)2; 89%. t) KOt-Bu. u) TsOH;
80%. v, w, x) See Ikegami; 48% (3 Steps). y) See Danishefsky; 58% (1 Step). (Overall yield : 0.7%).
Scheme 32. Schuda’s strategy for coriolin synthesis66

gathered most of the relevant contributions to this idea for
“making much chemistry in one step”.
This information theory approach is really not needed to
recognize these reactions in a given strategy. When we
first coined the term holosynthon we were not even aware
of Bertz’s approach. We believe, however, that the defini-
tion of holosynthon given in terms of complexity and or
similarity is more closely defined than the intuitive views
that we had at first. The graphical representation of strat-
egies complements the classical approaches in terms of
number of steps and key steps.78 It is particularly appro-
priate for polycyclic targets which are not too heavily
functionalized.
In Section 4.3. we saw that Bertz’s complexity was rather
good at describing carbon-skeleton complexity changes but
rather poor at describing the functional and stereochemical
aspects of complexity. The unsatisfactory treatment of co-
riolin strategies above demonstrated the limits of such a
narrow view. On the other hand, if one accepts the idea that
complexity could be sub-divided into three components the
holosynthon approach touches the heart of efficient synthe-
sis. Indeed the following examples show that a functional
holosynthon could be a structural unit for which large
changes of functional complexity and/or similarity can
be achieved in a one-pot reaction. Equations (4)79 and
(5)80 provide examples of such functional holosynthons:
1578 M. Chanon, R. Barone, C. Baralotto, M. Julliard, J. B. Hendrickson SYNTHESIS

Downloaded by: University of Texas at San Antonio. Copyrighted material.

a) 1. Li. 2. NH3 liq. 3. Ac2O; 80%. b) O3; 61%. c) Ph3PCH=CH(OEt)2, NaOEt; 55%. d) hν; 15%. e) PhSH, 100˚C; 72%. f) Li, NH3 liq; 80%.
g) 1. m-CPBA. 2. H2O; 67%. h) BF3. i) LDA, TMSCl. j) Pd(OAc)2. k) LDA, PhSSO2Ph; 42% (4 Steps). l) HOAc; 100%. m) 1. m-CPBA. 2.
77˚C; 64%. n) See Danishefsky; 58% (1 Step). (Overall yield : 0.4%).
Scheme 33. Wender’s strategy for coriolin synthesis67

a) NaH CO2; 63%. b) NaOMe, MeOH; 85%. c) TsOH; 51%. d) 120˚C. e) PhSeCl. f) H2O2; 57% (3 Steps). g) MeLi; 73%. h) O3. i) CrO3. j)
Ba(OH)2. k) Pb(OAc)4; 46% (4 Steps). l) KOt-Bu. m) TsOH; 71% (2 Steps). n) KOt-Bu. o) HOAc; 63% (2 Steps). p) DIBAH. q) Li, NH3 liq.
r) m-CPBA. s) PCC; 55% (4 Steps). t) 1. LDA. 2. Phenyl(thiophenyl)sulfonate; 40%. u) 1. m-CPBA. 2. 77˚C; 64%. v) H2O2, 0˚C; 58%. (Overall
yield : 0.2%).
Scheme 34. Danishefsky’s strategy for coriolin synthesis68
November 1998 Information Theory Description of Synthetic Strategies in the Polyquinane Series 1579

a-f) See Mehta’s synthesis of hirsutene ; 18% (6 Steps). g) MeMgI; 90%. h) POCl3, Pyr.; 75%. i) Li, NH3 liq. 63%. j) m-CPBA; 100%. k)
BF3-OEt2; 80%. l) LDA, TMSCl. m) Pd(OAc)2; 90% (2 Steps). n) LDA, PhSeBr. o) H2O2; 35% (2 Steps). p, q, r) See Ikegami; 48% (3 Steps).
s) See Danishefsky; 58% (1 Step). (Overall yield: 0.8%).
Scheme 35. Mehta’s strategy for coriolin synthesis54

Downloaded by: University of Texas at San Antonio. Copyrighted material.

