Fluoro-Carba-Sugars are Glycomimetic Activators of the glmS Ribozyme

Article abstract of DOI:10.1002/chem.201702371

The glmS ribozyme is a bacterial gene-regulating riboswitch that controls cell wall synthesis, depending on glucosamine-6-phosphate as a cofactor. Due to the presence of this ribozyme in several human pathogen bacteria (e.g., MRSA, VRSA), the glmS ribozyme represents an attractive target for the development of artificial cofactors. The substitution of the ring oxygen in carbohydrates by functionalized methylene groups leads to a new generation of glycomimetics that exploits distinct interaction possibilities with their target structure in biological systems. Herein, we describe the synthesis of mono-fluoro-modified carba variants of α-d-glucosamine and β-l-idosamine. (5aR)-Fluoro-carba-α-d-glucosamine-6-phosphate is a synthetic mimic of the natural ligand of the glmS ribozyme and is capable of effectively addressing its unique self-cleavage mechanism. However, in contrast to what was expected, the activity is significantly decreased compared to its non-fluorinated analog. By combining self-cleavage assays with the Bacillus subtilis and Staphylococcus aureus glmS ribozyme and molecular docking studies, we provide a structure–activity relationship for fluorinated carba-sugars.

Full text of DOI:10.1002/chem.201702371

A Journal of
Accepted Article
Title: Fluoro-carba-sugars are glycomimetic activators of the glmS
ribozyme
Authors: Günter Mayer, Daniel Matzner, Anna Schüller, Torben Seitz,
and Valentin Wittmann
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.201702371
Link to VoR: http://dx.doi.org/10.1002/chem.201702371
Supported by

10.1002/chem.201702371
Chemistry - A European Journal
FULL PAPER
Fluoro-carba-sugars are glycomimetic activators of the glmS
ribozyme
Daniel Matzner,[a] Anna Schüller,[a] Torben Seitz,[b,c] Valentin Wittmann,[b] Günter Mayer*[a]
Abstract: The glmS ribozyme is
a
bacterial gene-regulating
characteristic properties of reducing sugars like mutarotation.[7]
This dissimilarity enables the synthesis of analogs of the pure α-
and β-anomers of the respective carbohydrates.[8] Noteworthy is
the pioneering work of Ogawa et al. which is crucial in the field
of carba-sugars and their use as carbohydrate mimics.[9] More
recently carba-sugars were combined with fluorination strategies
to increase their stability, lipophilicity or to modulate acidity and
basicity.[10] Furthermore, fluoro-carba-sugars are advantageous
in mimicking the natural sugar conformations.[11] However, this
approach is still in its infancies as few examples are known
describing the synthesis of 5a-fluorinated carba-pyranoses[12]
and their biological activity[13] (reviewed in ref. 14).
riboswitch that controls cell wall synthesis, depending on
glucosamine-6-phosphate as a cofactor. Due to the presence of this
ribozyme in several human pathogen bacteria (e.g. MRSA, VRSA),
the glmS ribozyme represents an attractive target for the
development of artificial cofactors. The substitution of the ring
oxygen in carbohydrates by functionalized methylene groups leads
to a new generation of glycomimetics that exploits distinct interaction
possibilities with their target structure in biological systems. Herein,
we describe the synthesis of mono-fluoro modified carba-variants of
α-D-glucosamine and β-L-idosamine. (5aR)-Fluoro-carba-α-D-
glucosamine-6-phosphate is a synthetic mimic of the natural ligand
of the glmS ribozyme, glucosamine-6-phosphate and is capable of
Riboswitches are RNA elements mainly found in the 5’ non-
effectively addressing its unique self-cleavage mechanism. However, coding region of mostly bacterial mRNAs that specifically bind
in contrast to what was expected, the activity is significantly
decreased compared to its non-fluorinated analog. By combining
self-cleavage-assays with the Bacillus subtilis and Staphylococcus
aureus glmS ribozyme and molecular docking studies, we provide a
structure-activity relationship for fluorinated carba-sugars.
natural metabolites or ions and regulate gene expression.[15]
Since riboswitches are present in pathogenic bacteria, e.g.
Staphylococcus aureus, Clostridium difficile and Listeria
monocytogenes they represent an obvious target for
antibacterial drug development.[16] First steps have been taken
towards urgently needed alternative antibiotics through the
discovery of small synthetic molecules that mimic natural
riboswitch ligands and inhibit bacterial cell growth.[17] Among all
reported riboswitches, the glmS riboswitch represents a distinct
class as it utilizes D-glucosamine-6-phosphate (GlcN6P) as a
cofactor for the mechanistically unique induction of a catalytic
self-cleavage mechanism (Figure 1A).[18] The 5’-hydroxyl
cleavage product of the ribozyme is prone to hydrolysis by
RNase J1 leading to loss of the GlmS enzyme, which is
essential for the production of precursors for bacterial cell wall
synthesis.[19] Artificial activation of the glmS riboswitch thereby
provides a promising target to disturb the cell wall synthesis and
thereby inhibit bacterial cell growth.[18b] Carba-α-D-glucosamine-
6-phosphate (CGlcN6P), the carbocyclic mimic of GlcN6P,
represents such an artificial activator (Figure 1B).[20] As far as
carba-α-D-glucosamine (CGlcN) is concerned, an induction of
cell-envelope stress and bacterial growth inhibition for B. subtilis
and S. aureus strains could be shown. Its activity on B.subtilis,
thereby, depends on the GlcN-specific EII complex of the
Introduction
Carbohydrates are central elements in a variety of cellular
processes, e.g. cell-cell recognition, signaling, and mediation of
immune-responses.[2] In this regard, carbohydrate mimics
became an interesting but synthetically challenging compound
class with a broad range of potential applications.[3] Carba-
sugars are carbocyclic pseudo-sugars wherein the ring-oxygen
is replaced by carbon.[4] Because of the close resemblance to
their natural counterparts and their increased stability, these
sugar mimics found applications in numerous fields, e.g.
validamycins as potent antibiotics in agriculture[5], or acarbose
for the treatment of diabetes.[6] In contrast to otherwise
functional similarity, carba-sugars lack the most reactive
functional group of carbohydrates and thereby miss
21]
phosphoenolpyruvate-sugar transferase system (PTS).[18b,
[a]
D. Matzner, A. Schüller, Prof. Dr. G. Mayer
Life & Medical Sciences Institute
University of Bonn
This finding opens the route towards a novel class of antibiotics
with potential value in the treatment of multi-resistant pathogenic
bacterial strains.[22]
Gerhard-Domagk-Str. 1, 53121 Bonn (Germany)
E-mail: gmayer@uni-bonn.de
Dr. C. T. Seitz, Prof. Dr. V. Wittmann
Department of Chemistry
University of Konstanz
Universitätsstr. 10, 78457 Konstanz (Germany)
Dr. C. T. Seitz
The glmS ribozyme has a rigid cofactor binding pocket that does
not considerably change its conformation upon metabolite
binding.[23] Much effort has been undertaken aiming at the
identification of artificial glmS ribozyme activators, but even
high-throughput screening of drug-like molecule libraries[24] did
not yield novel sufficiently potent activators. Serial testing of
GlcN6P analogs with varying degrees of modification provided
detailed insight into the structure-activity relationship (SAR) of
[b]
[c]
Current address: Cilag AG
Schaffhausen (Switzerland)
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201702371
Chemistry - A European Journal
FULL PAPER
Figure 1. A) Metabolite-dependent self-cleavage mechanism of glmS ribozymes by the natural metabolite GlcN6P (orange hexagon) and artificial metabolite
fluoro-carba-α-D-GlcN6P (green hexagon). Binding of the respective cofactor leads to cleavage of the glmS ribozyme (the arrowhead depicts the cleavage site).
RNase J1 mediated hydrolysis of the 5’-OH cleavage product, including the glmS gene, results in a reduction of GlmS protein translation, thus reduced levels of
GlcN6P. B) Chemical structures of the natural glmS ribozyme metabolite α-D-GlcN6P and the carba-sugar analogs of GlcN6P and IdoN6P. C) Glc6P, an inhibitor
of glmS ribozyme cleavage, bound at the active site of the glmS ribozyme from Thermoanaerobacter tengcongensis (PDB 2Z74). Modified after Klein et al.[1]
Hydrogen bonds involving Glc6P, one water molecule (blue sphere) and one Mg2+ ion (purple sphere) are depicted as dotted lines. The scissile phosphate is
marked (★).
glmS ribozyme metabolites. The 2-amino group with its central
role as an acid-base-catalyst and a vicinal hydroxyl group could
be proven to be crucial for glmS ribozyme activation.[25] Minor
alterations, however, are tolerated by the riboswitch, so that
inversions or removal of functional groups at the ring result in
still active cofactors with reduced affinity.[25-26] The carba-sugar
variant of GlcN6P is the so far most potent artificial cofactor of
the glmS ribozyme with a similar activity as the natural
metabolite in vitro. From crystal structures, it can be assumed
that the substitution of the ring-oxygen by a carbon though leads
to a loss of a hydrogen acceptor and thereby one hydrogen
bond to the RNA in the binding pocket is lost (Figure 1C).[20]
Here, we report the synthesis of carbocyclic mimics of GlcN6P
and IdoN6P that bear fluorine at the carba-position (1 and 2 in
Figure 1B). We assessed the in vitro activity of the fluoro-carba
variants of α-D-glucosamine-6-phosphate 1 and β-L-idosamine-
6-phosphate 2 and their respective non-phosphorylated variants
18 and 14.
sugars 1 and 2 originate from the same fluorinated alkene 4. In
both cases, we used the same deprotection and phosphorylation
strategy. The installation of the hydroxy methylene group in
different configurations, leading to L-idosamine or D-glucosamine
analogs, were achieved through either direct hydroboration (4 →
2) or epoxidation (4 → 3) followed by the regioselective radical
epoxide opening. The alkene 4 was formed through Tebbe-
olefination after α-fluorination of cyclohexanone 5. The
cyclohexanone 5, in turn, was the product of the Ferrier-
rearrangement after elimination of the alcohol 6, which was the
known starting point of both routes.
Several approaches are described to replace oxygen by a
methylene group in the sugar-ring.[27] We made use of the
corresponding carbohydrate D-glucosamine for the synthesis of
the
modified
5a-carba-sugars.
Thus,
D-glucosamine
hydrochloride was readily transformed into 6[28] and selectively
benzylated (Scheme 2) to enable modification at the C6-OH of 7
(for carba-sugar numbering compare Figure S1). As a key-step
for the substitution of the ring-oxygen by carbon, the Ferrier
rearrangement was employed.[29] The synthesis of the substrate,
the unsubstituted terminal olefin 8, was achieved by conversion
of the 6-OH into an iodo-group followed by elimination with
Results and Discussion
sodium hydride.[30] The Ferrier carbocyclization of 8 to
5
Synthesis of mono-fluorinated amino carba-sugars
proceeded under catalytic conditions resulting in a C1-OH
axial/equatorial ratio of 77:23.[31]
Scheme 1 depicts the retrosynthetic analysis that underlies the
synthesis of the carbohydrate mimics 1 and 2. Both fluoro-carba-
This article is protected by copyright. All rights reserved.

10.1002/chem.201702371
Chemistry - A European Journal
FULL PAPER
been described. As far as mono-fluorinated carbapyranoses are
concerned, however, 5-fluoro-(+)-MK7607 and its C1-epimer,
the unsaturated carba-analogs of galactose, are the only
representatives described until now.[13] In the case of the gem-
difluoro analogs, the carbocyclization via a triisobutylaluminium
Scheme 1. Retrosynthetic analysis of fluoro-carba-sugar analogs 1 and 2.
(TIBAL)
mediated
rearrangement
of
the
respective
difluoroalkane was used as a key step that also introduced
fluorine at the carba-position. We utilized the electrophilic
fluorination of the preformed carbocyclic derivative of GlcN,
which has previously been used for the synthesis of the racemic
difluorocarba analog of thymidine.[34] The orientation of the
fluorine in 10 could not be validated by NMR at this point
because of unfavorable overlapping of the crucial proton-signals
and additional conformational flexibility of the carbocycle by the
carbonyl at C5.[35] The configuration of 17 could be verified after
the introduction of the equatorial hydroxy methylene group and
removal of the TBS-group. Small coupling constants for H5a
could be determined (3JH5a-H1 = 2.1 Hz, 3JH5a-H5 = 4.0 Hz) as clear
indicators for the axial configuration of the fluorine. The
Scheme 3. Synthesis of protected fluoro-carba-α-D-GlcN 12.
To introduce fluorine at the 5a-position of the cyclohexanone 5,
the hydroxyl group was protected as tert-butyldimethylsilyl (TBS)
ether (Scheme 3).[32] The silyl-group enabled easy separation of
diastereomers 9a and 9b. 9a was converted into the kinetic
enolate by treatment with lithium diisopropylamide (LDA) and
subsequently fluorinated using N-fluoro-benzene-sulfonimide
(NFSI) as
a
fluorinating agent.[33] The mono-fluorinated
cyclohexanone 10 was isolated in moderate yield (46%) as a
single diastereomer. In comparison, synthetic routes towards
gem-difluorocarba-sugar analogs of α-D-glucose[12a], α-, or β-D-
galactose, α-, or β-D-mannose[12b] and β-L-idose[12c] have already
Scheme 2. Synthesis of cyclohexanone 5.
Reagents and conditions: (a) TBDMSOTf (2.3 equiv), 2,6-lutidine (2.3 equiv),
CH2Cl2, 0 → 25 °C, 1.5 h, 89% (68:21) (b) LDA (1.1 equiv), THF, -74 °C, 3.5 h,
then NFSI (1.1 equiv), THF, -74 °C, 1.5 h, 46%; (c) Cp2TiMe2 (4.0 equiv),
toluene, 65 °C, 3 h, 61%; (d) 9-BBN (15.0 equiv), THF, 60 → 66 °C, 5.5 h,
then 3 N NaOH (30.0 equiv), H2O2 (30.0 equiv), THF, 0 °C → 25 °C, 2 h, 50%;
(e) mCPBA (8.8 equiv), CH2Cl2, 0 → 25 °C, 14 d, 59%; (f) Cp2TiCl2 (0.03
equiv), collidine hydrochloride (1.5 equiv), manganese (1.5 equiv), 1,4-C6H8
(4.4 equiv), THF, 50 → 58 °C, 3 h, 33%.
Reagents and conditions: (a) Trityl chloride (2.1 equiv), pyridine, 25 °C, 17 h;
(b) BnBr (6.0 equiv), NaH (6.0 equiv), DMF, 0 → 25 °C, 17 h; (c) CF3COOH
(3.87 equiv), tri-iso-propylsilan (1.5 equiv) CH2Cl2, 0 → 25 °C, 1.5 h, 51% over
3 steps; (d) iodine (1.2 equiv), PPh3 (1.7 equiv), imidazole (2.4 equiv), toluene,
60 → 80 °C, 6 h, 57%; (e) NaH (3.0 equiv), DMF, 0 → 25 °C, 3 h, 65%; (f)
HgSO4 (0.03 equiv), Dioxane/5 mM H2SO4 (2:1), 80 °C, 3 h, 86%.
This article is protected by copyright. All rights reserved.

10.1002/chem.201702371
Chemistry - A European Journal
FULL PAPER
12 in 33% yield and the recovery of 12% starting material. The
desired isomer formed with a high diastereoselectivity (dr 93:7),
whereby the absolute configuration of the main product was
confirmed by an NOE correlation between the proton at C5 and
the proton at C3 that indicates an equatorial configuration of the
hydroxy methylene group.
The diastereoselectivity was rationalized by comparison of the
energy of the two possible transition states TS-A and TS-B
(Figure 2). The difference in the Gibbs free energies of both
transition states (ΔΔGTS
) at 331 K was calculated to be
2.6 kcalmol-1, in favor for TS-A. Both optimized transition
structures (Figure 3) differ by the direction from which 1,4-
cyclohexadien approaches the radical intermediate. The carbon-
centered radical intermediate is formed after epoxide opening,
with the metal complex still bound to the oxygen of C6. Similar to
the model developed by Giese[42], describing the influence of
exocyclic substituents on the diastereoselectivity of cyclic
radicals, two opposing interactions with the bulky OTBS-group
are crucial. TS-A is destabilized through the interaction of the
Figure 2. Schematic energy diagram of the hydrogen atom transfer to the
radical intermediate of the epoxide opening via two competing transition states
TS-A and TS-B. Hydrogen transfer via TS-A is leading to the the hydroxy
methylene at C5 in equatorial orientation and TS-B to the axial configuration..
The calculated Gibbs free energies of activation (ΔGTS-A and ΔGTS-B) and the
difference in the Gibbs free energies of both transition states (ΔΔGTS) at 331 K
are indicated.
stereoselectivity of the fluorination reaction could be explained
by bulky substituents at positions 2, 4 and 5 that direct the
electrophilic attack of NFSI towards the β-face of the ring system.
Exposure of 10 to Petasis reagent[36] provided olefin 4 in a yield
of 61%. Hydroboration and subsequent oxidation afforded 5a-
fluoro-carba-β-L-idopyranose 11 stereoselectively in 50% yield.
The preference for the axial hydroxy methylene derivative is in
accordance with the results described for the unsubstituted
3
carba-analogs.[37] A JH4-H5 coupling of 5.4 Hz was indicative of
the axial configuration.
The α-D-glucose isomer 12, with the inverted configuration at C5,
was not accessible from 11 through isomerization of the
corresponding aldehyde at C5 because of a rapid loss of fluorine
38]
during the isomerization step (results not shown).[37b,
Therefore we developed an alternative synthetic route
comprising epoxidation of the olefin and subsequent
regioselective epoxid-opening. The epoxidation of 4 with meta-
chloroperoxybenzoic acid (mCPBA) required prolonged reaction
times of fourteen days and provided 3 in a yield of 59%. The
configuration of the epoxide was confirmed as 5-R through 2D-
ROESY NMR at -40°C (Figure S2).
For the next step, a regioselective reduction of the epoxide had
to be found. The epoxide-opening represented a crucial step
because attempts from other groups to open the epoxide next to
fluorine in a similar compound under multiple reaction conditions
had failed.[39] Our first approach, utilizing stoichiometric amounts
of TiCp2Cl in the radical opening of the epoxide, gave only a
complex mixture with low quantities of the desired product.[40] To
minimize side reactions associated with the radical titanium(III)
species we used a catalytic approach.[41] In this process,
collidine hydrochloride is added as an acid to protonate the
titanocene and regenerate titanocene dichloride. Thereby, the
amount of TiCp2Cl could be reduced to 0.3 equivalents. The
epoxide opening under an elevated temperature of 60 °C gave
Figure 3. Optimized structures of the transition states TS-A and TS-B
calculated at TPSS-D3/def2-TZVP level of theory. The distances between the
carbon-radical center and the hydrogen taken part in the transfer reaction are
shown (red dotted line). To reduce the calculation time all benzyl groups were
substituted by methyl. For each transition structures one negative vibrational
frequency was found, indicating that optimized structures were true transition
states.
This article is protected by copyright. All rights reserved.

10.1002/chem.201702371
Chemistry - A European Journal
FULL PAPER
Scheme 5. Synthesis of (5aR)-fluoro-carba-α-D-GlcN 18 and its
phosphorylated derivative 1.
Scheme 4. Synthesis of (5aR)-fluoro-carba-β-L-IdoN 14 and its
phosphorylated derivative 2.
Reagents and conditions: (a) TBAF (4.0 equiv), THF/K2HPO4-Buffer (100:1),
THF, 23 °C, 4 h, 90%; (b) 10% Pd/C (50% w/w), H2 (10 bar), CF3COOH (10.0
equiv), MeOH, 25 °C, 7 h, 53%; (c) (iPr)2NP(OBn)2 (2.5 equiv), 1H-tetrazole
(5.0 equiv), CH2Cl2, 25 °C, 2 h; then mCPBA (3.0 equiv), CH2Cl2, 0 °C, 1 h,
94%; (d) TBAF (4 equiv), THF/K2HPO4-Buffer (100:1), 23 °C, 1 h, 87%; (e)
10% Pd/C (50% w/w), H2 (10 bar), CF3COOH (10.0 equiv), MeOH, 25 °C, 3 h,
70%.
Reagents and conditions: (a) TAS-F (1.5 equiv), DMF, 23 °C, 1.5 h, 99%; (b)
10% Pd/C (50% w/w), H2 (10 bar), CF3COOH (10.0 equiv), MeOH, 25 °C, 7 h,
53%; (c) (iPr)2NP(OBn)2 (2.5 equiv), 1H-tetrazole (5.0 equiv), CH2Cl2, 25 °C,
2 h; then mCPBA (3.0 equiv), CH2Cl2, 0 °C, 1 h, 94%; (d) TBAF (4 equiv),
THF/K2HPO4-Buffer (100:1), 23 °C, 1 h, 87%; (e) 10% Pd/C (50% w/w), H2
(10 bar), CF3COOH (10.0 equiv), MeOH, 25 °C, 3 h, 70%.
obtained in three steps from 11. Treatment with dibenzyl N,N-
diisopropylphosphoramidite and subsequent oxidation with
mCPBA gave fully protected phosphate 15 in excellent yield
(99%).[45] Removal of the TBS-group by sub-stoichiometric
quantities of TBAF[46] (41% yield), followed by debenzylation
under similar conditions as described above gave 2 (68% yield).
The configuration of 14 and 2 was analyzed by NMR, which
incoming 1,4-cylclohexadien with the axial OTBS-group, while in
TS-B the bulky titanocene moiety interacts with TBS. However,
observations by Giese and Gansäuer on similar substrates
confirm that the repulsion between the CH2O[TiCp2Cl] group and
a nearby axial bulky substituent guides the radical trap to the
opposite face.[42-43] In the case of the fluoro-carba-glucosamine
precursor 3, this would lead to the observed equatorial
arrangement of the hydroxy methylene group through the axial
attack of 1,4-cyclohexadien. This approach shows how the
temporary insertion of the bulky titanocene reverses the
diastereoselectivity compared to the hydroboration with sterically
demanding boranes. With the sequence of epoxidation and
stereoselective radical epoxid opening, we found a novel
approach to carbocylic α-D-glucosamine derivatives from their
respective sugar. We thereby propose that this synthetic route
could be of general interest in the synthesis of carba-sugars
from base-sensitive precursors.
3
3
revealed coupling constants JH3-H4 = JH3-H2 = 6.1 Hz for 14 and
2. These values lie between the expected coupling constants for
the anti- or gauche-orientation of H2 and H3 protons in the chair-
4
1
conformations C1 or C4. Since three ring substituents occupy
the axial conformation in both cases, similar
a
thermodynamically stability can be assumed. This flexibility
leads to observed vicinal coupling constants that are averaged
over the constants of the corresponding conformers. Thus, the
coupling constants of 14 and 2 are significantly higher than the
calculated values for methyl β-L-idopyranoside (3JH3-H4 = 3JH3-H2
=
3.0 Hz) in the favored 1C4 conformation.[47] Similar
conformational flexibility could already be demonstrated for the
gem-difluorocarbasugar analog of methyl-β-L-idopyranoside.[12c]
For both 14 and 2 strong NOE-correlation between H2 and H4
as well as between H1 and H5 was observed, indicating a rapid
1C4↔4C1 exchange that is not distinguishable on the NMR
timescale at room-temperature. An NOE-correlation between
H5a and H2, however, was not observed which would indicate a
2S5a skew-boat conformation.[48] DFT-calculation with the ORCA
program package, on the two possible chair conformations of
As the 6-phosphate moiety is crucial for molecules to gain full
ribozyme-activating properties the main focus was on the
synthesis of the phosphorylated variants of the fluoro-carba-
sugars. Removal of the TBS protecting group of 11 was
achieved in near quantitative yield under mild reaction conditions
with
1.5
equivalents
of
tris(dimethylamino)sulfonium
difluorotrimethylsilicate (TAS-F) (Scheme 4).[44] In contrast, an
excess of TBAF led to the significant formation of side-product
by elimination most probably due to the high basicity of TBAF.
The final deprotection was achieved by heterogeneous
hydrogenolysis with 10% Pd/C in methanol and trifluoroacetic
acid under an H2 atmosphere at high pressure (10 bar) and
afforded 14 in 34% yield. (5aR)-Fluoro-carba-β-L-IdoN6P 2 was
fluoro-carba-β-L-idosamine-6-phosphate, revealed
a
relative
1
4
energy difference between C4 and C1 of 5.0 kcalmol-1 in favor
1
for the C4 conformation.[49] The structures were optimized at the
TPSS-D3/def2-TZVP level while the single point energy was
This article is protected by copyright. All rights reserved.

10.1002/chem.201702371
Chemistry - A European Journal
FULL PAPER
removal of the Z-protection group furnished (5aR)-fluoro-carba-
α-D-GlcN6P 1 (70% yield).
Table 1. EC50-values for 1, CGlcN6P and GlcN6P activating B. subtilis and
S. aureus glmS ribozymes.
1
CGlcN6P
EC50 [µM]
2.2 ± 0.4[51]
6.2 ± 0.7[20]
GlcN6P
B. subtilis
S. aureus
196 ± 17
312 ± 32
n.d.[a]
3.6 ± 0.4[20]
[a] n.d. (not determined)
Induction of glmS ribozyme self-cleavage by fluoro-carba-
sugars
The fluoro-carba-sugar analogs 1 and 2 of D-glucosamine-6-
phosphate and L-idosamine-6-phosphate and their non-
phosphorylated variants 14 and 18 were analyzed for their
capability to induce glmS ribozyme self-cleavage in vitro.
Therefore, we employed a metabolite-dependent self-cleavage
assay[20] and assessed the compound activity towards the glmS
ribozymes from B. subtilis and S. aureus (Figure 4). These
experiments revealed that (5aR)-fluoro-carba-α-D-GlcN6P 1 acts
as a cofactor of glmS ribozyme self-cleavage, whereas the non-
phosphorylated variant 18 and fluoro-carba-β-L-IdoN6P
2
showed only minor activity. Fluoro-carba-β-L-IdoN was
completely inactive as a cofactor. The tendencies among the
tested fluoro-carba-sugars are the same for the ribozymes of
both species while there is a slight deviation in the absolute
cleavage.
Figure 4. Characterization of carba-sugar derivatives in metabolite-dependent
glmS ribozyme cleavage assay. Cleavage of glmS ribozymes from B. subtilis
and S. aureus induced by 2000 µM of compounds 14 (blue), 2 (dark blue), 18
(green), 1 (dark green) compared to cleavage in the presence of 200 µM of
Having shown that
1 activates the glmS ribozyme most
efficiently, we next performed concentration dependent analyses.
These studies revealed EC50-values of 312 ± 32 µM and
196 ± 17 µM for S. aureus and B. subtilis, respectively (Table 1,
Figure S3A). We observed pseudo first-order rate constants
(kobs) (Table 2, Figure S3B) for induced ribozyme cleavage of
both species by 1 that are consistent with data for CGlcN6P and
GlcN6P.[20] In general, higher cleavage-rates could be detected
for the B. subtilis glmS ribozyme in the presence of 600 µM of 1
compared to S.aureus (kobs(B. subtilis) = 0.258 ± 0.036 and
GlcN6P and GlcN (black) and in the absence of metabolite and Mg2+
Experiments were carried out at least in triplicates.
.
calculated at the PW6B95-D3/def2-QVZP(-g,-f) level of theory
including the COSMO model.[50] At the same time, the
orientation of the fluorine in 2 could be confirmed to be R-
configured by determination of the coupling constants of H5a
(3JH5a-H5 = 3JH5a-H1 = 7.3 Hz).
kobs(S. aureus)
= 0.122 ± -
0.007 min-1. EC50- and kobs
Synthesis of fluoro-carba-α-D-GlcN 18 and its phosphorylated
version 1 proceeded with slight modifications compared to the
synthesis of fluoro-carba-β-L-idosamine (Scheme 5). TBS
removal from 12 succeeded with excess TBAF in excellent
yields of 90%. At this point, the equatorial configuration of the
hydroxy methylene group was clearly determined based on large
determination both show concentration dependency for the two
species, which implies that fluoro-carba-GlcN6P 1 acts as a
cofactor of both glmS ribozymes.[52]
Fluoro-carba-α-D-glucosamine-6-phosphate activates the
glmS ribozyme
coupling constants of H4 (3JH4-H3
=
3JH4-H5 = 10.0 Hz). Final
deprotection via heterogeneous hydrogenolysis of 17 resulted in
18 (53% yield). The phosphorylated variant 1 was accessible
through phosphorylation of 12 (94% yield), followed by TBS
removal giving 20 in 87% yield. Final debenzylation and the
This article is protected by copyright. All rights reserved.

10.1002/chem.201702371
Chemistry - A European Journal
FULL PAPER
In the case of 14 complete loss of activity could be observed.
This influence of the phosphate on cleavage-activity is
consistent with previous studies on the S. aureus glmS
riboswitch that showed a 50-fold decrease in activity of GlcN
compared to GlcN6P.[20]
Table 2. kobs values of 1 and carba-GlcN6P determined for cleavage of the
glmS riboswitches of B. subtilis and S. aureus.
1
carba-GlcN6P
Concentration-dependent analysis of S. aureus and B. subtilis
glmS riboswitch activation revealed EC50 values for 1 (S. aureus
312 ± 32 µM, B. subtilis 196 ± 17 µM) that are significantly
higher than the respective values for unmodified carba-GlcN6P
(S. aureus 6.2 ± 0.7 µM, B. subtilis 2.2 ± 0.4 µM).[20] Carba-
GlcN6P is so far the most active artificial cofactor of the glmS
ribozyme.[53] Similar binding to the catalytic core assumed, the
introduction of fluorine seems to add steric load to the ligand
neglecting potential attracting properties of fluorine as for
instance hydrogen acceptor.[54] That fluorination drastically
reduces effectivity is rather surprising as recent studies on
fluorinated pseudo-sugars showed improved capabilities in
resembling their natural counterparts compared to their
respective unmodified carba analogs.[11, 12c] However, in the case
of the glmS ribozyme the ligand binding pocket imposes high
steric demands that obviously conflict with possible attracting
interactions of newly introduced modifications.
Concentration
[µM]
kobs [min-1] B. subtilis
0.258 ± 0.036
0.204 ± 0.030
n.d.
n.d.
600
200
20
1.190 ± 0.302[51]
0.222 ± 0.027[51]
kobs [min-1] S. aureus
0.258 ± 0.036
0.081 ± 0.003
n.d.
n.d.
n.d.
600
300
200
30
0.153 ± 0.012[20]
n.d.
0.027 ± 0.002
One of the goals of this study was to synthesize pseudosugar
analogs of GlcN6P with modifications at the carba-position to
gain further insight into cofactor-dependent self-cleavage of
glmS ribozymes. From the crystal structure of the inactive Glc6P
bound to the ribozyme, it is known that only the α-anomer binds
in the binding pocket.[23] Further ligand requirements are evident
from comprehensive screenings of GlcN6P analogs or
commercially available libraries.[24] From these, ethanolamine
emerged as minimal required motif which itself shows very low
affinity to the riboswitch.[25] The amino group that all activators
have in common is crucial for acid-base catalysis of glmS
ribozyme self-cleavage.[52] The six-membered ring and multiple
hydrogen donors and acceptors together with a phosphate
group are critical for the recognition of the sugars by the glmS
ribozyme.[26] All final compounds synthesized in this work, fulfill
the minimal requirement for high-affinity binding to the glmS
ribozyme. Moreover, 1 and 18 exist exclusively as pseudo-α-
anomers. On this basis, the high activity of the GlcN6P-analog 1
compared to the minor activity of non-phosphorylated 18 and
phosphorylated IdoN6P-analog 2 can be expected due to the
high ligand discrimination by the glmS ribozyme binding pocket.
In previous studies it could already be shown that epimerization
of GlcN6P at C2 or C3, resulting in D-mannosamine-6-
phosphate and D-galactosamine-6-phosphate respectively,
reduced the activity on glmS riboswitch cleavage[26] The
alteration at C5 in case of idosamine is even more critical, since
it directly affects the orientation of the phosphate at C6, whose
correct positioning contributes to affinity. Also, idosamine
possesses increased flexibility leading to two interchanging chair
conformations. Recent studies revealed that gem-difluoro-
carbamethyl-L-idopyranosides mimic this natural behavior
confirming our findings on mono-fluoro-carba-L-idosamine.[12c]
The influence of the phosphate at C6 on activity is evident
(Figure 2), and is also reflected by the comparison of 18 and 1.
Figure 5. Docked structures of (A) CGlcN6P, (B) fluoro-carba-α-GlcN6P 1 and
(C) fluoro-carba-β-L-IdoN6P 2 (carba-sugar-ring shown in green, fluorine in
light blue) overlaid with the crystal structure of the glmS ribozyme from
T. tengcongensis and bound Glc6P (PDB 2Z74). (A) and (B) CGlcN6P and 1
both bind very similar to Glc6P with the amino group at the correct position to
promote catalytic activity of the glmS ribozyme. The axial fluorine of 1 points
towards G1 leading to a distortion of the ligand that could be a reason for
lower activity of the fluorine-modified carba-sugar compared to CGlcN6P. (C)
The phosphate of 2 is correctly positioned while the rest of the molecule is
moved out of the binding position. This binding mode shows an altered
orientation of the 2-amino group relative to 1-OH resulting from the ring-flip,
consistent with the low activity as a cofactor. A binding mode of the 4C1
conformation could not be found.
This article is protected by copyright. All rights reserved.

10.1002/chem.201702371
Chemistry - A European Journal
FULL PAPER
Molecular docking studies
Conclusions
To conclude structure-activity relationship (SAR) from the crystal
structure of the glmS ribozyme and the structure of the different
carba-sugar-glycomimetics, molecular docking studies were
carried out. There are no published crystal structures of the
bacterial strains we employed in this work. However, the
nucleotide identity at the ligand-binding site is highly conserved
among glmS ribozymes that have been discovered so far.[55]
Therefore we used the crystal structure of the glmS ribozyme
from Thermoanaerobacter tengcongensis bound to glucose-6-
phosphate (PDB: 2Z74).[1] To allow docking, the structures of the
dianionic, phosphorylated pseudo-sugars 1, 2 (1C4 and 4C1),
unmodified carba-GlcN6P, and GlcN6P were optimized using
the BP86 functional and the def2-TZVPP/J basis set[50c]
including the COSMO model[50a] with the dielectric constant and
refractive index of water. Each analog was docked to the glmS
ribozyme with the AutoDock Vina program.[56] The goodness of
the docking of the natural metabolite α-D-GlcN6P thereby
functions as a verification of molecular docking (Figure S4).
From the predicted ligand poses of each carba-sugar, the pose
that best resembles binding of Glc6P in the crystal was chosen
for further comparison. Comparison of the docked structures of 1
and carba-GlcN6P (Figure 5A and B) shows that their binding at
the catalytic core closely resembles that of Glc6P, but both
carba-sugars are turned slightly around their C5-C2 axis. Fluoro-
carba-α-D-GlcN6P is distorted to a larger extent, due to the
steric demand of the axial fluorine pointing towards G1. This
imperfect binding apparently overcompensates the attracting
effect of a potential hydrogen bond between the fluorine and C2.
The distance between the amine of C2 and fluorine is 2.7 Å,
compared to 3.0 Å for the amine and the ring-oxygen of Glc6P.
Though the distance corresponds to a hydrogen bond of fluorine,
it is currently assumed that organic fluorine has a very low
proton affinity.[57] Consequently, the docking explains an
increased EC50 of 1 compared to carba-GlcN6P and GlcN6P.
Synthesis and evaluation of a mono-fluorinated carba-sugar
variant with the fluorine in equatorial position or gem-difluoro-
carba-GlcN would allow deriving a more detailed structure-
The synthesis of new fluorinated glycomimetics of α-D-
glucosamine, α-D-glucosamine-6-phosphate, β-L-idosamine and
β-L-idosamine-6-phosphate is presented. In this process, a new
method for the introduction of mono-fluoro at the 5a-carba
position of pseudo-sugar analogs of 2-amino-sugars was found.
By utilizing titanocene(III)-catalyzed radical epoxide opening, a
new route for the synthesis of base-labile carbocyclic analogs of
α-D-glucosamine was established. The computation of the
transitions states involved in the hydrogen transfer step
suggests that the catalyst directs the hydrogen transfer to the
axial position leading to the observed diastereoselectivity. The
fluorinated carba-GlcN6P 1 showed glmS ribozyme activation
that is 50- to 100-fold decreased compared to the carba-sugar
analog. Docking studies supported the observed in vitro data
and allowed further statements on glmS ribozyme ligand
requirements. Modification at the 5a-carba position, in particular,
can be assumed to result in steric strain with the back of the
binding pocket. The in vitro results presented in this study
provide a promising step towards the application of fluorinated
carba-sugars on structures in biological systems. The
widespread applicability of this new generation of glycomimetics
goes far beyond ligands for the glmS ribozyme since the carba-
position as a tunable position in carbohydrate mimics has been
further established.
Experimental Section
Details of the molecular biological methods, synthesis of all
compounds along with the copies of NMR spectra, molecular
docking, and theoretical details are available in the supporting
Information.
Acknowledgements
activity-relationship. Docking of fluoro-carba-β-L-IdoN6P
2
We thank Andreas Meyer for support with DFT-calculations,
Anke Friemel, Johanna Fafinska and Thomas Hackl for support
on NMR-measurements and Daniel Wieland for the help with a
procedure for heterogenous hydrogenolysis of benzyl ethers and
removal of Z-protection groups. We also are very grateful to
Hans-Georg Sahl and all other members of the Forschergruppe
854 for discussion and valuable input.
(Figure 5C) was performed with both chair conformations (1C4
and 4C1) whereby only the 1C4 conformation resulted in a binding
mode at the binding pocket. The position of the predicted
phosphate of 2 resembles the phosphate of Glc6P while the rest
of the molecule is differently positioned. Because of the ring-flip
conformation, the distance between the 2-amino group and the
scissile A1-G1 phosphodiester bond is 5.2 Å compared to 3.1 Å
for the native ligand. This difference in positioning is consistent
with the low activity of 2 observed in ribozyme cleavage-assays.
Besides repulsive interaction of the fluorine explained by
molecular docking, reduced basicity of the amine caused by the
nearby fluorine could be another reason for reduced activity of
the fluoro-carba-sugars.[58] The lowered activity due to the
chemically small modification impressively reflects the
discriminative capacity of the glmS ribozyme that puts high
constrains on artificial ligand design.
Keywords: carba-sugars • fluorine • ribozymes • glycomimetics
• riboswitches
[1]
[2]
D. J. Klein, S. R. Wilkinson, M. D. Been, A. R. Ferre-D'Amare, J.
Mol. Biol. 2007, 373, 178-189.
a) P. M. Rudd, T. Elliott, P. Cresswell, I. A. Wilson, R. A. Dwek,
Science 2001, 291, 2370-2376; b) N. Sharon, H. Lis, Sci. Am.
1993, 268, 82-89; c) C. R. Bertozzi, L. L. Kiessling, Science 2001,
291, 2357-2364.
[3]
[4]
Y. Chapleur, John Wiley & Sons., Wiley-VCH,, Weinheim, 1998.
G. E. Mccasland, S. Furuta, L. J. Durham, J. Org. Chem. 1966, 31,
1516-&.
This article is protected by copyright. All rights reserved.

10.1002/chem.201702371
Chemistry - A European Journal
FULL PAPER
[5]
[6]
T. Iwasa, H. Yamamoto, M. Shibata, J. Antibiot. (Tokyo) 1970, 23,
595-602.
[30]
[31]
P. J. Garegg, B. Samuelsson, J. Chem. Soc., Perkin Trans. 1
1980, 2866-2869.
D. H. R. Barton, S. Augydorey, J. Camara, P. Dalko, J. M.
Delaumeny, S. D. Gero, B. Quicletsire, P. Stutz, Tetrahedron
1990, 46, 215-230.
F. D. Boyer, J. Y. Lallemand, Tetrahedron 1994, 50, 10443-10458.
D. Enders, S. Faure, M. Potthoff, J. Runsink, Synthesis 2001,
2307-2319.
H. Kumamoto, K. Haraguchi, M. Ida, K. T. Nakamura, Y. Kitagawa,
T. Hamasaki, M. Baba, S. S. Matsubayashi, H. Tanaka,
Tetrahedron 2009, 65, 7630-7636.
a) E. W. Garbisch, J. Am. Chem. Soc. 1964, 86, 1780-&; b) P. R.
Anizelli, J. D. Vilcachagua, A. C. Neto, C. F. Tormena, J. Phys.
Chem. A 2008, 112, 8785-8789.
N. A. Petasis, E. I. Bzowej, J. Am. Chem. Soc. 1990, 112, 6392-
6394.
a) R. Blattner, R. J. Ferrier, J. Chem. Soc., Chem. Commun. 1987,
1008-1009; b) Y. S. Sun, M. Nitz, J. Org. Chem. 2012, 77, 7401-
7410; c) D. H. R. Barton, S. D. Gero, S. Augy, B. Quicletsire, J.
Chem. Soc., Chem. Commun. 1986, 1399-1401.
B. Lopez-Mendez, C. Jia, Y. Zhang, L. H. Zhang, P. Sinay, J.
Jimenez-Barbero, M. Sollogoub, Chem. - Asian J. 2008, 3, 51-58.
A. Abad, C. Agullo, A. C. Cunat, A. Gonzalez-Coloma, D. Pardo,
Eur. J. Org. Chem. 2010, 2182-2198.
a) T. V. Rajanbabu, W. A. Nugent, M. S. Beattie, J. Am. Chem.
Soc. 1990, 112, 6408-6409; b) T. V. Rajanbabu, W. A. Nugent, J.
Am. Chem. Soc. 1994, 116, 986-997.
A. Gansauer, M. Pierobon, H. Bluhm, Angew. Chem. Int. Ed. 1998,
37, 101-103.
a) G. Derosa, P. Maffioli, Arch. Med. Sci. 2012, 8, 899-906; b) D.
D. Schmidt, W. Frommer, B. Junge, L. Muller, W. Wingender, E.
Truscheit, D. Schafer, Naturwissenschaften 1977, 64, 535-536.
T. Suami, S. Ogawa, Adv. Carbohydr. Chem. Biochem. 1990, 48,
21-90.
[7]
[8]
[9]
[32]
[33]
M. Kitaoka, S. Ogawa, H. Taniguchi, Carbohydr. Res. 1993, 247,
355-359.
[34]
[35]
a) S. Ogawa, Y. Shibata, T. Nose, T. Suami, Bull. Chem. Soc. Jpn.
1985, 58, 3387-3388; b) S. Ogawa, N. Chida, T. Suami, Chem.
Lett. 1980, 1559-1562.
K. Muller, C. Faeh, F. Diederich, Science 2007, 317, 1881-1886.
B. Xu, L. Unione, J. Sardinha, S. Wu, M. Etheve-Quelquejeu, A.
Pilar Rauter, Y. Bleriot, Y. Zhang, S. Martin-Santamaria, D. Diaz, J.
Jimenez-Barbero, M. Sollogoub, Angew. Chem. 2014, 53, 9597-
9602.
a) A. Deleuze, C. Menozzi, M. Sollogoub, P. Sinay, Angew. Chem.
Int. Ed. 2004, 43, 6680-6683; b) J. Sardinha, S. Guieu, A. Deleuze,
M. C. Fernandez-Alonso, A. P. Rauter, P. Sinay, J. Marrot, J.
Jimenez-Barbero, M. Sollogoub, Carbohydr. Res. 2007, 342,
1689-1703; c) L. Unione, B. X. Xu, D. Diaz, S. Martin-Santamaria,
A. Poveda, J. Sardinha, A. P. Rauter, Y. Bleriot, Y. M. Zhang, F. J.
Canada, M. Sollogoub, J. Jimenez-Barbero, Chem. - Eur. J. 2015,
21, 10513-10521.
J. Sardinha, A. P. Rauter, M. Sollogoub, Tetrahedron Lett. 2008,
49, 5548-5550.
E. Leclerc, X. Pannecoucke, M. Etheve-Quelquejeu, M. Sollogoub,
Chem. Soc. Rev. 2013, 42, 4270-4283.
M. Mandal, B. Boese, J. E. Barrick, W. C. Winkler, R. R. Breaker,
Cell 2003, 113, 577-586.
a) D. Matzner, G. Mayer, J. Med. Chem. 2015, 58, 3275-3286; b)
C. E. Lunse, A. Schuller, G. Mayer, Int. J. Med. Microbiol. 2014,
304, 79-92; c) K. F. Blount, R. R. Breaker, Nat. Biotechnol. 2006,
24, 1558-1564.
a) J. A. Howe, H. Wang, T. O. Fischmann, C. J. Balibar, L. Xiao, A.
M. Galgoci, J. C. Malinverni, T. Mayhood, A. Villafania, A. Nahvi,
N. Murgolo, C. M. Barbieri, P. A. Mann, D. Carr, E. Xia, P. Zuck, D.
Riley, R. E. Painter, S. S. Walker, B. Sherborne, R. de Jesus, W.
Pan, M. A. Plotkin, J. Wu, D. Rindgen, J. Cummings, C. G. Garlisi,
R. Zhang, P. R. Sheth, C. J. Gill, H. Tang, T. Roemer, Nature
2015, 526, 672-677; b) T. Hermann, Nature 2015, 526, 650-651.
a) W. C. Winkler, A. Nahvi, A. Roth, J. A. Collins, R. R. Breaker,
Nature 2004, 428, 281-286; b) A. Schuller, D. Matzner, C. E.
Lunse, V. Wittmann, C. Schumacher, S. Unsleber, H. Brotz-
Oesterhelt, C. Mayer, G. Bierbaum, G. Mayer, ChemBioChem
2017, 18, 435-440.
J. A. Collins, I. Irnov, S. Baker, W. C. Winkler, Genes Dev. 2007,
21, 3356-3368.
C. E. Lunse, M. S. Schmidt, V. Wittmann, G. Mayer, ACS Chem.
Biol. 2011, 6, 675-678.
a) I. Gaugue, J. Oberto, H. Putzer, J. Plumbridge, PLoS One 2013,
8, e63025; b) J. Reizer, S. Bachem, A. Reizer, M. Arnaud, M. H.
Saier, Jr., J. Stulke, Microbiology 1999, 145 ( Pt 12), 3419-3429.
E. D. Brown, G. D. Wright, Nature 2016, 529, 336-343.
D. J. Klein, A. R. Ferre-D'Amare, Science 2006, 313, 1752-1756.
a) G. Mayer, M. Famulok, ChemBioChem 2006, 7, 602-604; b) K.
Blount, I. Puskarz, R. Penchovsky, R. Breaker, RNA Biol. 2006, 3,
77-81.
[10]
[11]
[36]
[37]
[12]
[38]
[39]
[40]
[13]
[14]
[15]
[16]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
B. Giese, W. Damm, T. Witzel, H. G. Zeitz, Tetrahedron Lett. 1993,
34, 7053-7056.
A. Gansauer, H. Bluhm, B. Rinker, S. Narayan, M. Schick, T.
Lauterbach, M. Pierobon, Chem. - Eur. J. 2003, 9, 531-542.
K. A. Scheidt, H. Chen, B. C. Follows, S. R. Chemler, D. S. Coffey,
W. R. Roush, J. Org. Chem. 1998, 63, 6436-6437.
G. N. Wang, P. S. Lau, Y. F. Li, X. S. Ye, Tetrahedron 2012, 68,
9405-9412.
[17]
[18]
A. M. DiLauro, W. J. Seo, S. T. Phillips, J. Org. Chem. 2011, 76,
7352-7358.
B. M. Sattelle, B. Bose-Basu, M. Tessier, R. J. Woods, A. S.
Serianni, A. Almond, J. Phys. Chem. B 2012, 116, 6380-6386.
A. Canales, J. Angulo, R. Ojeda, M. Bruix, R. Fayos, R. Lozano, G.
Gimenez-Gallego, M. Martin-Lomas, P. M. Nieto, J. Jimenez-
Barbero, J. Am. Chem. Soc. 2005, 127, 5778-5779.
F. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2012, 2, 73-
78.
a) S. Sinnecker, A. Rajendran, A. Klamt, M. Diedenhofen, F.
Neese, J. Phys. Chem. A 2006, 110, 2235-2245; b) J. M. Tao, J.
P. Perdew, V. N. Staroverov, G. E. Scuseria, Phys. Rev. Lett.
2003, 91; c) F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys.
2005, 7, 3297-3305; d) Y. Zhao, D. G. Truhlar, J. Phys. Chem. A
2005, 109, 5656-5667; e) A. Klamt, G. Schuurmann, J. Chem.
Soc., Perkin Trans. 2 1993, 799-805.
C. E. Lünse, University of Bonn 2012.
J. Viladoms, M. J. Fedor, J. Am. Chem. Soc. 2012, 134, 19043-
19049.
J. J. Posakony, A. R. Ferre-D'Amare, J. Org. Chem. 2013, 78,
4730-4743.
J. A. K. Howard, V. J. Hoy, D. OHagan, G. T. Smith, Tetrahedron
1996, 52, 12613-12622.
[49]
[50]
[19]
[20]
[21]
[51]
[52]
[22]
[23]
[24]
[53]
[54]
[25]
[26]
[27]
T. J. McCarthy, M. A. Plog, S. A. Floy, J. A. Jansen, J. K. Soukup,
G. A. Soukup, Chem. Biol. 2005, 12, 1221-1226.
[55]
[56]
[57]
[58]
P. J. McCown, A. Roth, R. R. Breaker, RNA 2011, 17, 728-736.
O. Trott, A. J. Olson, J. Comput. Chem. 2010, 31, 455-461.
J. D. Dunitz, R. Taylor, Chem. - Eur. J. 1997, 3, 89-98.
M. Morgenthaler, E. Schweizer, A. Hoffmann-Roder, F. Benini, R.
E. Martin, G. Jaeschke, B. Wagner, H. Fischer, S. Bendels, D.
Zimmerli, J. Schneider, F. Diederich, M. Kansy, K. Muller,
ChemMedChem 2007, 2, 1100-1115.
J. Lim, B. C. Grove, A. Roth, R. R. Breaker, Angew. Chem. Int. Ed.
2006, 45, 6689-6693.
a) O. Arjona, A. M. Gomez, J. C. Lopez, J. Plumet, Chem. Rev.
2007, 107, 1919-2036; b) R. J. Ferrier, Glycoscience:
Epimerisation, Isomerisation and Rearrangement Reactions of
Carbohydrates, Vol. 215, Springer, 2001.
[28]
[29]
M. J. Sofia, R. Hunter, T. Y. Chan, A. Vaughan, R. Dulina, H. M.
Wang, D. Gange, J. Org. Chem. 1998, 63, 2802-2803.
R. J. Ferrier, J. Chem. Soc., Perkin Trans. 1 1979, 1455-1458.
This article is protected by copyright. All rights reserved.

10.1002/chem.201702371
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
Fluorinated sugar mimics: Mono-
fluorinated carba-sugar analogs of
glucosamine-6-phosphate and
idosamine-6-phosphate were
Daniel Matzner, Anna Schüller, Torben
Seitz, Valentin Wittmann, Günter
Mayer*
Page No. – Page No.
synthesized. Fluoro-carba-
glucosamine-6-phosphate acts as an
activator of the unique self-cleavage
mechanism of the glmS ribozyme.
Fluoro-carba-sugars are
glycomimetic activators of the glmS
ribozyme
This article is protected by copyright. All rights reserved.