Conjugation of a Ru(II) arene complex to neomycin or to guanidinoneomycin leads to compounds with differential cytotoxicities and accumulation between cancer and normal cells

Article abstract of DOI:10.1021/mp300723b

A straightforward methodology for the synthesis of conjugates between a cytotoxic organometallic ruthenium(II) complex and amino- and guanidinoglycosides, as potential RNA-targeted anticancer compounds, is described. Under microwave irradiation, the imidazole ligand incorporated on the aminoglycoside moiety (neamine or neomycin) was found to replace one triphenylphosphine ligand from the ruthenium precursor [(η⁶-p- cym)RuCl(PPh₃)₂]⁺, allowing the assembly of the target conjugates. The guanidinylated analogue was easily prepared from the neomycin-ruthenium conjugate by reaction with N,N′-di-Boc-N″- triflylguanidine, a powerful guanidinylating reagent that was compatible with the integrity of the metal complex. All conjugates were purified by semipreparative high-performance liquid chromatography (HPLC) and characterized by electrospray ionization (ESI) and matrix-assisted laser desorption-ionization time-of-flight (MALDI-TOF) mass spectrometry (MS) and NMR spectroscopy. The cytotoxicity of the compounds was tested in MCF-7 (breast) and DU-145 (prostate) human cancer cells, as well as in the normal HEK293 (Human Embryonic Kidney) cell line, revealing a dependence on the nature of the glycoside moiety and the type of cell (cancer or healthy). Indeed, the neomycin-ruthenium conjugate (2) displayed moderate antiproliferative activity in both cancer cell lines (IC ₅₀ ≈ 80 μM), whereas the neamine conjugate (4) was inactive (IC₅₀ ≈ 200 μM). However, the guanidinylated analogue of the neomycin-ruthenium conjugate (3) required much lower concentrations than the parent conjugate for equal effect (IC₅₀ = 7.17 μM in DU-145 and IC₅₀ = 11.33 μM in MCF-7). Although the same ranking in antiproliferative activity was found in the nontumorigenic cell line (3 a‰ 2 > 4), IC₅₀ values indicate that aminoglycoside-containing conjugates are about 2-fold more cytotoxic in normal cells (e.g., IC₅₀ = 49.4 μM for 2) than in cancer cells, whereas an opposite tendency was found with the guanidinylated conjugate, since its cytotoxicity in the normal cell line (IC₅₀ = 12.75 μM for 3) was similar or even lower than that found in MCF-7 and DU-145 cancer cell lines, respectively. Cell uptake studies performed by ICP-MS with conjugates 2 and 3 revealed that guanidinylation of the neomycin moiety had a positive effect on accumulation (about 3-fold higher in DU-145 and 4-fold higher in HEK293), which correlates well with the higher antiproliferative activity of 3. Interestingly, despite the slightly higher accumulation in the normal cell than in the cancer cell line (about 1.4-fold), guanidinoneomycin-ruthenium conjugate (3) was more cytotoxic to cancer cells (about 1.8-fold), whereas the opposite tendency applied for neomycin-ruthenium conjugate (2). Such differences in cytotoxic activity and cellular accumulation between cancer and normal cells open the way to the creation of more selective, less toxic anticancer metallodrugs by conjugating cytotoxic metal-based complexes such as ruthenium(II) arene derivatives to guanidinoglycosides.

Full text of DOI:10.1021/mp300723b

Subscriber access provided by HKU Libraries
Article
Conjugation of a Ru(II) arene complex to neomycin or to
guanidinoneomycin leads to compounds with differential
cytotoxicities and accumulation between cancer and normal cells
Ariadna Grau-Campistany, Anna Massaguer, Dolors Carrion-Salip, Barragan
Flavia, Gerard Artigas, Paula Lopez-Senin, Virtudes Moreno, and Vicente Marchán
Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/mp300723b • Publication Date (Web): 19 Mar 2013
Downloaded from http://pubs.acs.org on March 26, 2013
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Molecular Pharmaceutics is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 34
Molecular Pharmaceutics
1
2
3
4
5
6
7
8
Conjugation of a Ru(II) arene complex to neomycin or to
guanidinoneomycin leads to compounds with differential cytotoxicities
and accumulation between cancer and normal cells
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ariadna GrauꢀCampistany,a,& Anna Massaguer,b,& Dolors CarrionꢀSalip,b Flavia
Barragán,a,c Gerard Artigas,a Paula LópezꢀSenín,a Virtudes Morenoc and Vicente
Marchán*,a
aDepartament de Química Orgànica and IBUB, Universitat de Barcelona, Barcelona,
Eꢀ08028, Spain. Eꢀmail: vmarchan@ub.edu.
