Synthesis of aminoglycoside-DNA conjugates

Article abstract of DOI:10.1016/S0960-894X(02)00157-9

Development of aminoglycoside-nucleic acid conjugates is presented. Synthesis of a DNA dimer covalently linked to kanamycin and neomycin isothiocyanates has been carried out. The development of such conjugates will help couple the sequence specificity of nucleic acids to the electrostatic/shape complementarity of aminoglycoside antibiotics in binding nucleic acid targets.

Full text of DOI:10.1016/S0960-894X(02)00157-9

Bioorganic & Medicinal Chemistry Letters 12 (2002) 1259–1262
Synthesis of Aminoglycoside–DNA Conjugates
I. Charles, Liang Xue and Dev P. Arya*
Laboratory of Medicinal Chemistry, Department of Chemistry, Clemson University, Clemson, SC 29634, USA
Received 21 December 2001; accepted 26 February 2002
Abstract—Development of aminoglycoside-nucleic acid conjugates is presented. Synthesis of a DNA dimer covalently linked to
kanamycin and neomycin isothiocyanates has been carried out. The development of such conjugates will help couple the sequence
specificity of nucleic acids to the electrostatic/shape complementarity of aminoglycoside antibiotics in binding nucleic acid targets.
# 2002 Elsevier Science Ltd. All rights reserved.
Aminoglycoside antibiotics are bactericidal agents that
are comprised of two or more amino sugars joined in
glycosidic linkage to a hexose nucleus.1 Though they
exhibit a narrow toxic/therapeutic ratio, their broad
antimicrobial spectrum, rapid bactericidal action, and
ability to act synergistically with other drugs make them
highly effective in the treatment of nosocomial (hospital-
acquired) infections.2 They are clinically useful in the
treatment of urinary tract infections,3 lower respiratory
infections, bacteremias, and other superinfections by
resistant organisms.2,4 Their greatest potential has been
in the combination drug regimens in the treatment of
infections that are difficult to cure with single agents
and for use in patients who are allergic to other classes
of drugs.5 The bactericidal action of aminoglycosides is
attributed to the irreversible inhibition of protein
synthesis following their binding to the 30S subunit of
the bacterial ribosome. RNA affinity and discrimination
by aminoglycosides is modulated by the interplay of
nonspecific electrostatic forces, which are critical for
affinity and a few specific interactions. The flexible and
polycationic nature of the aminoglycoside antibiotics
not only allows them to preferentially bind to a pro-
karyotic ribosomal RNA, but also allows binding to a
increased affinity, at the price of increased promiscuity
and inefficient cellular uptake.
Our recent finding that aminoglycosides can stabilize
DNA/RNA triplexes,14,15 hybrid duplexes,16 and that
neomycin can even induce hybrid triplex formation16
suggested that aminoglycoside-DNA conjugates could
be effective models for targeting nucleic acids (including
rRNA) sequence specifically (via a hybrid duplex or
triplex formation). RNAꢁ DNA hybrids are transiently
formed in many vital biological processes, including
DNA replication,17 telomere replication by telomer-
ase,18 and the replication of HIV (and other retroviruses
by reverse transcription).17 In this paper, we report the
synthesis of neomycin and kanamycin isothiocyanates
as stable reagents that can be coupled to a variety of
amines. Their use in the synthesis of a DNA thymidine
dimer conjugated to neomycin and kanamycin
antibiotics is also presented.
The following recognition elements were kept in mind
for the design of the conjugate: The amino groups on
rings I, II, and IV (neomycin) and rings I–III (kanamy-
cin) are necessary in stabilizing and in recognizing vari-
ous nucleic acid forms (aminoglycosides without any of
these amines do not stabilize rRNA, DNA/RNA tri-
plexes as efficiently).14ꢀ16,19 The conjugates based on
aminoglycosides must then retain these amines. The 500-
OH on ring III (neomycin) and 600-OH on ring III
(kanamycin) were thus chosen to provide the linkage to
the nucleic acids. Our retrosynthetic analysis of DNA
dimer covalently attached to aminoglycoside (kanamy-
cin and neomycin) is shown in Scheme 1. The scheme
was devised to allow its extension to solid-phase DNA
synthesis.
1
variety of unrelated RNAs, groupI introns, a ham-
merhead ribozyme,6 the RRE transcriptional activator
region from HIV, (which contains the binding site for
the Rev protein),7ꢀ9 the 50-untranslated region of thy-
midylate synthase mRNA,10 a variety of RNA aptamers
from in vitro selection,11,12 and human rRNAs.13 Thus,
aminoglycoside charge is a necessary evil: leading to
*Corresponding author. Tel.: +1-864-656-1106; fax: +1-864-656-
6613; e-mail: dparya@clemson.edu
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00157-9

