Stereochemical control in the still-wittig rearrangement synthesis of cyclohexyl (Z)- alkene inhibitors of Pin1

Article abstract of DOI:10.1371/journal.pone.0139543

Three stereoisomeric inhibitors of Pin1: (2R,5S)-, (2S,5R)- and (2S,5S)-Ac-pSer-ψ[(Z) CH = C]-pipecolyl(Pip)-2-(2-naphthyl)ethylamine 1, that mimic L-pSer-D-Pro, D-pSer-LPro, and D-pSer-D-Pro amides respectively, were synthesized by a 13-step route. The newly formed stereogenic centers in the pipecolyl ring were introduced by Luche reduction, followed by stereospecific [2,3]-Still-Wittig rearrangement. The (Z)- to (E)-alkene ratio in the rearrangements were consistently 5.5 to 1. The stereochemistry at the original Ser α-carbon controlled the stereochemistry of the Luche reduction, but it did not affect the stereochemical outcome of the rearrangement, which consistently gave the (Z)-alkene. The epimerized by-product, (2S,5S)-10, resulting from the work-up after Na/NH3 debenzylation of (2S,5R)- 9, was carried on to the (2S,5S)-1 isomer. Compound (2S,5S)-10 was resynthesized from the Luche reduction by-product, (2R,3R)-3, and the stereochemistry was confirmed by comparison of the optical rotations. The IC50 values for (2R,5S)-1, (2S,5R)-1 and (2S,5S)-1 Pin1 inhibition were: 52, 85, and 140 μM, respectively.

Full text of DOI:10.1371/journal.pone.0139543

RESEARCH ARTICLE
Stereochemical Control in the Still-Wittig
Rearrangement Synthesis of Cyclohexyl (Z)-
Alkene Inhibitors of Pin1
Xingguo R. Chen, Shuang A. Fan, Rachel I. Ware, Felicia A. Etzkorn*
Department of Chemistry, Virginia Tech, Blacksburg, Virginia, 24061, United States of America
* fetzkorn@vt.edu
Abstract
Three stereoisomeric inhibitors of Pin1: (2R,5S)-, (2S,5R)- and (2S,5S)-Ac–pSer–Ψ[(Z)
CH = C]–pipecolyl(Pip)–2-(2-naphthyl)ethylamine 1, that mimic L-pSer–D-Pro, D-pSer–L-
Pro, and D-pSer–D-Pro amides respectively, were synthesized by a 13-step route. The
newly formed stereogenic centers in the pipecolyl ring were introduced by Luche reduction,
followed by stereospecific [2,3]-Still-Wittig rearrangement. The (Z)- to (E)-alkene ratio in the
rearrangements were consistently 5.5 to 1. The stereochemistry at the original Ser α-carbon
controlled the stereochemistry of the Luche reduction, but it did not affect the stereochemi-
cal outcome of the rearrangement, which consistently gave the (Z)-alkene. The epimerized
by-product, (2S,5S)-10, resulting from the work-up after Na/NH3 debenzylation of (2S,5R)-
9, was carried on to the (2S,5S)-1 isomer. Compound (2S,5S)-10 was resynthesized from
the Luche reduction by-product, (2R,3R)-3, and the stereochemistry was confirmed by com-
parison of the optical rotations. The IC50 values for (2R,5S)-1, (2S,5R)-1 and (2S,5S)-1 Pin1
inhibition were: 52, 85, and 140 μM, respectively.
OPEN ACCESS
Citation: Chen XR, Fan SA, Ware RI, Etzkorn FA
(2015) Stereochemical Control in the Still-Wittig
Rearrangement Synthesis of Cyclohexyl (Z)-Alkene
Inhibitors of Pin1. PLoS ONE 10(10): e0139543.
doi:10.1371/journal.pone.0139543
Editor: Joseph J Barchi, National Cancer Institute at
Frederick, UNITED STATES
Received: August 14, 2015
Accepted: September 14, 2015
Published: October 7, 2015
Copyright: © 2015 Chen et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any
medium, provided the original author and source are
credited.
Introduction
Pin1 (Peptidyl-prolyl isomerase (PPIase) interacting with never-in-mitosis A kinase 1) cata-
lyzes the isomerization of phospho-Ser/Thr–Pro (pSer/Thr-Pro) amides, and negatively regu-
lates the G2 to M transition in the cell cycle.[1,2] Pin1 plays an important role in cancer,[3]
Alzheimer’s disease,[4,5] and asthma,[6] and regulates the uncoating and replication processes
of human immunodeficiency virus type 1 (HIV–1).[7,8] Specific inhibitors for Pin1 are valu-
able for understanding its role in these diseases.[3,9] Inhibitors of Pin1 designed and synthe-
sized by several groups were recently reviewed. [9]
In our own work, we have synthesized competitive inhibitors of Pin1 that incorporated
phospho-Ser–C[(Z)CH = C]–Pro and phospho-Ser–C[(E)CH = C]–Pro into pentapeptides.
[10,11] Our peptidomimetics were used to elucidate the inhibition specificity,[11,12] structure,
[13] and dynamics[14,15] of the Pin1 catalytic and WW domains with these cis and trans
amide isosteres.
Data Availability Statement: Relevant data are
contained within the paper and its Supporting
Information files.
Funding: Funded by NIH Grant No. R01 CA110940
(FAE) and the NSF LC-MS Grant No. CHE0722638.
The funders had no role in study design, data
collection and analysis, decision to publish, or
preparation of the manuscript.
Competing Interests: The authors have declared
that no competing interests exist.
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
1 / 18

Stereochemical Control in Still-Wittig Rearrangement
By screening combinatorial peptide libraries containing unnatural amino acid residues,
the Fischer group identified several potent peptide inhibitors of Pin1.[16] Replacement of Pro
with pipecolate (Pip) in an octapeptide improved the inhibition by 100-fold. [16] Replacement
of L-Thr with D-Thr in the octapeptide improved the inhibition by 150-fold.[16] The
combination of D-Thr and Pip at the appropriate positions of the octapeptide gave the best
inhibitor for Pin1 to date, with a Ki value of 1.2 nM.[16] Zhang et al. reported the crystal struc-
tures of Pin1 in complex with Fisher’s pentapeptides, Ac–Phe–D/L-pThr–Pip–Nal–Gln–NH2
(Pip = piperidyl, Nal = 2-naphthylalanine).[17] Electrostatic contacts and hydrogen bonds
between the phosphate group and Pin1, and hydrophobic interactions between the Pip and Nal
residues and Pin1, accounted for the potent inhibitory activity.[17] These results were the start-
ing point for the design of the cyclohexyl alkene inhibitors that we now report (Fig 1).
In the present study, we wanted to see if we could control all aspects of the stereochemistry
in the Still-Wittig rearrangement. The key step in the synthesis of the Ser-cis-Pro (Z)-alkene
isostere was the Still-Wittig [2,3]-sigmatropic rearrangement.[10,18,19] In that native-like syn-
thesis, Felkin-Ahn reduction[20] of an L-Ser-based intermediate ketone set up the (S)-allylic
alcohol stereochemistry to introduce the cis-L-Pro mimic stereochemistry in the Still-Wittig
rearrangement to give the (Z)-alkene.[10] The Z/E selectivity was solvent dependent; THF
favored (Z)-selectivity, while toluene favored (E)-selectivity.[21] To produce the Ser-trans-
L-Pro mimic stereochemistry in the (E)-alkene,[10] Luche reduction[22] to the opposite (R)-
allylic alcohol preceeded an Ireland-Claisen rearrangement.[23]
In this work, strategic combinations of these stereospecific reactions were used to synthesize
three stereoisomeric inhibitors of Pin1. The Luche reduction[22] could be used to install the
anti-stereochemistry of the allylic alcohol. In addition to using such precursors in an Ireland-
Claisen rearrangement to make (E)-alkene isosteres from the Luche precursor,[10] we reasoned
that the Still-Wittig rearrangement would transfer the alcohol stereochemistry to the desired
stereogenic center in the ring. Thus, Ser-cis-Pro mimics could be made with opposite stereo-
genic centers at the Ser and Pro mimic alpha-carbons, i. e. L-Ser-D-Pro or D-Ser-L-Pro, which
had not been made before.
Materials and Methods
Synthesis
General. Unless otherwise indicated, all reactions were carried out under dry N2 in flame-
dried glassware. THF was distilled from Na-benzophenone, and CH2Cl2 was dried by passage
through dry alumina. Anhydrous DMF (99.8%), MeOH, and DIEA were used directly from
sealed bottles. Brine (NaCl), Na2S2O3, NaHCO3, and NH4Cl refer to saturated aqueous
Fig 1. Designed enantiomeric inhibitors synthesized.
doi:10.1371/journal.pone.0139543.g001
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
2 / 18