bDepartament de Biologia, Universitat de Girona, Girona, Eꢀ17071, Spain.
cDepartament de Química Inorgànica, Universitat de Barcelona, Barcelona, Eꢀ08028,
Spain.
&These two authors contributed equally to this work.
1
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 2 of 34
1
2
ABSTRACT
3
4
A straightforward methodology for the synthesis of conjugates between a cytotoxic
organometallic ruthenium(II) complex and aminoꢀ and guanidinoglycosides, as potential
RNAꢀtargeted anticancer compounds, is described. Under microwave irradiation, the
imidazole ligand incorporated on the aminoglycoside moiety (neamine or neomycin)
was found to replace one triphenylphosphine ligand from the ruthenium precursor [(η6ꢀ
pꢀcym)RuCl(PPh3)2]+, allowing the assembly of the target conjugates. The
guanidinylated analogue was easily prepared from the neomycinꢀruthenium conjugate
by reaction with N,N’ꢀdiꢀBocꢀN”ꢀtriflylguanidine, a powerful guanidinylating reagent
that was compatible with the integrity of the metal complex. All conjugates were
purified by semiꢀpreparative HPLC and characterized by ESI and MALDIꢀTOF MS and
NMR spectroscopy. The cytotoxicity of the compounds was tested in MCFꢀ7 (breast)
and DUꢀ145 (prostate) human cancer cells, as well as in the normal HEK293 (Human
Embryonic Kidney) cell line, revealing a dependence on the nature of the glycoside
moiety and the type of cell (cancer or healthy). Indeed, neomycinꢀruthenium conjugate
(2) displayed moderate antiꢀproliferative activity in both cancer cell lines (IC50 ≈ 80
ꢀM), whereas that of the neamine conjugate (4) was inactive (IC50 ≈ 200 ꢀM).
However, the guanidinylated analogue of the neomycinꢀruthenium conjugate (3)
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
required much lower concentrations than the parent conjugate for equal effect (IC50
=
7.17 ꢀM in DUꢀ145 and IC50 = 11.33 ꢀM in MCFꢀ7). Although the same ranking in
antiꢀproliferative activity was found in the nonꢀtumorigenic cell line (3 >> 2 > 4), IC50
values indicate that aminoglycosideꢀcontaining conjugates are about 2ꢀfold more
cytotoxic in normal cells (e.g. IC50 = 49.4 ꢀM for 2) than in cancer cells, whereas an
opposite tendency was found with the guanidinylated conjugate, since its cytotoxicity in
the normal cell line (IC50 = 12.75 ꢀM for 3) was similar or even lower than that found
in MCFꢀ7 and DUꢀ145 cancer cell lines, respectively. Cell uptake studies performed by
ICPꢀMS with conjugates 2 and 3 revealed that guanidinylation of the neomycin moiety
had a positive effect on accumulation (about 3ꢀfold higher in DUꢀ145 and 4ꢀfold higher
in HEK293), which correlates well with the higher antiꢀproliferative activity of 3.