Figure 17. Overall value of complexities (addition of the complexity of each intermediate) versus overall yield for coriolin.

Figure 18. Overall value of complexities (addition of the complexity of each intermediate) versus overall yield for hirsutene.
1580 M. Chanon, R. Barone, C. Baralotto, M. Julliard, J. B. Hendrickson
Equation 479
Equation 580
The same holds true for stereochemical holosynthons,
which would contain a structural unit in which large chang-
es of stereochemical complexity and/or similarity could be
brought about in a one-pot reaction. Many recent examples
illustrate this point. We selected only a recent one81 (equa-
tion 6) to save space. The endiandric acid cascade (equation
7), in a single operation, converts a simple achiral polyene
into the tetracyclic endiandric acid A methyl ester with
complete control over eight stereogenic centers.82 This is
probably the most spectacular example showing the syner-
gistic action of skeleton and stereochemical complexities in
the synthesis of natural products.
Equation 681
Equation 782
SYNTHESIS
The overall complexity of a structure is the result of these
three subcomponents. One should strive toward a treat-
ment making possible a semi-quantitative treatment of ho-
losynthons whose quality derives from their ability to un-
dergo one-pot reactions which drastically improve their
complexity in terms not only of skeleton but also of func-
tionality and stereochemistry, or any combination of the
three.
5.2 Good and Less Good Holosynthons
Fukumoto’s approach of hirsutene53 shows that good
strategies yield efficient results without involving a ho-
losynthetic key step. Others, clearly centered on a holo-
synthetic key step, were not outstanding in terms of
overall yields (Section 4.2). This leads to a consideration
of which qualities are to be expected for good holosyn-
thons and good holosynthetic transformations. A good
Downloaded by: University of Texas at San Antonio. Copyrighted material.
holosynthon must be of easy access. If the pleasure of
having a spectacular key step is to be paid for by many
supplementary steps, the apparent benefit of the holosyn-
thetic transformation has been wasted before the very start
of the reaction. The holosynthetic step should have a yield
which is not lower than the average yield of the overall
synthesis. Since the holosynthetic step yield is the product
of the yields of its component elementary reactions, this
condition would eliminate many holosynthetic reactions
from the box “good holosynthetic reactions”.
Mother Nature is expected to use the concept in biosyn-
thetic pathways if it is really efficient, and one should find
some examples of holosynthons in the biosynthesis of nat-
ural products. This is indeed the case. The enzyme cata-
lyzed conversion of squalene oxide to the plant triterpe-
noid dammaradienol (equation 8) illustrates this point
quite explicitly.83
Equation 883
Even more striking in this example is that this holosynthetic
transformation is enzymatically catalyzed. Nature makes a
formidable challenge to the chemists devoted to the synthe-
sis of natural products: to be able to reach the complexity of
these compounds by a series of catalytic reactions leading
to almost quantitative overall yields. Biosynthetic path-
ways, which are less and less studied because they demand
too much work, should pave the way to a new generation of
strategies dominantly biomimetic. The key to complexity
in natural products is, however, not necessarily enzymati-
cally driven as the formation of endiandric acid in nature
from polyunsaturated precursors proves.84
November 1998 Information Theory Description of Synthetic Strategies in the Polyquinane Series 1581

On the other hand, the critical examination of various holo- Table 2. Similarity of cyclic skeletons
synthons in this work suggests that the strategy yielding the
most impressive overall yield is not necessarily holosyn-
thetic in essence. A critical question before selecting an ef-
ficient strategy remains “would this kind of target be adapt-
ed to a holosynthetic approach?”. The question is still diffi-
cult to answer because new holosynthetic strategies keep
appearing in the litterature with a high frequency.

