Observations on the Influence of Precursor Conformations on Macrocyclization Reactions

Article abstract of DOI:10.1002/ejoc.201501445

Macrocycles hold great promise in drug discovery as an underutilized class of lead compounds. The low abundance of these molecules can, in part, be explained by the inherent difficulties in the synthesis of macrocycles and the lack of general methods for their rapid assembly. We have undertaken a research program aimed at developing methods for facile synthesis of macrocycles from simple precursors. The synthesis of two new cyclization precursors is described and the results of their reaction with thionyl chloride are presented and discussed. Whereas one acyclic diol smoothly underwent macrocyclization to afford a mixture of diastereomeric sulfites, subjection of the other precursor to identical reaction conditions resulted in the isolation of the linear dichloride. We hypothesize that there is a difference in the ability of the two molecules to adopt a conformation that is germane to macrocyclization, a proposition that is supported by conformational analyses using molecular mechanics. Macrocycles hold great promise in drug discovery; however, their synthetic generation through macrocyclization is nontrivial. Here, the synthesis of two new cyclization precursors is described and the results of the ensuing cyclization is reported and evaluated. Molecular modeling is used to explain the observed results and provide guidance for further work in this field.

Full text of DOI:10.1002/ejoc.201501445

DOI: 10.1002/ejoc.201501445
Full Paper
Macrocycle Cyclization
Observations on the Influence of Precursor Conformations on
Macrocyclization Reactions
Peter Hammershøj,[a,b] Klavs Beldring,[a,b] Anders R. Nielsen,[a,b] Peter Fristrup*[b] and
Mads H. Clausen*[a,b]
Abstract: Macrocycles hold great promise in drug discovery as cussed. Whereas one acyclic diol smoothly underwent macrocy-
an underutilized class of lead compounds. The low abundance
of these molecules can, in part, be explained by the inherent
difficulties in the synthesis of macrocycles and the lack of gen-
eral methods for their rapid assembly. We have undertaken a
research program aimed at developing methods for facile syn-
thesis of macrocycles from simple precursors. The synthesis of
two new cyclization precursors is described and the results of
their reaction with thionyl chloride are presented and dis-
clization to afford a mixture of diastereomeric sulfites, subjec-
tion of the other precursor to identical reaction conditions re-
sulted in the isolation of the linear dichloride. We hypothesize
that there is a difference in the ability of the two molecules to
adopt a conformation that is germane to macrocyclization, a
proposition that is supported by conformational analyses using
molecular mechanics.
Introduction
rings.[12] The success of using simple force-field based modeling
on the conformations of small and medium-sized rings was al-
ready achieved in the early and mid-1980s by Still and co-work-
ers,[13] and molecular mechanics calculations have been applied
to predict the most productive precursors of cyclic peptides.[14]
More recently, conformational and configurational preorganiza-
tion in macrocyclization has been studied, for example, folding
of oligoamides,[15] synthesis of cycloparaphenylens,[16] and
peptidomimetic macrocycles.[17]
To address the dearth of novel macrocyclic scaffolds for
screening campaigns, we have undertaken a research program
aimed at developing methods for the facile synthesis of macro-
cycles from simple precursors.[18,19] In the course of these inves-
Macrocycles are an underutilized class of molecules in drug dis-
covery,[1,2,3] despite their unique ability to address difficult tar-
gets,[4] such as protein–protein interactions, a fact that can be
attributed to the inherent difficulty in preparing large rings.[5]
Macrocyclic peptides and peptidomimetics are a subclass of
macrocycles, the synthesis[6] and easy of cyclization[7,8] of which
have been topics of excellent recent papers. The macrocycliza-
tion chemistry of particular challenging cyclophane natural
products has similarly been highlighted[9] and the efficiency of
such reactions in terms of the relation between chemical yield
and concentration was also addressed lately.[10]
To harvest the potential of macrocycles for screening cam-
paigns, one has to face the challenges posed by this compound
class: the chemistry of cyclization is also the chemistry of oli-
gomerization, which often means that high dilution conditions
are required. Furthermore, macrocyclization reactions are
known to be sensitive to conformational effects,[11] something
that is studied employing computational methods in this manu-
script. Indeed, the importance of configurational, conforma-
tional, and template-induced preorganization in macrocycliza-
tion reactions has been the topic of a number of recent reports
and was reviewed in 2015 by Luis and co-workers.[12]
In recent years, several groups have contributed towards ra-
tionalizing the success or failure of macrocyclizations of large
[a] Center for Nanomedicine and Theranostics, Technical University of
Denmark,
Building 207, 2800 Kgs. Lyngby, Denmark
http://www.kemi.dtu.dk/MHC
[b] Department of Chemistry, Technical University of Denmark,
Building 207, 2800 Kgs. Lyngby, Denmark
http://www.kemi.dtu.dk/PeterFristrup
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under http://dx.doi.org/10.1002/ejoc.201501445.
