Conformational study of tylosin A in water and full assignments of 1H and 13C spectra of tylosin A in D2O and tylosin B in CDCl3

Full text of DOI:10.1002/mrc.4537

Conformational study of tylosin A in water and full assignments
1
13
of H and C spectra of tylosin A in D O and tylosin B in CDCl
2
3
1
2
1
1
Biljana Arsic , Juan A. Aguilar , Richard A. Bryce and Jill Barber
1
Division of Pharmacy and Optometry, School of Health Sciences, University of
Manchester, Oxford Road, M13 9PT, Manchester, United Kingdom
2
School of Chemistry, University of Manchester, Oxford Road, M13 9PL, Manchester,
United Kingdom
Correspondence: Dr Biljana Arsic, Division of Pharmacy and Optometry, School of Health
Sciences, University of Manchester, Oxford Road, M13 9PT, Manchester, United Kingdom,
E-mail: ba432@ymail.com
Tylosin A and tylosin B
1
13
Keywords: NMR assignments; H; C; macrolides; molecular modelling; veterinary
medicine
Introduction
The first paper describing the use of a macrolide to prevent liver infections, and subsequent
blood infections caused by the malaria parasite (Plasmodium falciparum) was published in
1
995 where it was suggested, after a controlled phase II trial study, that azithromycin has the
[
1]
potential to be at least as effective as mefloquine and doxycycline. However, randomized
studies conducted at Mahidol University Hospital (Bangkok, Thailand) in 2005 concluded
[2]
that azithromycin is not efficient as a single agent, but that quinine, tafenoquine, artesunate
and/or primaquine have additive, synergistic qualities when used with azithromycin. The
[2,3]
study also reported that dihydroartemisinin tends to act as an antagonist.
Finally, the
azithromycin-chloroquine combination was suggested as a potential alternative to the
[
4]
sulphadoxine-pyrimethamine treatment.
McColm and McHardy reported in 1984 that tylosin A also has anti-malarial activity
-1
against the Liverpool strain of P. falciparum (IC =0.1 µg mL ). Novel C-9 oximes of tylosin
5
0
B also show anti-malarial activity against chloroquine-resistant P. falciparum K1 malaria
[
5]
parasites.
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1002/mrc.4537
This article is protected by copyright. All rights reserved.

The macrolide antibiotic tylosin A was first isolated from a strain of Streptomyces fradiae
[
6]
[7-11]
in 1961. Its chemical structure was determined by several groups over the next decade.
Although 14-membered ring macrolides, such as erythromycin A were assigned to human
use, tylosin A has been used in domestic animals, principally as an anti-inflammatory
[
12]
agent. There are few adverse side-effects.
Tylosin A acts by binding to the 50S ribosomal subunit of susceptible bacteria. Resistance
is mediated by the methylation of N1 of the nucleotide G748 within hairpin 35 of 23S rRNA,
[
13]
catalysed by the methyltransferase Rlm AII (formerly known as TlrB). Resistance in the
tylosin producing Streptomyces fradiae is mediated by TlrB, TlrD and TlrA. TlrB expression
gives very minor resistance to tylosin by itself (MIC=2), and requires co-expression of TlrD
(which N-methylates position A2058) for modest resistance to tylosin (MIC=256). High level
tylosin resistance (>2000) is mediated by TlrA, which dimethylates position A2058. TlrA has
[
14]
the same enzymatic activity as ErmE.
Remarkably, the first crystal structure of intact tylosin A has the drug bound to 50S
[15]
ribosomal subunits of Haloarcula marismortui.
The X-ray structure of 5-O-
mycarosyltylactone was determined earlier and represents the pioneering work in the
[
16]
crystallography of tylosin derivatives.
There are reports of NMR assignments of tylosin A in aprotic solvents (CDCl3,
[17,18]
[18]
CD CN).
Simova et al. reported that the spectral assignment for tylosin A in protic
3
solvents (CD OD, D O) is very difficult because of the existence of three different tylosin A
3
2
forms in CD OD (ketone, hemiacetal 1, hemiacetal 2) and two in D O (ketone, hydrate).
3
2
They found that, in D O at 300 K, the ketone:hydrate ratio is 65:35. It is known that
2
[19-22]
macrolides, such as erythromycins A and C often exist in several forms
Tylosin B is a degradation product of tylosin A produced by acid hydrolysis.
alkaline degradation product of tylosin A, known as tylosinaldol, has also been found as a
.
[
23]
(An
[
24]
contaminant of medicinal tylosin intended for injection.)
Tylosin B (desmycosin) is
important principally because it can be used as a starting material for the synthesis of more
active derivatives. The following modifications of tylosin B have been reported: 1)
replacement and modification of mycinose (labelled red); 2) modification of the dienone
moiety (labelled pink); 3) modification of mycaminose (labelled blue); 4) modification of the
[
25]
C-3 hydroxyl group (labelled green) (Figure 1 B).
Figure 1 should be here)
(
The most successful derivative of tylosin B, tilmicosin, contains the substitution at position
This article is protected by copyright. All rights reserved.