Appendix
Our approach to similarity is comparable to Bertz’s
complexity4 and adapted to be an efficient monitor of
skeletal differences for compounds containing only car-
bons and hydrogens. Therefore for compounds like corio-
lin the target is simplified by replacing all the functional-
ities by the appropriate number of hydrogens. Four struc-
tural parameters have been selected to describe every
intermediate in the synthetic tree:

Downloaded by: University of Texas at San Antonio. Copyrighted material.

(1) Number of carbon atoms present in the principal skeleton
(2) Types of bonds
(3) Types of rings
(4) Types of bicyclic sub-units. Table 3. Similarity of polycyclic skeletons (* see Fig. 19)

Figure 19. Types of bicyclic compounds.

To make clear the calculation we have treated typical ex-

amples in Tables 1–4. Every structural parameter yields a
value of similarity and the average of the four values in
taken as an overall index. To calculate the average two
methods are available. The first is just to make the ratio of
the number of relevant elements (for example atoms) be-
tween the two compared molecules. If one compares one
molecule with three carbon atoms in its skeleton with an-
other having four, the similarity index associated with this
element of comparison will be 3/4 × 100 = 75%:
Table 1. Similarity of acyclic skeletons

The second method is to add the number of elements

shared by the two molecules and to divide the obtained
number by the total number of elements. The preceding
case treated according to this scheme would yield a simi-
larity of 6/7 × 100 = 86%:
If one had compared a compound containing two rings
with one having three this second method would have giv-
en (2+2)/5 × 100 = 80%:
The first type of calculation has been adopted for the cal-
culation of similarity associated with the number of car-
1582 M. Chanon, R. Barone, C. Baralotto, M. Julliard, J. B. Hendrickson
Table 4. Similarity of polycyclic skeletons (* see Fig. 19)
bon atoms in the skeleton. The second type has been
adopted for the calculation of similarity associated with
the number of bonds and the number of rings. Since the
similarity is centered on the carbon skeleton, double
bonds are considered as totally equivalent to simple bonds
in the skeleton. This comes to considering that a totally
saturated skeleton is similar to the same skeleton in which
one or several CH2CH2 fragments have been replaced by
CHCH fragments. Heteroatoms present in the structure
are simply suppressed. The consequence is that the moni-
tor of similarity is blind to rings containing heteroatoms;
these rings become simple appendages linked to the car-
bon skeleton (epoxides are not seen).
The main merit of this crude similarity modelling is its “back