Scheme 1. Previously synthesized macrocyclic sulfites 2, 5, and 6, and the
variation in yields for acyclic diols 1, 3, and 4.
Eur. J. Org. Chem. 2016, 1533–1540
1533
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tigations, we have produced acyclic diols and tested their pro- looking for trends regarding the precyclization conformations
pensity to undergo macrocyclization, choosing the formation that could potentially be used to predict the propensity of sys-
of sulfites as a representative reaction. Examples of successful tems to macrocyclize efficiently. In this report, we discuss the
cyclizations of this type are shown in Scheme 1: diol 1 cyclizes synthesis of two precursors that show very different reactivity
in impressive 79 % yield to give sulfite 2 by using our standard in the formation of cyclic sulfites as a model reaction and dis-
conditions: dropwise addition of thionyl chloride to a 10 m
M
close the results of molecular mechanics simulations of their
solution of diol in dichloromethane, triethylamine, 4-(dimethyl- conformations.
amino)pyridine (DMAP).[19] Similarly, the ester 3 formed from 6-
hydroxycaproic acid cyclizes to form 5 in 55 % yield; however,
Results and Discussion
a similar ester from a sorbic acid derivative 4 was considerably
less willing to form the cyclic sulfite, which could be obtained
in just 25 % yield.[18]
Synthesis
To study the synthesis of different macrocycles, two novel build-
ing blocks, 11 and 21, were synthesized (Scheme 2).
To unlock the potential of macrocyclic scaffolds in lead dis-
covery it is important to develop general conditions for cycliza-
tion reactions, which is why we became interested in studying
A Grignard reaction with 2-{[(4-methoxybenzyl)oxy]methyl}-
the cyclization of different precursors. Furthermore, we were benzaldehyde (7) and allylmagnesium chloride afforded a race-
Scheme 2. Synthetic route to building blocks 11 and 21.
Eur. J. Org. Chem. 2016, 1533–1540
www.eurjoc.org
1534
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
mic mixture of 8.[20,21] Protection of the hydroxyl group was the same procedure described for 14 → 15, yielding aldehyde
achieved by using TIPSOTf, and 9 could be transformed into 18. Introduction of an alkyne moiety was achieved by using the
the corresponding alcohol 10 through anti-Markovnikov addi- commercially available tert-butyldimethyl(prop-2-yn-1-yloxy)-
tion[22] of a hydroxyl group to the alkene by addition of 9- silane and BuLi in THF at –78 °C, forming the reactive alkynelith-
borabicyclo[3.3.1]nonane (9-BBN) followed by hydrogen perox- ium reagent. The addition to give 19 afforded a single dia-
ide.[23] The newly formed alcohol 10 was first subjected to stereoisomer in 92 % yield, which was shown to be the Felkin
Swern oxidation conditions[24] to afford the corresponding product[28] by treatment of 19 with aqueous HCl in acetone,
crude aldehyde, followed by oxidation to acid 11 using NaClO2, deprotecting the silyl ethers, and generating the acetonide of
NaH2PO4, and 2-methylbut-2-ene[25] in 87 % yield.
the vicinal diols. The product was analyzed by NMR spectro-
scopy, and a NOESY spectrum confirmed the cis orientation of
the protons in the five-membered acetonide. Alcohol 19 was
protected with TIPS to give 20, which was then treated with
2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ)[29] to remove
the PMB group and provide 21, which was obtained in an over-
all yield of 16 % over eight steps.
The synthetic route to 21 started with treatment of commer-
cially available (4-methoxyphenyl)methanol (12) and 6-bromo-
hex-1-ene (13) with NaH in N,N-dimethylformamide (DMF).
