Synthesis and characterization of enantiomerically pure cis- and trans-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-oxides as γ- homoacetylcholine mimetics and inhibitors of acetylcholinesterase

Article abstract of DOI:10.1002/hlca.201100369

The title compounds, the P(3)-axially and P(3)-equatorially substituted cis- and trans-configured 8-benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-oxides (=8-benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphabicyclo[4.4.0]decane 3-oxides=2-fluorohexahydro-6-(phenylmethyl)-4H-1,3,2-dioxaphosphorino[5,4-c] pyridine 2-oxides) were prepared (ee>98%) and fully characterized (Schemes2 and 3). The absolute configurations were established from that of their precursors, the enantiomerically pure cis- and trans-1-benzyl-4- hydroxypiperidine-3-methanols which were unambiguously assigned. Being configuratively fixed and conformationally constrained phosphorus analogues of acetyl γ-homocholine (=3-(acetyloxy)-N,N,N-trimethylpropan-1-aminium), they are suitable probes for the investigation of molecular interactions with acetylcholinesterase. As determined by kinetic methods, all of the compounds are weak inhibitors of the enzyme. Copyright

Full text of DOI:10.1002/hlca.201100369

Helvetica Chimica Acta – Vol. 95 (2012)
17
Synthesis and Characterization of Enantiomerically Pure cis- and trans-3-
Fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-Oxides as g-Homoacetylcholine
Mimetics and Inhibitors of Acetylcholinesterase
by Christian Clerc and Peter Rꢀedi*
Organisch-chemisches Institut der Universitꢀt Zꢁrich,Winterthurerstrasse 190, CH-8057 Zꢁrich
(phone: þ 4144 8215579; e-mail: peru@oci.uzh.ch)
The title compounds, the P(3)-axially and P(3)-equatorially substituted cis- and trans-configured 8-
benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-oxides (¼ 8-benzyl-3-fluoro-2,4-dioxa-8-aza-3-phos-
phabicyclo[4.4.0]decane 3-oxides ¼ 2-fluorohexahydro-6-(phenylmethyl)-4H-1,3,2-dioxaphosphorino-
[5,4-c]pyridine 2-oxides) were prepared (ee > 98%) and fully characterized (Schemes 2 and 3). The
absolute configurations were established from that of their precursors, the enantiomerically pure cis- and
trans-1-benzyl-4-hydroxypiperidine-3-methanols which were unambiguously assigned. Being configu-
ratively fixed and conformationally constrained phosphorus analogues of acetyl g-homocholine (¼ 3-
(acetyloxy)-N,N,N-trimethylpropan-1-aminium), they are suitable probes for the investigation of
molecular interactions with acetylcholinesterase. As determined by kinetic methods, all of the
compounds are weak inhibitors of the enzyme.
1. Introduction. – In the preceding report [1], we have presented the synthesis and
characterization of the optically active P(3)-axially and P(3)-equatorially substituted
cis- and trans-configured 9-benzyl-3-fluoro-2,4-dioxa-9-aza-3-phosphadecalin 3-oxides
of type I (Scheme 1). Continuing our investigations on the inhibition of serine
hydrolases (chymotrypsin, acetylcholinesterase) by azadecalin-type organophosphates,
we discuss now the preparation and full characterization of the eight stereoisomeric,
enantiomerically pure 8-benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-oxides
(¼ 8-benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphabicyclo[4.4.0]decane 3-oxides ¼ 2-fluo-
rohexahydro-6-(phenylmethyl)-4H-1,3,2-dioxaphosphorino[5,4-c]pyridine 2-oxides)
10 and 11 representing type-II inhibitors (cf. Schemes 1 and 3) [2]. The compounds
are mimetics of g-homoacetylcholine (¼ 3-(acetyloxy)-N,N,N-trimethylpropan-1-ami-
nium)1) and as such are considered to be suitable probes for the investigation of
molecular interactions with acetylcholinesterase (AChE) [3][4] and the stereochem-
ical course of the inhibition reaction by 31P-NMR spectroscopy [5 – 7].
2. Synthesis and Characterization of the 3-Fluoro-2,4-dioxa-8-aza-3-phosphadeca-
lins. – 2.1. Precursor Diols 2 and 3. Following the protocol for the preparation of the
1
)
The trivial name ꢃg-homoacetylcholineꢄ is neither systematic nor is it common in the current
literature. We have introduced this expression for the homologue of acetylcholine [3] and made use
of it since then. Further common names used in the literature are acetyl homocholine and acetyl-g-
homocholine.
ꢂ 2012 Verlag Helvetica Chimica Acta AG, Zꢁrich

18
Helvetica Chimica Acta – Vol. 95 (2012)
Scheme 1a)
racemic compounds [3], the optically active (þ)- and (ꢀ)-trans-, and (þ)- and (ꢀ)-cis-
1-benzyl-3-(hydroxymethyl)piperidin-4-ols ((þ)- and (ꢀ)-2, and (þ)- and (ꢀ)-3, resp.)
were obtained after reduction of ethyl 1-benzyl-4-oxopiperidine-3-carboxylate (1) with
NaBH4 and LiAlH4 (Scheme 2). The resulting mixture (ca. 2 :1) of the diols (ꢁ)-2/(ꢁ)-
3 was transformed in situ into the acetonides (ꢁ)-4/(ꢁ)-5 that could easily be separated
by column chromatography (SiO2) into the pure trans- and cis-diastereoisomers (ꢁ)-4
and (ꢁ)-5, respectively [3][8]. Preparative HPLC (Chiralcelꢀ OD) of (ꢁ)-4 afforded
(þ)-4 ([a]D ¼ þ29.2, ee > 99%), and (ꢀ)-4 ([a]D ¼ ꢀ28.7, ee > 98%)2), and the
2
)
Although the method had been optimized (RS > 4) [2], enantiomerically pure (ꢀ)-4 ([a]D ¼ ꢀ29.3,
ee > 99%) could be obtained only in relatively small amounts, whereas bigger fractions of (ꢀ)-4 did
never exceed ee > 98%. As a consequence, the precursor (þ)-2 and the corresponding
phosphadecalins (ꢀ)-10a and (ꢀ)-10b had only ee > 98% [2]. The [a]D values were determined in
acetone (c ¼ 1).

Helvetica Chimica Acta – Vol. 95 (2012)
19
analogous procedure with (ꢁ)-5 gave (þ)-5 ([a]D ¼ þ66.1, ee > 99%), and (ꢀ)-5
([a]D ¼ ꢀ66.5, ee > 99%). Hydrolysis of the acetonides yielded the trans- and cis-1-
benzyl-3-(hydroxymethyl)piperidin-4-ols (þ)-2 ([a]D ¼ þ9.2, ee > 98%)2), (ꢀ)-2
([a]D ¼ ꢀ9.8, ee > 99%), (þ)-3 ([a]D ¼ þ16.7, ee > 99%), and (ꢀ)-3 ([a]D ¼ ꢀ16.6,
ee > 99%) (Scheme 2). As recently evidenced in detail [8], the absolute configurations
1
were inferred by means of the high-field H-NMR Mosher method [9] applied to the
(R)- and (S)-MTPA ester couples 6a/6b, 7a/7b, 8a/8b, and 9a/9b (Scheme 2).
Scheme 22)
a) NaBH4, Na2CO3, iPrOH, 108 (cis/trans ca. 1:2). b) LiAlH4, Et2O, r.t. c) Me2C(OMe)2, TsOH, CH2Cl2.
d) CC (SiO2, hexane/Et2O) (steps a – d [3]). e) Prep. HPLC (Chiralcelꢅ OD, hexane/EtOH). f) 1n HCl,
608. g) (ꢀ)-(R)- or (þ)-(S)-MTPA-Cl, resp., Et3N, N,N-dimethylpyridin-4-amine (DMAP), CH2Cl2, r.t.
(note: (R)-MTPA-Cl yields the (S)-MTPA-ester and vice versa). h) CC (SiO2, AcOEt).
2.2. 3-Fluoro-2,4-dioxa-8-aza-3-phosphadecalins 10 and 11. Applying the estab-
lished protocol [7], the trans-8-benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-
oxides 10 (Scheme 3) were prepared from (þ)- or (ꢀ)-2 by reaction with POCl2F, and
chromatographic separation of the resulting P(3)-epimer mixture (axial/equatorial ca.
1:1) gave the pure axial epimers (þ)-10a (ee > 99%) and (ꢀ)-10a (ee > 98%)2) and the
pure equatorial epimers (þ)-10b (ee > 99%) and (ꢀ)-10b (ee > 98%)2). Similarly,