1260
I. Charles et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1259–1262
The synthesis involves coupling of kanamycin iso-
thiocyanate 2 (Scheme 2) or neomycin isothiocyanate 11
(Scheme 5) with 50-amine of a DNA dimer 7 (Scheme 3).
solution however turns yellow after 24 h. The reactivity
of 2 and 11 with various aliphatic as well as aromatic
amines was first examined. The reaction with aliphatic
primary amines undergoes completion in less than 4 h.
Under similar conditions, even aromatic amines such as
aniline and p-methoxyaniline react with 2 and 11 to give
the corresponding thiourea derivatives in good yields.
The reaction times for the aromatic amines are 64 h
(aniline) and 48 h (p-methoxyaniline), respectively. The
isothiocyanates were then used for the synthesis of thy-
midine dimer–aminoglycoside conjugates.
The amino groups of kanamycin were selectively pro-
tected with a benzyloxycarbonyl (Cbz) group. Selective
protection of the primary hydroxyl group was achieved
using triisopropylbenzenesulfonyl chloride (TIBS-Cl)
considering the good leaving ability of the sulfonate
ester in the presence of a nucleophile such as amino-
thiolate as described in Scheme 2.20 Attempts to convert
the amino groupof modified kanamycin amine to iso-
thiocyanate 2 were unsuccessful using reported reagents
such as 1,10-thiocarbonyldiimidazole21 and CS2/HgO/
pyridine.21 The reaction of kanamycin amine with 1
equiv of 1,10-thiocarbonyldi-2-(1H)pyridone,22 gave the
expected isothiocyanate 2 as shown in Scheme 2. Neo-
mycin amine 10 was prepared from neomycin following
Tor’s reported literature procedure.20 Synthesis of neo-
mycin isothiocyanate 11 from neomycin amine 10 is
shown in Scheme 5. Isothiocyanates 2 and 11 are stable
in solution at room temperature. The isothiocyanates
were stable (as followed by TLC, IR and reactivity with
amines) for 7 days in anhydrous CH2Cl2. The colorless
Modeling studies showed that a linker size of 4–6 atoms
would allow neomycin/kanamycin to fold back and
stabilize a duplex/triplex groove, thus stabilizing a
DNA/RNA triplex by bridging the two pyrimidine
Scheme 3. Synthesis of DNA dimer. Reagents and conditions: (a) (i)
tetrachlorophthalimide, PPh3, DIAD and THF (97%); (ii) ethylene-
diamine and THF; (iii) MmTrCl, TEA, DMAP and pyridine (60% for
two steps); (iv) CNCH2CH2OP[N(iPr)2]2, bis(diisopropylammo-
nium)tetrazolide and CH2Cl2 (61%). (b) (i) DMTrCl, DMAP and
pyridine (90%); (ii) TBDMSCl, DMAP, Et3N and DMF (92%); (iii) I2
and MeOH (83%); (c) (i) 1H-tetrazole and CH3CN (84%); (ii) I2,
H2O/pyridine/THF (9.0:0.46:90.54 v/v/v %), (79%); (d) CCl3CO2H
and CH2Cl2 (88%).
Scheme 1. Retrosynthetic analysis of covalently linked aminoglyco-
side to DNA dimer.
Scheme 2. Synthesis of kanamycin isothiocyanate. Reagents and con-
ditions: (a) CbzCl, H2O and Na2CO3 (98%); (b) TIPSCl and pyridine
(56%); (c) Boc2O, H2, Pd/C and MeOH (76%); (d) HCl-
NH2CH2CH2SH, NaOEt and EtOH (58%); (e) 1,10-thiocarbonyldi-
2(1H)pyridone, DMAP and CH2Cl2 (55%).
Scheme 4. Coupling of DNA dimer and kanamycin isothiocyanate 2.
Reagents and conditions: (a) DMAP and pyridine, rt, 9 h (72%);
(b) 1,2-ethanedithiol, CF3CO2H and CH2Cl2 (48%).