After protection of the hydroxyl group with PMB, 14 was trans-
formed into aldehyde 15 in a one-pot reaction via the diol by
treatment first with OsO4/N-methylmorpholine N-oxide
(NMO),[26] followed by cleavage of the diol with NaIO4.[27]
A
The two exemplary macrocyclic target molecules were carb-
Grignard reaction using vinylmagnesium bromide in tetrahydro- onate 26 and sulfite 28 (Scheme 3). Starting from 11 and 21,
furan (THF) at –78 °C afforded the racemic mixture of alcohol the diastereomeric mixture of 22 was generated by using an
16.[20] After protection of the hydroxyl group with TIPS to afford equivalent amount of DMAP and an excess of EDCl. The primary
17, the alkene group was once more transformed by following alcohols in 22 then needed to be deprotected. Treatment with
Scheme 3. Synthetic route to macrocyclic carbonate 26 and sulfite 28 from compounds 11 and 21.
Eur. J. Org. Chem. 2016, 1533–1540
www.eurjoc.org
1535
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 4. Synthetic route to diol 34 and attempt to produce macrocyclic sulfite 35.
tetrabutylammonium fluoride (TBAF) under mild conditions in using TBAF in THF. Thus, the newly formed alcohol 31 was first
AcOH/THF led to selective deprotection of the TBDMS-pro- subjected to the Swern oxidation conditions to afford the corre-
tected hydroxyl group, yielding alcohol 23. The first part of the sponding crude aldehyde, followed by oxidation to acid 32 us-
synthetic route towards the target macrocyclic carbonate 26 ing NaClO2, NaH2PO4, and 2-methylbut-2-ene. Esterification of
was started with alcohol 23, which was treated with p-nitro- acid 32 with 4-(4-{[(tert-butyldimethylsilyl)oxy]methyl}-1H-1,2,3-
phenyl chloroformate and DMAP to form carbonate intermedi- triazol-1-yl)butan-1-ol was achieved with EDCl and DMAP in
ate 24 in high yield. Before the first cyclization attempt, the CH2Cl2 at 5 °C, affording ester 33. Before the first attempt to
other primary alcohol was unmasked by treatment with DDQ. cyclize to sulfite 35, the two hydroxyl groups had to be depro-
Base-catalyzed intramolecular cyclization of primary alcohol 25 tected; this was achieved by treatment of ester 33 with PPTS in
gave macrocyclic carbonate 26 in 76 % yield (40 % overall yield EtOH, affording the key intermediate diol 34. Under the same
starting from 11 and 21). The second part of the synthetic route reaction conditions described for the cyclization reaction
leads to sulfite 28 utilizing the same key intermediate com- 27 → 28, diol 34 did not undergo cyclization, affording instead
pound 23. Before cyclization to the sulfite, the second primary a dichlorinated product (S17, see the Supporting Information).
alcohol in 23 was deprotected with the same procedure de-
scribed for 24 → 25. Thus, treatment of the isolated diol 27
with an excess of thionyl chloride and triethylamine with a cata-
lytic amount of DMAP gave macrocyclic sulfite 28 in an overall
Molecular Modeling
yield of 43 % from building blocks 11 and 21.
With the rapid development of computational capability in the
The second synthetic route towards macrocyclic target mol- last two decades in mind, we speculated that it might be feasi-
ecules sulfite 35 was chosen to introduce a rigid biphenyl motif ble to generate a large number of possible conformations for
into the macrocycle structure; the approach is shown in the large and flexible cyclization precursors studied here. With
Scheme 4. The synthesis started with a one-pot boronation– the experimental data for the successful macrocyclizations of
Suzuki cross-coupling reaction initiated by in situ formation of diol precursors 1, 3, and 4 available, we carried out a simple
the alkylborate of 2-[(3-bromobenzyl)oxy]tetrahydro-2H-pyran conformational search using the OPLS-2005 force-field[31] and a
(29) followed by addition of 4-(bromobenzyloxy)(tert-butyl)di- GB/SA solvation model[32] for water as incorporated in Macro-
methylsilane along with Pd(OAc)2 and Ba(OH)2·8H2O, affording model version 9.6.[33] The conformational search was carried
biphenyl 30.[30] Selective deprotection of the TBDMS protected out by using a combination of the Monte Carlo multiple mini-
hydroxyl group was achieved in close to quantitative yield by mum (MCMM) algorithm[34,35] and the Low-Mode search algo-
Eur. J. Org. Chem. 2016, 1533–1540
www.eurjoc.org
1536
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
rithm.[36] We found conformations with a hydrogen bond estab-
lished between the two terminal OH as the global minimum
for all three compounds, indicating that macrocyclization is not
prohibited by strain (structures and energies included in the
Supporting Information).