Interestingly, despite the slightly higher accumulation in the normal cell than in the
cancer cell line (about 1.4ꢀfold), guanidinoneomycinꢀruthenium conjugate (3) was more
cytotoxic to cancer cells (about 1.8ꢀfold), whereas the opposite tendency applied for
neomycinꢀruthenium conjugate (2). Such differences in cytotoxic activity and cellular
2
ACS Paragon Plus Environment

Page 3 of 34
Molecular Pharmaceutics
1
2
3
4
5
6
7
accumulation between cancer and normal cells open the way to the creation of more
selective, less toxic anticancer metallodrugs by conjugating cytotoxic metalꢀbased
complexes such as ruthenium(II) arene derivatives to guanidinoglycosides.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
KEYWORDS: ruthenium, cytotoxicity, cellular accumulation, RNA ligands,
neomycin, guanidinoneomycin
3
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 4 of 34
1
2
INTRODUCTION
3
4
Organometallic complexes have emerged in recent years as promising anticancer
metallodrugs, which could well overcome cisplatin and platinumꢀrelated analogues’
disadvantages, mainly toxicity and drug resistance in cancer cells.1ꢀ5 Among them,
ruthenium(II) complexes bearing a πꢀbonded arene ligand are particularly interesting
since they have shown promising anticancer activities,6ꢀ8 even in cells that had become
resistant to cisplatin, such as Sadler’s compounds containing N,Nꢀchelating ligands.1,9,10
In addition, some of the Dyson’s RAPTA compounds containing pta ligand have shown
antimetastatic activity.11,12 Their soꢀcalled “pianoꢀstool” geometry includes an arene
unit that stabilizes the ruthenium +2 oxidation state and confers hydrophobicity on the
global metal complex, as well as monoꢀ or bidentate ligands, including one or two
leaving groups. In most cases, the release of labile chlorido ligands is triggered inside
the cell nucleus by the low chloride concentration (4 mM vs ∼100 mM in extracellular
fluids), allowing for the generation of the activated aqua species that possess the
capacity to react with the biological target.13,14
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ruthenium offers several advantages over platinum compounds, including reduced
toxicity and the possibility of controlling the shape and the chemical and
pharmacological properties of the complex by the adequate selection of the arene and
the ligands at the “legs” of the “pianoꢀstool” structure.15,16 Moreover, modification of
the nonꢀleaving ligands allows the metal complex to be anchored to a “tumorꢀtargeting
device” such as receptorꢀbinding peptides, folic acid or estrogens.17ꢀ24 This targeted
strategy has a tremendous potential in the development of more efficient, less toxic,
selective metallodrugs in chemotherapy because receptors for these carrier molecules
are overꢀexpressed in the membrane of tumoral cells.25ꢀ28
Another strategy to improve efficiency of a metalꢀbased drug is to increase its affinity
with its ultimate biological target. This strategy has been explored mainly with platinum
complexes through the covalent attachment to compounds with high affinity for DNA
(minor groove binders or intercalators) or to synthetic oligonucleotides or Peptide
Nucleic Acids complementary to specific DNA sequences.29 The cytotoxicity of
ruthenium(II) arene complexes, like that of platinum compounds, has been attributed
mainly to the binding of their aquation products to DNA.7,8 However, these complexes’
interaction with other potential cellular competitors such as the tripeptide glutathione
cannot be ruled out, since they are present in large intracellular concentrations and are
4
ACS Paragon Plus Environment

Page 5 of 34
Molecular Pharmaceutics
1
2
3
4
5
6
7
8
9
responsible for the detoxification of heavier transition metals.30 In fact, recent studies
have revealed that DNA is not always the primary target for some ruthenium anticancer
compounds and that they actually bind more strongly to proteins or enzymes than to
DNA.4,31 Hence, it seems important not only to develop efficient targeting strategies to
deliver metallodrugs selectively into cancer cells, but also to direct them towards a
particular biological target.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Although, traditionally, metalꢀbased drugs have been designed to target DNA, in
therapeutic terms, RNA offers several advantages over DNA as a drug target, since it is
involved in many cellular processes, from the regulation of gene expression to protein
synthesis.32ꢀ35 In addition, like proteins, RNA adopts complex threeꢀdimensional
structures that can be exploited to design specific small molecules to modulate its
functions.36ꢀ38 In recent years, microRNAs (miRNAs) have also emerged as new
therapeutic targets for cancer therapy since the abnormal expression of these nonꢀcoding
small RNAs is associated with the pathogenesis of human cancer.39,40 Like other RNAs,
miRNA precursors adopt secondary structures that can be targeted with small
molecules, to interfere with miRNA maturation and, for instance, to manipulate miRNA
levels.41,42
With the aim of developing new metalꢀbased anticancer drugs that could act at the level
of RNA, we focused on the conjugation of ruthenium(II) arene complexes with small
molecules that had the capacity to recognize selectively RNA over DNA or proteins.