Stereochemical Control in Still-Wittig Rearrangement
solutions, and HCl refers to a 1 N aqueous solution, unless otherwise noted. Flash chromatog-
raphy was performed on 230–400 mesh silica gel with reagent grade solvents. Analytical HPLC
were obtained on a 4.6 × 50 mm C18 column with 10% CH3CN/H2O for 3 min followed by a
10% to 90% CH3CN/H2O gradient over 6 min unless otherwise noted. HPLC results are
reported as retention time, integrated % purity. 1H-, 13C-, and 31P-NMR spectra were obtained
at ambient temperature in CDCl3, at 500, 125, and 162 MHz, respectively, unless otherwise
noted. Chemical shifts are reported in parts per million (ppm) downfield from tetramethylsi-
lane (TMS). NMR data are reported as follows: @ chemical shift, multiplicity: singlet (s), dou-
blet (d), triplet (t), multiplet (m), broad singlet (br s), coupling constants J in Hz, and
integration. Spectra and HPLC chromatograms are given in Fig C in S1 Dataset.
L- and D-Serine Weinreb amides. Synthesized by the method of Niel.[24]
1-Iodocyclohexene. Synthesized by the method of Barton.[25]
Ketone, (S)-2. A solution of 1-iodocyclohexene (9.8 g, 47 mmol) in THF (285 mL) was
cooled to –78°C and s-BuLi (1.4 M in cyclohexane, 67 mL, 94 mmol) was added dropwise over
15 min to generate 1-cyclohexenyl lithium. The resulting solution was stirred at −78°C for 3 h.
In another flask, a solution of Boc–L-Ser–N-(Me)-O-Me[10] Weinreb amide (9.9 g, 29 mmol)
in THF (82 mL) was cooled to −60°C, and i-PrMgCl (2.0 M in THF, 14 mL, 28 mmol) was
added dropwise and stirred for 55 min. The 1-cyclohexenyl lithium solution was added to the
solution of the deprotonated Weinreb amide via cannula. The mixture was stirred at −78°C
for 1 h and warmed slowly to rt. The mixture was stirred overnight, quenched with NH4Cl
(50 mL) at –30° to –40°C, diluted with EtOAc (300 mL), washed with NH4Cl (2 × 200 mL),
NaHCO3 (300 mL) and brine (300 mL). The organic solution was then dried with Na2SO4 and
evaporated at reduced pressure. The crude product was purified by flash chromatography with
EtOAc:hexanes (1:18), followed by EtOAc:hexanes (1:15) to give a colorless oil (6.4 g, 61%). 1H
NMR (400 MHz): δ 7.34–7.22 (m, 5H), 6.91 (m, 1H), 5.59 (d, J = 7.5, 1H), 5.13 (dt, J = 4.3, 8.4,
1H), 4.54 (d, J = 12.4, 1H), 4.42 (d, J = 12.4, 1H), 3.67 (d, J = 3.8, 2H), 2.39–2.10 (m, 4H), 1.65–
1.58 (m, 4H), 1.44 (s, 9H); 13C NMR (100 MHz): δ 197.8, 155.5, 141.8, 137.8, 137.4, 128.4,
127.8, 127.6, 79.8, 73.1, 71.3, 54.3, 28.4, 26.2, 23.4, 21.8, 21.5.
Ketone (R)-2. Yield 81%. The 1H and 13C NMR spectra were identical to (S)-2. FTIR 3448
cm–1 (NH), 1711 cm–1 (C = O).
Allyl alcohols, (2S,3R)-3 and (2S,3S)-3. A solution of ketone (S)-2 (6.2 g, 17 mmol) in
THF:CH3OH (2.5:1, 500 mL) was cooled in an ice bath. CeCl3 7H2O (9.6 g, 26 mmol) was
added and stirred for 15 min. NaBH4 (3.9 g, 0.10 mol) was added in three portions. The reac-
tion was stirred for 4.5 h, quenched with NH4Cl (50 mL), diluted with EtOAc (250 mL), and
washed with NH4Cl (2 × 250 mL) and brine (250 mL). The organic solution was dried with
Na2SO4 and evaporated under reduced pressure to give a colorless oil as an inseparable mixture
of two diastereomers, (2S,3R)-3 and (2S,3S)-3 (6:1 by 1H NMR, 6.2 g, 100%). The crude prod-
uct was used in the next step without further purification. 1H NMR: δ 7.37–7.29 (m, 5H), 5.72
(br, 1H), 5.26 (d, J = 8.6, 0.85H), 5.12 (d, J = 9.2, 0.15H), 4.55 (d, J = 12.4, 0.15H), 4.53 (d,
J = 12.0, 0.85H), 4.49 (d, J = 12.4, 0.15H), 4.44 (d, J = 12.0, 0.85H), 4.18 (br, 0.15H), 4.06 (t,
J = 6.3, 0.85H), 3.82 (m, 0.85H), 3.76 (dd, J = 2.4, 9.4, 1H), 3.64 (m, 0.3H), 3.56 (dd, J = 2.6, 9.2,
0.85H), 3.02 (d, J = 7.4, 0.85H), 2.02–1.93 (m, 4H), 1.63–1.48 (m, 4H), 1.44 (s, 7.65H), 1.43 (s,
1.35H).
Allyl alcohols, (2R,3S)-3 and (2R,3R)-3. Yield 98%. The 1H NMR spectrum was identical
to the (2S,3R)-3 and (2S,3S)-3 mixture. HRMS (ESI+, m/z): calcd for C21H31NO4Na [M+Na]+
384.2145, found 384.2126.
Dibenzyl amine, (2S,3R)-4. A mixture of (2S,3R)-3 and (2S,3S)-3, (0.50 g, 1.4 mmol) was
dissolved in CH2Cl2 (9 mL) and TFA (4.5 mL, 58 mmol) was added and stirred for 1.5 h. The
mixture was concentrated, and the residue was dissolved in CHCl3 (13 mL). DIEA (1.4 g,
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
3 / 18

Stereochemical Control in Still-Wittig Rearrangement
11 mmol) and BnBr (0.59 g, 3.4 mmol) were added, and the solution was stirred for 52 h. The
solution was diluted with EtOAc (25 mL), washed with NH4Cl (2 × 25 mL) and brine (25 mL).
The organic solution was dried with Na2SO4 and evaporated at reduced pressure. The crude
product was purified by flash chromatography with EtOAc:hexanes (1:25), followed by EtOAc:
hexanes (1:12) to give a colorless oil as a single diastereomer (0.40 g, 65%). HPLC: 18.8 min,
90%, λ = 210 nm; 1H NMR: δ 7.40–7.19 (m, 15H), 5.64 (m, 1H), 4.60 (d, J = 11.8, 1H), 4.54 (d,
J = 11.8, 1H), 4.32 (d, J = 8.1, 1H), 3.94 (dd, J = 4.6, 9.7, 1H), 3.86 (dd, J = 5.2, 9.7, 1H), 3.83 (d,
J = 13.8, 2H), 3.56 (d, J = 13.8, 2H), 2.93 (dt, J = 4.9, 8.1, 1H), 2.75 (br s, 1H), 2.09–1.99 (m,
2H), 1.80–1.75 (m, 1H), 1.65–1.47 (m, 4H), 1.42–1.37 (m, 1H); 13C NMR (100 MHz): δ 140.2,
138.7, 138.1, 129.2, 128.6, 128.2, 127.9, 127.8, 127.0, 125.0, 77.7, 73.6, 68.6, 58.0, 55.0, 25.3,
22.74, 22.72, 22.67; HRMS (ESI+, m/z): calcd for C30H36NO2 [M+H]+ 442.2741, found
442.2727.
Dibenzyl amine, (2R,3S)-4. Yield 64%. HPLC: 18.7 min, 94%, 254 nm. The 1H NMR spec-
trum was identical to (2S,3R)-4. FTIR (neat): 3676 cm–1 (OH), 2988 cm–1 (CH), 2901 cm–1
(CH), 1265 cm–1 (CO); HRMS (ESI+, m/z): calcd for C30H36NO2 [M+H]+ 442.2741, found
442.2750; [α]25D +6.6°(c 0.51, CH3OH).
Dibenzyl amine, (2R,3R)-4. Isolated yield 8% after dibenzylation of mixture (2R,3S)-3
and (2R,3R)-3 and chromatographic separation. 1H NMR (400 MHz): δ 7.39–7.22 (m, 15H),
5.59 (s, 1H), 4.56 (d, J = 12.0, 1H), 4.48 (d, J = 12.0, 1H), 4.36 (br, 1H), 3.94 (d, J = 13.0, 2H),
3.82 (d, J = 10.0, 1H), 3.68 (m, 3H), 3.52 (dd, J = 3.2, 10.5, 1H), 3.01 (ddd, J = 3.1, 8.1, 10.6, 1H),
1.96 (m, 2H), 1.82 (d, J = 16.5, 1H), 1.43 (m, 5H); 13C NMR (100 MHz): δ 139.3, 138.5, 137.2,
129.4, 128.6, 128.5, 127.8, 127.6, 127.3, 126.8, 73.5, 72.7, 67.8, 59.1, 54.5, 25.3, 22.8, 22.7, 22.4;
HRMS (ESI+, m/z): calcd. for C30H36NO2 [M+H]+ 442.2741, found 442.2712; [α]22D −68°(c
0.36, CH3OH).
Stannane, (2S,3R)-5. To a solution of dibenzyl amine, (2S,3R)-4, (3.5 g, 7.9 mmol) in
THF (115 mL), 18-crown–6 (2.7 g, 10 mmol) in THF (5 mL) was added, followed by the addi-
tion of KH (0.48 g, 12 mmol). A solution of n-Bu3SnCH2I (5.1 g, 12 mmol), prepared as previ-
ously reported,[26] was added and the mixture was stirred for 2.5 h. The reaction was
quenched with CH3OH (15 mL), and the resulting yellow solution was diluted with EtOAc
(350 mL) and washed with NH4Cl (2 × 200 mL). The organic solution was dried with Na2SO4
and evaporated under reduced pressure. The crude product was purified by flash chromatogra-
phy with hexanes, followed by EtOAc:hexanes (1:150) to give a colorless oil (4.9 g, 83%). 1H
NMR: δ 7.42–7.17 (m, 15H), 5.54 (br, 1H), 4.56 (d, J = 12.1, 1H), 4.51 (d, J = 12.1, 1H), 3.87
(dd, J = 2.7, 10.3, 1H), 3.82 (dd, J = 6.7, 10.3, 1H), 3.76 (d, J = 13.6, 2H), 3.70 (d, J = 13.6, 2H),
3.61 (d, J = 9.8, 1H), 3.54 (d, J = 8.0, 1H), 3.24 (d, J = 9.8, 1H), 2.90 (ddd, J = 2.6, 6.7, 8.0, 1H),
2.07 (m, 2H), 1.63–1.29 (m, 13H), 1.24 (app. sext., J = 7.4, 6H), 0.93–0.74 (m, 16H); 13C
NMR (100 MHz): δ 140.9, 139.3, 136.0, 129.4, 128.4, 128.0, 127.5, 127.4, 126.7, 126.5, 88.3,
73.4, 68.4, 58.4, 58.1, 55.0, 29.3, 27.5, 25.4, 23.0, 22.7, 22.4, 13.9, 9.0; HRMS (ESI+, m/z): calcd
for C43H64NO2Sn [M+H]+ 746.3954, found 746.3956; [α]22D –2.3°(c 2.9, CHCl3).
Stannane, (2R,3S)-5. Yield 58%. The 1H NMR spectrum was identical to (2S,3R)-5. FTIR
(neat) 1265 cm–1 (C-O); HRMS (ESI+, m/z): calcd for C43H64NO2Sn [M+H]+ 746.3954, found
746.3933.
Stannane, (2R,3R)-5. Yield 92%. 1H NMR: δ 7.40–7.16 (m, 15H), 5.59 (s, 1H), 4.42 (d,
J = 12.0, 1H), 4.32 (d, J = 12.0, 1H), 3.94 (d, J = 13.6, 2H), 3.86 (d, J = 13.7, 2H), 3.70 (d, J = 7.6,
1H), 3.67 (d, J = 9.9, 1H), 3.52 (dd, J = 5.8, 9.8, 1H), 3.45 (dd, J = 4.0, 9.8, 1H), 3.35 (d, J = 9.9,
1H), 2.96 (ddd, J = 4.2, 5.6, 8.6, 1H), 2.06–1.98 (m, 2H), 1.76 (m, 1H), 1.61–1.42 (m, 11H), 1.32
(sextet, J = 7.3, 6H), 0.95 (t, J = 8.2, 6H), 0.89 (t, J = 7.3, 9H); 13C NMR (100 MHz): δ 141.7,
139.0, 135.6, 129.1, 128.3, 128.0, 127.6, 127.4, 126.5, 126.1, 90.5, 73.2, 71.1, 58.3, 58.0, 55.7,
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
4 / 18