20
Helvetica Chimica Acta – Vol. 95 (2012)
starting from (þ)- or (ꢀ)-3, the cis-8-benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin
3-oxides (þ)-11a and (ꢀ)-11a, and (þ)-11b and (ꢀ)-11b were obtained (ee > 99%)
(Scheme 3). However, because the P(3)-equatorially substituted congeners 10b and
11b had a pronounced tendency to epimerize, the chromatographic separation of the
epimers was not as straightforward as for the type-I [1] and type-IV [10] compounds
[2], see Exper. Part. Moreover, due to both facile epimerization at the P-atom and
conformational changes of (þ)- and (ꢀ)-10b and (þ)- and (ꢀ)-11b [11][12] (see Section
2.3), the magnitude of the [a]D values differed in the enantiomer pairs despite of equal
ee (ee > 99% and ee > 98%, resp.)2).
Scheme 32)
a) P(O)Cl2F, pyridine, 08. b) CC (SiO2, Et2O). c) CC (SiO2, hexane/AcOEt/MeOH).
2.3. Conformations of 10 and 11 in Solution. The NMR data of the 2,4-dioxa-8-aza-
3-phosphadecalins 10 and 11 (see Exper. Part) display the same essential features as the
type-I [1] and type-IV [10] congeners (Scheme 1). They confirm the relative
configuration at the P-atom, the double chair conformations of the axial epimers 10a
and 11a and distorted conformations of the 2,4-dioxa-3-phospha moiety in the equatorial

Helvetica Chimica Acta – Vol. 95 (2012)
21
epimers 10b and 11b3). Due to the strongly electronegative F-substituent, the chemical
shift difference (Dd ¼ deq ꢀ dax) is small in the trans-couple 10a/10b (Dd ¼ þ0.5 ppm)
and negative in the cis-couple 11a/11b (Dd ¼ ꢀ0.4 ppm) as discussed earlier [3][11].
As previously presented [1][3][5][6] and directly evidenced [11], the anomeric
(stereoelectronic) effect is the predominant parameter that determines the conforma-
tion of the 3-substituted 2,4-dioxa-3-phosphadecalin 3-oxides. In the 3-axially
substituted 3-fluoro-2,4-dioxa-8-aza-3-phosphadecalins 10a and 11a, both the steric
and the anomeric effects act in the same direction. According to the vicinal couplings
3
(3J(P,HeqꢀC(5)) ¼ 24.8 Hz (10a) and J(P,HeqꢀC(5)) ¼ 24.5 Hz (11a), resp.)3), these
compounds adopt a double-chair conformation (C-1)4)5) (Scheme 4).
In the equatorial epimers 10b and 11b, the steric and the stereoelectronic effects are
opposite. Although the chair conformation is sterically favored, the anomeric
preference drives the strongly electronegative F-substituent into the stereoelectroni-
cally favored axial position, hence resulting in non-chair conformations such as boat or
twist-boats (i.e., B, TB-1, and TB-2)4) of the 2,4-dioxa-3-phospha moiety (Scheme 4)6).
According to the vicinal-coupling data (3J(P,HꢀC(1)) ¼ 2, 3J(P,HaxꢀC(5)) ¼ 11.5,
3J(P,HeqꢀC(5)) ¼ 10.2 Hz), and on the basis of our recent conformational studies on
the type-III [11 – 13] and type-IV congeners [10][12][13] (Scheme 1) that lead to the
conformational assignments of the type-I inhibitors [1], we conclude in analogy that the
trans-configured epimers (þ)- and (ꢀ)-10b exist in an equilibrium mixture of C-1 and
TB-2 (Scheme 4)7).
In the cis-configured equatorial epimers (þ)- and (ꢀ)-11b, the flexibility of the
decalin system combined with the anomeric preference of the P(3)-equatorial F-
substituent render the situation significantly more complex, and additional conforma-
tions must be envisaged. In particular, the bicyclic system can undergo complete ring
3
)
Generally, the 31P-NMR resonance of the axial epimer is shifted upfield with respect to the
equatorial one and the chemical shift difference (Dd ¼ deq ꢀ dax) is > 0, and its magnitude is
inversely proportional to the electronegativity of the substituent at the P-atom. However, the cyclic
phosphorofluoridates of the cis-series of the type I – IV compounds (Scheme 1, X ¼ F) display Dd <
0, a fact that is only explained by significant conformational changes. The magnitude of the 3J(P,H)
1
in the H-coupled 31P-NMR is indicative of the conformation of the 2,4-dioxa-3-phospha moiety:
Diagnostically relevant values for the axial epimers are 3J(P,HeqꢀC(5)) ca. 25 Hz and 3J(P,HaxꢀC(5))
3
3
ca. 0, whereas the equatorial ones display J(P,HaxꢀC(5)) ꢂ J(P,HeqꢀC(5)) ꢂ 10 – 15 Hz. Hence, the
axial epimers exhibit a d-type and the equatorial ones a m-type splitting pattern (see [11]).
The short terms for the conformations (IUPAC convention) were introduced in [11]: C ¼ chair, B ¼
boat, E ¼ envelope, TB ¼ twist-boat (see also [1])
4
5
6
)
)
)
There is no evidence for the completely inverted C-2 conformation in 11a. Hence, we assume that
the strong anomeric preference prevents this interconversion
In such systems, the resulting minimum-energy conformations represent a balance between the
anomeric effect favoring the axial orientation in the twist-boat and the 1,3-steric and eclipsing
interactions favoring the chair conformation. This fact explains the unusual stabilization of non-
chair conformations.
Since 3J(P,HꢀC(1)) is ca. 0, conformations B and TB-1 are excluded. At room temperature, the 31P-
NMR spectra are well resolved and coalescence phenomena are not observed, i.e., the
interconversion C-1 > TB-2 is fast on the NMR time scale. In contrast to [11 – 13], the
conformational assignments are tentative and not corroborated by variable-temperature NMR
experiments nor by X-ray crystallographic analyses.
7
)

22
Helvetica Chimica Acta – Vol. 95 (2012)
Scheme 4
inversion to yield the prominent C-2 arrangement (Scheme 4), which seems to be
favored by both the anomeric and the steric effects. Moreover, although the piperidine
moiety lacks stereoelectronic impact, the conformational equilibrium is also influenced
by the N-benzyl group. According to crystal structures (see [1][11], and Fig. 2 below),
the piperidine moiety adopts a chair conformation in solution with the N-benzyl group
in an equatorial position. However, its steric demand certainly affects the equilibrium
population of the conformers, and it may interfere with a complete ring inversion to C-
2. With regard to the experimental 3J(P,HeqꢀC(5)) ¼ 3J(P,HꢀC(1)) ¼ 18.5 and
3J(P,HaxꢀC(5)) ¼ 7.5 Hz, none of the conformations depicted in Scheme 4 can be
excluded, nor can any be assigned8). Hence, we assume that 11b exists in solution as a
complex mixture of C-1/C-2, and/or TB-1/TB-2, and most likely, also B and envelope E
have to be considered9). A full account of our detailed conformational studies on the 3-
fluoro-2,4-dioxa-3-phosphadecalin 3-oxides (types I – IV) will be presented later [12].
8
)
The most striking argument for the exclusion of a prominent conformation is that none of the
3J(P,H) is ca. 0, see [11].
As discussed for 10b, the conformational assignments are tentative, and the interconversions must
be fast on the NMR time scale at room temperature.
9
)