of an envelope” character. It was compared with much more
elaborate approaches in the case of coriolin. The results ob-
tained with these more elaborate models did not bring much
more insight for the perspective considered here.
We thank Dr. C. Meyer from Tripos Inc. for applying the algorithm
“Unity” from Sybyl to coriolin and its precursors) and Professor G.
Maury for stimulating discussions.
(1) a) Bertz, S. H. J. Am. Chem. Soc. 1981, 103, 3599.
b) Bertz, S. H.; Sommer, T. J. in Organic Synthesis: Theory and
Application, Ed. T. Hudlick, Jai Press Inc.: N.Y. 1993, Vol. 2,
p. 67–92.
(2) Shannon, C. A.; Weaver, W. The Mathematical Theory of Com-
munication, University of Illinois Press: 1949, p. 18.
(3) Balaban, A. T., Ed. Chemical Applications of Graph Theory,
Academic: London, 1976. Available on request.
SYNTHESIS
(4) Hendrickson, J. B.; Huang, P.; Toczko, A. G. J. Chem. Inf.
Comput. Sci. 1987, 27, 63. Available on request.
(5) Mehta, G. Chem. Rev. 1997, 97, 671.
(6) Hendrickson, J. B.; Braun-Keller, E. J. Comput. Chem. 1980,
1(4), 323.
Jochum, C.; Gasteiger, J.; Ugi, I. Angew. Chem. 1980, 92, 503;
Angew. Chem., Int. Ed. Engl. 1980, 19, 495.
Wochner, M.; Brandt, J.; von Scholley, A.; Ugi, I. Chimia 1988,
6, 217.
Ugi, I.; Bauer, J.; Bley, K.; Dengler, A.; Dietz, A.; Fontain, E.;
Gruber, R.; Knauer, M.; Reitsam, K.; Stein, N. Angew. Chem.
1993, 105, 210; Angew. Chem., Int. Ed. Engl. 1993, 32, 201.
(7) Denis, J. D.; Greene, A. E.; Guénard, D.; Guéritte-Voegelein,
F.; Mangatal, L.; Potier, P. J. Am. Chem. Soc. 1988, 110, 5917.
(8) Wipke, W.T.; Rogers, D. J. Chem. Inf. Comput. Sci. 1984, 24, 71.
(9) Chastrette, M.; Rajzmann, M.; Chanon, M.; Purcell, K. F. J. Am.
Chem. Soc. 1985, 107, 1.
(10) Hendrickson, J. B.; Chanon, M.; Chastrette, M. to be published
(11) Atlan, H. Organisation Biologique et Théorie de l'Information,
Hermann, Paris, 1992.
(12) Hendrickson, J. B.; Braun-Keller, E.; Toczko, G. A. Tetrahe-
Downloaded by: University of Texas at San Antonio. Copyrighted material.
dron 1981, 37, 359.
Hendrickson, J. B. Angew. Chem. 1990, 102, 1328; Angew.
Chem., Int. Ed. Engl. 1990, 29, 1286.
Hendrickson, J. B. ACS Symposium Series 1996, 626, 214.
Hendrickson, J. B.; Huang, P. J. Chem. Inf. Comput. Sci. 1989,
29, 145.
(13) Barone, R.; Chanon, M. Heterocycles 1981, 16, 1357.
Winter, J. H. J. Chem. Inf. Comput. Sci. 1985, 25, 389.
Hamm, P.; Jauffret, P.; Kaufmann, G. Rec. Trav. Chim. pays
Bas 1992, 111, 317.
Mehta, G.; Barone, R.; Azario, P.; Barberis, F.; Arbelot, M.;
Chanon, M. Tetrahedron 1992, 48, 8953.
Wender, P. A.; Miller, B. L. Organic Synthesis Theory and Ap-
plications. Hudlicky, T., Ed., JAI Press: 1993, 27.
Barberis, F.; Barone, R.; Chanon, M. Tetrahedron 1996, 52, 14625.
(14) Schelhaas, M.; Waldmann, H. Angew. Chem. 1996, 108, 2192;