Stereochemical Control in Still-Wittig Rearrangement
29.4, 27.6, 25.3, 23.6, 22.9, 22.8, 13.9, 9.0; HRMS (ESI+, m/z): calcd. for C43H64NO2Sn [M+H]+
746.3959, found 746.3937; [α]22D −27°(c 0.52, CH2Cl2).
(Z)-Alkene, (2R,3Z,5S)-6. The intermediate (2S,3R)-5 (2.45 g, 3.29 mmol) was dissolved
in THF (35 mL) and dried with 4 Å molecular sieves for 2 h. The solution was transferred to
another flask via cannula and cooled to –78°C. n-BuLi (2.5 M in hexanes, 1.7 mL, 4.3 mmol)
was added slowly and stirred for 2.5 h. (The reaction time was critical. If the reaction was
quenched before completion, the remaining starting material was converted into the corre-
sponding methyl ether, and could not be recovered. Prolonged reaction time resulted in
removal of benzyl protecting groups. The color change from pale yellow to red was a reason-
ably good indicator for the completion of the reaction.) The reaction was quenched with
CH3OH (8 mL), diluted with CH2Cl2, washed with NH4Cl (150 mL) and brine (150 mL). The
organic solution was dried with Na2SO4 and evaporated under reduced pressure. The ratio of
(Z)- to (E)-alkene was 5.5:1 as calculated by the NMR of the crude product. The crude product
was purified by flash chromatography with EtOAc:hexanes (1:25), followed by EtOAc:hexanes
(1:12) to give a colorless oil (0.94 g, 62%). HPLC: 100% H2O for 3 min, then 0% to 100%
CH3CN/H2O gradient over 15 min, 100% CH3CN for 15 min, flow rate 1.0 mL/min, λ = 254
nm, 21.0 min, 98%. 1H NMR: δ 7.34–7.18 (m, 15H), 5.40 (dd, J = 1.4, 10.4, 1H), 4.49 (d,
J = 12.6, 1H), 4.44 (d, J = 12.6, 1H), 3.76–3.67 (m, 5H), 3.48 (t, J = 8.8, 1H), 3.44 (d, J = 14.2,
2H), 3.33 (ddd, J = 4.8, 8.2, 10.5, 1H), 2.58 (dd, J = 3.4, 8.2, 1H), 2.52 (m, 1H), 2.32 (m, 1H),
2.18 (d, J = 13.7, 1H), 1.89 (m, 1H), 1.68(d, J = 13.4, 1H), 1.61–1.52 (m, 2H), 1.50–1.38 (m,
2H); 13C NMR (100 MHz): δ 144.9, 140.5, 137.8, 128.5, 128.4, 128.3, 128.0, 127.8, 127.0, 122.0,
73.2, 72.3, 63.6, 54.8, 54.5, 39.1, 33.4, 29.9, 28.9, 22.2; HRMS (ESI+, m/z): calcd. for C31H38NO2
[M+H]+ 456.2897, found 456.2916; [α]25D +49°(c 0.30, CH3OH).
(Z)-Alkene, (2S,3Z,5R)-6. The ratio of (Z)- to (E)-alkene was 5.5:1 as calculated by the
NMR of the crude product. Yield 58%. The 1H NMR spectrum was identical to (2R,3Z,5S)-6.
FTIR (neat): 3488 cm–1 (OH), 3057 cm–1 (sp2 CH), 2933–2851 cm–1 (sp3 CH), 1265 cm–1
(CO); HRMS (ESI+, m/z): calcd for C31H38NO2 [M+H]+ 456.2897, found 456.2874; [α]25
49°(c 0.33, CH3OH).
–
D
(Z)-Alkene, (2S,3Z,5S)-6. The ratio of (Z)- to (E)-alkene was 5.5:1 as calculated by the
NMR of the crude product. Yield 59%. 1H NMR: δ 7.39–7.20 (m, 15H), 5.42 (dd, J = 10, 1, 1H),
4.57 (d, J = 12, 1H), 4.52 (d, J = 12, 1H), 3.83 (m, 2H), 3.73 (m, 1H), 3.61 (m, 4H), 3.43 (m,
2H), 2.21 (m, 2H), 2.06 (m, 1H), 1.75 (m, 1H), 1.46 (m, 3H), 1.24 (m, 2H); 13C NMR: δ 144.8,
139.1, 138.6, 129.9, 128.5, 128.3, 127.8, 127.7, 127.1, 123.0, 73.4, 70.1, 63.1, 54.7, 53.3, 39.1,
33.0, 28.02, 27.99, 21.7; 1D nOe Hf−Hm; HRMS (ESI+, m/z): calcd for C31H37NO2 [M+H]+
456.2897, found 456.2878; cald for C31H37NO2Na [M+Na]+ 478.2717 found 478.2678;
[α]22D +36°(c 1.3 CH3OH).
Benzylamino alcohol, (2R,5S)-7. To a flask containing (Z)-alkene, (2R,5S)-6 (303 mg,
0.665 mmol) was added 20% Pd(OH)2/C (25.9 mg). CH3OH (21 mL) was added, followed by
the addition of HCOOH (4.59 g, 99.8 mmol). The reaction mixture was stirred and monitored
by TLC with EtOAc:hexanes (1:4). When all starting material was consumed, approximately
15 min, the mixture was filtered through Celite immediately because the other benzyl protect-
ing groups could be removed with prolonged reaction time. The Celite was washed with
CH3OH (250 mL), and the combined filtrate was concentrated under reduced pressure. The
residue was neutralized with NaHCO3 and extracted with CH2Cl2 (7 × 80 mL). The combined
organic solution was dried with Na2SO4 and evaporated to give a colorless oil (237 mg, 97%). It
was critical to remove formic acid completely to prevent the formation of formamide by-prod-
uct in the subsequent acylation step. The resulting secondary amine was not stable on silica gel,
and no further purification was performed. 1H NMR: δ 7.35–7.22 (m, 10H), 5.13 (dd, J = 1.9,
9.6, 1H), 4.49 (s, 2H), 3.85 (d, J = 13.4, 1H), 3.75–3.70 (m, 2H), 3.65 (d, J = 13.4, 1H), 3.50 (dd,
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
5 / 18