Helvetica Chimica Acta – Vol. 95 (2012)
23
2.4. Crystallographic Analyses of (ꢀ)-10a and (ꢀ)-11a10). The structures of (ꢀ)-10a
and (ꢀ)-11a were solved and refined successfully. Both compounds in the crystal were
enantiomerically pure, and the absolute configurations of the molecules were
determined independently by the diffraction experiments. The refinement of the
absolute structure parameter confidently confirmed that the refined coordinates (cf.
Table 2, Exper. Part) represent the true enantiomorphs with the expected (1S,3S,6R)-
configuration for (ꢀ)-10a (Fig. 1), and the (1R,3R,6R)-configuration for (ꢀ)-11a
(Fig. 2). These results independently corroborate the configurational assignments of
the precursors (þ)-(3R,4S)-1-benzyl-3-(hydroxymethyl)piperidin-4-ol ((þ)-2) and
(ꢀ)-(3R,4R)-1-benzyl-3-(hydroxymethyl)piperidin-4-ol ((ꢀ)-3) by means of the high-
field 1H-NMR Mosher method. As discussed above, the six-membered ring containing
the P-atom has an undistorted chair conformation with the F-atom in the axial position,
and the N-benzyl group occuppies the sterically favored equatorial position.
Fig. 1. Molecular structure of (ꢀ)-10a ((1S,3S,6R)). Trivial numbering; 50% probability ellipsoids.
All attempts to crystallize one of the equatorial epimers (þ)- or (ꢀ)-10b and/or (þ)-
or (ꢀ)-11b resulted in complete epimerization. Hence, no X-ray crystallographic
analysis of the equatorial compounds is available.
3. Enzyme Kinetics. – 3.1. General. The inhibitory potency and the mode of action
of the enantiomerically pure 3-fluoro-2,4-dioxa-8-aza-3-phosphadecalins 10 and 11 was
determined according to the general considerations and procedure explicitely
described in the precedent article [1]. In particular, the data acquisition was based
on the Ellman assay [14]) and the mathematical evaluation was performed according to
[15]. For irreversible inhibition, the simplified overall processes were considered that
directly yield the covalently phosphorylated enzyme EꢀI (Scheme 5, mechanism a) or
involve a preceding reversible step via the associative complex E · I* (Scheme 5,
mechanism b) [1]. However, since several congeners of the 3-fluoro-2,4-dioxo-8-aza-3-
phosphadecalins were reversible inhibitors of AChE, the basic equilibria that govern
the formation of the adsorptive complexes EI, ES, and ESI had to be considered, too
(Scheme 5, mechanism c) [16].
10
)
The full data sets are summarized in Table 2 (see Exper. Part). CCDC-838432 ((ꢀ)-10a) and CCDC-
838433 ((ꢀ)-11a) contain supplementary crystallographic data. These can be obtained free of
charge from the Cambridge Crystallographic Data Centre via www.code.cam.ac.uk/data_request/cif.

24
Helvetica Chimica Acta – Vol. 95 (2012)
Fig. 2. Molecular structure of (ꢀ)-11a ((1R,3R,6R)). Trivial numbering; 50% probability ellipsoids.
3.2. Data Analysis and Results. Determination of the apparent rate constants (kobs
)
from the progress curves (A ¼ f(t), see Eqn. 1 in the Exper. Part) and the mathematical
evaluation of the dependence kobs ¼ f([I]) to evaluate the irreversible inhibition
parameters (KD, ka, ki, and kp) are performed according to [15]. The reversible
inhibition parameters (KI and KI’) were determined from the linear primary plots (A ¼
f(t)) by application of approved linearization procedures [16][17]. The mathematical
equations (Eqns. 1 – 3) underlying Scheme 5 and further details are summarized in the
Exper. Part. The secondary plot (kobs ¼ f([I])) exhibited a linear dependence for (ꢁ)-
10a, (þ)-10a, and mechanism a was assigned. In the case of (þ)- and (ꢀ)-10b, (ꢀ)-11a,
and (þ)- and (ꢀ)-11b, the secondary plot depended hyperbolically upon [I], and
mechanism b was assigned for these compounds. Since the primary plots (A ¼ f(t)) for
(ꢀ)-10a and (þ)-11a clearly displayed a linear dependence, these compounds are
reversible inhibitors of AChE. According to the comparison of the results of the
different linearization methods [2], the linear mixed inhibition mechanism c could be
assigned (Scheme 5)11). The experimental results are summarized in Table 1.
Compared to the potent standard reference diisopropyl phosphorofluoridate
(P(O)F(OiPr)2) and both the type-I [1] and type-IV [10] congeners, the investigated
3-fluoro-2,4-dioxa-8-aza-3-phosphadecalins 10 and 11 are very weak inhibitors of
AChE, the most potent compound (þ)-10a being only half as active as the weakest
type-I inhibitor. The P(3)-equatorially subsituted congeners (þ)- and (ꢀ)-10b and (þ)-
and (ꢀ)-11b react according to mechanism b and do not display significant stereo-
selectivity. In contrast, the P(3)-axially substituted pairs (þ)- and (ꢀ)-10a and (þ)- and
(ꢀ)-11a are remarkably enantioselective. Whereas (þ)-10a phosphorylates AChE
directly (mechanism a), its enantiomer (ꢀ)-10a is a reversible inhibitor (mechanism c).
11
)
However, the results of the linearization processes were not fully distinct [2], see Exper. Part.

Helvetica Chimica Acta – Vol. 95 (2012)
25
Scheme 5
Table 1. Kinetic Data of the Inhibition of AChE with the Enantiomerically Pure 3-Fluoro-2,4-dioxa-8-aza-
3-phosphadecalins (þ)- and (ꢀ)-10 and (þ)- and (ꢀ)-11 (P(O)F(OiPr)2 as reference)
Kinetic parameters
Mechanism
Kinetic parameters
Mechanism
(þ)-10a ka ¼ 4.3 ꢁ 0.1mꢀ1sꢀ1
(ꢀ)-10a Ki ¼ 1’600 ꢁ 20 mm
Kꢁi ¼ 1’750 ꢁ 20 mm
a
c
(þ)-11a
(ꢀ)-11a
Ki ¼ 930 ꢁ 5 mm
c
K’i ¼ 7’600 ꢁ 30 mm
ki ¼ 18.5 ꢁ 0.7 mꢀ1sꢀ1
KD ¼ 40 ꢁ 5 mm
b
b
b
a
(ꢁ)-10a ka ¼ 1 ꢁ 0.1mꢀ1sꢀ1
(þ)-10b ki ¼ 31.7 ꢁ 0.3 mꢀ1sꢀ1
KD ¼ 105 ꢁ 6 mm
a
b
kp ¼ 0.0007 ꢁ 0.00003 sꢀ1
ki ¼ 75.0 ꢁ 0.5 mꢀ1sꢀ1
KD ¼ 100 ꢁ 5 mm
(þ)-11b
(ꢀ)-11b
kp ¼ 0.0033 ꢁ 0.0004 sꢀ1
(ꢀ)-10b ki ¼ 23.3 ꢁ 0.4 mꢀ1sꢀ1
KD ¼ 73 ꢁ 5 mm
b
kp ¼ 0.0075 ꢁ 0.0001 sꢀ1
ki ¼ 48.3 ꢁ 0.5 mꢀ1sꢀ1
KD ¼ 52 ꢁ 5 mm
kp ¼ 0.0017 ꢁ 0.0001 sꢀ1
kp ¼ 0.0025 ꢁ 0.0003 sꢀ1
P(O)F(OiPr)2 ka ¼ 195 ꢁ 3 mꢀ1sꢀ1

26
Helvetica Chimica Acta – Vol. 95 (2012)
Interestingly, its racemate (ꢁ)-10a reacts irreversibly (mechanism a), thus confirming
the finding that the inhibitory activity of a racemic inhibitor is roughly the arithmetic
mean of the enantiomers [13]12). Similarly, (ꢀ)-11a is an irreversible inhibitor of AChE
(mechanism b), whereas (þ)-11a reacts reversibly with the enzyme (mechanism c).
However, since all compounds are very weak inhibitors of AChE, a clear differentiation
of the primary plot type (A ¼ f(t)) was not always straightforward11). In particular, in
some cases its curvature was not too pronounced and rather converged towards a
straight line as is characteristic for reversible inhibitors. Therefore, the interpretation
and assignment of mechanism c) is not fully assured13).
4. Remarks. – Besides the investigation of the stereochemical implications of the
inhibition reaction by 31P-NMR spectroscopy [5 – 7][13], it is the aim of our research
project to study the molecular interactions of AChE with acetylcholine (ACh). In
particular, kinetic studies are performed to deliver insight into the physiologically
active (recognition) conformation of ACh that governs the enzymatic hydrolysis
process14). Being g-homoacetylcholine mimetics, the 8-benzyl-3-fluoro-2,4-dioxa-8-
aza-3-phosphadecalin 3-oxides 10 and 11 (type-II inhibitors) clearly differ from the
natural substrate. They cannot be compared directly with ACh and are not able to
adopt a respective recognition conformation. Hence, the disclosed very weak inhibitory
potency is not surprising. Moreover, the N-benzyl group as additional structural feature
is considered to influence both the steric demand and the basicity of the compound. In
particular, the cationic site of ACh is substituted by an electron donating group.
However, in spite of these restrictions, the actual results provide some evidence of the
viability of our concept15), but they do not allow to draw significant further
conclusions.
The authors are indebted to PD Dr. A. Linden, head of the X-ray department of our institute, for the
high-quality X-ray crystallographic analyses. The financial support of the project by the Swiss National
Science Foundation is gratefully acknowledged.
Experimental Part
1. General. See [1][3][6]. For the particular precautions in preparing and handling the organo-
phosphates, see [3]. Determination of ee: based on the integration of the peak areas of the anal. HPLC
separations of the precursor diols with optimized resolution (Rs > 4). NMR Assignments: based on
extensive 2D-NMR (see [3]) and selective 1H-decoupling experiments.
12
)
)
Although apparently obvious, it cannot be concluded a priori that the inhibition constants are
simply additive.
13
Meanwhile, we have presented a novel, integrated approach with a reappraisal of kinetic
mechanisms and diagnostic methods that enables distinct as well as subtle differentiations of
generalized inhibition mechanisms [18]. Because the kinetic experiments and the data evaluation
for the (þ)- and (ꢀ)-9-benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-oxides 10 and 11 were
performed earlier [2] than the recent findings [18], the analysis of the assay data follows the
simplified procedure according to [15]. A complete re-interpretation of the kinetic data of 10 and 11
is in process [19].
14
15
)
)
Concerning a discussion on the recognition conformation of ACh, see [1].
At least, our working hypothesis is not discredited by this negative evidence.