Aminoglycoside antibiotics are possibly the most commonly studied RNA
ligands,33,36,43,44 since they have a relatively high affinity with RNA structures and are
able to discriminate Aꢀtype from Bꢀtype duplexes.43,44 Apart from their selectivity for
RNA, aminoglycosides possess several amino functions that are mostly protonated
under physiological conditions, which would confer some drugꢀlike properties on the
metal complex, such as aqueous solubility. However, aminoglycoside antibiotics are
known to have a more inefficient uptake by eukaryotic than by prokaryotic cells.45 This
problem has been solved by replacing their amine functions with guanidinium groups to
generate guanidinoglycosides,46 since these derivatives have higher efficient uptake by
eukaryotic cells than their aminoglycoside precursors.45 In fact, guanidinylated
neomycin is known to transport bioactive, high molecular weight cargo into the interior
of cells.47ꢀ49 Like naturally occurring aminoglycosides, guanidinoglycosides bind RNA
5
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 6 of 34
1
2
3
4
5
6
7
8
9
over DNA preferentially, but they have shown higher binding affinity and selectivity
against several RNA targets.46,50 This strategy was first explored by Tor and
collaborators by conjugating a platinum(II) complex with neomycin B and
guanidinoneomycin B.51 The fact that these compounds were able to selectively crossꢀ
link an RNA structure derived from HIV Rev Response Element demonstrated that
therapeutically relevant RNA structures could be targeted with metal complexes by
using the appropriate glycoside.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Herein, we describe for the first time the synthesis, characterization, cellular uptake and
antiproliferative activity of conjugates between a ruthenium(II) arene complex, [(η6ꢀpꢀ
cym)RuCl(ImꢀBzCOOMe)(PPh3)]+ (1),24 where ImꢀBzCOOMe refers to the methyl
ester of the 4ꢀ(1Hꢀimidazolꢀ1ꢀyl)benzoic acid, and neomycin B (2), guanidinoneomycin
B (3) and neamine (4) (Scheme 1). In all cases, the metal complex was attached to the
aminoꢀ or the guanidinoglycoside through the phenyl ring linked to the imidazole
ligand. The triphenylphosphineꢀcontaining ruthenium(II) complex 1 was recently
described as a promising anticancer lead compound, since its antiꢀproliferative activity
is comparable to that of cisplatin in two tumoral cell lines, MCFꢀ7 and DUꢀ145.24
Interestingly, a direct correlation was found between the cytotoxicity of its octreotide
conjugate and the level of expression of the receptors for this peptide in the membrane
of tumoral cells.24
Scheme 1. Structure of the amino(guanidino)glycosides and their ruthenated conjugates.
6
ACS Paragon Plus Environment

Page 7 of 34
Molecular Pharmaceutics
1
2
MATERIALS AND METHODS
3
4
General Procedures. Unless otherwise stated, common chemicals and solvents (HPLC
grade or reagent grade quality) were purchased from commercial sources and used
without further purification. Peptide grade DMF was from Scharlau. MilliꢀQ water was
directly obtained from a MilliꢀQ system equipped with a 5000ꢀDa ultrafiltration
cartridge. Bocꢀprotected amino derivative of neomycin (5)52 and tritylꢀ and pꢀ
methoxybenzylꢀprotected amino derivative of neamine (8)53 were prepared as reported
elsewhere. Metal complexes [(η6ꢀpꢀcym)RuCl(PPh3)2][PF6]54 and [(η6ꢀpꢀcym)RuCl(Imꢀ
BzCOOMe)(PPh3)][PF6] (1)24 were synthesized according to procedures described
elsewhere.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NMR spectra were recorded at 25ºC on a Varian Mercury 400 MHz, Bruker 500 MHz
or 600 MHz spectrometer, using deuterated solvents. Tetramethylsilane (TMS) was
1
used as an internal reference (δ 0 ppm) for H spectra recorded in CDCl3 and the
residual signal of the solvent (δ 77.16 ppm) for 13C spectra. For CD3OD, the residual
signal of the solvent was used as a reference.