Although this was a very satisfactory and interesting first re-
sult, we were interested in constructing a computational ap-
proach that could guide our synthetic efforts. We speculated
that a long conformational search (10,000 steps) with a large
energy window of 100 kJ mol–1 would allow the generation of
a representative conformational ensemble (typically 5,000 to
7,000 structures). From this ensemble we could then compare
both the number and the relative energy of the conformations
amendable to cyclization (i.e., short end-to-end distance) in
comparison to the data gathered from the entire ensemble (en-
ergy vs. end-to-end distance). Ultimately, our goal is to be able
to use such a conformational search as a first, rough indication
of whether or not the chosen backbone is suitable for synthesis.
Application of this extended conformational search con-
firmed the existence of a stabilizing hydrogen bond, and this
conformation was found to be the global minimum for all three
compounds (1, 3, and 4). We carried out the same conforma-
tional search for the two possible intermediates where one of
the alcohols had reacted with SOCl2 to form a SOCl moiety. This
analysis was carried out for all six possible intermediates
formed from diols 1, 3, and 4, and, although the global mini-
mum did not always have a short end-to-end distance, there
were always conformations with energies only a few kJ/mol
higher than the global minimum that did (see the Supporting
Information for details).
In this first study, we apply the methodology to two series
of cyclization precursors, one that undergoes smooth macrocy-
clization and one that does not cyclize. Although it is clearly
advantageous to limit the computational investigation to the
cyclization precursors (diols in this study), we also include the
two possible intermediates in which the cyclization reagent
(SOCl2) has reacted with one of the two possible termini as has
been done for the previously studied molecules 1, 3, and 4.
Figure 1. (A) Results from the conformational search of 27. (B) Global mini-
mum obtained in the conformational search of 27. (C) Conformation no. 3,
with a relative energy of 1.5 kJ mol–1 and a hydrogen bond between the two
alcohols. Images were constructed with the XYZ-viewer.[37]
The cyclization precursor, in this case both of the stereoiso-
mers of diol 27, was built and subjected to a conformational
search for 10,000 steps. Given that both stereoisomers gave
very similar results, we only include the results from the (R)-
enantiomer here. In total, 7,052 conformations were generated
and the relative energy was plotted as a function of the O–O
distance of the two free alcohols (Figure 1, A).
gen–oxygen distance in the final, cyclized sulfite macrocycle
(ca. 2.6 Å). Results from the other stereoisomer were similar
except for the fact that here even the global minimum had a
conformation with short oxygen–oxygen distance (both stereo-
isomers are included in the Supporting Information).
Whereas the global minimum was characterized by having a
For compound 34, a similar conformational search was car-
rather long distance between the two free alcohols (11.5 Å, ried out that resulted in 5,190 structures, wherein the global
Figure 1, part A, red square and part B), the conformation with minimum had an O–O distance of 7.8 Å (Figure 2, A and B).
third lowest energy had a very short oxygen–oxygen distance Here, none of the most energetically favored structures had a
(1.9 Å, Figure 1, part A green triangle and part C). This clearly similar short O–O distance, and only two structures could be
shows that a structure that is preorganized for cyclization is found in total. These two had O–O distances of 2.9 Å and rela-
indeed energetically accessible at room temperature, thus pro- tive energies of 7.4 (Figure 2, C) and 13.8 kJ mol–1, compared
viding theoretical support for the high yield obtained in the with the global minimum, respectively. A simple two-state
cyclization reaction. A simple two-state Boltzmann expression Boltzmann expression shows that an energy difference of
shows that an energy difference of 1.5 kJ mol–1 corresponds 7.4 kJ mol–1 corresponds to approximately 5 % population at
to approximately 35 % population at room temperature. The room temperature. However, judging by the graph it is clear
oxygen–oxygen distance of ca. 1.8 Å is shorter than the oxy- that there are many more low-energy conformations with long
Eur. J. Org. Chem. 2016, 1533–1540
www.eurjoc.org
1537
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
O–O distances, so the real population of conformations with and 34, which have the same functional groups with similar
short O–O distances will be much lower.
reactivities (benzylic and propargylic alcohols for 27, benzylic
and heterobenzylic for 34), reacted under identical conditions
to give different products, it seems plausible that the major
reason why 27 underwent cyclization and 34 was instead chlor-
inated is the ease with which the precursors can adopt a confor-
mation that is germane to cyclization.