Angew. Chem., Int. Ed. Engl. 1996, 35, 2056.
(15) Complexity being a probabilistic dimension, one could count ste-
reochemical complexity in these terms. A structure with one con-
trolled asymmetric center would have a stereochemical complexity
of 21.With two controlled asymmetric centers, this complexity
would be 22. And so on, taking into account the classical rules of
counting stereoisomers (Eliel, E.L.,Wilen,S. H. Stereochemistry of
Organic Compounds Wiley: New York 1994. p 50).
(16) Ihlenfedt, W-D.; Gasteiger, J. Angew. Chem. 1995, 107, 2807;
Angew. Chem., Int. Ed. Engl. 1995, 34, 2613.
(17) Hanessian, S.; Franco, J.; Gagnon, G.; Laramée, D.; Larouche,
B. J. Chem. Inf. Comput. Sci. 1990, 30, 413.
(18) Willet, P. Chemometrics Series, Bawden D., Letchworth, 1987.
J. Chem. Inf. Comput. Sci. 1992, 32, all issue 6.
Herndon, W. C.; Bertz, S. H. J. Comput. Chem. 1987, 8, 367.
Petitjean, M. J. Comput. Chem. 1995, 16, 80.
Hall, L. H.; Kier, L. B.; Brown, B. B. J. Chem. Inf. Comput. Sci.
1995, 35, 1074.
Cheng, C.; Maggiora, G.; Lajiness, M.; Johnson, M. J. Chem.
Inf. Comput. Sci. 1996 36, 909.
Thorner, D. A.; Wild, D. J.; Willett, P.; Wright, P. M. J. Chem.
Inf. Comput. Sci. 1996, 36, 900.
(19) Lehn, J. M. Supramolecular Chemistry, VCH: Weinheim, 1995.
(20) a) Wharton, S. A.; Martin, S. R.; Ruigrok, R. W. H.; Skehel, J.
J.; Steinhauser, D. A.; Wiley D. C., EMBO J. 1995, 14, 240.
b) Gallaher, W. R.; Ball, J. M.; Garry, R. F.; Martin-Amedee, A.
M.; Montelaro, R. C. Aids Res. Hum. Retroviruses, 1995, 11, 191.
(21) Hudlicky, T.; Chem. Rev. 1996, 96, 3.
(22) Paquette, L. A. Top. Curr. Chem. 1979, 79, 41.
Paquette, L. A. Top. Curr. Chem. 1984, 119, 1.
Trost, B. M. Chem. Soc. Rev. 1982, 11, 141.
November 1998 Information Theory Description of Synthetic Strategies in the Polyquinane Series
(23) Trost, B. M.; Science, 1991, 254, 1471.
Trost, B. M. Angew. Chem. 1995, 107, 285; Angew. Chem., Int.
Ed. Engl. 1995, 34, 259.
(24) Crimmins, M. T.; Mascarella, S. W. J. Am. Chem. Soc. 1986,
108, 3435.
(25) Sternbach, D. D.; Hugues, J. W.; Burdi, D. F.; Banks, B. A.
J. Am. Chem. Soc. 1985, 107, 2149.
(26) Wender, P. A.; Ternansky, R. J. Tetrahedron Lett. 1985, 26,
2625.
(27) Paquette, L. A.; Leone-Bay, A. J. Am. Chem. Soc. 1983, 105,
7352.
(28) Tsunoda, T.; Kodama, M.; Itô, S. Tetrahedron Lett. 1983, 24, 83.
(29) Franck-Neumann, M.; Miesch, M.; Gross, L. Tetrahedron Lett.
1991, 32, 2135.
Miesch, M.; Miesch, L.; Franck-Neumann, M. Tetrahedron
1997, 53, 2103
(30) Yamamura, S.; Ohkubo, M.; Shizuri, Y. Tetrahedron Lett.
1989, 30, 3798.
(31) Koteswar Rao, Y.; Nagarajan, M. Tetrahedron Lett. 1988, 29, 107.
(32) Bertz, S. H.; Herndon, W. C. in “Artificial Intelligence. Appli-
cations in Chemistry", Pierce, T. H.; Hohne, B. A. Ed. ACS