HighꢀResolution MALDIꢀTOF mass spectra were recorded on a 4800 Plus MALDIꢀ
TOF/TOF spectrometer (Applied Biosystems) in the positive mode, using 2,4ꢀ
dihydroxybenzoic acid as a matrix. ESI mass spectra (ESIꢀMS) were recorded on a
Micromass ZQ instrument with single quadrupole detector coupled to an HPLC. Highꢀ
Resolution Electrospray mass spectra (HR ESI MS) were obtained on an Agilent 1100
LC/MSꢀTOF instrument.
Analytical reversedꢀphase HPLC analyses were carried out on a GraceSmart RP C18
column (250x4 mm, 5 ꢀm, flow rate: 1 mL/min), using linear gradients of 0.045% TFA
in H2O (solvent A) and 0.036% TFA in ACN (solvent B). Largeꢀscale purification was
carried out in a Jupiter Proteo semipreparative column (250 x 10 mm, 10 ꢀm, flow rate:
3ꢀ4 mL/min), using linear gradients of 0.1% TFA in H2O (solvent A) and 0.1% TFA in
ACN (solvent B). After several runs, pure fractions were combined and lyophilized.
A [Vydac C18]ꢀfilled glass column (22x2 cm, 15ꢀ20 mm, 300 Å) was used for mediumꢀ
pressure liquid chromatography (MPLC), using aqueous and ACN solutions containing
0.05% TFA (flow rate: 3 mL/min). Elution was carried out by connecting a piston pump
to the mixing chamber of a gradientꢀforming device and to the top of the glass column.
The mixing chamber of the gradientꢀforming device was the flask containing solvent A,
which was connected through a stopcock to the flask containing solvent B. The bottom
7
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 8 of 34
1
2
3
4
5
6
7
8
9
of the preparative column was connected to an automatic fraction collector through a
UV/Vis detector, which was also connected to a chart recorder using the appropriate
ports. The column was equilibrated with 200 mL of solvent A, and 600 mL of each
mobile phase was introduced into the appropriate compartments of the gradientꢀforming
device.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis and characterization of conjugates
Boc-protected imidazole-neomycin derivative 6: 4ꢀ(1HꢀImidazolꢀ1ꢀyl)benzoic acid
(233 mg, 1.24 mmol) and PyBOP (643 mg, 1.24 mmol) were suspended in anhydrous
DMF (8 mL) under Ar. Upon addition of DIPEA (840 ꢀL, 4.94 mmol), the suspension
rapidly dissolved (within 5 min with stirring at rt) and acquired an intense yellow color.
This solution was added over the Bocꢀprotected amino derivative of neomycin 5 (300
mg, 0.25 mmol) dissolved in anhydrous DMF (10 mL). Then, the mixture was stirred at
rt for 2 h under Ar. After evaporation in vacuo the residue was partitioned between
AcOEt (100 mL) and H2O (100 mL). The organic phase was taken up and washed with
H2O (2 x 100 mL) and 10 % aqueous NaHCO3 (2 x 100 mL), dried over anhydrous
MgSO4 and filtered, and the solvent was removed in vacuo. Purification by silica gel
flashꢀcolumn chromatography (gradient: 0ꢀ15 % of MeOH in DCM) afforded the
desired product as a white solid (120 mg, 35%). Rf (10% MeOH in DCM): 0.44; HR
ESI MS, positive mode: m/z 1384.6985 (calcd mass for C63H102N9O25 [M+H]+:
1384.6988); HR MALDIꢀTOF MS, positive mode: m/z 1406.91 (calcd mass for
1
C63H101N9NaO25 [M+Na]+: 1406.6806); H NMR (500 MHz, CD3OD) δ (ppm): 8.23
(2H, d, J=8.5 Hz), 7.93 (1H, s br), 7.79 (3H, d, J=8.5 Hz), 7.51 (1H, s br), 5.40 (1H, s
br), 5.17 (1H, s), 4.94 (1H, s), 4.32 (1H, m), 4.24 (1H, m), 4.06 (1H, m), 3.96 (1H, m),
3.92 (2H, m), 3.79 (1H, m), 3.75 (2H, m), 3.70 (1H, m), 3.63 (2H, m), 3.55 (3H, m),
3.50 (2H, m), 3.32 (4H, m), 3.16 (1H, m), 1.96 (1H, m), 1.41 (55H, Boc+1H, m).