With this computational methodology in hand we also went
back and analyzed the cyclizations of precursors 1, 3, and 4
reported previously. Here, 1 cyclized in impressive 79 % yield
to sulfite 2. Similarly, ester 3, formed from 6-hydroxycaproic
acid, cyclized to form 5 in 55 % yield – but a similar ester from
a sorbic acid derivative 4 was considerably less willing to form
the cyclic sulfite, which could be obtained in only 25 % yield.
The conformational search was carried out on precursor 1,
which resulted in 7,379 conformations. Of these conformations,
92 were preorganized for cyclization in that they had a
Figure 2. (A) Results from the conformational search of 34. (B) Global mini-
mum obtained in the conformational search of 34. (C) Conformation no.
1314, which has a relative energy of 7.4 kJ mol–1 and a hydrogen bond be-
tween the two free alcohols.
When considering that carbonate 26 (and the sulfite ana-
logue 28) were formed successfully, whereas diol 34 failed to
cyclize, it is interesting to attempt to draw some general conclu-
sions regarding the feasibility for cyclization as a consequence
of the conformational energies.
For compound 27, nine conformations with short O–O dis-
tances (<3 Å) exist that are within 10 kJ mol–1 of the global
minimum. Furthermore, for 27, we found a total of 249 struc-
tures with very short O–O distances (<3 Å) indicative of a pro-
ductive ring-closing reaction. For 34, only two conformations
that satisfy the constraint of having a short O–O distance within
the entire ensemble of ca. 5,000 structures were generated dur-
ing the conformational search. Given that the precursors 27
Figure 3. (A) Results from the conformational search of 1. (B) Global minimum
obtained in the conformational search of compound 1. (C) Conformation no.
46 from the conformational search of 1, which has a relative energy of
3.5 kJ mol–1 and a hydrogen bond between the two alcohols.
Eur. J. Org. Chem. 2016, 1533–1540
www.eurjoc.org
1538
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
hydrogen bond between the two termini, which was confirmed Experimental Section
by a short oxygen–oxygen distance (<3 Å). Conformation no.
Formation of Macrocyclic Sulfite 28 and Dichloride 35. General
Procedure: The diol cyclization precursor (0.14 mmol) was dis-
solved in anhydrous CH2Cl2 (14 mL), and DMAP (9 mg, 0.07 mmol)
and Et3N (0.05 mL, 0.42 mmol) were added under nitrogen atmos-
phere. The reaction mixture was stirred and SOCl2 (11 μL,
0.16 mmol) dissolved in anhydrous CH2Cl2 (3 mL) was added drop-
wise over a period of 10 min at room temperature. After 2 h, addi-
tional SOCl2 (6 μL, 0.08 mmol) dissolved in anhydrous CH2Cl2 (1 mL)
was added. After 18 h, Et3N (16 μL, 0.14 mmol) and SOCl2 (6 μL,
0.08 mmol) dissolved in anhydrous CH2Cl2 (1 mL) were added and
the reaction mixture was stirred for an additional 2 h. After full
conversion (evaluated by TLC) the reaction mixture was evaporated
to dryness in vacuo. The residue was purified by column chroma-
tography (heptane/EtOAc) on silica gel to give the pure product.
46 was the most favorable of these and it had a relative energy
of 3.5 kJ mol–1 compared with the global minimum (see the
Supporting Information, Figure S1). This is in line with the ob-
served ease of cyclization and excellent yield (Figure 3).
For compound 3 (saturated) and 4 (unsaturated) a similar
approach was taken, which resulted in 7,996 and 6,787 confor-
mations, respectively (Figure S4). For both of these structures,
the global minimum displayed a conformation that was prone
to cyclization stabilized by a hydrogen bond interaction be-
tween the two alcohols (Figure S7). When we analyze the num-
ber of conformations that have a sufficiently short oxygen–oxy-
gen distance (<3.0 Å) we find 250 conformations for 3 com-
pared with 121 conformations for 4. The higher number of pre-
reactive conformations found for 3 provides a possible explana-
tion for the observed higher yield in the macrocyclization of 3
compared with 4.