Symposium Series, American Chemiscal Society: Washington,
D.C. 1986, 306, 169.
(33) a) Seto, H.; Fijimoto, Y.; Tatsuno, T.; Yoshioka, H. Synth. Com-
mun. 1985, 15, 1217.
b) Kasturi, T. R.; Chandra, R. J. Org. Chem. 1988, 53, 3178.
c) Hudlicky, T.; Fleming, A.; Radesca, L. J. Am. Chem. Soc.
1989, 111, 6691.
(34) Dickson, Jr. J. K. Fraser-Reid, B. J. Chem. Soc., Chem. Com-
mun. 1990, 1440.
(35) Norris, D.; Corrigan, J. F.; Sun, Y.; Taylor, N. J. Can. J. Chem.
1993, 71, 1029.
(36) Dauben, W. G.; Walker, D. M. J. Org. Chem. 1981, 46, 1103.
Snider, B. B.; Beal, R. J. Org. Chem. 1988, 53, 4508.
Fitjer L.; Majewski, M.; Monzó-Oltra, H. Tetrahedron 1995,
51, 8835.
(37) Tanabe, M.; Suzuki, K. T.; Jankowski, W. C. Tetrahedron Lett.
1974, 2271.
Feline, T. C.; Mellow, G.; Jones, R. B.; Phillips, L. J. Chem.
Soc., Chem. Commun. 1974, 63.
Nozoe, S.; Furukawa, J.; Sankawa, U.; Shibata, S. Tetrahedron
Lett. 1976, 195.
(38) Cohen, T.; McNamara, K.; Kuzemko, M. A.; Ramig, K.; Landi,
Jr. J. J.; Dong, Y. Tetrahedron 1993, 49, 7931.
(39) Hudlicky, T.; Kutchan, T. M. J. Am. Chem. Soc. 1980, 102,
6353.
(40) Iyoda, M.; Kushida, T.; Kitami, S.; Oda, M. J. Chem. Soc.,
Chem. Commun. 1986, 1049.
(41) Moriarty, K. J.; Shen, C. C.; Paquette, L. A. Synlett 1990, 263.
(42) Sternbach, D. D.; Ensinger, C. L. J. Org. Chem. 1990, 55, 2725.
(43) Funk, R. L.; Bolton, G. L. J. Org. Chem. 1984, 49, 5021.
(44) Curran, D. P.; Rakiewicz, D. M. Tetrahedron 1985, 41, 3958.
(45) Franck-Neumann, M.; Miesch, M.; Lacroix, E.; Metz, B.; Kern,
J. M. Tetrahedron 1992, 48, 1911.
(46) Rawal, V. H.; Fabre, A.; Iwasa, S. Tetrahedron Lett. 1995, 36,
6851.
(47) Wender, P. A.; Howbert, J. J. Tetrahedron Lett. 1982, 23, 3983.
(48) Ley, S. V.; Murray, P. J.; Palmer, B. D. Tetrahedron 1985, 41,
4769.
(49) Cossy, J.; Belotti, D.; Pete, J. P. Tetrahedron 1990, 43, 1859.
(50) Disanayaka, B. W.; Weedon, A. C. J. Chem. Soc., Chem. Com-
mun. 1985, 1282.
(51) Hewson, A. T.; MacPherson, D. T. J. Chem. Soc., Perkin Trans.
I 1985, 2625.
(52) Little, R. D.; Higby, R. G.; Moeller, K. D. J. Org. Chem. 1983,
48, 3140.
(53) Toyota, M.; Nishikawa, Y.; Motoki, K.; Yoshida, N.; Fukumo-
to, K. Tetrahedron 1993, 49, 11189.
1583
(54) Mehta, G.; Murthy, A. N.; Reddy, D. S.; Reddy, A. V. J. Am.
Chem. Soc. 1986, 108, 3443.
(55) Magnus, P.; Quagliato, D. J. Org. Chem. 1985, 50, 1621.
(56) Majetich, G.; Defauw, J. Tetrahedron 1988, 44, 3833.
(57) Sarkar, T. K.; Ghosh, S. K.; Subba Rao, P. S. V.; Satapathi, T.
K.; Mamdapur, V. R. Tetrahedron 1992, 48, 6897.
(58) Castro, J.; Sorensen, A.; Riera, A.; Morin, C.; Moyano, A.; Per-
iras, M. A.; Greene, A. E. J. Am. Chem. Soc. 1990, 112, 9388.
(59) Hua, D. H.; Sinai-Zindge, G.; Venkataraman, S. J. Am. Chem.
Soc. 1985, 107, 4088.
Hua, D. H.; Venkataraman, S.; Ostander, R. A.; Sinai, G.-Z.; Mc-