Ruthenium-neomycin conjugate 2:
6
(100 mg, 0.07 mmol) and [(η6ꢀpꢀ
cym)RuCl(PPh3)2][PF6] (68 mg, 0.07 mmol) were introduced into a microwave reactor.
After addition of a 0.4 M LiCl solution in anhydrous DMF (217 ꢀL, 1.2 mol eq LiCl)
and triethylamine (103 ꢀL, 10 mol eq), the reaction mixture was diluted with
DCM/MeOH 9:1 until a final volume of 10 mL was reached. Microwave irradiation (40
W) was performed at 60ºC under Ar atmosphere for 40 min with continuous magnetic
stirring. The reaction mixture was diluted with AcOEt (25 mL) and the organic phase
8
ACS Paragon Plus Environment

Page 9 of 34
Molecular Pharmaceutics
1
2
3
4
5
was washed with H2O (3 x 25 mL), dried over anhydrous MgSO4 and filtered, and the
solvent was removed in vacuo.
6
7
8
9
The crude was treated at 0ºC with a 30% TFA solution in DCM (10 mL) and allowed to
come to rt. After 25 min at rt, the solvent was evaporated and the residue was reꢀ
dissolved in a 1:1 H2O/MeOH mixture, lyophilized and purified by MPLC. Pure
fractions by analytical HPLC were combined and lyophilized, affording the desired
product as the trifluoroacetateꢀsalt of a yellow solid (20.5 mg, 54%).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Characterization: Rt= 14.6 min (analytical gradient: 0 to 100 % in 30 min); HR ESI MS,
positive mode: m/z 1316.4510 (calcd mass for C61H82ClN9O13PRu [M]+: 1316.4502),
m/z 1054.3580 (calcd mass for C43H67ClN9O13Ru [MꢀPPh3]+: 1054.3590), m/z
1018.3827 (calcd mass for C43H66N9O13Ru [MꢀClꢀPPh3ꢀH]+: 1018.3824); HR MALDIꢀ
TOF MS, positive mode: m/z 1280.4735 (calcd mass for C61H81N9O13PRu [MꢀClꢀH]+:
1280.47). 1H NMR (500 MHz, CD3OD) δ (ppm): 8.28 (1H, Im, s), 8.05 (2H, 2Ar Ph, d,
J=7.5 Hz), 7.53 (2H, Im, d, J=7.0 Hz), 7.48ꢀ7.41 (17H, 2Ar Ph +15Ar PPh3, d, m), 6.12
(1H, Ar pꢀcym, d, J=6.0 Hz), 5.94 (1H, Ar pꢀcym, t, J=5.5 Hz), 5.87 (1H, H1´´´, d, J=4.0
Hz), 5.76 (1H, Ar pꢀcym, d, J=6.0 Hz), 5.47 (1H, H1´´, d, J=5.5 Hz), 5.36 (1H, H1´, s),
5.17 (1H, Ar pꢀcym, d, J=6.0 Hz), 4.57 (1H, H3´´, t, J=4.5 Hz), 4.37 (1H, H4´´, m), 4.31
(1H, H5´, m), 4.21 (1H, H4, m), 4.15 (1H, H3´, m), 4.08ꢀ3.97 (5H,
H2´´+H3´´´+H4´´´+H5´´´+1H5´´, m), 3.91 (1H, H5, m), 3.69ꢀ3.62 (4H,
H2´´´+1H5´´+H4´+H6, m), 3.51 (1H, H3, m), 3.44 (1H, H2´, m), 3.40ꢀ3.35 (2H,
i
1H6´+1H6´´´, m), 3.29ꢀ3.20 (3H, 1H6´+1H6´´´+H1, m), 2.52 (1H, Pr, m), 2.45 (1H,
i
1H2, m), 2.08 (1H, 1H2, m), 1.77 (3H, CH3, m), 1.20 (3H, CH3 Pr, d, J=6.5 Hz), 1.17
i
(3H, CH3 Pr, d, J=6.5 Hz).