Helvetica Chimica Acta – Vol. 95 (2012)
27
2. (þ)-(3R,4S)- and (ꢀ)-(3S,4R)-, and (þ)-(3S,4S)- and (ꢀ)-(3R,4R)-1-Benzyl-3-(hydroxymethyl)-
piperidin-4-ols (¼corresponding 4-Hydroxy-1-(phenylmethyl)piperidine-3-methanols; (þ)- and (ꢀ)-2,
and (þ)- and (ꢀ)-3, resp.). The precursor diols were prepared and characterized as described earlier
[3][8]. The absolute configurations were assigned by means of the high field 1H-NMR Mosher method
[9] applied to the (R)- and (S)-MTPA ester couples 6a/6b, 7a/7b, 8a/8b, and 9a/9b; for explicit details, see
[2][8].
3. trans- and cis-8-Benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-Oxides (¼2-Fluorohexahy-
dro-6-(phenylmethyl)-4H-1,3,2-dioxaphosphorino[5,4-c]pyridine 2-Oxides) (þ)- and (ꢀ)-10a, (þ)- and
(ꢀ)-10b, (þ)- and (ꢀ)-11a, and (þ)- and (ꢀ)-11b, resp. To a cooled soln. (ꢀ 58) of (ꢀ)-2 (136 mg,
0.61 mmol) in anh. CH2Cl2 (2.5 ml) in a glove box (N2 atmosphere), a cooled soln. (ꢀ 58) of POCl2F [20]
(62 ml, 0.68 mmol, 1.11 equiv.) in Et2O (2 ml) was added with a syringe, and the mixture was stirred for
5 min at 08 when the mixture was withdrawn and quickly passed through SiO2 (SiO2 60, 15 –40 mm (Merck,
115111), pH 5.6 [6], Et2O). The eluate was gently evaporated (N2 stream, < 308), and the residue ((þ)-
10a/(þ)-10b, ax/eq ca. 1:1; 167 mg) was subjected to CC (SiO2, pH 5.6 [6], Et2O): (þ)-10a (48 mg, 28%)
and (þ)-10b (43 mg, 25%). Applying the same procedure, phosphorylation of (þ)-2 (98 mg, 0.44 mmol)
afforded (ꢀ)-10a (34 mg, 27%) and (ꢀ)-10b (22 mg, 17%); the axial epimers were less polar.
Similarly, the cis-configured compounds were obtained after phosphorylation and CC (SiO2, pH 5.6
[6], hexane/AcOEt 1:2 containing 3.5% MeOH): from (þ)-3 (132 mg, 0.59 mmol), (þ)-11a (62 mg,
37%) and (ꢀ)-11b (39.2 mg, 41%); from (ꢀ)-3 (162 mg, 0.61 mmol), (ꢀ)-11a (81 mg, 43%) and (þ)-11b
(64 mg, 37%); the equatorial epimers were less polar.
In contrast to the racemic compounds [3], the optically active P(3)-equatorially substituted 3-fluoro-
2,4-dioxa-8-aza-3-phosphadecalins partially epimerized during CC. Therefore, the contact time to SiO2
had to be minimized by an optimized adsorbent/substance ratio (ca. 25 mg on a 8 ꢃ 1 cm column) and
high flow rates (ca. 4 ml/min). However, although the operating conditions were optimized, (þ)- and
(ꢀ)-10b contained ca. 4 – 5% of (þ)- and (ꢀ)-10a [2]. The axial epimers were stable when stored as such
(<08, Ar), but in soln. they slowly decomposed. Although being resonably stable when stored as such
(<08, Ar), the equatorial epimers gradually epimerized; in particular, the trans-congeners (þ)- and (ꢀ)-
10b completely epimerized during crystallization to yield (þ)- and (ꢀ)-10a, resp. [2].
(þ)-(1R,3R,6S)-8-Benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-Oxide (¼(þ)-(2R,4aS,
8aR)-2-Fluorohexahydro-6-(phenylmethyl)-4H-1,3,2-dioxaphosphorino[5,4-c]pyridine 2-Oxide; (þ)-
10a): Colorless prisms. M.p. 174 – 175.58. Rf (Et2O) 0.37. [a]D ¼ þ30.3 (c ¼ 1.00, acetone; ee > 99%).
IR (KBr): 3008w, 3060w, 3026m, 2961m, 2938m 2869w, 2820m, 2805m, 2778m, 2728w, 1493m, 1486w,
1471w, 1449m, 1393w, 1368w, 1322s, 1264w, 1218m, 1182m, 1142m, 1134w, 1105m, 1074w, 1045s, 997s,
968m, 915m, 899s, 846w, 799m, 786m, 741m, 698m, 663m, 535m. 1H-NMR (400 MHz, CDCl3): 7.36 – 7.25
(m, PhCH2); 4.29 (A of ABX-P, 2J ¼ 11.0, 3J(5eq,P) ¼ 24.5, 3J(5eq,6) ¼ 4.5, HeqꢀC(5)); 4.18 (td, 3J(1,6) ¼
3J(1,10ax) ¼ 12.0, 3J(1,10eq) ¼ 4.8, HꢀC(1)); 4.11 (B of ABX-P, 2J ¼ 3J(5ax,6) ¼ 11.0, 3J(5ax,P) ¼ 1.0,
2
2
3
3
HaxꢀC(5)); 3.58, 3.47 (AB, J ¼ 13.2, PhCH2); 3.01 (dddd, J ¼ 12.0, J(9eq,10ax) ¼ 4.5, J(9eq,10eq) ¼
2.5, 4J(9eq,7eq) ¼ 2.2, HeqꢀC(9)); 2.81 (ddd, 2J ¼ 11.2, 3J(7eq,6) ¼ 3.8, 4J(7eq,9eq) ¼ 2.2, HeqꢀC(7));
2.34 (X of ABX-P, 3J(6,1) ¼ 12.0, 3J(6,5ax) ¼ 11.0, 3J(6,5eq) ¼ 4.5, 3J(6,7ax) ¼ 11.2, 3J(6,7eq) ¼ 3.8,
HꢀC(6)); 2.12 (td, 2J ¼ 3J(9ax,10ax) ¼ 12.0, 3J(9ax,10eq) ¼ 2.5, HaxꢀC(9)); 2.08 (ddt, 2J ¼ 12.0,
3J(10eq,1) ¼ 4.8, 3J(10eq,9ax) ¼ 3J(10eq,9eq) ¼ 2.5, HeqꢀC(10)); 1.94 (qd, 2J ¼ 3J(10ax,1) ¼
3J(10ax,9ax) ¼ 12.0, 3J(10ax,9eq) ¼ 4.5, HaxꢀC(10)); 1.70 (t, 2J ¼ 3J(7ax,6) ¼ 11.2, HaxꢀC(7)). 13C-NMR
(100.6 MHz, CDCl3): 137.4 (C(1’)); 128.7 (C(2’), C(6’)); 128.3 (C(3’), C(5’)); 127.4 (C(4’)); 83.0 (d,
2J(1,P) ¼ 6.5, C(1)); 71.8 (d, 2J(5,P) ¼ 7.6, C(5)); 62.9 (PhCH2); 51.1 (C(9)); 51.0 (C(7)); 39.6 (d,
3J(6,P) ¼ 6.2, C(6)); 31.6 (d, 3J(10,P) ¼ 9.2, C(10)). 31P-NMR (161.9 MHz, CDCl3): ꢀ 15.9 (dd, 1J(P,F) ¼
1010, 3J(P,HeqꢀC(5)) ¼ 24.5. 19F{1H}-NMR (376.5 MHz, CDCl3): ꢀ 85.3 (d, 1J(F,P) ¼ 1010). EI-MS: 285
(25, Mþ), 208 (8), 194 (10, [M ꢀ PhCH2]þ), 186 (14), 185 (13), 172 (12), 146 (3), 132 (8), 126 (17), 118
(4), 94 (23), 92 (25), 91 (100, PhCHþ2 ), 67 (5), 65 (17), 56 (3).
(ꢀ)-(1S,3S,6R)-8-Benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-Oxide (¼(ꢀ)-(2S,4aR,8aS)-
2-Fluorohexahydro-6-(phenylmethyl)-4H-1,3,2-dioxaphosphorino[5,4-c]pyridine 2-Oxide; (ꢀ)-10a):
[a]D ¼ ꢀ28.9 (c ¼ 1.00, acetone; ee > 98%)2). All other data: identical with those of (þ)-10a.
(þ)-(1R,3S,6S)-8-Benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-Oxide (¼(þ)-(2S,4aS,8aR)-
2-Fluorohexahydro-6-(phenylmethyl)-4H-1,3,2-dioxaphosphorino[5,4-c]pyridine 2-Oxide; (þ)-10b):