Stereochemical Control in Still-Wittig Rearrangement
J = 5.5, 10.6, 1H), 3.47 (dd, J = 6.2, 8.7, 1H), 3.31 (dd, J = 7.2, 8.7, 1H), 2.78 (m, 1H), 2.25 (ddt,
J = 1.7, 4.4, 13.5, 1H), 2.12 (m, 3H), 1.82 (m, 1H), 1.72 (m, 1H), 1.56 (m, 1H), 1.52–1.41 (m,
2H), 1.40–1.29 (m, 1H); 13C NMR: δ 143.3, 140.5, 137.9, 128.6, 128.5, 128.2, 128.1, 127.9,
127.1, 126.5, 73.33, 73.31, 63.7, 53.6, 51.1, 39.9, 33.3, 29.3, 28.5, 22.1.
Benzylamino alcohol, (2S,5R)-7. Yield 94%. The 1H and 13C NMR spectra were identical
to (2R,5S)-7. FTIR (neat): 3054 cm–1 (sp2 CH), 1265 cm–1 (CO); MS (ESI+, m/z): calcd for
C24H32NO2 [M+H]+ 366.2, found 366.5.
Benzylamino alcohol, (2S,5S)-7. Yield 86%. 1H NMR: δ7.37–7.20 (m, 10H), 5.39 (dd,
J = 1.8, 7.3, 1H), 4.56 (d, J = 12.1, 1H), 4.53 (d, J = 12.0, 1H), 3.77–3.66 (m, 3H), 3.61 (m, 2H),
3.56 (dd, J = 5.1, 10.0, 1H), 3.50 (dd, J = 6.9, 10.6, 1H), 2.80 (m, 1H), 2.23 (m, 1H), 2.06 (d,
J = 13.8, 1H), 1.77–1.66 (m, 2H), 1.50–1.25 (m, 4H).
Acetylbenzylamino alcohol, (2R,5S)-8. Benzylamino alcohol, (2R,5S)-7 (237 mg, 0.648
mmol) was dissolved in CH2Cl2 (7 mL), and Et3N (197 mg, 1.94 mmol) and Ac2O (132 mg,
1.30 mmol) were added and stirred for 30 min. The mixture was washed with NH4Cl (30 mL),
NaHCO3 (30 mL), and water (30 mL). The organic solution was dried with Na2SO4 and evapo-
rated at reduced pressure. The crude product was purified by flash chromatography with
EtOAc:hexanes (1:2), followed by EtOAc:hexanes (1:1) to give a colorless oil (215 mg, 81%).
HPLC: 15.7 min, 94%, λ = 210 nm; (2R,5S)-8 exists as a pair of rotamers, in CDCl3, the ratio
was ca. 7:3, in DMSO-d6, the ratio was ca. 1:1; 1H NMR (CDCl3): δ 7.34–7.17 (m, 10H), 5.58
(dd, J = 1.3, 9.9, 0.7H), 5.25 (dt, J = 6.3, 9.9, 0.7H), 5.18 (d, J = 8.9, 0.3H), 4.89 (t, J = 7.7, 0.3H),
4.59 (d, J = 15.6, 0.3H), 4.56 (d, J = 15.6, 0.3H), 4.51 (s, 1.3H), 4.48 (d, J = 12.2, 0.7H), 4.45 (d,
J = 12.2, 0.7H), 4.37 (d, J = 12.0, 0.3H), 4.33 (d, J = 12.0, 0.3H), 3.75 (t, J = 9.8, 0.3H), 3.71 (t,
J = 10.6, 0.7H), 3.62–3.49 (m, 2.3H), 3.34 (dd, J = 7.2, 9.5, 0.3H), 3.31 (dd, J = 7.0, 9.0, 0.3H),
2.91–2.88 (m, 0.7H), 2.80–2.78 (m, 0.3H), 2.30 (br, 0.3H), 2.28 (s, 0.9H), 2.20–2.14 (m, 1H),
2.03 (s, 2H), 1.98–1.94 (m, 1H), 1.71–1.69 (m, 2.7H), 1.50–1.48 (m, 1H), 1.42–1.29 (m, 2H),
1.13–1.05 (m, 1H); 1H NMR (DMSO-d6): δ 7.37–7.14 (m, 10H), 5.46 (m, 0.5H), 5.22 (d, J = 8.7,
0.5H), 5.03 (d, J = 7.3, 0.5H), 4.89 (dt, J = 4.3, 8.8, 0.5H), 4.63 (d, J = 15.8, 0.5H), 4.60 (t, J = 5.6,
0.5H), 4.57 (s, 1H), 4.53 (t, J = 5.5, 0.5H), 4.45 (d, J = 12.4, 0.5H), 4.43 (d, J = 12.4, 0.5H), 4.37
(d, J = 12.0, 0.5H), 4.34 (d, J = 12.0, 0.5H), 4.28 (d, J = 15.8, 0.5H), 3.56 (dd, J = 7.6, 10.0, 0.5H),
3.49–3.34 (m, 3.5H), 2.81 (m, 0.5H), 2.65 (m, 0.5H), 2.19 (s, 1.5H), 2.06–1.99 (m, 1H), 1.86 (s,
1.5H), 1.84–1.76 (m, 2H), 1.63–1.55 (m, 1H), 1.40–1.31 (m, 2H), 1.21–1.13 (m, 1H), 0.98 (m,
0.5H), 0.85 (m, 0.5H); 13C NMR: δ 171.6, 144.9 (m), 144.7, 139.6 (m), 138.0, 137.9, 128.8, 128.6
(m), 128.5, 128.4 (m), 128.01 (m), 127.98, 127.94 (m), 127.8, 127.5, 127.4, 126.8 (m), 126.3,
121.5, 121.2 (m), 73.3 (m), 73.2, 71.3, 63.7, 54.7 (m), 52.1, 50.4, 45.1 (m), 40.1 (m), 39.7, 33.2,
33.1 (m), 28.8, 27.8, 27.2 (m), 22.7, 21.9, 21.8 (m); MS (ESI+, m/z): calcd for C26H34NO3
[M+H]+ 408.25, found 408.65.
Acetylbenzylamino alcohol, (2S,5R)-8. Yield 86%. The 1H and 13C NMR spectra were
identical to (2R,5S)-8. HPLC: 15.9 min, 96%, λ = 210 nm; FTIR (neat): 3422 cm–1 (OH), 1630
cm–1 (C = O), 1266 cm–1 (CO); HRMS (ESI+, m/z): calcd for C26H34NO3 [M+H]+ 408.2533,
found 408.2542; [α]25D –37°(c 0.99, CH3OH).
Acetylbenzylamino alcohol, (2S,5S)-8. Yield 78%. 1H NMR (400 MHz): δ 7.34–7.21 (m,
10H), 5.73 (dt, J = 4.8, 9.1, 0.8H), 5.18 (d, J = 9.5, 1H), 4.94 (dt, J = 5.4, 8.6, 0.2H), 4.62 (d,
J = 17.7, 1H), 4.53 (d, J = 17.6, 1H), 4.39 (d, J = 11.9, 0.8H), 4.32 (d, J = 11.9, 0.8H), 4.28 (d,
J = 11.1, 0.2H), 4.23 (d, J = 11.9, 0.2H), 3.69 (m, 1.6H), 3.53 (dd, J = 8.9, 10.1, 1H), 3.39 (dd,
J = 4.8, 10.6, 1H), 3.31 (m, 0.4H), 3.11 (m, 1H), 3.00 (m, 0.8H), 2.78 (m, 0.2H), 2.33 (s, 0.6H),
2.28 (m, 0.2H), 2.21 (dt, J = 3.2, 13.6, 1H), 2.01 (s, 2.4H), 1.93 (d, J = 13.7, 1H), 1.78 (d, J = 11.7,
2H), 1.49–1.34 (m, 3H), 1.28–1.17 (m, 1H). 13C NMR: δ 172.8, 172.0 (m), 145.1, 145.0 (m),
139.7 (m), 138.1, 138.0 (m), 128.7, 128.4, 127.8, 127.7, 127.2, 126.9 (m), 126.2, 119.9 (m), 119.5,
73.0 (m), 72.8, 71.3 (m), 70.7, 63.36, 62.73 (m), 55.3 (m), 50.8, 48.6, 44.8 (m), 40.3 (m), 39.8,
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
6 / 18

Stereochemical Control in Still-Wittig Rearrangement
33.3 (m), 33.0, 29.1, 28.3, 28.0 (m), 22.5, 21.7, 21.4 (m). HRMS (ESI+, m/z): calcd for
C26H34NO3 [M+H]+ 408.2533, found 408.2516; calcd for C52H66N2O6Na [2M+Na]+ 837.4813,
found 837.4767; [α]25D +36°(c 0.82, CH3OH).
Acetylbenzylamino acid, (2R,5S)-9. Acetylbenzylamino alcohol (2R,5S)-8 (266 mg, 0.653
mmol) was dissolved in acetone (36 mL) and cooled with an ice bath. A solution of 24% CrO3
in aq. H2SO4 (0.73 mL, 2.7 M) was added dropwise, and the mixture was stirred for 30 min. i-
PrOH (8 mL) was added and the mixture was stirred for 30 min. Water (50 mL) was added to
the solution and the mixture was extracted with CH2Cl2 (12 × 25 mL). The combined organic
solution was dried with Na2SO4 and evaporated at reduced pressure. The crude product was
purified by flash chromatography with 1% AcOH in EtOAc:hexanes (1:3) to give a colorless oil
(246 mg, 89%). Compound (2R,5S)-9 exists as a pair of rotamers, with a ratio of ca. 4:1 in
CDCl3. 1H NMR: δ 7.34–7.15 (m, 10H), 5.50 (d, J = 9.8, 0.8H) 5.22 (m, 0.8H), 5.15 (d, J = 8.5,
0.2H), 4.81 (q, J = 7.3, 0.2H), 4.58–4.42 (m, 3.5H), 4.36 (d, J = 12.0, 0.2H), 4.30 (d, J = 12.0,
0.2H), 3.68–3.58 (m, 2.2H), 3.46 (br, 0.2H), 3.40–3.32 (m, 0.5H), 2.28 (s, 0.8H), 2.22–2.02 (m,
5.25H), 1.72–1.55 (m, 3H), 1.25–1.04 (m, 2H); HRMS (ESI+, m/z): calcd for C26H32NO4
[M+H]+ 422.2326, found 422.2332.
Acetylbenzylamino acid, (2S,5R)-9. Yield 77%. HPLC: 10.7 min, 98%, λ = 210 nm. The
1H NMR spectrum was identical to (2R,5S)-9. FTIR (neat): 3422 cm–1 (OH), 1630 cm–1
(C = O), 1266 cm–1 (CO); MS (ESI+, m/z): calcd for C26H32NO4 [M+H]+ 422.2, found 422.3;
MS (ESI–, m/z): calcd for C26H30NO4 [M–H]– 420.2, found 420.4; HRMS (ESI+, m/z): calcd for
C26H31NO4 [M+H]+ 422.2326, found 422.2305; calcd for C26H31NO4Na+ [M+Na]+ 447.2315,
found 447.2235; [α]25D –86°(c 0.37, CH3OH).
Acetylbenzylamino acid, (2S,5S)-9. Yield 68%. 1H NMR (400 MHz): δ 7.35–7.15 (m,
10H), 5.60 (dt, J = 5.2, 8.3, 0.8H), 5.23 (d, J = 8.1, 0.8H), 5.14 (d, J = 7.4, 0.2H), 4.91 (d, J = 15.9,
0.2H), 4.85 (dt, J = 5.6, 8.1, 0.2H), 4.59 (d, J = 17.9, 0.8H), 4.51 (d, J = 17.9, 0.8H), 4.39 (d,
J = 11.9, 0.8H), 4.34 (d, J = 12.0, 0.8H), 4.22 (d, J = 15.8, 0.2H), 4.18 (s, 0.4H), 3.80 (d, J = 3.9,
0.8H), 3.51 (dd, J = 7.8, 10.2, 0.8H), 3.48 (m, 0.2H), 3.41 (dd, J = 5.0, 10.3, 0.8H), 3.26 (dd,
J = 5.0, 9.7, 0.2H), 3.19 (t, J = 9.2, 0.2H), 2.37 (s, 0.6H), 2.32–2.13 (m, 2H), 2.09–2.14 (m, 1H),
2.03 (s, 2.4H), 1.80–1.57 (m, 3H), 1.41–1.19 (m, 2H); 13C NMR: δ 175.2, 173.3, 173.0 (m),
143.6 (m), 143.0, 139.2 (m), 137.9, 128.8, 128.5, 127.9, 127.8 (m), 126.2, 120.9, 119.8 (m), 73.0,
71.4 (m), 70.7, 55.6 (m), 51.4, 49.0, 45.1 (m), 43.1, 42.9 (m), 34.5, 34.3 (m), 29.3, 29.2 (m), 28.0,
27.3 (m), 22.6, 22.4 (m), 22.2 (m), 22.1; HRMS (ESI+, m/z): calcd for C26H32NO4 [M+H]+
422.2326, found 422.2329; calcd for C52H62N2O8Na [2M+Na]+ 865.4398, found 865.4369;
[α]22D +67°(c 0.84, CH3OH).
Acetylamino acid, (2R,5S)-10. NH3 (19 mL) was distilled into a flask and Na (139 mg,
6.04 mmol) was added at –78°C. A solution of acetyl-benzylamino acid, (2R,5S)-9 (212 mg,
0.503 mmol) in dry THF (9.5 mL) was added dropwise to the Na/NH3 solution. The reaction
was warmed to reflux for 3 h, and quenched with solid NH4Cl (ca. 4 g). AcOH (40 mL) was
added slowly to the solution at –78°C, followed by the addition of CH3OH (30 mL). The mix-
ture was filtered, and the filtrate was concentrated. The residue was purified by flash chroma-
tography with 1% AcOH in CH3OH:CH2Cl2 (1:12) to give a colorless oil (84 mg, 69%). HPLC:
λ = 210 nm, 9.48 min, 100%. 1H NMR (CD3OD): δ 5.21 (d, J = 9.0, 1H), 4.70 (dt, J = 5.9, 9.0,
1H), 3.76 (br, 1H), 3.54 (dd, J = 5.1, 11.0, 1H), 3.49 (dd, J = 6.6, 11.0, 1H), 2.28–2.23 (m, 2H),
2.13 (m, 1H), 1.95 (s, 3H), 1.82–1.77 (m, 1H), 1.64–1.59 (m, 2H), 1.53–1.43 (m, 1H), 1.36–1.27
(m, 1H); 13C NMR (CD3OD): δ 172.6, 142.5, 123.6, 65.2, 50.5, 35.5, 30.4, 28.9, 24.0, 22.7; MS
(ESI+, m/z): calcd for C12H19NNaO4 [M + Na]+ 264.12, found 264.51.
Acetylamino acid, (2S,5R)-10. Yield 81%. The 1H NMR contained a small amount of
(2S,5S)-10. MS (ESI–, m/z): calcd for C12H18NO4 [M–H]– 240.1, found 240.4; [α]23D –0.27°(c
0.84, CH3OH).
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
7 / 18