Ruthenium-guanidinoneomycin conjugate 3: The trifluoroacetate salt of conjugate 2
(2 mg, 0.95 ꢀmol) was dissolved in methanol (1 mL) and reacted with N,N’ꢀdiꢀBocꢀN”ꢀ
triflylguanidine (78 mg, 0.20 mmol) and triethylamine (140 ꢁL, 1 mmol) for 20 days at
rt under an Ar atmosphere. After evaporation in vacuo, the residue was treated with a
1:1 TFA/DCM mixture and stirred for 4 h at room temperature. The reaction mixture
was concentrated under vacuum and, after several coꢀevaporations from toluene, the
crude was dissolved in 0.1% trifluoroacetic acid in water and lyophilized. Purification
by analytical HPLC afforded the TFA salt of conjugate 3 as a yellow solid (0.77 mg,
34%).
Characterization: Rt= 17.0 min (analytical gradient: 0 to 100% in 30 min); HR ESI MS,
positive mode: m/z 784.7942 (calcd mass for C67H95ClN21O13PRu [M+H]2+: 784.7946).
9
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 10 of 34
1
2
3
4
5
6
7
8
9
1H NMR (600 MHz, D2O) δ (ppm): 7.93 (1H, s), 7.87 (1H, s), 7.78ꢀ7.75 (2H, m), 7.48
(1H, d, J=8.4 Hz), 7.33ꢀ7.23 (17H, m), 5.97 (1H, Ar pꢀcym, d, J=6.0 Hz), 5.68 (1H,
H1´´´, m), 5.67 (1H, Ar pꢀcym, m), 5.64 (1H, Ar pꢀcym, d, J=6.0 Hz), 5.07 (1H, H1´´,
m), 5.04 (1H, Ar pꢀcym, d, J=6.0 Hz), 4.91 (1H, H1´, m), 4.30ꢀ4.24 (2H, m), 4.01ꢀ3.90
(3H, m), 3.70ꢀ3.59 (3H, m), 3.54ꢀ3.30 (10H, m), 3.26ꢀ3.16 (2H, m), 3.05ꢀ2.98 (2H, m),
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
i
2.30 (1H, Pr, sp, J=6.5 Hz), 1.63 (3H, CH3, s), 1.50 (1H, 1H2, q, J=12 Hz), 1.11 (1H,
i
i
1H2, t, J=7.2 Hz), 0.98 (3H, CH3 Pr, d, J=6.5 Hz), 0.93 (3H, CH3 Pr, d, J=6.5 Hz).
(Trt)4(PMB)3-protected imidazole-neamine derivative 9: 4ꢀ(1HꢀImidazolꢀ1ꢀyl)benzoic
acid (54 mg, 0.29 mmol) and PyBOP (149 mg, 0.29 mmol) were suspended in
anhydrous DMF (1 mL) under Ar. On addition of DIPEA (194 ꢀL, 1.14 mmol), the
suspension rapidly dissolved (within 5 min with stirring at rt) and acquired an intense
yellow color. This solution was added over the (Trt)4(PMB)3ꢀprotected amino derivative
of neamine 8 (100 mg, 0.057 mmol) dissolved in anhydrous DMF (3 mL). Then, the
mixture was stirred at rt for 2 h under Ar. After evaporation in vacuo the residue was
partitioned between AcOEt (10 mL) and H2O (10 mL). The organic phase was taken up
and washed with H2O (2 x 10 mL), brine (1 x 10 mL) and 10% aqueous NaHCO3 (1 x
10 mL), dried over anhydrous MgSO4 and filtered, and the solvent was removed in
vacuo. Purification by silica gel flashꢀcolumn chromatography (gradient: 0ꢀ20% of
MeOH in DCM) afforded a white solid (33 mg) that corresponds to the desired product.