Although the analysis performed here is rudimentary and
does not, for example, involve detailed analysis of the transition
state for the cyclization reaction, it does have the benefit of
being extremely fast and simple to perform. We recommend
that programs aimed at generating macrocyclic products, espe-
cially those in a parallel synthesis setting, include a simple anal-
ysis of cyclization precursors similar to that outlined herein. An
up-front investigation of the degrees of freedom involved, the
number of conformations amendable to cyclization and their
relative energies can give a good indication of whether a given
precursor should be included in a library design.
Acknowledgments
The authors thank the Novo Nordisk Foundation (Biotechnology
Based Synthesis and Production Research) and the Danish
Council for Independent Research [Sapere Aude research leader
grant (to P. F.); grant no. 11-105487] for financial support.
Keywords: Synthetic methods · Drug discovery ·
Cyclization · Molecular modeling · Conformation analysis ·
Macrocycles
[1] C. M. Madsen, M. H. Clausen, Eur. J. Org. Chem. 2011, 3107–3115.
[2] E. M. Driggers, S. P. Hale, J. Lee, N. K. Terrett, Nat. Rev. Drug Discovery
2008, 7, 608–624.
[3] E. Marsault, M. L. Peterson, J. Med. Chem. 2011, 54, 1961–2004.
[4] W. Brandt, V. J. Haupt, L. A. Wessjohann, Curr. Top. Med. Chem. 2010, 10,
1361–1379.
[5] L. A. Wessjohann, E. Ruijter, D. Garcia-Rivera, W. Brandt, Mol. Diversity
2005, 9, 171–186.
Conclusions
[6] C. J. White, A. K. Yudin, Nature Chem. 2011, 3, 509–524.
[7] A. Thakkar, T. B. Trinh, D. Pei, ACS Comb. Sci. 2013, 15, 120–129.
[8] J. Becerril, M. Bolte, M. I. Burguete, F. Galindo, E. García-España, S. V. Luis,
J. F. Miravet, J. Am. Chem. Soc. 2003, 125, 6677–6686.
[9] T. Gulder, P. S. Baran, Nat. Prod. Rep. 2012, 29, 899–934.
[10] J. C. Collins, K. James, MedChemComm 2012, 3, 1489–1495.
[11] a) J. Blankenstein, J. Zhu, Eur. J. Org. Chem. 2005, 1949–1964; b) V. Martí-
Centelles, M. D. Pandey, M. I. Burguete, S. V. Luis, Chem. Rev. 2015, 115,
8736–8834.
[12] V. Martí-Centelles, M. D. Pandey, M. I. Burguete, S. V. Luis, Chem. Rev.
2015, 115, 8736–8834.
[13] a) W. C. Still, I. Galynker, Tetrahedron 1981, 37, 3981–3996; b) W. C. Still,
I. Galynker, J. Am. Chem. Soc. 1982, 104, 1774–1776.
[14] F. Cavelier-Frontin, G. Pepe, J. Verducci, D. Siri, R. Jacquier, J. Am. Chem.
Soc. 1992, 114, 8885–8890.
[15] L. H. Yuan, H. Q. Zeng, K. Yamato, A. R. Sanford, W. Feng, H. Atreya, D. K.
Sukumaran, T. Szyperski, B. Gong, J. Am. Chem. Soc. 2004, 126, 16528–
16537.
[16] Y. Segawa, S. Miyamoto, H. Omachi, S. Matsuura, P. Senel, T. Sasamori, N.
Tokitoh, K. Itami, Angew. Chem. Int. Ed. 2011, 50, 3244–3248; Angew.
Chem. 2011, 123, 3302.
[17] M. Bru, I. Alfonso, M. I. Burguete, W. V. Luis, Tetrahedron Lett. 2005, 46,
7781–7785.
[18] M. J. Wingstrand, C. M. Madsen, M. H. Clausen, Tetrahedron Lett. 2009,
50, 693–695.