Stereochemical Control in Still-Wittig Rearrangement
Acetylamino acid, (2S,5S)-10. Yield 75%. 1H NMR (CD3OD): δ 5.38 (d, J = 7.9, 1H), 4.74
(q, J = 6.6, 1H), 3.74 (br, 1H), 3.64 (dd, J = 5.8, 11.0, 1H), 3.58 (dd, J = 6.0, 11.0, 1H), 2.43–2.35
(m, 2H), 2.25 (m, 1H), 2.01 (s, 3H), 1.89 (m, 1H), 1.70–1.57 (m, 3H), 1.47–1.41 (m, 1H). 13C
NMR: δ 172.7, 143.8, 123.3, 65.5, 50.1, 35.5, 30.7, 29.0, 24.0, 22.6. [α]22D +180°(c 0.59,
CH3OH).
NEA amide, (2R,5S)-11. Acetyl amino acid, (2R,5S)-10 (42.0 mg, 0.174 mmol) was dis-
solved in DMF:CH2Cl2 (1:2, 18 mL). 2-(2-naphthyl)ethylamine (89.4 mg, 0.522 mmol), DIEA
(67.5 mg, 0.522 mmol), DMAP (ca. 3 mg), HOBt (79.9 mg, 0.522 mmol) and DCC (108 mg,
0.522 mmol) were added, and the mixture was stirred for 24 h. The reaction was diluted with
EtOAc (75 mL), washed with water (30 mL), HCl (1M, 30 mL), NaHCO3 (30 mL) and brine
(30 mL), dried with Na2SO4 and concentrated. The crude product was purified by flash chro-
matography with CH3OH:CHCl3 (1:8) to give a colorless oil (67 mg, 98%). HPLC: λ = 254 nm,
16.0 min, 95%. 1H NMR: δ 7.79 (d, J = 7.4, 1H), 7.77 (d, J = 8.0, 1H), 7.76 (d, J = 5.7, 1H), 7.60
(s, 1H), 7.45 (dt, J = 1.4, 7.0, 1H), 7.42 (dt, J = 1.4, 7.0, 1H), 7.31 (dd, J = 1.6, 8.7, 1H), 6.75 (t,
J = 5.7, 1H), 5.89 (d, J = 7.3, 1H), 5.13 (d, J = 9.5, 1H), 4.56 (m, 1H), 3.57 (quintet, J = 6.8, 1H),
3.55 (t, J = 6.6, 1H), 3.49 (quintet, J = 6.6, 1H), 3.43 (dd, J = 4.6, 10.0, 1H), 3.21 (br, 1H), 3.16
(dd, J = 7.6, 10.0, 1H), 2.94 (dt, J = 2.0, 7.0, 2H), 2.33 (d, J = 13.2, 1H), 1.99 (m, 2H), 1.93 (s,
3H), 1.66 (d, J = 12.0, 1H), 1.56–1.45 (m, 2H), 1.36–1.28 (m, 1H), 1.23 (m, 1H); 13C NMR: δ
172.1, 170.1, 143.3, 136.8, 133.6, 132.2, 128.2, 127.7, 127.5, 127.40, 127.38, 126.2, 125.6, 123.6,
64.9, 49.1, 43.6, 40.8, 35.7, 34.7, 28.4, 27.7, 23.4, 22.8; [α]25D +100°(c 1.6, CH3OH).
NEA amide, (2S,5R)-11. Yield 61%. HPLC: λ = 210 nm, 16.2 min, 99%. The 1H and 13
NMR spectra were identical to (2R,5S)-11. MS (ESI+, m/z): calcd for C24H31N2O3 [M+H]+
422.2, found 422.3; HRMS (ESI+, m/z): calcd for C19H30N4O5 [M+H]+ 395.2289, found
C
395.2315; calcd for C19H30N4O5Na+ [M+Na]+ 417.2108, found 417.2129; [α]25D –100°(c 1.8,
CH3OH).
NEA amide, (2S,5S)-11. Yield 66%. HPLC: λ = 210 nm, 16.0 min, 91%; 1H NMR: δ 7.79–
7.74 (m, 4H), 7.62 (s, 1H), 7.42 (m, 2H), 7.36 (dd, J = 1.4, 8.5, 1H), 6.24 (m, 1H), 5.05 (dd,
J = 1.5, 9.0, 1H), 4.34 (tt, J = 4.8, 9.2, 1H), 3.68–3.56 (m, 2H), 3.52–3.45 (m, 2H), 3.42 (d,
J = 3.8, 1H), 3.27–3.15 (m, 1H), 3.05–2.96 (m, 2H), 2.40 (d, J = 13.6, 1H), 2.00–1.97 (m, 1H),
1.90–1.85 (m, 2H), 1.77 (s, 3H), 1.70–1.53 (m, 2H), 1.26–1.13 (m, 2H); 13C NMR: δ 171.8,
171.1, 142.4, 137.3, 133.7, 132.2, 127.9, 127.8, 127.7, 127.5, 127.4, 126.0, 125.3, 123.4, 64.7, 50.7,
42.7, 41.1, 35.8, 34.3, 27.4, 27.9, 23.1, 22.8; [α]22D +97°(c 0.59, CH3OH); As synthesized from
the epimerized (2S,5S)-10 [α]22D +99°(c 0.32, CH3OH).
Dibenzylphosphate, (2R,5S)-12. To a flask containing amide, (2R,5S)-11, (24.0 mg,
0.0610 mmol) and 5-ethylthio-1H-tetrazole (31.8 mg, 0.244 mmol) under N2, THF (8 mL) was
added. The mixture was stirred for 5 min, and P(OBn)2N(i-Pr)2 (63.2 mg, 0.183 mmol) was
added via syringe. The mixture was stirred for 18 h, cooled to –40°C with a CH3CN/dry ice
bath, and t-BuOOH (49 μL, 5.0–6.0 M in decane) was added. The mixture was warmed slowly
to rt, and the stirring was continued for 30 min. Na2S2O5 (15 mL) was added. The mixture was
stirred for 20 min, extracted with CH2Cl2 (4 × 20 mL), dried with Na2SO4 and evaporated at
reduced pressure. The crude product was purified by flash chromatography with EtOAc to give
(2R,5S)-12 (24.5 mg, 62%) as a colorless oil. 1H NMR: δ 7.75 (dd, J = 1.2, 7.1, 1H), 7.74 (d,
J = 8.2, 2H), 7.59 (s, 1H), 7.42–7.29 (m, 13H), 6.10 (t, J = 5.6, 1H), 5.87 (d, J = 7.9, 1H), 5.09
(dd, J = 1.4, 9.0, 1H), 5.02 (dd, J = 5.2, 11.6, 1H), 5.00 (dd, J = 5.4, 11.7, 1H), 4.98 (dd, J = 5.4,
11.6, 1H), 4.97 (dd, J = 5.1, 11.7, 1H), 4.66 (app. septet, J = 4.2 Hz, 1H), 3.70 (ddd, J = 4.2, 8.5,
10.7, 1H), 3.67 (ddd, J = 4.6, 9.8, 10.4, 1H), 3.55 (tt, J = 6.8, 13.4, 1H), 3.51 (tt, J = 6.5, 12.8, 1H),
6.92 (d, J = 3.8, 1H), 2.96 (t, J = 7.0, 2H), 2.33 (d, J = 13.1, 1H), 2.03–1.90 (m, 2H), 1.87 (s, 3H),
1.65 (d, J = 12.1, 1H), 1.54–1.51 (m, 1H), 1.47 (app. tq, J = 3.6, 13.1, 1H), 1.31 (app. tt, J = 4.8,
13.1, 1H), 1.20 (app. tq, J = 4.1, 12.7, 1H); 13C NMR: δ 171.6, 169.4, 143.8, 136.8, 132.8, 131.9,
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
8 / 18

Stereochemical Control in Still-Wittig Rearrangement
131.1, 129.0, 128.8, 128.3, 128.19, 128.18, 127.7, 127.6, 127.43, 127.40, 126.2, 125.5, 121.9, 77.4,
69.81, 69.77, 46.9, 44.0, 40.7, 35.7, 34.5, 28.9, 27.7, 23.4, 22.7; 31P NMR: δ 0.63; MS (ESI+, m/z):
calcd for C38H44N2O6P [M+H]+ 655.29, found 655.98; [α]25D +84°(c 0.57, CH3OH).
Dibenzylphosphate, (2S,5R)-12. Yield 58%. The 1H and 13C NMR spectra were identical
to (2R,5S)-12. MS (ESI+, m/z): calcd for C38H44N2O6P [M+H]+ 655.3, found 655.4; [α]25
79°(c 0.82, CH3OH).
–
D
Dibenzylphosphate, (2S,5S)-12. Yield 47%. HPLC: λ = 254 nm, 20.8 min, 96%; 1H NMR
δ 7.78–7.72 (m, 3H), 7.65 (t, J = 5.6, 1H), 7.61 (s, 1H), 7.43–7.33 (m, 13H), 6.70 (d, J = 5.3, 1H),
5.10–5.00 (m, 4H), 4.92 (d, J = 9.2, 1H), 4.53–4.48 (m, 1H), 3.88–3.78 (m, 2H), 3.67 (dq, J = 6.9,
13.6, 1H), 3.47 (ddt, J = 5.4, 7.2, 13.2, 1H), 3.34 (d, J = 3.5, 1H), 3.00 (t, J = 7.4, 2H), 2.42 (d,
J = 12.7, 1H), 1.94–1.84 (m, 2H), 1.74 (s, 3H), 1.71–1.68 (m, 2H), 1.55–1.53 (m, 1H), 1.16–1.09
(m, 2H); 13C NMR: δ 171.1, 170.7, 137.4, 144.1, 135.5, 135.3, 133.6, 132.2, 129.08, 129.06,
128.9, 128.24, 128.22, 127.86, 127.82, 127.69, 127.56, 127.3, 125.9, 125.2, 121.2, 77.4, 70.1, 70.0,
68.56, 68.51, 48.74, 48.71, 42.68, 41.01, 35.9, 34.4, 28.6, 28.1, 23.0, 22.7; [α]25D +98°(c 0.61,
CH3OH).
Phosphate, (2R,5S)-1. Dibenzylphosphate, (2R,5S)-12, (22 mg, 0.034 mmol) was dis-
solved in iPr3SiH:H2O:TFA (2.5:2.5:95, 5 mL), and the mixture was stirred for 3 h. The reaction
solution was concentrated, and the residue was purified by HPLC on a Waters XBridge C18
5 μm 19 × 100 mm column, gradient 0% to 100% CH3CN/H2O over 12 min, 100% CH3CN for
4 min, at 12 mL/min, λ = 254 nm, to give a white solid (5.3 mg, 33%). (Note: Higher yields may
be obtained by neutralizing the TFA of the reaction, and eliminating the use of TFA in the
mobile phase.) HPLC with 0.1% TFA in the mobile phase: 14.5 min, 99%; 1H NMR (CD3OD):
δ 7.68 (m, 3H), 7.59 (s, 1H), 7.31 (m, 3H), 5.21 (br, 1H), 4.66 (br, 1H), 3.71 (br, 2H), 3.52 (br,
2H), 3.44 (br, 1H), 2.91 (br, 2H), 2.14 (d, J = 9.2, 1H), 1.83 (m, 4H), 1.48–1.08 (m, 6H); 13C
NMR (CD3OD): δ 174.7, 172.3, 142.2, 138.3, 135.1, 133.7, 129.0, 128.9, 128.6, 128.45, 128.38,
126.8, 126.2, 125.6, 67.6, 44.7, 44.6, 41.8, 36.6, 35.3, 29.8, 28.7, 23.8, 22.7; 31P NMR (CD3OD):
δ 0.2 (br, overlapped with external H3PO4 standard); MS (ESI+, m/z): calcd for C24H32N2O6P
[M+H]+ 475.2, found 475.2; HRMS (ESI–, m/z): calcd for C24H30N2O6P [M–H]– 473.1842,
found 473.1837; [α]25D +62°(c 0.12, CH3OH).
Phosphate, (2S,5R)-1. Yield 38%. HPLC with 0.1% TFA in the mobile phase: 14.5 min,
95%; The 1H, 13C, and 31P NMR spectra were identical to (2R,5S)-1; MS (ESI+, m/z): calcd for
C24H32N2O6P [M+H]+ 475.2, found 475.2; [α]25D –61°(c 0.22, CH3OH).
Phosphate, (2S,5S)-1. Yield 54%. HPLC: λ = 254 nm, 20.8 min, 96%. 1H NMR (CD3OD):
δ 7.80 (d, J = 8.1, 1H), 7.76 (d, J = 8.1, 2H), 7.62 (s, 1H), 7.44 (t, J = 6.8, 1H), 7.40 (t, J = 6.8,
1H), 7.36 (d, J = 8.4, 1H), 5.09 (d, J = 8.3, 1H), 4.62 (br, 1H), 3.84 (br, 2H), 3.66 (dt, J = 7.2,
13.1, 1H), 3.56 (br, 1H), 3.42 (app. dt, J = 6.8, 13.3, 1H), 2.97 (t, J = 6.8, 2H), 2.31 (d, J = 13.1,
1H), 1.94–1.91 (m, 1H), 1.80 (m, 1H), 1.78 (s, 3H), 1.63–1.51 (m, 3H), 1.36–1.31 (m, 1H),
1.24–1.16 (m, 1H); 13C NMR: 171.1, 169.8, 140.7, 137.3, 133.1, 131.6, 127.6, 127.5, 127.4, 127.3,
126.7, 125.9, 125.2, 123.8, 48.3, 41.8, 35.2, 33.6, 30.7, 27.9, 27.4, 22.5, 22.3; 31P NMR (DMSO-
d6): δ 1 (br, overlapped with external H3PO4 standard); [α]25D +180°(c 0.14, CH3OH).
Oxazolidinones, (2R,3S)-13 and (2R,3R)-13. A mixture of allyl alcohol (2R,3S)-3 and
(2R,3R)-3 (22 mg, 0.059 mmol) was dissolved in THF (3 mL). KH (3.5 mg, 0.088 mmol) was
added, and the mixture was stirred for 1 h. The reaction was quenched with CH3OH, diluted
with NH4Cl (15 mL) and extracted with EtOAc (15 mL). The organic solution was washed
with brine, dried with Na2SO4, and evaporated under reduced pressure. The crude product was
purified by flash chromatography with EtOAc:hexanes (1:4), followed by EtOAc:hexanes (1:3)
to give a colorless oil as a mixture of two diastereomers, (13 mg, 81%). 1H NMR: δ 7.38–7.26
(m, 5H), 5.84 (m, 0.87H), 5.76 (m, 0.13H), 5.30 (s, 0.87H), 5.24 (s, 0.13H), 4.96 (d, J = 7.7,
0.87H), 4.55 (s, 0.26H), 4.51 (d, J = 5.5, 0.13H), 4.49 (s, 1.74H), 4.02 (m, 0.87H), 3.78 (m,
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
9 / 18

Stereochemical Control in Still-Wittig Rearrangement
0.13H), 3.51 (dd, J = 4.0, 9.3, 0.13H), 3.45 (dd, J = 7.7, 9.3, 0.13H), 3.39 (dd, J = 3.7, 9.3, 0.87H),
3.34 (app t, J = 9.3, 0.87H), 2.06 (m, 2H), 1.95–1.86 (m, 2H), 1.69–1.50 (m, 4H).
Oxazolidinones, (2S,3R)-13 and (2S,3S)-13. The 1H spectrum was identical to (2R,3S)-
13 and (2R,3R)-13.
Bicyclic alkene, (2S,5R)-14. The (Z)-alkene, (2S,5R)-6, (34 mg, 0.075 mmol) was dis-
solved in CH2Cl2, and the solution was cooled with an ice bath. Pyridine (88 mg, 1.1 mmol)
was added, followed by the addition of CH3SO2Cl (100 mg, 0.90 mmol). The mixture was stir-
red and warmed slowly to rt. The stirring was continued for 16 h. The reaction was washed
with H2O, dried with Na2SO4 and evaporated to give (S,R)-Z-alkene methylsulfonate as a col-
orless oil (29 mg, 73%). The crude product was used for next step without further purification.
Pd(OH)2/C (2.5 mg, 20%) was added to a flask containing (S,R)-Z-alkene methylsulfonate (18
mg, 0.034 mmol). CH3OH (2 mL) was added, followed by the addition of HCOOH (235 mg,
5.10 mmol). The mixture was stirred and monitored by TLC. When the reaction was complete,
the mixture was filtered through Celite immediately and washed with CH3OH (25 mL). The fil-
trate was concentrated at reduced pressure, and the residue was neutralized with NaHCO3 and
extracted with CH2Cl2 (2 × 25 mL). The combined organic solution was dried with Na2SO4
and evaporated. The crude product was purified by flash chromatography with EtOAc:hexanes
(1:12) to give (2S,5R)-14 (9.3 mg, 79%). 1H NMR: δ 7.35–7.21 (m, 10H), 5.36 (m, 1H), 4.57 (d,
J = 12.1, 1H), 4.53 (d, J = 12.1, 1H), 4.18 (d, J = 13.8, 1H), 3.72 (dd, J = 4.6, 9.6, 1H), 3.46 (dd,
J = 5.8, 9.6, 1H), 3.24 (d, J = 13.8, 1H), 3.15 (br, 1H), 2.83 (dd, J = 5.6, 11.1, 1H), 2.36–2.23 (m,
1H), 2.13 (br, 1H), 2.04–1.96 (m, 1H), 1.87 (dd, J = 9.5, 11.2, 1H), 1.76–1.67 (m, 2H), 1.64–1.59
(m, 2H), 1.34–1.13 (m, 2H), 0.87 (dq, J = 3.5, 12.5, 1H); 13C NMR: δ 140.3, 139.6, 138.5, 129.1,
128.5, 128.3, 127.8, 127.7, 126.9, 120.2, 73.8, 73.4, 60.8, 59.4, 56.4, 37.0, 34.0, 31.8, 27.5, 25.5;
HRMS (ESI+, m/z): calcd. for C24H30NO [M+H]+ 348.2322, found 348.2329.
Bicyclic alkene, (2R,5S)-14. The 1H and 13C NMR spectra were identical to (2S,5R)-14.
Bicyclic alkene, (2S,5S)-14. Yield 6.2 mg, 79%. 1H NMR: δ 7.39–7.21 (m, 10H), 5.36 (d,
J = 1.8, 1H), 4.53 (d, J = 12.2, 1H), 4.49 (d, J = 12.2, 1H), 3.77 (s, 2H), 3.66 (dd, J = 6.8, 10.0,
1H), 3.42 (dd, J = 5.1, 10.0, 1H), 3.31 (br, 1H), 2.70 (dd, J = 5.6, 12.8, 1H), 2.54 (dd, J = 9.2,
12.8, 1H), 2.22 (d, J = 13.5, 1H), 2.14 (br, 1H), 2.02 (t, J = 12.3, 1H), 1.76 (t, J = 12.2, 2H), 1.64
(d, J = 12.3, 1H), 1.36 (tq, J = 3.2, 13.2, 1H), 1.24 (tq, J = 3.7, 13.1, 1H), 0.93 (dq, J = 3.3, 12.5,
1H); 13C NMR: δ 141.1, 140.2, 138.8, 128.8, 128.4, 128.3, 127.6, 127.5, 126.8, 118.6, 73.1, 72.1,
58.6, 58.3, 51.8, 34.8, 33.8, 32.2, 27.8, 26.0; HRMS (ESI+, m/z): calcd. for C24H30NO [M+H]+
348.2322, found 348.2330.
Bioassay
IC50 Determination of Inhibitors (2R,5S)-1, (2S,5R)-1, and (2S,5S)-1. The concentra-
tions of the inhibitor stock solutions in DMSO:H2O (2:1) were determined by UV at 286 nm
(log ξ = 3.59 for the naphthyl group). The assays were performed as published.[11] The stock
solution was diluted to prepare final concentrations of: 0.45, 1.8, 7.3, 29, 58, 116, 232 and
465 μM for (2R,5S)-1; 6.0, 12, 24, 48, 96, 191 and 382 μM for (2S,5R)-1; and 6.0, 12, 24, 47, 94,
113, 189 and 378 μM for (2S,5S)-1. The assay was performed in duplicate for each concentra-
tion. The inhibitors were pre-equilibrated in a cuvette at 4°C for 12 min. The concentration of
released pNA was recorded by UV at 390 nm for 90 s. The % inhibitions were plotted against
log [I] with Table Curve, Version 3. The IC50 value of an inhibitor was calculated as the con-
centration of the inhibitor at 50% inhibition derived from the fitted equation (S2 Dataset).
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
10 / 18