28
Helvetica Chimica Acta – Vol. 95 (2012)
Colorless prisms. M.p. 113 – 114816). Rf (Et2O): 0.29. [a]2D5 ¼ þ47.5 (c ¼ 1.00, acetone; ee > 99%). IR
(KBr): 3065w, 3028w, 2978w, 2920w, 2869w, 2823m, 2780w, 1605w, 1498m, 1479w, 1455m, 1393m, 1368m,
1345s, 1327s, 1280w, 1205w, 1183w, 1178w, 1148w, 1131w, 1109m, 1085m, 1069m, 1042s, 999m, 975m, 959m,
915m, 897s, 849m, 772m, 705w, 635m, 575m. 1H-NMR (400 MHz, CDCl3): 7.36 – 7.25 (m, PhCH2); 4.37 (A
of ABX-P, 2J ¼ 11.5, 3J(5eq,P) ¼ 10.2, 3J(5eq,6) ¼ 5.7, HeqꢀC(5))17); 4.31 (tdd, 3J(1,10ax) ¼ 3J(1,6) ¼ 11.5,
3J(1,10eq) ¼ 4.5, 4J(1,F) ¼ 2.0, HꢀC(1)); 4.13 (B of ABX-P, 2J ¼ 3J(5ax,P) ¼ 11.5, 3J(5ax,6) ¼ 10.5,
4J(5ax,F) ¼ 3.7, HaxꢀC(5))17); 3.58, 3.46 (AB, 2J ¼ 13.2, PhCH2); 3.01 (dddd, 2J ¼ 11.5, 3J(9eq,10ax) ¼
3
4
2
3
4.5, J(9eq,10eq) ¼ 2.5, J(9eq,7eq) ¼ 2.2, HeqꢀC(9)); 2.84 (ddd, J ¼ 11.0, J(7eq,6) ¼ 3.8, 4J(7eq,9eq) ¼
2.2, HeqꢀC(7)); 2.48 (X of ABX-P, 3J(6,1) ¼ 11.5, 3J(6,5ax) ¼ 10.5, 3J(6,5eq) ¼ 5.7, 3J(6,7ax) ¼ 11.0,
3J(6,7eq) ¼ 3.8, HꢀC(6)); 2.15 (dtd, 2J ¼ 11.5, 3J(10eq,1) ¼ 3J(10eq,9eq) ¼ 4.5, 3J(10eq,9ax) ¼ 2.5,
HeqꢀC(10)); 2.12 (td, J ¼ 3J(9ax,10ax) ¼ 11.5, J(9ax,10eq) ¼ 2.5, HaxꢀC(9)); 1.90 (qd, J ¼ 3J(10ax,1) ¼
3J(10ax,9ax) ¼ 11.5, 3J(10ax,9eq) ¼ 4.2, HaxꢀC(10)); 1.72 (t, 2J ¼ 3J(7ax,6) ¼ 11.0, HaxꢀC(7)). 13C-NMR
(100.6 MHz, CDCl3): 137.4 (C(1’)); 128.7 (C(2’), C(6’)); 128.3 (C(3’), C(5’)); 127.4 (C(4’)); 82.3 (d,
2J(1,P) ¼ 6.6, C(1)); 71.4 (d, 2J(5,P) ¼ 7.1, C(5)); 62.0 (PhCH2); 52.0 (C(7)); 51.0 (C(9)); 38.5 (d,
3J(6,P) ¼ 12.6, C(6)); 32.0, 3J(10,P) ¼ 7.1 (C(10)). 31P-NMR (161.9 MHz, CDCl3): ꢀ 15.4 (dddt, 1J(P,F) ¼
998, 3J(P,HaxꢀC(5)) ¼ 11.5, 3J(P, HeqꢀC(5)) ¼ 10.2, 3J(P,HꢀC(1)) ¼ 4J(P,HꢀC(6)) ¼ 2)17). 19F{1H}-NMR
(376.5 MHz, CDCl3): ꢀ 69.3 (d, 1J(F,P) ¼ 998). EI-MS: 285 (22, Mþ), 208 (10), 194 (32, [M ꢀ PhCH2]þ),
186 (6), 172 (3), 160 (3), 132 (8), 126 (16); 112 (18), 94 (15), 92 (25), 91 (100, PhCHþ2 ), 67 (10), 65 (15),
55 (5).
2
3
2
(ꢀ)-(1S,3R,6R)-8-Benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-Oxide (¼(ꢀ)-(2R,4aR,8aS)-
2-Fluorohexahydro-6-(phenylmethyl)-4H-1,3,2-dioxaphosphorino[5,4-c]pyridine 2-Oxide; (ꢀ)-10b):
[a]D ¼ ꢀ42.3 (c ¼ 1.00, acetone; ee > 98%)2). All other data: identical with those of (þ)-10b.
(þ)-(1S,3S,6S)-8-Benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-Oxide (¼(þ)-(2S,4aS,8aS)-2-
Fluorohexahydro-6-(phenylmethyl)-4H-1,3,2-dioxaphosphorino[5,4-c]pyridine 2-Oxide; (þ)-11a): Col-
orless prisms. M.p. 111.5 – 1138. Rf (AcOEt/MeOH 98 :2) 0.20. [a]D ¼ þ47.0 (c ¼ 1.00, acetone; ee >
99%). IR (KBr): 3088w, 3065w, 3029w, 2968m, 2920m, 2908m, 2833w, 2808m, 2779m, 1500w, 1479w,
1456w, 1424w, 1386w, 1372w, 1366w, 1355w, 1338m, 1311s, 1278w, 1212m, 1161w, 1125w, 1099m, 1077s,
1048m, 1028s, 987s, 973m, 952m, 915m, 893s, 835m, 790w, 773m, 739m, 726m, 642m, 509m. 1H-NMR
(400 MHz, CDCl3): 7.35 – 7.23 (m, PhCH2); 4.87 (qd-like, 3J(1,6) ¼ 3J(1,10ax) ¼ 3J(1,10eq) ¼ 2.6, 4J(1,P) ¼
2
3
3
1.5, HꢀC(1)); 4.53 (A of ABX-P, J ¼ 11.7, J(5ax,P) < 1, J(5ax,6) ¼ 2.9, HaxꢀC(5)); 4.22 (B of ABX-P,
2J ¼ 11.7, 3J(5eq,P) ¼ 24.8, 3J(5eq,6) < 1, HeqꢀC(5)); 3.60, 3.50 (AB, 2J ¼ 13.0, PhCH2); 2.72 (dquint.-like,
3
2J ¼ 11.8, 3J(9eq,10ax) ꢂ J(9eq,10eq) ¼ 2.4, 4J(9eq,7eq) ¼ 1.0, HeqꢀC(9)); 2.69 (ddd, 2J ¼ 11.8,
3J(7eq,6) ¼ 4.6, 4J(7eq,9eq) ¼ 1.0, HeqꢀC(7)); 2.54 (t, 2J ¼ 3J(7ax,6) ¼ 11.8, HaxꢀC(7)); 2.39 (td, 2J ¼
3J(9ax,10ax) ¼ 11.8, 3J(9ax,10eq) ¼ 3.2, HaxꢀC(9)); 2.06 (X of ABX-P, 3J(6,1) ¼ 2.6, 3J(6,5ax) ¼ 2.9,
3J(6,5eq) < 1, 3J(6,7ax) ¼ 11.8, 3J(6,7eq) ¼ 4.6, 4J(6,P) ¼ 1.5, HꢀC(6)); 2.01 (br. dddd, 2J ¼ 14.5,
3
3
4
2
3J(10eq,1) ¼ 2.6, J(10eq,9ax) ¼ 3.2, J(10eq,9eq) ¼ 2.4, J(10eq,P) ꢂ 1, HeqꢀC(10)); 1.96 (qdd-like, J ¼
14.5, 3J(10ax,1) ¼ 2.6, 3J(10ax,9ax) ¼ 11.8, 3J(10ax,9eq) ¼ 2.4, 4J(10ax,P) ¼ 6.0, HaxꢀC(10)). 13C-NMR
(100.6 MHz, CDCl3): 137.9 (C(1’)); 129.0 (C(2’), C(6’)); 128.3 (C(3’), C(5’)); 127.3 (C(4’)); 78.5 (dd,
2J(1,P) ¼ 7.3, J(1,F) ¼ 1.5, C(1)); 72.3 (d, J(5,P) ¼ 6.9, C(5)); 62.9 (PhCH2); 49.5 (C(7)); 46.7 (C(9));
35.5 (d, 3J(6,P) ¼ 5.9, C(6)); 31.0 (d, 3J(10,P) ¼ 8.8, C(10)). 31P-NMR (161.9 MHz, CDCl3): ꢀ 16.6 (ddd,
1J(P,F) ¼ 1009, 3J(P, HeqꢀC(5)) ¼ 24.8, 4J(P,HaxꢀC(10)) ¼ 6.0, 4J(P,HꢀC(1)) ¼ 4J(P,HꢀC(6) ¼ 1.5).
19F{1H}-NMR (376.5 MHz, CDCl3): ꢀ 85.1 (d, 1J(F,P) ¼ 1009). EI-MS: 285 (22, Mþ), 208 (11), 194
(28, [M ꢀ PhCH2]þ), 172 (6), 144 (5), 133 (10), 118 (7), 94 (20), 92 (23), 91 (100, PhCHþ2 ), 67 (10), 65
(16).
3
2
16
)
)
The specified m.p. corresponds to (ꢁ)-10b [3]. As discussed above, (þ)- and (ꢀ)-10b epimerize
during the crystallization process to yield (þ)- and (ꢀ)-10a, resp. (colorless needles, m.p. ca. 170 –
1758).
The descriptors ꢃaxꢄ and ꢃeqꢄ for the H-atoms at C(5) are based on their relative positions in the
chair conformation of the 2,4-dioxa-3-phospha moiety. As discussed (Scheme 4), the conformation is
rather a twist-boat (TB-2) than a chair in the P(3)-equatorially substituted compounds. For reasons
of simplicity, the notation ꢃaxꢄ and ꢃeqꢄ is maintained. HaxꢀC(5) is always cis to HꢀC(1) and
HeqꢀC(5) trans to HꢀC(1), see [11].
17

