Structure-based design of highly crowded ribostamycin/kanamycin hybrids as a new family of antibiotics

Article abstract of DOI:10.1002/chem.200903003

Sugar sugar! The synthesis and evaluation of ribostamycin/kanamycin hybrids incorporating a highly crowded trisubstituted aminocyclitol unit that should provide maximum complementation with the RNA receptor are reported (see picture). Analysis shows that the existing conflicts between the different sugar rings can be significantly alleviated by a simple chemical modification, leading to an improvement in activity. (Figure Presented)

Full text of DOI:10.1002/chem.200903003

DOI: 10.1002/chem.200903003
Structure-Based Design of Highly Crowded Ribostamycin/Kanamycin
Hybrids as a New Family of ACHUTNGRNEGUN ntibiotics
Julia Revuelta,[a] Tatiana Vacas,[a] Francisco Corzana,[b] Carlos Gonzalez,[c]
Agatha Bastida,[a] and Juan Luis Asensio*[a]
Aminoglycoside antibiotics are highly potent, wide-spec-
trum bactericidals. These drugs bind to the aminoacyl-tRNA
site (A-site) in the ribosome decoding region inducing
codon misreading and inhibiting translocation, which even-
tually results in cell death.[1] Unfortunately, their use in clini-
cal practice has been seriously limited as a result of their
toxicity and susceptibility to enzymatic inactivation[2]
prompting the search for new active derivatives.[3]
Most aminoglycosides incorporate an aminocyclitol unit
called 2-deoxystreptamine (2-DOS) that can exhibit two al-
ternative glycosidation patterns, giving rise to the so-called
4,5-DOS and 4,6-DOS subfamilies (Scheme 1). The rele-
vance of the different sugar units for biological activity has
been firmly established from structural and functional stud-
ies.[4–6] Thus, the pseudodisaccharide fragment I/II (see
Scheme 1), with slightly different patterns of OH/NH2 distri-
bution in unit I, is present in most aminoglycosides and has
been shown to be essential for specific complex formation
with the prokaryotic 16S RNA.[4] Despite their secondary
importance, sugar units at the 5- or 6-positions in the 2-DOS
ring provide additional contacts with the lower and upper
stems, respectively, of the RNA receptor as shown by de-
tailed NMR and crystallographic studies carried out in
recent years (represented in Scheme 1 and Figure 1).[5]
[a] Dr. J. Revuelta, Dr. T. Vacas, Dr. A. Bastida, Dr. J. L. Asensio
Departamento de Quꢀmica Bio-orgꢁnica
Instituto de Quꢀmica Orgꢁnica General (CSIC)
Juan de la Cierva 3, 28006 Madrid (Spain)
Scheme 1. The general strategy employed in the design of a new family
Fax : (+34)91-5644853
of antibiotics. The pseudodisaccharide fragment common to most amino-
E-mail: iqoa110@iqog.csic.es
glycosides is represented at the top. Addition of a sugar ring (grey) at the
[b] Dr. F. Corzana
5- or 6-position of the aminocyclitol unit gives rise to the 4,5-DOS (left)
Departamento de Quꢀmica
and 4,6-DOS (right) subfamilies, respectively. The functional relevance of
Universidad de La Rioja. UA-CSIC
the additional drug/RNA contacts established through ring III (indicated
Madre de Dios 51, 26006 LogroÇo, La Rioja (Spain)
in each case) is reflected in the antibiotic activities (typical ranges for the
[c] Dr. C. Gonzalez
MIC values are also presented for comparison). We propose the synthesis
and evaluation of the highly crowded new derivatives 5 and 6, which in-
corporate the trisubstituted 2-DOS scaffold. The ribose unit, numbered
as III in ribostamycin (2), has been renumbered as IV in the non-natural
pseudotetrasaccharides.
Instituto de Quꢀmica-Fꢀsica Rocasolano (CSIC)
Serrano 11,. 28006 Madrid (Spain)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200903003.
2986
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 2986 – 2991

COMMUNICATION
considered that the unfavourable contacts between the dif-
ferent sugar units might be alleviated by adequate modifica-
tions of the drug.
Preliminary conformational studies carried out in our
group support the idea that the incorporation of a xylos-
amine unit, instead of glucosamine, as ring III should drasti-
cally reduce the contacts between the different sugar rings,
providing an improved pseudoACHTUNGTRNEtNNUG etAHCTNURTGEGrNNUN aACHTNUGTRENNsUGN accharide structure for
the RNA binding pocket. According to the X-ray data the
glucose hydroxymethyl group is not involved in any interac-
tion with the RNA receptor and therefore this substitution
should not destabilize the complex. To prove the viability of
the proposed new family of aminoglycosides, we have ad-
dressed the synthesis and evaluation of pseudotetrasacchar-
ides 5 and 6 (Scheme 1). From a synthetic perspective, the
preparation of such derivatives is challenging as it involves
the glycosidation of an extremely hindered hydroxyl func-
tion (at the 5-position of the aminocyclitol ring). Despite
this potential limitation, we have demonstrated that such re-
actions can be successfully accomplished in moderate yields.
Compound 5 was prepared from kanamycin-A (4) follow-
ing the five-step procedure in Scheme 2. The free amine
groups were transformed into azides by means of a diazo-
transfer reaction with freshly prepared triflyl azide (TfN3).[7]
The subsequent acetylation with acetic anhydride led to de-
rivative 9 in excellent yield.
Figure 1. As shown by X-ray studies, the disaccharides I/II, present in
both kanamycin-A and ribostamycin, occupy nearly identical positions
within the RNA binding pocket. A superposition of kanamycin-A and
ribostamycin I/II fragments in the complex is shown. Aminoglycoside/
RNA contacts established through sugar unit III in both cases are high-
lighted.
AHCTUNGTRENNUNG
Bearing in mind these experimental findings, we analysed
the viability of a new family of antibiotics incorporating the
pseudodisaccharide fragment common to most aminoglyco-
sides (rings I and II in Scheme 1) with a neutral ribose unit
and a pyranose unit at the 5- and 6-positions, respectively, of
the 2-DOS moiety (see Scheme 1). These novel derivatives
could be considered as hybrids between those antibiotics
pertaining to the 4,5- and 4,6-DOS subfamilies and simulta-
neously exploit the ligand/RNA contacts typical of them
both. In principle, their lower
Glycosyl donor 10 was prepared by using a well-estab-
lished procedure.[8] Glycosidation of 9 with 10 was carried
flexibility and more optimized
occupancy of the RNA binding
pocket should translate into an
increased specificity and re-
duced toxicity.
It could be hypothesised that
steric crowding, inherent to the
4,5-/4,6-DOS hybrid scaffold,
could have a negative influence
on its RNA binding properties
and antibiotic activity. Howev-
er, extensive molecular dynam-
ics simulations on model pseu-
dotetrasaccharides in complex
with RNA (see Supporting In-
formation) show that, despite
their increased dimensions,
these ligands should, with
a
minor conformational adjust-
ment, be tolerated within the
RNA binding pocket. More-
over, according to these data,
the ligand/receptor contacts es-
tablished through the three
sugar units (I, II and III), previ-
ously observed in natural com-
plexes, are compatible with
Scheme 2. Methodology employed for the synthesis of 5. a) 1. Amberlite 400 (OHÀ), H2O, RT, 3 h; 2. TfN3,
CuSO4, toluene/MeOH/py, RT, 16 h; b) Ac2O, DMAP, py, RT, over night, yield: 80% (global yield for steps a
and b); c) BF3·Et2O, CH2Cl2, À25 to À208C, glycosyl donor 10, yield: 30%; d) MeONa, MeOH/THF, 2 h; e)
each other. Finally, it should be Me3P, THF, NaOH, 24 h, 558C, yield: 37% (global yield for steps d and e).
Chem. Eur. J. 2010, 16, 2986 – 2991
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2987

J. L. Asensio et al.
out under conventional condi-
tions (À20/À258C, BF3·OEt2
catalyst and CH2Cl2 solvent).
Close inspection of minimized
models of the acceptor 9 sug-
gests that the reactive OH func-
tion at the 5-position of the 2-
DOS ring is, to a large extent,
hidden by the protected vicinal
units I and III. We were pleased
to discover that, despite its pre-
sumably low accessibility, the
stereospecific glycosidation pro-
ceeded successfully to give the
pseudotetrasaccharide 11 in
moderate yields. Finally, hydrol-
ysis of the acetyl groups fol-
lowed by reduction of the
azides, employing the Stauding-
er reaction, yielded derivative
5.
Pseudotetrasaccharide 6 was
prepared from the neamine ac-
ceptor 12 employing two con-
Scheme 3. Methodology employed for the synthesis of 6. a) Glycosyl donor 14, TfOH, NIS, CH2Cl2/Et2O,
À408C, yield: 89%; b) BF3·Et2O, CH2Cl2, À25 to À208C, glycosyl donor 10, yield 41%; c) MeONa, MeOH/
THF, 2 h; d) Me3P, THF, NaOH, 24 h, 558C; e) H2, Pd/C, H2O/AcOH; yield 35% (global yield for steps c–e);
secutive glycosidation reactions f) 1. NaN3/DMF; 2. benzylation; yield 42%.
(Scheme 3). First, the readily
accessible 6-position of the ami-
nocyclitol was regioselectively glycosidated with donor 14,
stacking of the 2AP base that accompanies complex forma-
employing the methodology described by Chang and co-
tion causes a significant increase in the fluorescence emis-
sion at 370 nm that can be monitored by titration experi-
ments to derive the Kd values (Figure 2b). Keeping in mind
that, for this family of antibiotics, RNA binding and biologi-
cal activity are frequently poorly correlated[12] we also mea-
sured minimal inhibitory concentration (MIC) values against
the bacterial strain E. coli (DH5a). Typical binding curves
obtained for aminoglycosides 3–6 are shown in Figure 2.
The RNA binding affinities and biological activities mea-
sured for the hybrid aminoglycosides 5 and 6, together with
those exhibited by related natural antibiotics, are shown in
Table 1 and Figure 3.
workers, to give the corresponding pseudotrisacchaACTHNUTRGNEUNG
ride 15.[9]
The reaction proceeded stereospecifically and in good
yields. Donor 14 was obtained from epoxide 13[10] in just
two steps: opening of the oxirane ring with sodium azide in
DMF followed by the protection of the resulting OH func-
tions with benzyl groups. As a final step, a second glycosida-
tion reaction was performed at the 5-position of the 2-DOS
unit employing the glycosyl donor 10. After hydrolysis of
the acetyl groups the final product was obtained from the
Staudinger reaction followed by hydrogenolysis of the
benzyl protecting groups. The final pseudotetrasaccharide
was purified by ion-exchange chromatography.
It can be observed that, for the natural antibiotics, the in-
troduction of a ribose moiety at the 5-position of the amino-
Subsequently, the binding properties of the non-natural
derivatives, 5 and 6, and the
parent natural compounds, 2–4,
to ribosomal RNA were ana-
lysed. For this we employed the
fluorescence-based
approach
developed independently by
Pilch and Hermann.[11] This
methodology makes use of a
modified 27-mer RNA frag-
ment (A-siteACHTUNGTRENNUNG(2AP)) consisting
of the sequence of the riboso-
mal A-site with the fluorescent
base analogue, 2-amino
(2AP), replacing key residue
ACHUTGTNRNEUGpN urCAHTUNGTRENiNUGN ne
Figure 2. a) Schematic representation of the modified RNA fragment employed in our binding studies and the
structural change induced by the ligand binding. b) Some representative binding curves obtained for natural
A1492 (Figure 2a). The de- aminoglycosides 3 and 4 (black) and the related pseudotetrasacchꢁrides 5 and 6 (grey).
2988
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 2986 – 2991

Ribostamycin/Kanamycin Hybrids
COMMUNICATION
Table 1. RNA-binding affinities (Kd) and biological activities (MIC) of
aminoglycosides 1–6.
complexation and not with the complex stability itself. Ac-
cording to this view, the additional ligand/RNA contacts es-
tablished by the ribose unit in ribostamycin (2) may further
decrease the internal mobility of the RNA. Although
merely speculative, the entropic penalty associated with the
extra freezing of the receptor may compensate for the en-
thalpic benefit derived from the additional RNA/ligand con-
tacts (enthalpy/entropy compensation).[12,13]
[b]
Kd [mm][a]
–/
(7.39)[c]
15.16/
0.88
6.39
67.62
2.68
MIC [mgmLÀ1
]
1
2
3
4
5
6
G
–/(64)[c]
5–10/(8)[c]
2.0–2.5
2.0–2.5
25–35
(16.35)[c]
ACHTUNGTRENNUNG
1.5–2.0
Different behaviour was observed for the non-natural
aminoglycosides. Thus, the introduction of a ribose ring to
kanamycin-A, to yield derivative 5 (kanamycin-A, 4, vs de-
rivative 5), has a negative impact on both complex forma-
tion and antibiotic activity. More specifically, the Kd value
increases by one order of magnitude. This represents a
+1.40 kcalmolÀ1 destabilisation of the aminoglycoside/RNA
complex at 258C, almost an extra 1.0 kcalmolÀ1 energy pen-
alty in comparison with that caused by the ribose ring in
natural ribostamycin (2). As a consequence the antibiotic
activity is now drastically reduced (from ca. 2.5 mgmLÀ1 in
kanamycin-A, 4, to ca. 30 mgmLÀ1 in derivative 5). A specu-
lative but plausible explanation for this behaviour is that the
close proximity of the III and IV sugar units in 5 prevents
an optimal adaptation of the ligand to the receptor binding
pocket.
[a] Dissociation constants, Kd were measured at 258C in 100 mm NaCl,
10 mm phosphate and pH 7.0. [b] Experimental MIC values were mea-
sured against E. coli (DH5a). [c] Kd and MIC values in brackets were
taken from the literature.[13]
cyclitol ring of neamine (ribostamycin, 2, vs. neamine, 1, see
Table 1 and Figure 3) causes a twofold increase in the disso-
ciation constant Kd, which represents a 0.41 kcalmolÀ1 un-
favourable contribution to DG. Despite this reduction in the
aminoglycoside/RNA complex stability, ribostamycin exhib-
its an enhanced biological activity (almost an eightfold de-
crease in the MIC value). Piltch and co-workers have shown
that aminoglycoside biological activity strongly correlates
with the degree of conformational restriction that the drug
imposes on the key RNA residues, A1492 and A1493, on
Figure 3. The aminoglycoside/RNA complex stabilities (DG) and biological activities (MIC values) measured for several natural and non-natural amino-
glycosides, with (right) and without (left) a neutral ribose unit (grey) at the 5-position of the 2-DOS ring, are represented.
Chem. Eur. J. 2010, 16, 2986 – 2991
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2989

J. L. Asensio et al.
This hypothesis is fully supported by the experimental
data measured for the xylose derivative (6). In this particu-
lar case, the RNA binding affinity is reduced by only a
factor of three with respect to the natural compound kana-
mycin-B (3), implying a destabilisation of the aminoglyco-
side/RNA complex of only 0.68 kcalmolÀ1. This energy pen-
alty is similar to that caused by the ribose ring in natural ri-
bostamycin (2, +0.40 kcalmolÀ1). Moreover, the biological
activity is not reduced by the presence of the additional
ring IV. In contrast, derivative 6 exhibits a slightly improved
MIC value (ca. 1.7 mgmLÀ1). It seems that the removal of
the hydroxymethyl function from ring III alleviates the
steric conflict between rings III and IV allowing a more opti-
mal adaptation of the antibiotic to the RNA binding pocket.
In summary, our analysis shows that the simultaneous
presence of ribose and glucose sugar units at 5- and 6-posi-
tions, respectively, of the aminocyclitol ring leads to a signif-
icant increase in both the Kd (dissociation constant for the
ligand/RNA complex) and MIC values. However, the steric
conflict between sugar rings III and IV can be drastically re-
duced by the removal of a single hydroxymethyl function in
unit III (replacement of glucose by xylose) that is not in-
volved in direct drug/RNA contacts. According to our data,
this simple chemical modification produces a dramatic im-
provement in both the ligand/RNA binding strength and
biological activity.
Overall, our study demonstrates the viability of the pro-
posed new family of 4,5-/4,6-DOS hybrid aminoglycosides.
Considering its improved complementarity with the RNA
A-site, the 4,5-/4,6-DOS hybrid scaffold constitutes an inter-
esting basis for further modifications aiming to increase
RNA binding affinity by establishing simultaneous contacts
with the RNA upper and lower stems. Thus, the incorpora-
tion of a 2,6-diaminoidose at the 3-position of the ribose
would lead to neomycin/kanamycin hybrids with high RNA
binding affinity. Moreover, the replacement of the OH func-
tion at 2-position of the ribose ring by a smaller hydrogen-
bonding acceptor, such as a fluorine atom, might further
reduce the interaction between rings III and IV leading to
additional improvements in biological activity. The larger
steric volume of these derivatives, together with their limit-
ed flexibility and better occupation of the RNA binding
pocket might translate into an improved selectivity for the
prokaryotic ribosome and a reduction in toxicity. Current ef-
forts to explore the potential of the 4,5-/4,6-DOS scaffold in
the design of new aminoglycosides are currently underway
in our laboratory.
Acknowledgements
This investigation was supported by research grants from the Spanish
“Plan Nacional” (MCYT) CTQ2007-67403/BQU, CTQ2004-04994/BQU
and from CAM, S2009/ppq-1752. T.V. thanks the CSIC for a JAE fellow-
ship. J.R and F.C. thank the Ministerio de Educaciꢃn y Ciencia for a Juan
de la Cierva fellowship and a Ramꢃn y Cajal contract, respectively. The
authors also thank CESGA for computer support.
Keywords: antibiotics
·
oligosaccharide recognition
·
oligosaccharides · RNA recognition
[1] a) F. Walter, Q. Vicens, E. Westhof, Curr. Opin. Chem. Biol. 1999, 3,
694–704; b) T. Hermann, Angew. Chem. 2000, 112, 1962–1979;
Angew. Chem. Int. Ed. 2000, 39, 1890–1905; c) R. Schroeder, C.
Waldsich, H. Wank, EMBO J. 2000, 19, 1–9; d) J. S. Sucheck, C. H.
Wong, Curr. Opin. Chem. Biol. 2000, 4, 678–686.
[2] a) D. Greenwood in Antimicbrobial Chemotherapy, Oxford Univer-
sity Press, Oxford, 1995, pp. 32–48; b) S. Magnet, J. S. Blanchard,
Chem. Rev. 2005, 105, 477–497; c) C. A. Smith, E. N. Baker, Curr.
Drug Targets 2002, 3, 143–160; d) E. Azucena, S. Mobashery, Drug
Resist. Updates 2001, 4, 106–117; e) A. E. Talaska, J. Schacht in
Aminoglycoside Antibiotics: From Chemical Biology to Drug Dis-
covery (Ed.: D. P. Arya), Wiley, New York, 2007, pp. 255–266.
[3] a) F. Zhao, Q. Zhao, K. F. Blount, Q. Han, Y. Tor, T. Hermann,
Angew. Chem. 2005, 117, 5463–5468; Angew. Chem. Int. Ed. 2005,
44, 5329–5334; b) K. G. Blount , F. Zhao, T. Hermann, Y. Tor, J.
Am. Chem. Soc. 2005, 127, 9818–9829; c) A. Bastida, A. Hidalgo,
J. L. Chiara, M. Torrado, F. Corzana, J. M. CaÇadillas, P. Groves, E.
Garcia-Junceda, C. Gonzalez, J. Jimenez-Barbero, J. L. Asensio, J.
Am. Chem. Soc. 2006, 128, 100–116; d) J. L. Asensio, A. Hidalgo,
A. Bastida, M. Torrado, F. Corzana, E. G. Junceda, J. Canada, J. L.
Chiara, J. Jimꢄnez-Barbero, J. Am. Chem. Soc. 2005, 127, 8278–
8279; e) J. Roestamadji, I. Graspas, S. Mobashery, J. Am. Chem.
Soc. 1995, 117, 11060–11069; f) J. Haddad, S. Vakulenko, S. Mo-
bashery, J. Am. Chem. Soc. 1999, 121, 11922–11923; g) B. Francois,
J. Szychowski, S. Sekhar A. K. Pachamuthu, E. E. Swayze, R. H.
Griffey, M. T. Migawa, E. Westhof and S. Hanessian, Angew. Chem.
2004, 116, 6903; Angew. Chem. Int. Ed. 2004, 43, 6735–6738;
h) K. G. Blount , F. Zhao, T. Hermann, Y. Tor, J Am Chem. Soc.
2005, 127, 9818–9829; i) I. Nudelman, A. Rebibo-Sabbah, M. Cher-
niavsky, V. Belakhov, M. Hainrichson, F. Chen, J. Schacht, D. S.
Pilch, T. Ben-Yosef, T. Baasov, J. Med. Chem. 2009, 52, 2836–2845;
j) V. Pokrovskaya, V. Belakhov, M. Hainrichson, S. Yaron, T.
Baasov, J. Med. Chem. 2009, 52, 2243–2254; k) F. S. Liang, S. K.
Wang, T. Nakatani, C. H. Wong, Angew. Chem. 2004, 116, 6658;
Angew. Chem. Int. Ed. 2004, 43, 6496–6500.
[4] D. Fourmy, M. I. Recht, J. D. Puglisi, J. Mol. Biol. 1998, 277, 347–
362.
[5] a) D. Fourmy, M. I. Recht, S. C. Blanchard, J. D. Puglisi, Science
1996, 274, 1367–1371; b) A. P. Carter, W. M. Clemons, D. E. Bro-
dersen, R. J. Morgan-Warren, B. T. Wimberly, V. Ramakrishnan,
Nature 2000, 407, 340–348; c) Q. Vicens, E. Westhof, Chem. Biol.
2002, 9, 747–755; d) B. Francois, R. J. M. Rusell, J. B. Murray, F.
Aboul-ela, B. Masquida, Q. Vicens, E. Westhof, Nucleic Acids Res.
2005, 33, 5677–5690; e) J. Kondo, R. J. M. Russell, J. B. Murray, E.
Westhof, Biochimie 2006, 88, 1027–1031; f) Q. Vicens, E. Westhof,
Structure 2001, 9, 647–658.
[6] S. N. Hobbie, P. Pfister, C. Bruell, P. Sander, B. Francois, E. Westhof,
E. C. Bçttger, Antimicrob. Agents Chemother. 2006, 50, 1489–1496.
[7] A. Titz, Z. Radic, O. Schawardt, D. Ernst, Tetrahedron Lett. 2006,
47, 2383–2385.
[8] I. Chiu-Machado, J. C. Palomino-Castro, O. Madrazo-Alornso, C.
Lopetegui-Palacios, V. Verez-Bencomo, J. Carbohydr. Chem. 1995,
14, 551.
Experimental Section
A detailed description of the synthetic protocols together with the char-
acterisation of products and intermediates is included in the Supporting
Information. Binding experiments and biological activities were per-
formed as previously described.[3c]
2990
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 2986 – 2991

Ribostamycin/Kanamycin Hybrids
COMMUNICATION
[9] a) J. Wang, J. Li, H. Cheng, H. Chang, C. T. Tanifum, H. Liu, P. G.
Czyryca, C. T. Chang, J. Med. Chem. 2004, 47, 1381–1384; b) J. Li, J.
Wang, P. G. Czyryca, H. Chang, T. W. Orsak, R. Evanson, C. T.
Chang, Org. Lett. 2004, 6, 1381–1384.
[10] N. Rehnberg, G. Magnusson, J. Org. Chem. 1990, 55, 5467–5476.
[11] a) S. Shandrick, Q. Zhao, Q. Han, B. K. Ayida, M. Takahashi, G. C.
Winters, K. B. Simonsen, D. Vourloumis, T. Hermann, Angew.
Chem. 2004, 116, 3239–3244; Angew. Chem. Int. Ed. 2004, 43, 3177–
3182; b) M. Kaul, C. M. Barbieri, D. S. Pilch, J. Am. Chem. Soc.
2004, 126, 3447–3453.
[12] M. Kaul, C. M. Barbieri, D. S. Pilch, J. Am. Chem. Soc. 2006, 128,
1261–1271.
[13] a) C. H. Wong, M. Hendrix, E. S. Priestley, W. A. Greenberg, Chem.
Biol. 1998, 5, 397–406; b) G. Yang, J. Tryiska, Y. Tor, J. A. McCam-
mon, J. Med. Chem. 2006, 49, 5478–5490.
[14] P. M. Waterworth in Laboratory Methods in Antimicrobial Chemo-
therapy (Ed.: L. Garrod), Churchill Livingstone, Edinburgh, 1978,
pp. 31–40.
Received: October 30, 2009
Published online: February 16, 2010
Chem. Eur. J. 2010, 16, 2986 – 2991
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2991