I. Charles et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1259–1262
1261
strands together. A model of neomycin covalently
bound to a triplex pyrimidine strand is shown in the
supporting information. The synthesis of DNA dimer 7
from thymidine 3 is presented in Scheme 3. 50-Amino-50-
deoxythymidine was prepared from thymidine, with
regioselective introduction of tetrachlorophthalimide
standard phosphoramidite chemistry gave 4 in good
yields.24
30-TBDMS (t-butyldimethylsilyl) protected thymidine 5
was prepared from 50-O-DMTr protected thymidine
and TBDMS-Cl in the presence of Et3N and catalytic
amount of 4-N,N-dimethylaminopyridine (DMAP).25
Selective deprotection of di(p-methoxyphenyl) phe-
0
groupat the 5 position under mitsunobu condition,
followed by deprotection with ethylenediamine.23 Pro-
tection of 50-amine by p-methoxyphenyldiphenylmethyl
(MmTr) group followed by phosphitylation with
nylmethyl (DMTr) groupusing I
compound 5 in 83% yield.
in MeOH26 gave
2
Coupling of phosphoramidite 4 and 30-TBDMS pro-
tected thymidine 5 in the presence of 1H-tetrazole gave
DNA dimer 6 in 84% yield, which was oxidized with I2
in H2O/pyridine/THF (9.0:0.46:90.54 v/v/v %) to give 7.
The above conversion was supported by an upfield shift
in 31P NMR from 140 to ꢀ1.72 ppm. Deprotection of
MmTr groupwas achieved with CCl CO2H in CH2Cl2.
3
The final coupling of kanamycin isothiocyanate 2/neo-
mycin isothiocyanate 11 with 50-amino DNA dimer 7
was achieved with catalytic amount of DMAP and pyr-
idine as a solvent (Schemes 4 and 5). While the coupling
of kanamycin isothiocyanate to the DNA dimer pro-
ceeds efficiently and in good yields, neomycin iso-
thiocyanate requires a much longer time to undergo
completion, suggesting the importance of crowding/
linker size on the electrophilic isothiocyanate carbon.
Scheme 5. Coupling of DNA dimer and neomycin isothiocyanate 11.
Reagents and conditions: (a) 1,10-Thiocarbonyldi-2(1H)pyridone,
DMAP and CH2Cl2 (60%); (b) (i) compound 7, DMAP and pyridine,
rt, 9 h (43%); (ii) 1,2-ethanedithiol, 4 M HCl/Dioxane (79%).
1
The H NMR spectra of the conjugates showed Boc to
thymidine base methyl protons integration ratio of 9:1
(neomycin) and 6:1 (kanamycin). The 31P NMR spectra
of neomycin dimer 12 as well as kanamycin dimer 9 are
shown in Figure 1, confirming the presence of a single
product.
The mass spectrum of kanamycin dimer shows a m/z
peak at 1173.5 [M+2H++Na+] (Fig. 2). The MS of
neomycin dimer is available in the supporting information.
In conclusion, an efficient synthesis of aminoglycoside
isothiocyanates and their DNA dimer conjugate has
been achieved. The synthesis is being extended to the
development of neomycin/kanamycin–DNA conjugates
Figure 1. 31P NMR of kanamycin dimer 9 (top) and neomycin dimer
12 (bottom).
Figure 2. Mass spectra of kanamycin DNA dimer 9.

1262
I. Charles et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1259–1262
using solid-phase DNA chemistry and will be reported
in due course. These conjugates can be used for target-
ing genes of interest via triplex formation, and can also
be used to bring sequence specificity to aminoglycoside–
rRNA interactions. The potential of these conjugates in
targeting RNA remains an attractive option, because of
the highly conserved nature of the A-site sequence of
the 30S ribosomal subunit aminoglycoside binding site.
Additionally, such nucleic acid conjugates can allow
us to overcome traditional enzymatic/efflux mediated
methods of antibiotic resistance.
6. Stage, T. K.; Hertel, K. J.; Uhlenbeck, O. C. RNA 1995, 1,
95.
7. Wang, Y.; Hamasaki, K.; Rando, R. R. Biochemistry 1997,
36, 768.
8. Cho, J.; Rando, R. R. Biochemistry 1999, 38, 8548.
9. Park, W. K. C.; Auer, M.; Jaksche, H.; Wong, C.-H. J. Am.
Chem. Soc. 1996, 118, 10150.
10. Tok, J. B. H.; Cho, J.; Rando, R. R. Biochemistry 1999,
38, 199.
11. Wang, Y.; Killian, J.; Hamasaki, K.; Rando, R. R. Bio-
chemistry 1996, 35, 12338.
12. Hamasaki, K.; Killian, J.; Cho, J.; Rando, R. R. Bio-
chemistry 1998, 37, 656.
13. Hamasaki, K.; Rando, R. R. Biochemistry 1997, 36,
12323.
14. Arya, D. P.; Coffee, R. L., Jr. Bioorg. Med. Chem. Lett.
2000, 10, 1897.
Acknowledgements
Financial support for the above work was provided by
Clemson University. The support of NMR facility by
NSF (CHE-9700278) is gratefully acknowledged. We
thank Dr. Lisa Marzilli (Genetics Institute, MA, USA)
for helpwith MS analysis.
15. Arya, D. P.; Coffee, R. L., Jr.; Willis, B.; Abramovitch, A.
J. Am. Chem. Soc. 2001, 123, 5385.
16. Arya, D. P.; Coffee, R. L., Jr.; Charles, I. J. Am. Chem.
Soc. 2001, 123, 11093.
17. Mathews, C. K.; van Holde, K. E.; Ahern, K. G. Bio-
chemistry, 3rd ed; Benjamin/Cummings: San Fransisco, 2000.
18. McEachern, M. J.; Krauskopf, A.; Blackburn, E. H.
Annu. Rev. Genet. 2000, 34, 331.
19. Xue, L.; Charles, I.; Arya, D. P. Chem. Commun. 2002, 70.
20. Michael, K.; Wang, H.; Tor, Y. Bioorg. Med. Chem. 1999,
7, 1361.
Supporting information available: Experimental pro-
cedures and characterization data for the compounds
prepared.
21. Zehl, A.; Cech, D. Liebigs Ann. 1996, 595.
22. Kim, S.; Yi, K. Y. J. Org. Chem. 1986, 51, 2613.
23. Tetzlaff, C. N.; Schwope, I.; Bleczinski, C. F.; Steinberg,
J. A.; Richert, C. Tetrahedron Lett. 1998, 39, 4215.
24. Bannwarth, W. Helv. Chim. Acta 1988, 71, 1517.
25. Yang, O.; Wu, H.; Fraser-Smith, E.; Walker, K. Tetra-
hedron Lett. 1992, 33, 37.
References and Notes
1. Chow, C. S.; Bogdan, F. M. Chem. Rev. 1997, 97, 1489.
2. Mohan, C. In Calbiochem Catalog and Technical Resource;
Calbiochem: San Diego, 2000; p252.
3. Santucci, R.; Krieger, J. J. Urol. 2000, 163, 1076.
4. Forge, A.; Schacht, J. Audiol. Neurootol 2000, 5, 3.
5. Gerding, D. Infect. Control Hosp. Epidemiol. 2000, 21, S12.
26. Ramasamy, K.; Bandaru, R.; Averett, D. Synth. Commun.
1999, 29, 2881.