[
26]
2
0 (20-deoxo-20-(3,5-dimethylpiperidin-1-yl) desmycosin)
Figure 2 should be here)
There have been attempts to determine the three dimensional structure of tylosin in aprotic
(Figure 2).
(
[
27,28]
solvents using molecular modelling and different force fields.
The molecular modelling
data were combined with NMR data that included NOE experiments as well as the
measurement of vicinal proton-proton coupling constants. These studies gave structures
[
29]
whose dominant conformation does not resemble that of the crystal structure of tylosin A.
We have previously shown that the larger macrolides do not behave as small molecules in
[
30-32]
standard molecular modelling calculations.
Overall, however, there is surprisingly little
published work on the three-dimensional structure of tylosin A. In particular, the
conformation of tylosin A in water has not been determined. In order to understand the
conformation of tylosin A we carried out both molecular modelling and NMR study.
Results and Discussion
Determination of the carbonyl:hydrate ratio of tylosin A in deuterated water, using
NMR spectroscopy
[18]
Simova et al. reported that at 300 K tylosin A exists as a mixture of carbonyl and hydrate
forms in a 65:35 ratio. We investigated the effect of temperature on this ratio in D O by NMR
2
1
spectroscopy monitoring the integral of the aldehyde H signal at 9.67 ppm, and comparing
its value with that of proton H-11. Within the 298-338 K temperature range, an increase in
temperature correlated with an increase of the aldehyde form. At 338 K, tylosin A exists
mostly as an aldehyde (approximately 80 % of the total tylosin A). Here we see an interesting
[
19,33,34]
parallel with erythromycin A
, which exists as a tautomeric mixture of keto and
hemiacetal forms in aqueous solution (Figure 1).
The three-dimensional structure of tylosin A
An unconstrained conformational search on the aldehyde form of tylosin A in water was
performed using molecular mechanics (Macromodel 8) with the MM2 force field.
As the starting point in the process of modelling the aldehyde form of tylosin A we have
used the only available structure of tylosin A (in the form of co-crystal structure with 50 S
[
15]
ribosomal subunit of Haloarcula marismortui). It is noteworthy that the crystal structure of
tylosin A alone has not been published, although the drug was isolated for the first time in the
mid-1960s.
This article is protected by copyright. All rights reserved.

Using Macromodel 8, we found the global minimum of tylosin A (Figure 3) 17 times
-1
(
energy E=280.9 kJ mol ). The global minimum satisfies all of the NMR constraints derived
from the ROESY experiments that were acquired to assist the molecular mechanics studies.
The exceptions were H11-H13, H 20-H4′, and H 6″-H2″b (the ROESY correlation table is
3
3
available in Supplementary Data as Table 1S). The H 20-H4′ correlations are vanishingly
3
small, while those of H 6″-H2″s were attributed to spin diffusion.
3
The H3-H11 and H4-H11 contacts, (characteristic of 14- and 15-membered macrolides and
necessary for grouping into various types of folded-in and folded-out structures), in the global
minimum structure are 3.34 Å and 5.04 Å with no indications of folded-in or folded-out
conformers. Other characteristic contacts for folded-in conformer (H3-H8, H8-H11) and H5-
H6Me are 4.88 Å, 4.68 Å and 2.84 Å respectively, so tylosin A does not adopt a folded-in
conformation. H5-H6Me Distance is short and characteristic of a folded-out conformation.
The H4-H6Me, distance, characteristic of folded-in conformers is, in tylosin A, 4.83 Å,
which indicates that tylosin A is certainly neither in a folded-in conformation nor in a folded-
out one. It seems that the presence of double bonds inside the lactone ring, as well of certain
sugar components make the conformation of tylosin A different from that of 14- and 15-
membered macrolide antibiotics.
(Figure 3 should be here)
When the global minimum for tylosin A (green) was superimposed on the crystal structure
obtained from co-crystallizing tylosyn A with H. marismortui ribosomes (pink), it was found
that the two structures were almost identical, except for the positions of the mycinose rings
that differ significantly (Figure 3). In particular, there is very good agreement in the
macrolide ring. This is interesting because of the difficulty in obtaining agreements between
the experimental and the modelled structures of azithromycin. This might be due to the fact
that the double bonds in tylosin impose some rigidity, thus limiting the range of possible
conformations.
2
D ROESY NMR experiments on tylosin A
NOESY experiments are seldom appropriate for macrolides in solution. The NOE is
normally positive for small molecules, goes through zero for medium-sized molecules (MW
range 700-1500), and becomes negative for large molecules (MW>1500). The 2D-ROESY is
therefore to be preferred in the present case (the molecular weight is 916 Da), although such
spectra are often contaminated by TOCSY peaks. The mixing time was optimised
This article is protected by copyright. All rights reserved.