Stereochemical Control in Still-Wittig Rearrangement
Results and Discussion
Ac–D-pSer–C[(Z)CH = C]–L-Pip–2-(2-naphthyl)ethylamine (NEA), (2S,5R)-1, with a D-
pSer, a cis-locked alkene, a 6-membered ring, and a naphthyl side chain, was designed as an
inhibitor for Pin1 (Fig 1). The D-pSer–C[(Z)CH = C]–L-Pip core mimics the D-pSer–L-Pip,
and the NEA group mimics the Nal in Fischer’s peptides.[16] An acetyl group was attached to
the Ser analogue to avoid a charged terminus; no further N-terminal residues were included
because the N-terminal residues were disordered in electron density maps of several Pin1-inhi-
bitor complexes.[13,17] As a test case, the enantiomeric (2R,5S)-1 was synthesized first as a
model compound (Fig 1). The desired enantiomer (2S,5R)-1, with the more expensive Boc–
D-Ser(Bn)–OH as the starting material, was then synthesized (Fig A in S1 Dataset). The
(2S,5S)-1 diastereomer arose from an accidental epimerization during work-up after ammonia
debenzylation.
In the synthesis of (2R,5S)-1, Luche reduction was used to set up for the synthesis of the
D-Pro mimic in the key Still-Wittig rearrangement (Fig 2). We were not certain if the initial
Ser stereochemistry would affect the outcome of the Still-Wittig rearrangement. It did not––
the stereochemistry depended only upon the stereochemistry of the allylic alcohol resulting
from the Luche reduction.[10,22] The Weinreb amide of Boc–L-Ser(OBn)–OH was synthe-
sized as reported.[24,27] 1-Iodocyclohexene was prepared from cyclohexanone by the method
of Barton.[25] Nucleophilic addition of cyclohexenyl lithium, prepared from 1-iodocyclohex-
ene in situ, to the Weinreb amide afforded the new cyclohexenyl ketone (S)-2 (Fig 2). Luche
reduction of (S)-2 gave two inseparable diastereomers (2S,3R)-3 and (2S,3S)-3 in a ratio of
6-to–1. The stereochemistry and the ratio of the intermediate alcohols (2S,3R)-3 and (2S,3S)-3
from the Luche reduction[22] was determined by 1D coupling constants of cyclic oxazolidi-
none derivatives (2S,3R)-13 and (2S,3S)-13 (Fig 3). The yield and diastereoselectivity were
comparable to the Luche reduction step in the synthesis of the (E)-alkene 5-membered ring
analogue.[10]
Without separation, the mixture was converted to the dibenzyl protected amine, and the
desired single diastereomer, (2S,3R)-4, was isolated by chromatography. The bulky dibenzyl
amine was necessary to obtain high stereoselectivity in the Still-Wittig rearrangement. The pre-
cursor for the Still-Wittig rearrangement, (2S,3R)-5, was synthesized by treating (2S,3R)-4
with n-Bu3SnCH2I, prepared as previously reported (Fig 2).[26] In the presence of n-BuLi at –
78°C, Still-Wittig rearrangement of (2S,3R)-5 afforded the (Z)-alkene (2R,3Z,5S)-6 as the
major product, with a (Z):(E) ratio of 5.5:1. The alkene geometries were determined by 1D
nuclear Overhauser effect (nOe) spectra (Fig 4). The desired stereoselectivity was higher than
that obtained with the Ser−Pro (3:1), or the Ala−Pro (2:1), alkene isosteres, probably due to the
greater bulk of the 6-membered ring.[10,19] The (3R)-alcohol stereocenter from the Luche
reduction was successfully transferred to the (5S)-cyclohexyl stereocenter. The stereoselectivity
in the rearrangements was not affected by the stereocenter at the original Ser α-carbon.
The relative stereochemistry in the cyclohexyl rings of (2R,5S)-6 formed in the Still-Wittig
rearrangement, and the stereochemistry of the enantiomer (2S,5R)-6, were determined by 1D
nOe in bicyclic derivatives (2R,5S)-14 and (2S,5R)-14 (Fig 5). Bicyclic derivative (2R,5S)-14
was synthesized in three steps without purification of intermediates. The primary alcohol was
converted to the mesylate, one benzyl of the amine was deprotected, and NaHCO3 was used as
base to cyclize the amine (Fig 5A). The enantiomer (2S,5R)-14 was prepared the same way. For
both derivatives, the 1H NMR coupling constants between Hi-Hh and Hi’-Hh were 5.6 Hz and
9.5 Hz, respectively, which showed Hi was syn to Hh, and Hi’ was anti to Hh. The nOe correla-
tions Hi—Hh and Hg—Hi’ of (2R,5S)-14 demonstrated that Hh is anti to Hg (Fig 5B). For
(2S,5R)-14, the nOe Hg—Hi’ also showed that Hh is anti to Hg (Fig 5C). Thus, the relative
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
11 / 18

Stereochemical Control in Still-Wittig Rearrangement
Fig 2. Synthesis of the Ac–L-pSer–Ψ[(Z)CH = C]-D-Pip–NEA inhibitor (2R,5S)-1.
doi:10.1371/journal.pone.0139543.g002
relationship between Hg and Hh was shown to be anti in both enantiomers, confirming the
assigned relative stereochemistry.
One N-benzyl protecting group of (2R,5S)-6 was selectively removed with formic acid
with 20% Pd(OH)2/C catalyst (Fig 2). Selective acylation of the amino group of (2R,5S)-7 with
acetic anhydride gave (2R,5S)-8, without affecting the primary hydroxyl group. The primary
hydroxyl was converted to the carboxylic acid with Jones reagent to afford (2R,5S)-9. The
Fig 3. Synthesis and determination of the stereochemistry of derivative (2S,3R)-13. Cyclic compounds
13 were synthesized. 1H NMR coupling constants (J) were used to determine the relative stereochemistry at
the carbons to which Ha and Hb are attached.
doi:10.1371/journal.pone.0139543.g003
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
12 / 18

Stereochemical Control in Still-Wittig Rearrangement
Fig 4. Determination of the (Z)-alkene stereochemistry of intermediate (2R,5S)-6. The 1D nOe 1H NMR. Irradiation of 1Hf shows an nOe at 1Hm and not
at 1Hh.
doi:10.1371/journal.pone.0139543.g004
remaining N- and O-benzyl protecting groups were removed in one step with Na/NH3 to give
(2R,5S)-10 (Fig 2).
Partial epimerization occurred during the Na/NH3 deprotection of (2R,5S)-9 to produce
(2S,5S)-10, which was used to synthesize (2S,5S)-1. To determine which stereocenter was epi-
merized, (2R,3R)-4 was used to resynthesize (2S,5S)-11 (Fig B in S1 Dataset), and the optical
rotations were compared (Materials and Methods). The 2D nuclear Overhauser effect spectros-
copy (NOESY) of derivative (2S,5S)-14 was used to determine the relative stereochemistry of
the Still-Wittig rearrangement product (2S,5S)-6 (Fig 6). The 1H NMR coupling constants
between Hi—Hh and Hi’—Hh were 5.6 Hz and 9.2 Hz, respectively, which indicated that on
this 6-membered ring, Hi and Hh were syn to each other, while Hi’ and Hh were anti to each
other. In the NOESY spectrum of (2S,5S)-14, the nOe correlation He—Hi’ indicated that the
CH2OBn group and Hi’ were syn to each other (Fig 6). The NOESY correlation Hi—Hh indi-
cated that Hi and Hh were syn to each other. Therefore the relative position of Hh and the
CH2OBn group was confirmed to be anti, and the configuration of the stereogenic center in the
6-membered ring was determined to be (S).
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
13 / 18

Stereochemical Control in Still-Wittig Rearrangement
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
14 / 18

Stereochemical Control in Still-Wittig Rearrangement
Fig 5. Determination of the stereochemistry of Still-Wittig intermediates 6. (A) Compound (2S,5R)-14 was synthesized to rigidify intermediate (2R,5S)-6
for nOe determination. (B) Structure of (2S,5R)-14 with lettering of the protons, structure of the major conformation showing nOe interactions, the 1H NMR
and 1D nOe spectra in CDCl3 (400 MHz) are shown. Irradiation of Hi shows an nOe at Hh. Irradiation of Hg shows an nOe at Hi’. (C) The stereochemistry of
enantiomer (2R,5S)-14 was determined. Structure with lettering of the protons, structure of the major conformation showing nOe interactions, and a 1D nOe
spectrum in CDCl3 (400 MHz) are shown. Irradiation of Hg shows an nOe at Hi’.
doi:10.1371/journal.pone.0139543.g005
Enzymatic assays to evaluate inhibition of Pin1 by the three stereoisomeric target com-
pounds were performed by the protease-coupled method as previously reported.[11] The
results of bioassay showed that all three stereoisomers were poor inhibitors of Pin1. The IC50
value of (2R,5S)-1 was 52 4 μM (Fig A in S2 Dataset), the IC50 value of (2S,5R)-1 was
Fig 6. Determination of the stereochemistry of of (2S,5S)-14. The 2D NOESY spectrum is shown with lettering of the major conformation showing key
nOe interactions. Crosspeaks between Hi and Hh, and between He and Hi’ show the stereochemistry given.
doi:10.1371/journal.pone.0139543.g006
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
15 / 18