Cann, P. J.; Coulter, M. J.; Xu, M. R. J. Org. Chem. 1988, 53, 507.
Inoue, T.; Hosomi, K.; Araki, M.; Nishide, K.; Node, M. Tetra-
hedron Asymm. 1995, 6, 31.
Weinges, K.; Reichert, H.; Huber-Patz, U.; Irngartinger, H. Lie-
bigs Ann. Chem. 1993, 403.
(60) Oppolzer, W.; Robyr, C. Tetrahedron 1994, 35, 415.
(61) Iseki, K.; Yamazaki, M.; Shibasaki, M.; Ikegami, S. Tetrahe-
dron 1981, 37, 4411.
(62) Ito, T.; Tomiyoshi, N.; Nakamura, N.; Azuma, S.; Izawa, M.
Tetrahedron Lett. 1982, 23, 1721.
Downloaded by: University of Texas at San Antonio. Copyrighted material.
(63) Fevig, T. L.; Elliott, R. L.; Curran, D. P. J. Am. Chem. Soc.
1988, 110, 5064.
(64) Tatsuta, K.; Akimoto, K.; Kinoshita, M. Tetrahedron 1981, 37,
4365.
(65) Trost, B. M.; Curran, D. P. J. Am. Chem. Soc. 1981, 103, 7380.
(66) Schuda, P. F.; Heimann, M. R. Tetrahedron 1984, 40, 2365.
(67) Wender, P. A.; Howbert, J. J. Tetrahedron Lett. 1983, 24, 5325.
(68) Danishefsky, S.; Zamboni, R.; Kahn, M.; Etheredge, S. J. J. Am.
Chem. Soc. 1981, 103, 3460.
(69) Domon, K.; Masuya, K.; Tanino, K.; Kuwajima, I. Tetrahedron
Lett. 1997, 38, 465.
(70) Singh, V.; Thomas, B. J. Org. Chem. 1997, 62, 5310.
(71) Chanon, M.; Barone, R. “Computer Aids in Chemistry", Ver-
non, G.; Chanon, M. Eds. E. Horwood: Chichester, 1986, Chap-
ter 1, p. 68.
(72) Nicolaou, K. C.; Xiao, X.-Y.; Parandoosh, Z.; Senyei, A.; Nova,
M. P. Angew. Chem. 1995, 107, 2476; Angew. Chem.; Int. Ed.
Engl. 1995, 34, 2289.
(73) Posner, G. H. Chem. Rev. 1986, 86, 831.
(74) Ho, T. “Tandem Reactions", Wiley, N. Y., 1992.
(75) Hoffmann, H. M. R. Angew. Chem. 1992, 104, 1361; Angew.
Chem., Int. Ed. Engl. 1992, 31, 1332.
(76) Tietze, L. F.; Beifuss, U. Angew. Chem. 1993, 105, 137; Angew.
Chem., Int. Ed. Engl., 1993, 32, 131.
Tietze, L. F. J. Heterocycl. Chem. 1990, 27, 47.
Ziegler, F. Chem. Rev. 1988, 88, 1423.
(77) Chem. Rev. 1996, 96, n˚ 1, Wender, P. A. Ed.
(78) Nicolaou, K. C.; Sorensen, E. J."Classics in Total Synthesis,
VCH. Weinheim, 1996.
Corey, E. J.; Cheng, X.-M. The Logic of Chemical Synthesis,
Wiley: New York, 1989.
(79) Allerton, R.; Fletcher, H. G. J. Am. Chem. Soc. 1954, 76, 1757.
(see also ref. 74 p. 57)
(80) Gras, J. L.; Pelissier, H.; Nouguier, R. J. Org. Chem. 1989, 54,
5675.
(81) Bogen, S.; Malacria, M. J. Am. Chem. Soc. 1996, 118, 3992.
Bogen, S.; Fensterbank, M.; Malacria, M. J. Am. Chem. Soc.
1997, 119, 5037.
(82) Nicolaou, K. C.; Petasis, N. A. in Strategies and Tactics in Or-
ganic Synthesis, Lindberg, T. Ed. Academic: San Diego, 1984,
vol. 1, p. 155.
(83) Ref. 78 (a), chapter 6.
(84) Bandaranayake, W. M.; Banfield, J. E.; Black, D. St. C.; Fallon,
G. D.; Gatehouse, B. M. J J. Chem. Soc., Chem. Commun. 1980,
162.
Bandaranayake, W. M.; Banfield, J. E.; Black, D. St. C.
J. Chem. Soc., Chem. Commun. 1980, 902.