experimentally as 200 ms. The 2D ROESY NMR spectrum is presented in Figure 1S.
Superposition of structures of azithromycin and tylosin A
[
35]
Baram et al.
published a crystal structure containing two molecules of azithromycin
bound to the Deinococcus radiodurans ribosome. We have superimposed a tylosin molecule
on top of these two azithromycin molecules, as shown in Figure 4. The structure of the two
molecules of azithromycin were extracted from the reported co-crystal structure with the
ribosome of D. radiodurans, while that of tylosin A was extracted from its co-crystal
structure with the ribosome of H. marismortui. The operation was carried out using the MOE
and the Macromodel 8 programmes. Figure 4 shows that a tylosin molecule has the size of
[
36]
two azithromycin molecules. It was suggested by Hansen et al. (2002)
that the mycinosyl
moiety of tylosin interacts with A841 (A748, E. coli numbering) in domain II of the 50S
ribosome. The mycarose unit is very important to the anti-bacterial activity; its removal
[
37]
reduces the anti-bacterial activity. This is consistent with the observation that the mycarose
sugar contributes up to 2/3 of the interaction surface energy, which is an important
[
36]
component of the binding free energy.
(Figure 4 should be here)
1
13
-1
Full H and C assignments of tylosin A in phosphate buffered solution (0.047 mol L ,
uncorrected pD=7)
1
13
1
The proton and carbon assignments of tylosin A were based on H, C{ H} DEPT, COSY,
1
13
1
13
H- C HMBC and H- C HSQC.
The assignments are shown in Table 1. Detailed explanation of the assignment procedure
together with the spectra used in the analysis is available in Supplementary Material.
Table 1 should be here)
H and C assignments of tylosin B (desmycarosyl tylosin) in CDCl are shown in Table 2.
(
1
13
3
Detailed description of the assignment procedure, as well as the spectra used is placed in
Supplementary material.
(Table 2 should be here)
Experimental
Synthesis of tylosin B (desmycarosyl tylosin)
The mixture of tylosin A tartrate (1.0000 g) and 0.2 M HCl (20 mL) was stirred for 4 h. After
This article is protected by copyright. All rights reserved.