Stereochemical Control in Still-Wittig Rearrangement
85 10 μM (Fig B in S2 Dataset), and the IC50 value of (2S,5S)-1 was 140 20 μM (Fig C in S2
Dataset). We note that these alkene isosteres are ground-state analogues of the Pin1 substrate.
Compound (2R,5S)-1, which mimics the L-pSer–D-Pip peptide, initially synthesized as a
model compound, was the best inhibitor among them. This was surprising since the most
potent peptidic Pin1 inhibitor has the D-pThr–L-Pip stereochemistry.[16] Compound
(2S,5R)-1, which mimics D-pSer–L-Pip, gave slightly weaker inhibition, and (2S,5S)-1, which
mimics D-pSer–D-Pip, was the weakest inhibitor. The relatively small differences in the inhibi-
tion implied that the stereochemistry of the Ser or Pro of these (Z)-alkene isosteres affects Pin1
inhibition only slightly––Pin1 can accommodate a variety of stereoisomers in its active site.
In the Zhang crystal structures, the two pentapeptide inhibitors, D-peptide and L-peptide,
bound to Pin1 in approximately trans (ω angle = 183°) and cis (ω angle = −19°) conformations,
respectively.[17] In our crystal structure of the Pin1 complex with the (Z)-alkene pentapeptide,
the phosphate and the 5-membered ring of the inhibitor were found to bind to the same sites
of Pin1 as the Zhang L-peptide.[13,17] Our (Z)-alkene pentapeptide inhibitor with both natu-
ral L-stereocenters, was 65-fold less potent than the Zhang L-peptide.[17] Upon changing Pro
to Pip in Fischer’s peptide series, the IC50 value improved by 100-fold.[16] So, we thought that
the 6-membered ring analogues of Pip could significantly improve the inhibitory activity.
However, this was not the case; inhibition was worse than any of our previous (Z)-alkene iso-
steres,[11,12] probably because the D-peptide binds to Pin1 in the trans conformation.[17]
Conclusions
The Still-Wittig [2,3]-sigmatropic rearrangement has proven to be a reliable method to synthe-
size Xaa–Pro and Ser–Pip alkene isosteres, predictably achieving the desired (Z) geometry of
the double bond and the stereogenic center in the 6-membered ring. We developed practical
methods to determine the configurations of the newly formed stereogenic centers in the Luche
reduction, and in the [2,3]-Still-Wittig rearrangement. None of our final compounds were
potent Pin1 inhibitors. We conclude that the (S)-configuration, at either D-Ser or D-Pro
mimic site, were not optimal in any combination in the cyclohexyl (Z)-alkene inhibitors. Our
analysis leads us to suggest that the inhibitory activity could be improved by using either D-
pSer–[(E)CH = C]–L-Pip or L-pSer–[(Z)CH = C]–L-Pip as core structures.
Supporting Information
S1 Dataset. Fig A. Synthesis of (2S,5R)-1. Fig B. Synthesis of (2S,5S)-1. Fig C. NMR and IR
spectra, HPLC chromatograms for compounds 1–14.
(PDF)
S2 Dataset. Pin1 inhibition plots for (2R,5S)-1, (2S,5R)-1, and (2S,5S)-1. Fig A. Inhibition of
Pin1 by (2R,5S)-1, IC50 = 52 4 μM. Fig B. Inhibition of Pin1 by (2S,5R)-1, IC50 = 85 10 μM.
Fig C. Inhibition of Pin1 by (2S,5S)-1, IC50 = 140 20 μM.
(PDF)
Acknowledgments
We thank Guoyan G. Xu for helpful advice.
Author Contributions
Conceived and designed the experiments: FAE XRC. Performed the experiments: XRC SAF
RIW. Analyzed the data: FAE XRC SAF RIW. Wrote the paper: FAE XRC.
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
16 / 18

Stereochemical Control in Still-Wittig Rearrangement
References
1. Lu KP, Hanes SD, Hunter T (1996) A human peptidyl-prolyl isomerase essential for regulation of mito-
sis. Nature 380: 544–547. PMID: 8606777
2. Yeh ES, Means AR (2007) PIN1, the cell cycle and cancer. Nat Rev Cancer 7: 381–388. PMID:
17410202
3. Xu GG, Etzkorn FA (2009) Pin1 as an anticancer drug target. Drug News Perspect 22: 399–407. doi:
10.1358/dnp.2009.22.7.1381751 PMID: 19890497
4. Lu PJ, Wulf G, Zhou XZ, Davies P, Lu KP (1999) The prolyl isomerase Pin1 restores the function of Alz-
heimer-associated phosphorylated tau protein. Nature 399: 784–788. PMID: 10391244
5. Ma SL, Pastorino L, Zhou XZ, Lu KP (2012) Prolyl isomerase Pin1 promotes amyloid precursor protein
(APP) turnover by inhibiting glycogen synthase kinase-3beta (GSK3beta) activity: novel mechanism for
Pin1 to protect against Alzheimer disease. J Biol Chem 287: 6969–6973. doi: 10.1074/jbc.C111.
298596 PMID: 22184106
6. Esnault S, Rosenthal LA, Shen ZJ, Sedgwick JB, Szakaly RJ, Sorkness RL, et al. (2007) A critical role
for Pin1 in allergic pulmonary eosinophilia in rats. J Allergy Clin Immunol 120: 1082–1088. PMID:
17720236
7. Watashi K, Khan M, Yedavalli VR, Yeung ML, Strebel K, Jeang KT (2008) Human immunodeficiency
virus type 1 replication and regulation of APOBEC3G by peptidyl prolyl isomerase Pin1. J Virol 82:
9928–9936. doi: 10.1128/JVI.01017-08 PMID: 18684817
8. Misumi S, Inoue M, Dochi T, Kishimoto N, Hasegawa N, Takamune N, et al. (2010) Uncoating of
human immunodeficiency virus type 1 requires prolyl isomerase Pin1. J Biol Chem 285: 25185–25195.
doi: 10.1074/jbc.M110.114256 PMID: 20529865
9. Moore JD, Potter A (2013) Pin1 inhibitors: Pitfalls, progress and cellular pharmacology. Bioorg Med
Chem Lett 23: 4283–4291. doi: 10.1016/j.bmcl.2013.05.088 PMID: 23796453
10. Wang XJ, Hart SA, Xu B, Mason MD, Goodell JR, Etzkorn FA (2003) Serine-cis-proline and Serine-
trans-proline Isosteres: Stereoselective Synthesis of (Z)- and (E)-Alkene Mimics by Still-Wittig and Ire-
land-Claisen Rearrangements. J Org Chem 68: 2343–2349. PMID: 12636401
11. Wang XJ, Xu B, Mullins AB, Neiler FK, Etzkorn FA (2004) Conformationally Locked Isostere of Phos-
phoSer-cis-Pro Inhibits Pin1 23-Fold Better than PhosphoSer-trans-Pro Isostere. J Am Chem Soc 126:
15533–15542. PMID: 15563182
12. Zhao S, Etzkorn FA (2007) A phosphorylated prodrug for the inhibition of Pin1. Bioorg Med Chem Lett
17: 6615–6618. PMID: 17935986
13. Zhang M, Wang XJ, Chen X, Bowman ME, Luo Y, Noel JP, et al. (2012) Structural and kinetic analysis
of prolyl-isomerization/phosphorylation cross-talk in the CTD code. ACS Chem Biol 7: 1462–1470. doi:
10.1021/cb3000887 PMID: 22670809
14. Namanja AT, Wang XJ, Xu B, Mercedes-Camacho AY, Wilson BD, Wilson KA, et al. (2010) Toward
Flexibility-Activity Relationships by NMR Spectroscopy: Dynamics of Pin1 Ligands. J Am Chem Soc
132: 5607–5609. doi: 10.1021/ja9096779 PMID: 20356313
15. Namanja AT, Wang XJ, Xu B, Mercedes-Camacho AY, Wilson KA, Etzkorn FA, et al. (2011) Stereospe-
cific gating of functional motions in Pin1. Proc Nat Acad Sci USA 108: 12289–12294. doi: 10.1073/
pnas.1019382108 PMID: 21746900
16. Wildemann D, Erdmann F, Alvarez BH, Stoller G, Zhou XZ, Fanghanel J, et al. (2006) Nanomolar inhib-
itors of the peptidyl prolyl cis/trans isomerase Pin1 from combinatorial peptide libraries. J Med Chem
49: 2147–2150. PMID: 16570909
17. Zhang Y, Daum S, Wildemann D, Zhou XZ, Verdecia MA, Bowman ME, et al. (2007) Structural basis
for high-affinity peptide inhibition of human Pin1. ACS Chem Biol 2: 320–328. PMID: 17518432
18. Still WC, Mitra A (1978) A Highly Stereoselective Synthesis of Z-Trisubstituted Olefins via [2,3]-Sigma-
tropic Rearrangement. Preference for a Pseudoaxially Substituted Transition State. J Am Chem Soc
100: 1927–1928.
19. Hart SA, Sabat M, Etzkorn FA (1998) Enantio- and regioselective synthesis of a (Z)-alkene cis-proline
mimic. J Org Chem 63: 7580–7581.
20. Chérest M, Felkin H, Prudent N (1968) Torsional strain involving partial bonds. The stereochemistry of
the lithium aluminum hydride reduction of some simple open-chain ketones. Tetrahedron Lett: 2199–
2204.
21. Hart SA, Trindle CO, Etzkorn FA (2001) Solvent-Dependent Stereoselectivity in a Still-Wittig Rear-
rangement: an Experimental and Ab Initio Study. Org Lett 3: 1789–1791. PMID: 11405712
22. Luche JL, Rodriguez-Hahn L, Crabbe P (1978) Reduction of natural enones in the presence of cerium
trichloride. J Chem Soc, Chem Commun: 601–602.
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
17 / 18

Stereochemical Control in Still-Wittig Rearrangement
23. Ireland RE, Mueller RH (1972) The Claisen Rearrangement of Allyl Esters. J Am Chem Soc 94: 5897–
5898.
24. Niel G, Roux F, Maisonnasse Y, Maugras I, Poncet J, Jouin P (1994) Substrate-controlled Croylbora-
tion from N-(tert-Butoxycarbonyl)amino Aldehydes. J Chem Soc Perkin Trans 1 10: 1275–1280.
25. Barton DHR, Bashiardes G, Fourrey J-L (1983) An Improved Preparation of Vinyl Iodides. Tetrahedron
Lett 24: 1605–1608.
26. Seitz DE, Carroll JJ, Cartaya M. CP, Lee S-H, Zapata A (1983) Efficient Synthesis of Derivatives of Tri-
butylstannylmethanol. Synth Comm 13: 129–134.
27. Nahm S, Weinreb SM (1981) N-Methoxy-N-Methylamides as Effective Acylating Agents. Tetrahedron
Lett 22: 3815–3818.
PLOS ONE | DOI:10.1371/journal.pone.0139543 October 7, 2015
18 / 18