Helvetica Chimica Acta – Vol. 95 (2012)
29
(ꢀ)-(1R,3R,6R)-8-Benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-Oxide (¼(ꢀ)-(2R,4aR,
8aR)-2-Fluorohexahydro-6-(phenylmethyl)-4H-1,3,2-dioxaphosphorino[5,4-c]pyridine 2-Oxide; (ꢀ)-
11a): [a]D ¼ ꢀ46.7 (c ¼ 1.00, acetone; ee > 99%). All other data: identical with those of (þ)-11a.
(þ)-(1R,3S,6R)-8-Benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-Oxide (¼(þ)-(2S,4aR, 8aR)-
2-Fluorohexahydro-6-(phenylmethyl)-4H-1,3,2-dioxaphosphorino[5,4-c]pyridine 2-Oxide; (þ)-11b):
Colorless viscous oil. Rf (AcOEt/MeOH 98 :2) 0.32. [a]D ¼ þ18.3 (c ¼ 1.00, acetone; ee > 99%). IR
(KBr): 3029w, 2965m, 2818m, 1495m, 1476w, 1451m, 1389w, 1330s, 1238w, 1203w, 1174m, 1133m, 1058s,
1
1029s, 1005m, 942m, 906w, 850m, 800w, 782w, 741m, 696m, 625w, 512m. H-NMR (400 MHz, CDCl3):
2
3
3
7.35 – 7.22 (m, PhCH2); 4.82 (A of ABX-P, J ¼ 11.5, J(5ax,P) ¼ 7.5, J(5ax,6) ¼ 4.6, HaxꢀC(5))17); 4.71
(quint.d-like, 3J(1,P) ¼ 18.5, 3J(1,10ax) ¼ 9.8, 3J(1,6) ꢂ J(1,10eq) ꢂ J(1,F) ꢂ 5, HꢀC(1)); 4.36 (B of ABX-
P, 2J ¼ 11.5, 3J(5eq,P) ¼ 18.5, 3J(5eq,6) ¼ 4.6, HeqꢀC(5))17); 3.50, 3.40 (AB, 2J ¼ 13.0, PhCH2); 2.85 (m, br.
3
4
d-like, w1/2 ꢂ 22, 2J ¼ 11, HeqꢀC(9)); 2.71 (ddd, 2J ¼ 12.5, 3J(7eq,6) ¼ 3.8, J(7eq,9eq) ¼ 1.6, HeqꢀC(7));
4
2.51 (X of ABX-P, quint.-like, w1/2 ꢂ 22, HꢀC(6)); 2.32 (dd, 2J ¼ 12.5, 3J(7ax,6) ¼ 3.5, HaxꢀC(7)); 2.25 (br.
qd-like, 2J ꢂ 12, 3J(10ax,9ax) ꢂ J(10ax,1) ꢂ 10, 3J(10ax,9eq) ꢂ 5, HaxꢀC(10)); 2.17 (br. t-like, 2J ꢂ
3
3J(9ax,10ax) ꢂ 11, HaxꢀC(9)); 2.08 (m, w1/2 ꢂ 25, HeqꢀC(10)). 13C-NMR (100.6 MHz, CDCl3): 137.8
(C(1’)); 128.6 (C(2’), C(6’)); 128.4 (C(3’), C(5’)); 127.4 (C(4’)); 80.7 (dd, 2J(1,P) ¼ 6.5, 3J(1,F) ¼ 2, C(1));
2
3
70.1 (d, J(5,P) ¼ 6.9, C(5)); 62.1 (PhCH2); 51.7 (C(7)); 50.9 (C(9)); 35.1 (d, J(6,P) ¼ 8.1, C(6)); 28.5
(C(10)). 31P-NMR (161.9 MHz, CDCl3): ꢀ 16.1 (br. dtd, 1J(P,F) ¼ 1001, 3J(P,HꢀC(1)) ¼
3J(P,HeqꢀC(5))) ¼ 18.5, 3J(P, HaxꢀC(5)) ¼ 7.5 )17). 19F{1H}-NMR (376.5 MHz, CDCl3): ꢀ 74.8 (d,
1J(F,P) ¼ 1001). EI-MS: 285 (22, Mþ), 208 (11), 194 (28, [M ꢀ PhCH2]þ), 172 (6), 133 (10), 132 (14),
94 (20), 92 (23), 91 (100, PhCHþ2 ), 67 (10), 65 (16).
(ꢀ)-(1S,3R,6S)-8-Benzyl-3-fluoro-2,4-dioxa-8-aza-3-phosphadecalin 3-Oxide (¼(ꢀ)-(2R,4aS,8aS)-
2-Fluorohexahydro-6-(phenylmethyl)-4H-1,3,2-dioxaphosphorino[5,4-c]pyridine 2-Oxide; (ꢀ)-11b):
[a]D ¼ ꢀ18.4 (c ¼ 1.00, acetone; ee > 99%). All other data: identical with those of (þ)-11b.
4. X-Ray Crystal-Structure Determinations of (ꢀ)-10a and (ꢀ)-11a10). 4.1. General. All measure-
ments were made by means of a Nonius KappaCCD area-detector diffractometer [21] with graphite-
monochromated MoKa radiation (l 0.71073 ꢆ) and an Oxford-Cryosystems-Cryostream-700 cooler.
Data reduction was performed with HKL DENZO and SCALEPACK [22]. The intensities were
corrected for Lorentz and polarization effects, and an absorption correction based on the multi-scan
method [23] was applied. Neutral-atom scattering factors for non-H-atoms were taken from [24], and the
scattering factors for H-atoms were taken from [25]. Anomalous dispersion effects were included in Fc
[26]; the values for f’ and f’’ were those of [27]. The values of the mass attenuation coefficients are those
of [28]. All calculations were performed with the SHELXL97 program [29], and the crystallographic
diagrams were drawn with ORTEPII [30].
4.2. Determination of (ꢀ)-10a. The unit cell constants and an orientation matrix for data collection
were obtained from a least-squares refinement of the setting angles of 48331 reflections in the range 48 <
2q < 608. The mosaicity was 0.968(2)8. A total of 804 frames were collected by using f and w scans with k
offsets, 17 s exposure time, a rotation angle of 0.78 per frame, and a crystalꢀdetector distance of 38.6 mm.
The space group was uniquely determined by the systematic absences. Equivalent reflections were
merged. The data collection and refinement parameters are given in Table 2. A view of the molecule is
shown in Fig. 1. The structure was solved by direct methods with SIR92 [31], which revealed the positions
of all non-H-atoms. The non-H-atoms were refined anisotropically. All of the H-atoms were placed in
geometrically calculated positions and refined by using a riding model where each H-atom was assigned a
fixed isotropic displacement parameter with a value equal to 1.2 Ueq of its parent atom. The refinement of
the structure was carried out on F2 by using full-matrix least-squares procedures, which minimized the
function Sw(F2o ꢀ Fc2 )2. The weighting scheme was based on counting statistics and included a factor to
downweight the intense reflections. Plots of Sw(F2o ꢀ Fc2 )2 vs. Fc/Fc(max) and resolution showed no
unusual trends. A correction for secondary extinction was applied. Refinement of the absolute structure
parameter [32] yielded a value of ꢀ 0.06(16), which confirms that the refined model corresponds with
the true enantiomorph, although the precision of the determination is low.
4.3. Determination of (ꢀ)-11a. The unit cell constants and an orientation matrix for data collection
were obtained from a least-squares refinement of the setting angles of 11546 reflections in the range 48 <
2q < 558. The mosaicity was 1.016(3)8. A total of 322 frames were collected by using f and w scans with k

30
Helvetica Chimica Acta – Vol. 95 (2012)
Table 2. Crystallographic Data of (ꢀ)-10a and (ꢀ)-11a
(ꢀ)-10a
(ꢀ)-11a
Crystallized from
Empirical formula
Mr
pentane/Et2O
C13H17FNO3P
285.25
pentane/Et2O
C13H17FNO3P
285.25
Crystal color, habit
Crystal dimensions [mm]
Temperature [K]
colorless, prism
0.12 ꢃ 0.20 ꢃ 0.32
160(1)
colorless, prism
0.17 ꢃ 0.28 ꢃ 0.40
160(1)
Crystal system
Space group
Z
orthorhombic
P212121 (#19)
4
triclinic
P1 (#1)
1
Reflections for cell determination
48331
11546
2q Range for cell determination [8]
4 – 60
4 – 55
Unit cell parameters:
a [ꢆ]
b [ꢆ]
c [ꢆ]
a [8]
b [8]
5.6609(2)
10.0766(4)
23.0798(8)
90
90
90
1316.53(8)
600
1.439
0.224
f and w
60
5.5530(2)
7.7169(4)
8.7115(4)
104.506(2)
91.269(3)
107.293(3)
343.18(3)
150
1.380
0.215
f and w
55
g [8]
V [ꢆ3]
F(000)
Dx [g cmꢀ3
]
m(MoKa) [mmꢀ1
Scan type
]
2q(max) [8]
Transmission factors (min; max)
Total reflections measured
Symmetry independent reflections
Rint
0.702; 0.974
16761
3817
0.105
0.873; 0.965
8012
2854
0.046
Reflections with I > 2s(I)
Reflections used in refinement
Parameters refined
2715
3817
173
2769
2854
172; 3
Final R(F) (I > 2s(I) reflections)
0.0586
0.1503
0.0442
0.1079
wR(F2) (all data)
Weights
w ¼ [s2(F2o ) þ (0.073P)2
þ 0.4328P]ꢀ1
where P ¼ (F2o þ 2Fc2 )/3
1.034
w ¼ [s2(F2o ) þ (0.0543P)2
þ 0.1646P]ꢀ1 where
P ¼ (F2o þ 2F2c )/3
1.096
Goodness of fit
Secondary extinction coefficient
Final Dmax/s
0.018(3)
0.001
–
0.001
D1 (max; min) [e ꢆꢀ3
s(d(CꢀC)) [ꢆ]
]
0.45; ꢀ 0.33
0.37; ꢀ 0.24
0.003 – 0.005
0.003 – 0.004
offsets, 46 s exposure time, a rotation angle of 2.08 per frame, and a crystalꢀdetector distance of 30.0 mm.
The space group was determined from packing considerations, a statistical analysis of intensity
distribution, and the successful solution and refinement of the structure. Equivalent reflections, other
than Friedel pairs, were merged. The data collection and refinement parameters are given in Table 2. A
view of the molecule is shown in Fig. 2. The structure was solved by direct methods with SIR92 [31],
which revealed the positions of all non-H-atoms. The non-H-atoms were refined anisotropically. All of
the H-atoms were placed in geometrically calculated positions and refined by using a riding model where

Helvetica Chimica Acta – Vol. 95 (2012)
31
each H-atom was assigned a fixed isotropic displacement parameter with a value equal to 1.2Ueq of its
parent atom. The refinement of the structure was carried out on F2 by using full-matrix least-squares
procedures, which minimized the function Sw(F2o ꢀ Fc2 )2. The weighting scheme was based on counting
statistics and included a factor to downweight the intense reflections. Plots of Sw(F2o ꢀ Fc2 )2 vs. Fc/Fc(max)
and resolution showed no unusual trends. A correction for secondary extinction was not applied.
Refinement of the absolute structure parameter [32] yielded a value of ꢀ 0.07(10), which confidently
confirms that the refined model corresponds with the true enantiomorph.
5. Enzyme Kinetics. 5.1. General. For the detailed experimental procedure and the methods of data
analysis, see the preceding report [1]. The following parameters are as in [1]: Apparatus and general
experimental conditions, phosphate buffer pH 7.00, AChE soln., ATC and DTNB solns., inhibitor solns.,
and the determination of Km. P(O)F(OiPr)2 (diisopropyl fluorophosphate) was used as the standard
reference.
5.2. Ellman Assay [14]. In a polystyrene cell (4 ml, d ¼ 1 cm), phosphate buffer pH 7.00 (2 ml),
DTNB soln. (100 ml), and ATC soln. (20 ml) were mixed and thermostatted at 258 (ca. 5 min). Then
inhibitor soln. (x ml ¼ 25, 20, 15, 10, and 5 ml; known [I]), and MeCN ((25 – x) ml) were added. At t ¼ 0,
the AChE soln. (1 ml) was added and the mixture gently mixed for 5 s. After 10 s, the monitoring of the
absorption at 412 ꢁ 2 nm (liberated bis-anion of 5-sulfanyl-2-nitrobenzoic acid) automatically started,
and 900 data points were collected for 15 min at various concentrations of the inhibitor. As in the Km
determinations, the total volume was 3.15 ml, the concentration of the substrate [S] was 500 mm. Since all
of the compounds were weak inhibitors, relatively high [I] were necessary, and due to solubility reasons,
[I] < 300 mm had to be maintained. Per inhibitor, at least five measurements with different inhibitor
concentrations were performed.
5.3. Data Analysis. a) Irreversible Inhibitors [15]. The integrated rate equation describing product
generation (monitored by the absorbance A at l 412 ꢁ 2 nm) and the apparent rate constants (kobs) is
given by Eqn. 1. It is fitted (R > 0.999) to progress curves recorded at fixed [S] and variable [I] (primary
plot, A ¼ f(t)) to obtain a series of kobs values and their standard errors (SE). The inhibition parameters
are obtained from the secondary plots (kobs ¼ f([I])) that result from weighted (SEꢀ2) linear or nonlinear
regression according to Eqns. 2 or 3. The analysis of these plots enables a differentiation between the
inhibition mechanisms: kobs depends linearly upon the inhibitor concentration for mechanism a and
hyperbolically for mechanism b (see Scheme 5). For mechanism a, the ka values are calculated according
to Eqn. 2, its slope kobs/[I] is obtained from the linear regression (Scheme 5,a). The decisive plot for
mechanism b is doubly reciprocal (1/kobs ¼ f(1/[I]), Eqn. 3) and the KD and kp values are calculated by
linear regression, the slope being (KD/kp)(1 þ [S]/Km) and the intercept 1/kp. The overall inhibitory
potency (ki) is expressed by kp/KD (see Scheme 5,b).
vz
kobs
obsꢄt
A ¼
ð1 ꢀ eꢀk
Þ
(1)
(2)
ꢀ
ꢁꢀ
ꢁ
kobs
½Iꢅ
½Sꢅ
ka ¼
1 þ
Km
ꢀ
ꢁꢀ
ꢁꢀ
ꢁ
1
kobs
KD
kp
½Sꢅ
1
½Iꢅ
1
kp
18
)
¼
1 þ
þ
(3)
Km
b) Reversible Inhibitors [16]. Since all the compounds were weak inhibitors, a clear differentiation of
the primary plot type was not always straightforward. In particular, in some cases, the curvature of the
primary plot (A ¼ f(t)) was not too pronounced and rather converged towards a straight line as is
characteristic for reversible inhibitors. When reversible behavior was assumed, the assays were
performed each at five different inhibitor ([I]) and substrate concentrations ([S]) resulting in 25 data
points per inhibitor. The inhibition parameters were evaluated from the secondary plots that were
obtained by linearization methods according to LineweaverꢀBurk (1/kobs ¼ f(1/[S])) [16], HanesꢀWoolf
18
)
The last term in Eqn. 3 was erroneously printed as a subtrahend in [1].

32
Helvetica Chimica Acta – Vol. 95 (2012)
([S]/kobs ¼ f([S])) [16], Dixon (1/kobs ¼ f([I])) [16], and Chan (1 þ [S]/Km)(v0 ꢀ vi)/vi[I] ¼ f([S]/Km))
[17]. Comparative analysis of the individual results [2] lead to the differentiation of the various reversible
inhibition mechanisms (competitive, noncompetitive, uncompetitive, linear mixed type) [16][17] (see
Scheme 5,c).
5.4. Results (Table 1). The secondary plot (kobs ¼ f([I])) exhibited a linear dependence for (ꢁ)-10a,
(þ)-10a, and P(O)F(iPr)2. Hence, mechanism a was assigned. In the case of (þ)- and (ꢀ)-10b, (ꢀ)-11a,
and (þ)- and (ꢀ)-11b, the secondary plot depended hyperbolically upon [I], and mechanism b was
assigned for these compounds. Since the primary plots (A ¼ f(t)) for (ꢀ)-10a and (þ)-11a clearly
displayed a linear dependence, these compounds are reversible inhibitors of AChE. According to the
comparison of the different linearization methods [2], the linear mixed inhibition mechanism c could be
assigned [16] (Scheme 5). However, the results of the linearization processes were ambiguous to some
extent as exemplified for (þ)-11a: Data evaluation according to LineveaverꢀBurk suggested a linear
mixed inhibition mechanism (intersection above the abscissa; from the intercepts (Dx, Dy) the modifying
factors a and a’ were calculated (a ¼ꢀ a’/DxKm, a’ ¼ vmaxDy, and KI ¼ [I]/(a ꢀ 1), K’I ¼ [I]/(a’ ꢀ 1)) [16]:
KI ¼ 947 mm, KI’¼ 7050 mm (a ¼ 7.44) [2]. A similar result was obtained from the approach according to
Chan (KI ¼ 1/Dy, slope ¼ 1/aKI, KI’¼ aKI) [17]: KI ¼ 919 mm, KI’¼ 8100 mm (a ¼ 8.81) [2]. The value in
Table 1 is the rounded mean of these two approaches. The interpretation of the HanesꢀWoolf diagram
[16] resulted in a borderline case between noncompetitive and linear mixed inhibition (intersection just
above the abscissa) [2]. Linearization according to Dixon [16] was ambiguous, too: The parallel straight
lines are indicative for competititve inhibition, while others suggest a linear mixed mechanism [2].
Moreover, the rather close similarity of KI and K’I of (ꢀ)-10a may suggest a noncompetitive mechanism
for this compound.
REFERENCES
[1] P. Lorenzetto, M. Wꢀchter, P. Rꢁedi, Helv. Chim. Acta 2011, 94, 746.
[2] C. Clerc, ꢃSynthese und Charakterisierung von enantiomerenreinen N-heterocyclischen Phospha-
decalinen als Inhibitoren von Acetylcholinesteraseꢄ, Ph.D. Thesis, University of Zurich, 2008.
[3] S. Furegati, W. Ganci, F. Gorla, U. Ringeisen, P. Rꢁedi, Helv. Chim. Acta 2004, 87, 2629.
[4] S. Furegati, F. Gorla, A. Linden, P. Rꢁedi, Chem.-Biol. Interact. 2005, 157 – 158, 415.
[5] W. Ganci, E. J. M. Meier, F. A. Merckling, G. Przibille, U. Ringeisen, P. Rꢁedi, Helv. Chim. Acta
1997, 80, 421; S. Furegati, W. Ganci, G. Przibille, P. Rꢁedi, Helv. Chim. Acta 1998, 81, 1127.
[6] M. J. Stçckli, P. Rꢁedi, Helv. Chim. Acta 2001, 84, 106; M. J. Stçckli, P. Rꢁedi, Helv. Chim. Acta 2007,
90, 2058.
[7] S. Furegati, O. Zerbe, P. Rꢁedi, Chem.-Biol. Interact. 2005, 157 – 158, 418.
[8] C. Clerc, I. Matarazzo, P. Rꢁedi, Helv. Chim. Acta 2009, 92, 14.
[9] J. A. Dale, D. L. Dull, H. S. Mosher, J. Org. Chem. 1969, 34, 2543; G. R. Sullivan, J. A. Dale, H. S.
Mosher, J. Org. Chem. 1973, 38, 2143; J. A. Dale, H. S. Mosher, J. Am. Chem. Soc. 1973, 95, 512; I.
Ohtani, T. Kusumi, Y. Kahman, H. Kakisawa, J. Am. Chem. Soc. 1991, 113, 4092.
[10] M. Wꢀchter, P. Rꢁedi, Chem. Biodiversity 2009, 6, 283.
[11] S. Furegati, M. Binder, A. Linden, P. Rꢁedi, Helv. Chim. Acta 2006, 89, 1351.
[12] M. Wꢀchter, S. Jurt, A. Linden, P. Rꢁedi, in preparation.
[13] M. Wꢀchter, ꢃHerstellung von optisch aktiven Organophosphaten mit cis- und trans-Decalingerꢁst
zur Untersuchung der Inhibition von Acetylcholinesterase mittels Enzymkinetik und 31P-NMR
Spektroskopieꢄ, Ph.D. Thesis, University of Zurich, 2009.
[14] G. L. Ellman, K. D. Courtney, V. Andres Jr., R. M. Featherstone, Biochem. Pharmacol. 1961, 7, 88.
[15] A. Baici, Biol. Chem. 1998, 379, 1007.
[16] I. H. Segel, ꢃEnzyme Kinetics: Behavior and Analysis of Rapid Equilibrium and Steady-State
Enzyme Systemsꢄ, Wiley, New York, 1993; H. Bisswanger, ꢃEnzymkinetik, Theorie und Methodenꢄ,
3rd edn., VCH-Verlagsgesellschaft Weinheim, 2000.
[17] W. W. Chan, Biochem. J. 1993, 311, 981.
[18] A. Baici, P. Schenker, M. Wꢀchter, P. Rꢁedi, Chem. Biodiversity 2009, 6, 261.

Helvetica Chimica Acta – Vol. 95 (2012)
[19] M. Wꢀchter, P. Rꢁedi, in preparation.
33
[20] C. J. Wilkins, J. Chem. Soc. 1951, 2726.
[21] R. Hooft, KappaCCD Collect Software, Nonius BV, Delft, The Netherlands, 1999.
[22] Z. Otwinowski, W. Minor, ꢃMacromolecular Crystallography, Part Aꢄ, in ꢃMethods in Enzymologyꢄ,
Vol. 276, Eds. C. W. Carter Jr., R. M. Sweet, Academic Press, New York, 1997, p. 307.
[23] R. H. Blessing, Acta Crystallogr., Sect. A 1995, 51, 33.
[24] E. N. Maslen, A. G. Fox, M. A. OꢄKeefe, in ꢃInternational Tables for Crystallographyꢄ, Ed. A. J. C.
Wilson, Kluwer Academic Publishers, Dordrecht, 1992, Vol. C, Table 6.1.1.1, p. 477.
[25] R. F. Stewart, E. R. Davidson, W. T. Simpson, J. Chem. Phys. 1965, 42, 3175.
[26] J. A. Ibers, W. C. Hamilton, Acta Crystallogr. 1964, 17, 781.
[27] D. C. Creagh, W. J. McAuley, in ꢃInternational Tables for Crystallographyꢄ, Ed. A. J. C. Wilson,
Kluwer Academic Publishers, Dordrecht, 1992, Vol. C, Table 4.2.6.8, p. 219.
[28] D. C. Creagh, J. H. Hubbell, in ꢃInternational Tables for Crystallographyꢄ, Ed. A. J. C. Wilson,
Kluwer Academic Publishers, Dordrecht, 1992, Vol. C, Table 4.2.4.3, p. 200.
[29] G. M. Sheldrick, ꢃSHELXL97, Program for the Refinement of Crystal Structuresꢄ, University of
Gçttingen, Germany, 1997.
[30] C. K. Johnson, ꢃORTEPIIꢄ, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge,
Tennessee, 1976.
[31] A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, M. C. Burla, G. Polidori, M. Camalli,
SIR92, J. Appl. Crystallogr. 1994, 27, 435.
[32] H. D. Flack, G. Bernardinelli, Acta Crystallogr., Sect. A 1999, 55, 908; H. D. Flack, G. Bernardinelli,
J. Appl. Crystallogr. 2000, 33, 1143.
Received September 20, 2011