the reaction was finished, extraction with chloroform (4x15 mL) was performed in order to
remove mycarose. The pH of the solution was adjusted to 8.0 using a saturated solution of
sodium bicarbonate and a universal indicator, and was extracted with chloroform (4x15 mL).
The extract was dried over anhydrous magnesium sulphate, the solid removed by filtration,
and the filtrate evaporated to dryness. White crystalline tylosin B was obtained (0.5101 g,
1
yield 70.45 %). Mp. 113-116 °C. The compound was identified using a combination of H,
1
3
1
1
1
1
1
1
13
1
13
C{ H} H- H DQF-COSY, H- H ROESY, H- C HMBC and H- C HMQC and low and
high resolution mass spectrometry. m/z (ES+): 772.6; HRMS (EI): m/z calculated for
C H O N : 772.4478. Found 772.4467.
39
66 14
1
Molecular modelling
[
15]
Tylosin A was constructed from the crystal structure of tylosin A
using Macromodel 8
[
38]
software.
The conformational analysis was performed via the Monte Carlo/multiple
[39]
minimum (MC/MM) approach, to identify the lowest energy conformation in solution. The
[40]
solute was described using the all-atom MM2 potential.
The effect of the solvent was
incorporated into the MC/MM calculations using the generalised Born/surface area (GB/SA)
[
41]
continuum solvent model for water.
Cut-offs of 12.0 Å and 7.0 Å were employed for
electrostatic and van der Waals non-bonded interactions, respectively. The MC simulation
4
involved 10 steps at 300 K, applied to all rotatable bonds, with random torsional rotations of
3
up to ±180°. This was combined with 10 steps of energy minimization. All conformational
[
38]
calculations were performed using the Macromodel 8.0 and BatchMin suite of programs.
NMR analysis
Tylosin A tartarate was obtained from Aldrich and used without further purification while
tylosin B was synthesized as previously described. All measurements were carried out in
-
1
deuterated water or chloroform. A 0.047 mol L phosphate buffer was used in the former
case that give a pD of 7 (values provided without correction). The measurements were
performed using BrukerAvance 400 MHz and 500 MHz spectrometers (Karlsruhe, Germany).
NMR experiments of tylosin A in phosphate buffered D O (pD=7, values provided
2
1
without correction). The operating frequency for the H measurements was 400.132 MHz;
the spectral width (SWH) was 8223.685 Hz, the acquisition time (AQ) 3.98 s, the high-power
pulse (P1) 90˚ pulse was 7.79 μs, 65536 complex data points were collected (TD). The
number of scans was 16, the relaxation delay (d1) was 2.4 s. The temperature was 303.7 K.
This article is protected by copyright. All rights reserved.

1
3
1
Exponential functions were used with a line broadening (LB) of 0.3 Hz. For the C{ H}
spectrum, the operating frequency was 125.77 MHz, the P1=8.80 μs, AQ=1.1 s, 2 dummy
scans were used, SWH=29761.904 Hz, TD=65536; for the DEPT 90 spectrum, the operating
frequency was 125.77 MHz, P1=8.80 μs, AQ=1.1 s, dummy scans=4, SWH=29761.904 Hz,
TD=65536; for DEPT 135 spectrum, operating frequency was 125.77 MHz, P1=8.80 μs,
AQ=1.1 s, dummy scans=4, SWH=29761.904 Hz, TD=65536. Exponential function was used
1
3
as a window function for processing C spectrum with LB=1.50 Hz. The same window
function was used for DEPT 90 and DEPT 135 processing, but LB values in both cases were
1
.00 Hz. In case of the COSY spectrum, the operating frequency was 400.132 MHz,
AQ=0.25 s, SWH=4157.4 Hz, TD=2048x512, and QSINE window functions using a Fourier
number (SI) of 4096 and 2048 for the direct and indirect dimensions were performed,
respectively, with SSB=1. The acquisition parameters for the HSQC spectrum were AQ=0.15
s, SW(F1)=165.6 ppm, SWH(F2)=3378.4 Hz, TD=1024x256. The Echo-Antiecho procedure
1
was used to produce phase sensitive data (FnMODE) with operating frequencies for H and
1
3
C equal to 400.132 MHz and 100.62 MHz, respectively. The processing parameters for
HSQC were: window function-QSINE, SI=4098x1024, SSB=2. Regarding the HMBC, the
acquisition parameters were: AQ=0.61 s, SW(F1)=221.8 ppm, SWH (F2)=3378.4 Hz,
1
13
TD=4096x128 with operating frequencies for H and C equal to 400.13 MHz and 100.61
MHz, respectively. The ROESY experiments were recorded using the following acquisition
parameters: operating frequency 400.13 MHz, TD=2048x512, AQ=0.23 s, D1=2 s, P15=200
ms, FnMODE-States-TPPI; and the following processing parameters: SI=2048x1024,
window function was QSINE, SSB=2.
1
NMR experiments of tylosin B in deuterated chloroform. The operating frequency for H
measurements was 300.13 MHz; spectral width was SWH=6172.839 Hz, acquisition time
AQ=5.3 s, high-power pulse (P1)=8.3 μs, data points (TD)=65536, number of scans=16,
relaxation delay D1=2.4 s; T=294.4 K. Exponential function was used as a window function
1
3
for processing with LB=0.3 Hz. For the C spectrum, the operating frequency was 75.48
MHz, P1=7.80 μs, AQ=1.74 s, dummy scans=4, SWH=18832.393 Hz, TD=65536; for DEPT
9
0 spectrum, the operating frequency was 75.48 MHz, P1=7.80 μs, AQ=1.82 s, dummy
scans=4, SWH=17985.611 Hz, TD=65536; for DEPT 135 spectrum, operating frequency was
5.48 MHz, P1=7.80 μs, AQ=1.82 s, dummy scans=4, SWH=17985.611 Hz, TD=65536.
7
1
3
Exponential function was used as a window function for processing C spectrum with
LB=2.00 Hz. The same window function was used for DEPT 90 and DEPT 135 processing
This article is protected by copyright. All rights reserved.

and the same LB value (LB=2.00 Hz). In case of COSY spectrum, operating frequency was
00.13 MHz, AQ=0.30 s, SWH=3360.215 Hz, TD=2048x128, and QSINE window function
3
was used for processing with SI=4096x1024 and SSB=1. The acquisition parameters for the
HMQC spectrum were: AQ=0.15 s, SW(F1)=165.6 ppm, SWH(F2)=3360.215 Hz,
1
13
TD=1024x128, FnMODE-QF with operating frequencies for H and C equal to 300.13
MHz and 75.47 MHz, respectively. Processing parameters for HMQC were: window
function-QSINE, SI=4096x1024, SSB=1 and MC2-QF. Regarding the HMBC, the
acquisition parameters were: AQ=0.64 s, SW(F1)=222.3 ppm, SWH (F2)=3205.128 Hz,
1
13
TD=4096x128 with operating frequencies for H and C equal to 300.13 MHz and 75.48
MHz, respectively.
Conclusions
1
13
We report the full H and C NMR assignments on tylosin A in deuterated water, as well as
those of tylosin B in deuterated chloroform. NMR investigations on the equilibrium of
carbonyl to hydrate ratio shows that the percentage of the aldehyde form increases with the
rise of temperatures. Molecular mechanics calculations performed using the Macromodel 8
package gave, as a global minimum, a conformation of tylosin A which satisfies almost all
’
2
D ROESY NMR constraints, except those of H11-H13, H 20-H4 , and H 6”-H2”b. Keeping
3
3
in mind that H 20-H4’ is vanishingly small and that H 6”-H2”s is probably due to spin
3
3
diffusion, we can say that modelling represents an excellent tool for the prediction of the
conformation of tylosin A in aqueous solution. It was found that the predicted structure is
almost the same as the crystal structure (differences occurred only in the positions of the
mycinose rings). Two azithromycin molecules are almost completely superimposable on one
tylosin A structure, which suggests that tylosin A can potentially be used as anti-malarial.
Acknowledgements
We would like to thank Dr Andrew Regan for his valuable advice regarding calculations in
Macromodel 8. Also, special thanks to Mr Roger Speak, Dr Elena Bichenkova, Professor
Gareth A. Morris and Dr Steven M. Miles for their guidance through different NMR
experiments.
This article is protected by copyright. All rights reserved.