We have presented the synthesis of two diol precursors for cy-
clization to form, for example, sulfites. Although the alcohol
groups have comparable reactivity based on their stereoelec-
tronic properties, one was shown to afford a cyclic sulfite
smoothly, whereas the other precursor gave an acyclic dichlo-
ride exclusively under identical conditions. The propensity of
the precursors to adopt a conformation that was deemed fit for
a productive cyclization was examined through a long confor-
mational search (10,000 steps) that was carried out with a large
energy window of 100 kJ mol–1 to allow generation of a repre-
sentative conformational ensemble (typically 5,000 to 7,000
structures).
The proportion of conformations with a short end-to-end
distance was compared and was shown to be in line with the
observed experimental results. In addition, the conformational
space of the two possible intermediates in which one of the
alcohol groups had reacted with SOCl2 was investigated. Inter-
estingly, we found that this change did not result in major dif-
ferences in the overall distribution of conformers. Our observa-
tions underline the importance of conformational analysis for
macrocyclization reactions and offer a facile modeling method
for performing an initial examination of conformational space
of cyclization precursors.
[19] C. M. Madsen, M. Hansen, M. V. Thrane, M. H. Clausen, Tetrahedron 2010,
66, 9849–9859.
[20] a) V. Grignard, C. R. Acad. Sci. 1900, 1322.1324; b) V. Grignard, Ann. Chim.
1901, 7, 433–490.
Eur. J. Org. Chem. 2016, 1533–1540
www.eurjoc.org
1539
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[21]
[31] G. A. Kaminski, R. A. Friesner, J. Tirado-Rives, W. J. Jorgensen, J. Phys.
Chem. B 2001, 105, 6474–6487.
M. Jay-Smith, D. P. Furkert, J. Sperry, M. A. Brimble, Synlett 2011, 1395–
1398.
[32] W. C. Still, A. Tempczyk, R. C. Hawley, T. Hendrickson, J. Am. Chem. Soc.
1990, 112, 6127–6129.
[33] MacroModel, v. 9.9 from Schrödinger Inc.: F. Mohamadi, N. G. J. Richards,
W. C. Guida, R. Liskamp, M. Lipton, C. Caulfield, G. Chang, T. Hendrickson,
W. C. Still, J. Comput. Chem. 1990, 11, 440–467. For current versions, see:
http://www.schrodinger.com.
[22]
[23]
W. Markownikoff, Ann. Chim. 1870, 228–259.
a) R. Liotta, H. C. Brown, J. Org. Chem. 1977, 42, 2836–2839; b) H. C.
Brown, J. V. N. V. Prasad, J. Org. Chem. 1985, 50, 3002–3005.
[24]
[25]
D. Swern, A. J. Mancuso, Synthesis 1981, 165–185.
B. O. Lindgren, T. Nilsson, Acta Chem. Scand. 1973, 27, 888–890.
[26] V. VanRheenen, R. C. Kelly, D. Y. Cha, Tetrahedron Lett. 1976, 17, 1973–
[34] G. Chang, W. C. Guida, W. C. Still, J. Am. Chem. Soc. 1989, 111, 4379–
4386.
1976.
[27] B. Sklarz, Q. Rev. Chem. Soc. 1967, 21, 3–28.
[28] M. Chérest, H. Felkin, N. Prudent, Tetrahedron Lett. 1968, 9, 2199–2204.
[29] Y. Oikawa, T. Yoshioka, O. Yonemitsu, Tetrahedron Lett. 1982, 23, 885–
888.
[35] M. Saunders, K. N. Houk, Y.-D. Wu, C. W. Still, M. Lipton, G. Chang, W. C.
Guida, J. Am. Chem. Soc. 1990, 112, 1419–1427.
[36] I. Kolossváry, W. C. Guida, J. Am. Chem. Soc. 1996, 118, 5011–5019.
[37] S. A. de Marothy, XYZ-viewer, v. 0.970, Stockholm, Sweden, 2010.
[30] a) N. Miyaura, A. Suzuki, J. Chem. Soc., Chem. Commun. 1979, 866–867;
b) N. Miyaura, K. Yamada, A. Suzuki, Tetrahedron Lett. 1979, 20, 3437–
3440.
Received: November 14, 2015
Published Online: February 15, 2016
Eur. J. Org. Chem. 2016, 1533–1540
www.eurjoc.org
1540
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim