Enantioselective Synthesis of a Tricyclic, sp3-Rich Diazatetradecanedione: an Amino Acid-Based Natural Product-Like Scaffold

Article abstract of DOI:10.1002/chem.201905144

6-, 7-, and 8-membered rings are assembled from a linear precursor by successive cyclisation reactions to construct a tricyclic diazatricyclo[6.5.1.0^{4, 9}]-tetradecanedione scaffold. Advanced building blocks based on d-aspartic acid and l-pyroglutamic acid were combined by a sp³?sp² Negishi coupling. A carbamate-guided syn-diastereoselective epoxidation followed by an intramolecular epoxide opening allowed the construction of the piperidine ring. An efficient one-pot hydroxyl-group protection twofold deprotection reaction prepared the ground for the cyclisation to the bicycle. A final deprotection of the orthogonal protecting groups and lactamisation led to the novel, sp³-rich tricycle. The final compound is a substrate mimic of peptidyl-prolyl cis-trans isomerases featuring a locked trans-amide bond. Cheminformatic analysis of 179 virtual derivatives indicates favourable physicochemical properties and drug-like characteristics. As proof of concept we, show a low micromolar activity in a fluorescence polarisation assay towards the FK506-binding protein 12.

Full text of DOI:10.1002/chem.201905144

A Journal of
Accepted Article
Title: Enantioselective Synthesis of a Tricyclic, sp3-Rich
Diazatetradecanedione: an Amino Acid-Based Natural Product-
Like Scaffold
Authors: Matthias Bischoff, Peter Mayer, Christian Meyners, and Felix
Hausch
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201905144
Link to VoR: http://dx.doi.org/10.1002/chem.201905144
Supported by

Chemistry - A European Journal
10.1002/chem.201905144
3
Enantioselective Synthesis of a tricyclic, sp -rich Diazatetradecanedione:
an Amino Acid-based Natural Product-like Scaffold
a
b
c
c
Matthias Bischoff , Peter Mayer , Christian Meyners and Felix Hausch *
a
Compound Management and Screening Center (COMAS), Max Planck Institute of Molecular
Physiology, Otto-Hahn-Strasse 11, 44227 Dortmund, Germany
b
Department of Chemistry, Ludwig-Maximilians-University München, Butenandtstrasse 5-13, 81377
München, Germany
c
Department of Chemistry, Institute of Chemistry and Biochemistry, Darmstadt University of
Technology, Alarich-Weiss-Strasse 4, 64287 Darmstadt, Germany
Abstract: 6-, 7-, and 8-membered rings are assembled from a linear precursor by
4,9
successive cyclisation reactions to construct
a
tricyclic diazatricyclo[6.5.1.0 ]-
tetradecanedione scaffold. Advanced building blocks based on D-aspartic acid and L-
3
2
pyroglutamic acid were combined by a sp -sp Negishi coupling. A carbamate-guided syn-
diastereoselective epoxidation followed by an intramolecular epoxide opening allowed the
construction of the piperidine ring. An efficient one-pot hydroxyl-group protection twofold
deprotection reaction prepared the ground for the cyclisation to the bicycle. A final
3
deprotection of the orthogonal protecting groups and lactamisation led to the novel, sp -rich
tricycle. The final compound is a substrate mimic of peptidyl-prolyl cis-trans isomerases
featuring a locked trans-amide bond. Cheminformatic analysis of 179 virtual derivatives
indicates favourable physicochemical properties and drug-like characteristics. As proof of
concept we, show a low micromolar activity in a fluorescence polarisation assay towards the
FK506-binding protein 12.
A recent scaffold analysis of more than 24 million organic compounds from the CAS Registry
concluded that a ‘very small percentage of frameworks are found in a large percentage of
compounds’. Concretely, more than 75% of the compounds can be described by only 5% of
the most common frameworks. Thus, the structural diversity of the organic chemical space
has been explored to a very limited extend.[1]
Obviously, in the context of medicinal chemistry, the question arises which kind of
compounds have to be synthesized in order to find small molecule modulators of biological
targets.[2] In particular sp -rich, natural product-like scaffolds have shifted into focus as new
biologically active compounds with novel properties. Small molecule secondary metabolites
have been selected and optimized during evolution to interact with biological targets and
3
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201905144
[3]
represent prevalidated chemical structures explored by nature. In accordance with that is
the fact that natural products and natural product scaffolds continue to contribute to a large
extent in the area of drug discovery and development. More than 50 percent of all new
approved drugs in the time period 1981 to 2014 are natural product related.[4]
3
Representative examples for heterocyclic, sp -rich scaffolds are shown in Scheme 1.
Furopyrane A was synthesized from a (D)-ribose derivative by an intramolecular hetero Diels-
Alder reaction.[5] The keystep for the construction of spiro lactam B consists of a domino
nitrone formation / intramolecular [3+2] cycloaddition[6] and phenol C was built by a domino
Petasis / intramolecular hetero Diels-Alder reaction.[7]
3
Scheme 1. Examples of heterocyclic, sp -rich and natural product-like scaffolds.
All three examples have in common that the complexity-generating step consists of a
multicomponent reaction, a cycloaddition or a combination of both. Undoubtedly, these are
very elegant and efficient methods to establish molecular complexity. Nevertheless, there is
a considerable relevant chemical space, which cannot be addressed by that means. The use
of naturally occurring building blocks like amino acids benefit from the intrinsic chirality
available in both enantiomeric forms. It offers the possibility to build up complex scaffolds
with a maximum degree of flexibility regarding size, shape and stereochemistry. Here, we
3
like to present a synthesis of a novel, sp -rich natural product-like scaffold based on amino
acids.
3
In the light of our general interest in sp -enriched scaffolds and in constrained amino acid
analogs as the key to efficiently target protein-protein interactions, we have previously
demonstrated that the conformationally preorganized diazabicyclo[4.3.1]decanone scaffold
[8]
shows superior binding affinities and ligand efficiencies compared to more flexible motifs. In
our quest to further rigidify this bioactive scaffold, we here describe a novel, highly rigidified
4,9
diazatricyclo[6.5.1.0 ]tetradecanedione core, which can be synthesized in
a
stereochemcially defined and substrate-controlled manner using simple amino acids as chiral
starting material.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201905144
Scheme 2. Retrosynthesis of the diazatricyclo[6.5.1.04,9]tetradecanedione motif 1.
Retrosynthetic analysis of 1, based on an intramolecular epoxide opening and two amide
bond formations, leads to epoxide 2, which in turn is accessible from advanced vinyl iodide
building block 3 and -iodinated amino acid derivative 4 by a Negishi coupling (Scheme 2). 3
can be derived from D-aspartic acid (5), whereas 4 can be derived from L-pyroglutamic acid
(
6). The synthesis of 3 was accomplished in 7 steps (Scheme 3). Treatment of D-aspartic
acid (5) with thionyl chloride in methanol followed by Boc protection afforded methyl ester 8.[9]
Borane reduction of the carboxylic acid yielded alcohol 9 in 68 % yield. Parikh-Doehring
oxidation[10] delivered an -amino aldehyde, which was directly subjected to Takai olefination
conditions to give vinyl iodide 10 as a single E-isomer in 46 % over 2 steps. Reduction of the
methyl ester with DIBAL in toluene at –78 °C afforded the desired aldehyde 12 in 74 % yield
along with the corresponding alcohol 11, which could be reoxidized to the aldehyde in 83 %
yield using IBX in DMSO. Finally, a HWE reaction with Methyl diethylphosphonoacetate
delivered advanced vinyl iodide 3 in 79 % yield.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201905144
Scheme 3. Synthesis of advanced vinyl iodide 3 starting from D-aspartic acid.
For the completion of the synthesis of the tricyclic scaffold 1, we coupled vinyl iodide 3 with 4
3
2
[11]
in a sp -sp Negishi reaction in 71 % yield (Scheme 4). Compound 4 can be prepared in 4
steps from L-pyroglutamic acid.[8b] Based on previous reports about stereoselective
epoxidations of allylic carbamates in favour of the syn-isomer[12], we were pleased to obtain a
diastereomeric ratio of 4:1 and a yield of 79 % when treating alkene 13 with mCPBA.
Hydrogenation of epoxide 2 with Pd/C as catalyst allowed the intramolecular opening of the
epoxide and the formation of the piperidine ring, with simultaneous reduction of the double
bond. Reprotection of the amine with Benzyl chloroformate afforded 14 in 64 % yield over
two steps. Formation of the second ring was accomplished by simultaneous TES protection
of the hydroxyl function and tBu- and Boc deprotection using TESOTf in lutidine[13], followed
by lactamisation by means of HATU to give bicycle 15 in 74 % over two steps. The TES
group was cleaved by TBAF and the alcohol and the amide functionalities were methylated
with MeI and silver(I)oxide as base in 74 % over two steps. Ultimately, the methyl ester in 16
was hydrolyzed with lithium hydroxide, the Cbz group was cleaved with H and Pd/C, and the
2
third ring was closed smoothly using HATU to provide the target tricycle 1 in 72 % yield over
three steps.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201905144
*
1
.) Zn , DMF
2
.) 3, Pd (dba) ,
2
3
P(otol) , DMF
3
CO tBu
CO tBu
2
2
MeO C
I
2
71 %
NHBoc
NHCbz
4
NHCbz
13
mCPBA, CH Cl , 79 %
1
.) H , Pd/C, EtOH,
rt, 2 h
2
2
2
d.r. 4:1
NaHCO , rt, 24 h
3
2
.) CbzCl, Na CO ,
2
3
O
dioxane/H O, rt, 20 h
2
CO tBu
2
MeO C
2
64 %
2
NHBoc
NHCbz
(
2 steps)
1.) TESOTf, lutidine,
CH Cl , 0 °C to rt, 20 h
2
2
O
2
.) HATU, (iPr) NEt,
2
NH
HO
TESO
CH Cl , rt, 20 h
N
CO tBu
2
2
2
CO Me
2
Cbz
MeO C
2
7
4 %
NHBoc
NCbz
1
5
14
(
2 steps)
1
.) TBAF, THF, rt, 1 h
74 %
2 steps)
(
2.) Ag O, MeI, DMF, rt, 20 h
2
1
.) LiOH, THF/H O, 0 °C, 4 h
2
O
2.) H , Pd/C, EtOH, rt, 2 h
O
2
NMe
NMe
MeO
3.) HATU, (iPr)
NEt, CH
2
Cl
2
2
,
MeO
rt, 18 h
CO Me
2
N
NCbz
7
2 %
16
1
(
3 steps)
O
Scheme 4. Synthesis of tricyclic scaffold 1.
Pipecolate derivatives such as 1 have been suggested to mimic substrates of peptidyl-prolyl
cis-trans isomerases. In order to understand the precise architecture of this novel framework,
we prepared crystals of 1, which were suitable for single-crystal X-ray diffraction (Scheme
5a). This showed an undistorted chair conformation of the piperidine ring featuring a trans
conformation of the N1-C1-O1 amide bond. Importantly, the geometry of compound 1
perfectly matched the active conformation of the natural product FK506, when bound to
FKBPs (Scheme 5b).
This highly rigid ring system can be regarded as a structural scaffold with well defined
vectors of the functional groups, i.e. amide and methyl ether. Instead of the methyl groups
(
C13 and C14), these functionalities offer two independent attachment point for
straightforward further derivatization. Additionally, this framework possesses a high degree of
3
3
complexity, which is expressed through the fractional sp character (Fsp = 0.86) and the
number of stereogenic centers (4), and should therefore be suitable for the exploration of
drug-relevant chemical space.[14]
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201905144
a)
b)
Scheme 5. a) Crystal structure of tricyclic scaffold 1. One molecule of H O is omitted for clarity. b)
2
Crystal structures of compound 1 (green) superimposed on FK506 (cyan) bound to the FK1 domain of
FKBP51 (grey surface, PDB-ID: 3O5R[ ).
15]
To assess the molecular properties of this new scaffold we used the open-access
cheminformatic tool LLAMA (Lead-Likeness and Molecular Analysis).[16] Based on medicinal
chemistry capping groups the tricycle was virtually decorated once or twice by amide
alkylation/arylation, alcohol alkylation/arylation, esterification and carbamate formation
resulting in 179 derivatives (Figure 1). The mean PMI (principal moment of inertia)
coordinates (I1 = 0.416, I2 = 0.842), as indicated by the ‘+‘, show a substantial shift from the
linear - disc-like axis towards three-dimensionality. 22 of the 179 derivatives (12 %) fall within
the lead-like space, whereas 152 of the 179 derivatives (85 %) fall within the Lipinski ‘rule of
5‘ space. Taken this data together, these derivatives have favourable physicochemical
properties and occupy a drug-like chemical space.
a)
b)
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201905144
Figure 1. a) PMI plot and b) AlogP vs. molecular mass plot of 179 virtual derivatives of scaffold 1
based on the decoration of the hydroxyl group and/or the amide group.
We were interested to see if tricycle 1 is able to modulate emerging biological targets such
as protein-protein interactions. As a first proof of concept, we tested binding of 1 to the
FK506-binding proteins FKBP12, FKBP51 and FKBP52,[17] which have raised substantial
interest as potential targets for neurological and psychiatric diseases.[18] Indeed, 1 bound to
FKBP12 with a K
of 5 ± 1 µM, making it one of the smallest FKBP12 ligands known.[19] 1 did
i
not bind to the larger FKBP51 or FKBP52 up to the solubility limits.
Figure 2. Binding of compound 1 to human FKBP12 (1 nM), determined by a fluorescence
[16]
[8d]
polarization
using 0.5 nM of the TAMRA-labelled compound [4.3.1]-16g
as a tracer. The data
points represent the mean of three independent replicates and the standard deviation is indicated by
the error bars. The continuous line represents a least square fit to a competitive binding model yielding
a K-value of 5±1 µM for compound 1.
i
3
In summary, we have developed an enantioselective synthesis of a tricyclic, sp -rich scaffold
based on simple amino acids as starting material. D-aspartic acid was elaborated in 7 steps
3
2
into an advanced vinyl iodide building block. Sp -sp Negishi coupling with a L-pyroglutame
derived building block and a diastereoselective epoxidation of the resulting alkene
constituted the key steps of the synthesis. A cheminformatic analysis of 179 virtual
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201905144
derivatives of the novel tricycle, which were generated by the LLAMA tool, indicated
favourable physicochemical properties and drug-like characteristics. Finally, we showed the
biological relevance of this novel scaffold by confirming an undistorted, locked trans-
conformation of the key amide bond and a low micromolar inhibition of the FK506-binding
protein 12.
Acknowledgement
This work was supported by the M4 award 2011 to F.H. and the ERC-PoC grant
FKBP2Drug. We are indebted to Claudia Dubler (Ludwig-Maximilians-University, München),
Elisabeth Weyher (Max Planck Institute of Biochemistry, Martinsried) and Claudia Sippel
(
Max Planck Institute of Psychiatry, München) for measurement of the NMR spectra, high-
resolution mass spectra and FKBP binding assay, respectively.
3
2
Keywords: natural products • amino acids • sp -sp Negishi coupling • diastereoselective
epoxidation • FK506-binding protein
References
[1]
[2]
[3]
A. H. Lipkus, Q. Yang, K. A. Lucas, S. A. Funk, W. F. Bartelt III, R. J. Schenk, A. J.
Trippe, J. Org. Chem. 2008, 73, 4443–4451.
W. P. Walters, J. Green, J. R. Weiss, M. A. Murcko, J. Med. Chem. 2011, 54, 6405–
6
416.
S. Wetzel, R. S. Bon, K. Kumar, H. Waldmann, Angew. Chem. Int. Ed. 2011, 50,
0800–10826.
1
[
[
[
4]
5]
6]
D. J. Newman, G. M. Cragg, J. Nat. Prod. 2016, 79, 629−661.
R. Messer, A. Schmitz, L. Moesch, R. Häner, J. Org. Chem. 2004, 69, 8558–8560.
J. Rawling, T. E. Storr, W.A. Bawazir, S. J. Cully, W. Lewis, M. S. I. T. Makki, I. R.
Strutt, G. Jones, D. Hamza, R. A. Stockman, Chem. Comm. 2015, 51, 12867–12870.
T. Flagstad, G. Min, K. Bonnet, R. Morgentin, D. Roche, M. H. Clausen, T. E. Nielsen,
Org. Biomol. Chem. 2016, 14, 4943–4946.
[
7]
8]
[
a) Y. Wang, A. Kirschner, A.-K. Fabian, R. Gopalakrishnan, C. Kress, B. Hoogeland,
U. Koch, C. Kozany, A. Bracher, F. Hausch, J. Med. Chem. 2013, 56, 3922–3935; b)
M. Bischoff, C. Sippel, A. Bracher, F. Hausch, Org. Lett. 2014, 16, 5254–5257; c) S.
Pomplun, Y. Wang. A. Kirschner, C. Kozany, A. Bracher, F. Hausch, Angew. Chem.
Int. Ed. 2015, 54, 345–348; d) S. Pomplun, C. Sippel, A. Hähle, D. Tay, K. Shima, A.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201905144
Klages, C.M. Ünal, B. Rieß, H.T. Toh, G. Hansen, H.S. Yoon, A. Bracher, P. Preiser,
J. Rupp, M. Steinert, F. Hausch, J. Med. Chem. 2018, 61, 3660–3673.
[9]
a) F. K. Brown, P. J. Brown, D. M. Bickett, C. L. Chambers, G. Davies, D. N. Deaton,
D. Drewry, M. Foley, A. B. McElroy, M. Gregson, G. M. McGeehan, P. Myers, D.
Norton, J. M. Salovich, J. Frank, F. J. Schoenen, P. Ward, J. Med. Chem. 1994, 37,
6
74–688; b) K. Ramalingam, R. W. Woodard, J. Org. Chem. 1988, 53, 1900–1903.
Y. Xie, M. Sun, H. Zhou, Q. Cao, K. Gao, C. Niu, H. Yang, J. Org. Chem. 2013, 78,
0251–10263.
R. F. W. Jackson, R. J. Moore, C. S. Dexter, J. Elliott, C. E. Mowbray, J. Org. Chem.
998, 63, 7875−7884.
a) A. Jenmalm, W. Berts, K. Luthman, I. Csoregh, U. Hacksell, J. Org. Chem. 1995,
0, 1026–1032; b) D. Wiktelius, W. Berts, A. Jenmalm Jensen, J. Gullbo, S. Saitton,
[10]
[11]
[12]
1
1
6
I. Csoregh, K. Luthman, Tetrahedron 2006, 62, 3600–3609.
[13]
Y.-P. He, H. Tan, L. Arve, S. Baumann, H. Waldmann, H.-D. Arndt, Chem. Asian J.
2011, 6, 1546−1556.
[
14]
15]
F. Lovering, J. Bikker, C. Humblet, J. Med. Chem. 2009, 52, 6752−6756.
[
A. Bracher, C. Kozany, A. K. Thost, F. Hausch, Acta Crystallogr. D Biol. Crystallogr.
2011, 67, 549−559.
[16]
I. Colomer, C. J. Empson, P. Craven, Z. Owen, R. G. Doveston, I. Churcher, S. P.
Marsden, A. Nelson, Chem. Commun. 2016, 52, 7209−7212.
[
17]
18]
C. Kozany, A. März, C. Kress, F. Hausch, ChemBioChem 2009, 10, 1402–1410.
[
a) S. Gaali, R. Gopalakrishnan, Y. Wang, C. Kozany, F. Hausch, Curr. Med. Chem.
2011, 18, 5355–5379; b) X. Feng, S. Pomplun, F. Hausch, Curr. Mol. Pharmacol.
2
9
2
015, 9, 27 36; c) B. M. Dunyak, J. E. Gestwicki, J. Med. Chem. 2016, 59,
−
622−9644; d) J. Kolos, A. M. Voll, M. Bauder, F. Hausch, Front. Pharmacol.
018, 9, 1425.
[19]
J. L. Stebbins, Z. Zhang, J. Chen, B. Wu, A. Emadadi, M. E. Williams, J. Cashman,
M. Pellecchia, J. Med. Chem. 2007, 50, 6607−6617.
COMMUNICATION
Matthias Bischoff, Peter Mayer,
Christian Meyners, Felix Hausch*
Page No. – Page No.
Enantioselective Synthesis of a
3
tricyclic, sp -rich Diazatetradecane-
dione: an Amino Acid-based Natural
product-like Scaffold
Using simple amino acids, an elaborated synthesis was designed to construct a
rigid tricyclic scaffold consisting of six, seven and eight membered rings.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201905144
This article is protected by copyright. All rights reserved.