References
[
1] S. L. Anderson, J. Berman, R. Kuschner, D. Wesche, A. Magill, B. Wellde, I. Schneider,
M. Dunne, B. Schuster. Ann. Intern. Med. 1995, 123, 771-773.
2] ClinicalTrials.gov, Azithromycin Combination
http://clinicaltrials.gov/ct2/show/NCT00299208 [04/12/2015]
3] C. Ohrt, G. D. Willingmyre, P. Lee, C. Knirsch, W. Milhous. Antimicrob. Agents
Chemother. 2002, 46, 2518-2524.
[
Therapy
for
Malaria.
[
[
[
4] R. M. Chico, R. Pittrof, B. Greenwood, D. Chandramohan. Malar. J. 2008, 7, 255-268.
5] A. Tsutsui, T. Hirose, A. Ishiyama, M. Iwatsuki, A. Yokota, H. Maruyama, H. Matsui, K.
Otoguro, H. Hanaki, S. Omura, T. Sunazuka. J. Antibiot. 2013, 66, 191-194.
6] R. L. Hamill, M. E. Haney Jr, M. Stamper, P. F. Wiley. Antibiot. Chemother. 1961, 11,
28-334.
[
3
[
[
[
7] H. Achenbach, W. Regel, W. Karl. Chem. Ber. 1975, 108 (7), 2481-2483.
8] S. Omura, A. Nakagawa, I. H. Machida. Tetrahedron Lett. 1977, 18 (12), 1045-1048.
9] S. Omura, A. Nakagawa, A. Neszmelyi, S. D. Gero, A. M. Sepulchre, F. Piriou, G.
Lukacs. J. Am. Chem. Soc. 1975, 97(14), 4001-4009.
[
[
[
10] A. Neszmelyi, S. Omura, G. Lukacs. J. Chem. Soc. Chem. Comm. 1976, 3, 97-98.
11] R. B. Morin, M. Gorman. Tetrahedron Lett. 1964, 5 (34), 2339-2345.
12] X-Y. Cao, M. Dong, J.-Z. Shen, B.-B.Wu, C.-M. Wu, X.-D. Du, Z. Wang, Y.-T. Qi, B.-
Y. Li. Int. J. Antimicrob. Ag. 2006, 27, 431-438.
[
(
[
[
13] I. Lebars, C. Husson, S. Yoshizawa, S. Douthwaite, D. Fourmy. J. Mol. Biol. 2007, 372
2), 525-534.
14] M. Liu, S. Douthwaite. PNAS 2002, 99 (23), 14658-14663.
15] S. Omura, H. Matsubara, A. Nakagawa, A. Furusaki, T. Matsumoto. J. Antibiot. 1980,
3
3 (8), 915-917.
16] N. D. Jones, M. O. Chaney, H. A. Kirst, G. M. Wild, R. H. Baltz, R. L. Hamill, J. W.
Paschal. J. Antibiot. 1982, 35 (4), 420-425.
17] C. Rossi, A. Donati, M. P. Picchi, M. R. Sansoni, G. Corbini, P. Corti. Magn. Reson.
Chem. 1992, 30 (10), 954-957.
18] S. Simova, G. Ivanova. Magn. Reson. Chem. 1996, 34 (4), 255-260.
[
[
[
This article is protected by copyright. All rights reserved.

[
19] J. Barber, J. I. Gyi, L. Lian, G. A. Morris, D. A. Pye, J. K. Sutherland. J. Chem. Soc.,
Perkin Trans. 2, 1991, 1489-1494.
[
[
20] J. R. Everett, E. Hunt, J. W. Tyler. J. Chem. Soc., Perkin Trans. 2. 1991, 10, 1481-1487.
21] P. Tyson, H. Barjat, W. Mann, J. Barber. J. Chem. Res. Synop. 1998, 727 (12), 3146-
3
158.
[
[
22] R. Tosukano. Jpn. Pat., 1985, 60038394.
23] J. Paesen, W. Cypers, K. Pauwels, E. Roets, J. Hoogmartens. J. Pharmaceut. Biomed.
1
995, 13 (9), 1153-1159.
24] H. A. Kirst, J. E. Toth, M. Debono, K. E. Willard, B. A. Truedell, J. L. Ott, F. T.
Counter, A. M. Felty-Duckworth, R. S. Pekarek. J. Med. Chem. 1988, 31, 1631-1641.
25] H. A. Kirst, G. M. Wild, R. H. Baltz, R. L. Hamill, J. L. Ott, F. T. Counter, E. E. Ose. J.
Antibiot. (Tokyo) 1982, 35, 1675-1682.
26] M. Debono, K. E. Willard, H. A. Kirst, J. A. Wind, G. D. Crouse, E. V. Tao, J. T.
Vicenzi, F. T. Counter, J. L. Ott, E. E. Ose, S. Omura. J. Antibiot. 1989, 42 (8), 1253-1267.
27] P. M. Ivanov. J. Mol. Struct. 1998, 440 (1-3), 121-130.
28] P. M. Ivanov. J. Mol. Struct. 2002, 606 (1-3), 217-229.
29] J. L. Hansen, J. A. Ippolito, N. Ban, P. Nissen, P. B. Moore, T. A. Steitz. Mol. Cell 2002,
[
[
[
[
[
[
1
0 (1), 117-128.
[
30] B. Arsic, A. Awan, R. J. Brennan, J. A. Aguilar, R. Ledder, A. J. McBain, A. C. Regan,
J. Barber. Med. Chem. Commun. 2014, 5, 1347-1354.
[
31] P. Novak, I. Tatic, P. Tepeš, S. Koštrun, J. Barber. J. Phys. Chem. A 2006, 110, 572-579.
32] P. Novak, J. Barber, A. Čikoš, B. Arsic, J. Plavec, G. Lazarevski, P. Tepeš, N. Kosutic-
[
Hulita. Bioorg. Med. Chem. Lett. 2009, 17, 5857-5867.
[
33] D. A. Pye, J. I. Gyi, J. Barber. J. Chem. Soc. Chem. Commun. 1990, 17, 1143-1145.
34] J. Barber, J. I. Gyi, D. A. Pye. J. Chem. Soc. Chem. Commun. 1991, 18, 1249-1252.
35] D. Baram, A. Yonath. FEBS Lett. 2005, 579, 948-954.
[
[
[
36] J. Hansen, J. A. Ippolito, N. Ban, P. Nissen, P. Moor, T. Steitz. Mol. Cell 2002, 10, 117-
1
28.
[
37] (a) S. Mutak, N. Marsic, M. D. Kramaric, D. Pavlovic. J. Med. Chem. 2004, 47, 411-
4
31.
This article is protected by copyright. All rights reserved.

(
b) H. Fu, S. Marquez, X. Gu, L. Katz, D. C. Myles. Bioorg. Med. Chem. Lett. 2006, 16,
259-1266.
c) T. Miura, K. Kanemoto, S. Natsune, K. Atsumi, H. Fushimi, T. Yoshida, K. Ajito. Bioorg.
Med. Chem. 2008, 16, 10129-10156.
1
(
[
38] F. Mohamadi, N. G. J. Richards, W. C. Guida, R. Liskamp, M. Lipton, C. Caufield, G.
Chang, T. Hendrickson, W. C. Still. J. Comput. Chem. 1990, 11 (4), 440-467.
[
39] G. Chang, W. C. Guida, W. C. Still. J. Am. Chem. Soc. 1989, 111, 4379-4386.
[
40] J.-H. Lii, S. Gallion, C. Bender, H. Wikstrom, N. L. Allinger, K. M. Flurchick, M. M.
Teeter. J. Comp. Chem. 1989, 10, 503-513.
41] W. C. Still, A. Tempczyk, R. C. Hawley, T. J. Hendricksen. J. Am. Chem. Soc. 1990,
12, 6127-6129.
[
1
This article is protected by copyright. All rights reserved.

1
13
Table 1. The full H and C NMR assignments of tylosin A in phosphate buffered D O
2
(pD=7, values provided without correction)
1
13
13
1
Position Multiplicity H (ppm) JHH (Hz)
C (ppm) HMBC connectivities ( C H)
1
-
m
m
m
m
m
m
m
m
m
-
d
d
-
dd
m
dt
m
m
t
-
-
-
-
-
-
-
-
-
181.3
H2a, H2b
2.18
2.67
3.83
1.67
3.65
1.94
1.50
1.90
2.55
-
2
42.5
H 18
3
3
4
5
6
71.3
43.6
83.6
34.0
H 18
3
-
H 18
3
H 21
3
7
34.6
H 21
3
-
-
-
8
9
48.0
212.2
121.6
152.5
139.1
146.8
47.6
H 21
3
H10, H11, H 21
3
1
0
1
2
3
4
5
6.59
7.32
-
15.5
15.5
-
10.4
-
-
1
1
1
1
1
H 22
3
H11, H 22
3
5.96
3.14
4.89
1.85
1.60
0.94
0.98
2.93
2.60
9.67
1.24
1.87
4.00
3.71
4.45
3.73
3.79
3.06
3.69
3.42
1.36
5.19
2.05
1.99
-
H10, H11, H 22
3
-
10.4; 2.6
78.8
H 17
3
-
-
7.3
3.9
-
-
-
6.5
1
6
27.6
H 17
3
1
1
7
8
11.8
11.3
-
-
d
m
m
s
d
s
dd
m
d
m
m
s
m
m
d
1
9
46.6
H 21
3
2
2
2
0
1
2
210.9
19.9
18.2
-
H7
H11
-
10.2; 3.4
2
3
72.3
H14
-
7.2
-
-
-
1
2
3
′
′
′
104.8
77.2
46.6
43.9
75.5
76.7
20.7
99.3
H2′
H 6′
3
H 7′/8′
6
7
′/8′
H 7′/8′
6
4
5
6
′
′
′
-
-
H 6′
3
H 6′
3
8.6
2.5
-
4.1
-
9.1
-
2.7
-
H2′
1
″
d
H 6″, H 7″
3
3
m
dd
-
d
m
d
2
″
42.9
H 7″
3
3
4
5
6
7
″
″
″
″
″
72.7
77.9
69.8
20.1
28.0
103.5
H1″, H2″a, H2″b, H 6″, H 7″
3 3
3.23
3.86
1.33
1.28
4.62
H 6″, H 7″
3 3
H4″, H 6″
3
H4″
H4″
H2″′, H3″′
s
d
1
″′
8.1
This article is protected by copyright. All rights reserved.

2
″′
dd
s
t
3.16
3.47
3.95
3.57
3.34
3.75
1.24
8.4;2.8
-
2.7
-
9.8;2.7
-
83.1
60.9
81.9
64.5
72.7
72.2
19.6
H3″′, HOCH 2″′
H2″′, H4″′, H5″′
3
OCH 2″′
3
3
″′
HOCH 3″′, H2″′, H4″’
3
OCH 3″′
s
H3″′, H4″′, H5″′
H3″′, H6″′
H1″′, H4″′, H6″′
-
3
4
5
6
″′
″′
″′
dd
m
d
6.5
This article is protected by copyright. All rights reserved.

Table 2. Multiplicities, proton and carbon chemical shifts and HMBC connectivities
1
3
1
(
C→ H) for tylosin B (desmycosine tylosin) in CDCl3
1
13
13
1
Position Multiplicity H (ppm)
C (ppm) HMBC connectivities ( C→ H)
1
-
m
d
-
173.9
H2a, H2b, H3
2.45
1.94
3.84
1.56
3.71
2.27
1.47
1.64
2.57
-
6.25
7.30
-
5.91
2.92
4.98
1.86
1.53
0.94
1.1
2
39.4
H 18
3
3
4
5
6
d
70.1
75.1
81.9
H2a
-
m
m
m
m
m
m
-
d
d
-
d
m
t
m
m
t
H 18, H1′,H2′
3
30-35*
H 21
3
7
30-35*
H 21
3
8
9
44.7
203.2
118.4*
148.1
134.9
142.3
45.0
H 21
3
H10, H11
-
1
0
1
2
3
4
5
1
1
1
1
1
H13, H 22
3
H10, H11, H 22
3
H11, H 22, H23a, H23b
3
H23a, H23b
75.1
H16b, H 17, H 22, H23a
3
3
1
6
25.5
H 17
3
1
1
7
8
9.7
9.0
H16a
-
d
m
m
s
d
s
dd
m
d
m
m
m
m
d
s
d
m
s
m
s
m
-
d
2.93
2.36
9.70
1.21
1.79
4.01
3.57
4.25
3.60
2.47
3.52
3.30
1.26
2.50
4.56
3.02
3.49
3.75
3.62
3.27
3.75
1.26
1
9
43.7
H 20
3
2
2
2
0
1
2
203.0
17.4
13.0
H19a, H 21
3
H7a, H7b
H11, H13, H16b
2
3
69.0
H2″′
1
2
3
4
5
6
′
′
′
′
′
′
104.0*
71.0
70.6
75.1
73.3
17.8
41.7
101.1
81.9
59.7
79.8
61.8
72.6
70.8
17.8
H5, H3′, H4′
H3′, H4′, H6 7′/8′
H4’, H5’, H6 7’/8’
H2′, H36′
-
H2′
7
′/8′
H3′, H 6′, H 7′/8′
3
6
1
2
″′
″′
H15, H23a, H23b, H2″′, H3″′
H3″′, H4″′
OCH -2″′
H2″′, H3″′, H4″′
H1″′
3
3
″′
OCH -3″′
H (OCH -2″′), H3″′
3
3
3
4
5
6
″′
″′
″′
H2″′, H5″′
H1″′, H3″′, H 6″′
3
H2″′
This article is protected by copyright. All rights reserved.

*
Signals are very broad due to chemical exchange. In the case of C1′ and C10, it was
possible to see the apex of the peak (values are presented in the table). Regarding C6 and C7,
however, peaks are merged and broad, so it is not possible to determine the position of the
apex.
This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

Conformational study of tylosin A in water and full assignments
1
13
of H and C spectra of tylosin A in D O and tylosin B in CDCl
2
3
1
2
1
1
Biljana Arsic , Juan A. Aguilar , Richard A. Bryce and Jill Barber
1
13
Full H and C NMR assignments of the macrolide antibiotic tylosin A in deuterated water
are reported for the first time. A combination of molecular modelling and ROESY NMR
experiments has been used to investigate the three-dimensional structure of this macrolide in
solution. Its antibacterial activity and of that the closely related tylosin B have been
determined in vitro
This article is protected by copyright. All rights reserved.