However, 9 was slightly contaminated with tris(pyrrolidinꢀ1ꢀyl)phosphine oxide
1
according to the H NMR spectrum (overall yield of 9: 39.3%). Rf (10% MeOH in
DCM): 0.45; HR ESI MS, positive mode: m/z 1920.9595 (calcd mass for C128H126N7O10
[M+H]+: 1920.9566), m/z 1678.8498 (calcd mass for C109H112N7O10 [MꢀTrt+H]+:
1678.8470), m/z 1436.7402 (calcd mass for C90H98N7O10 [Mꢀ2Trt+H]+: 1436.7375); H
1
NMR (400 MHz, CDCl3) δ (ppm): 7.91 (1H, m), 7.85 (2H, m), 7.68 (6H, m), 7.45 (1H,
m), 7.43 (1H, m), 7.38 (9H, m), 7.30 (17H, m), 7.25 (1H, m), 7.23 (2H, m), 7.15 (6H,
m), 7.08 (5H, m), 7.00 (12H, m), 6.90 (8H, m), 6.64 (8H, m), 5.30 (1H, s), 4.51 (4H,
m), 4.23 (1H, d, J = 10.4 Hz), 4.08 (1H, m), 3.78 (3H, s), 3.76 (3H, s), 3.73 (1H, m),
3.69 (3H, s), 3.55 (2H, m), 3.35 (3H, m), 3.06 (1H, s), 2.82 (3H, m), 2.65 (1H, m), 2.55
(1H, m), 2.44 (2H, m), 2.21 (2H, m), 1.48 (5H, m), 1.26 (9H, m). **signals in the
spectrum from tris(pyrrolidinꢀ1ꢀyl)phosphine oxide: 3.15 ppm (10H, m) and 1.81 ppm
(10H, m).
Ruthenium-neamine conjugate 4:
9
(33 mg, 0.017 mmol) and [(η6ꢀpꢀ
cym)RuCl(PPh3)2][PF6] (24 mg, 0.026 mmol) were dissolved in a 9:1 mixture of
10
ACS Paragon Plus Environment

Page 11 of 34
Molecular Pharmaceutics
1
2
3
4
5
6
7
8
9
DCM/EtOH (1 mL) in a microwave reactor. After addition of a 0.4 M LiCl solution in
anhydrous DMF (78 ꢀL, 1.8 mol eq LiCl) and triethylamine (25 ꢀL, 10 mol eq), the
reaction mixture was irradiated (40 W) at 60ºC under Ar atmosphere for 1 h with
continuous magnetic stirring. The reaction mixture was diluted with AcOEt (10 mL)
and washed with H2O (3 x 10 mL) and brine (3 x 10 mL), dried over anhydrous MgSO4
and filtered, and evaporated in vacuo.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The crude was dissolved in a 1:1 mixture of TFA/anisole (25 mL) under argon. After
stirring at rt for 15 h, MeOH (50 mL) was added and evaporated in vacuo. The resulting
oil was partitioned between DCM (50 mL) and MilliꢀQ water (50 mL) and the aqueous
phase was taken up and washed with DCM (2 x 50 mL) and lyophilized. Purification by
semiꢀpreparative HPLC (gradient 0 to 100% in 30 min) afforded the TFA salt of 4 as a
yellow solid (13.4 mg, 46%).
Characterization: Rt= 17.5 min (analytical gradient: 0 to 100% in 30 min); HR ESI MS,
positive mode: m/z 1124.4146 (calcd mass for C56H74ClN7O7PRu [M]+: 1124.4119); HR
MALDIꢀTOF MS, positive mode: m/z 1090.5 (calcd mass for C56H73N7O7PRu [MꢀClꢀ
1
H]+: 1088.43). H NMR (500 MHz, CD3OD) δ (ppm): 8.26 (1H, Ar Im, m), 7.98 (2H,
2Ar Ph, d, J=9.0 Hz), 7.53 (2H, Ar Im, dt, J=6.0 Hz, J’=1.5 Hz), 7.48ꢀ7.39 (17H, 2Ar
Ph +15H Ar PPh3, m), 6.10 (1H, Ar pꢀcym, dd, J=6.0 Hz, J’=1.0 Hz), 5.94 (1H, Ar pꢀ
cym, dd, J=6.0 Hz, J’=1.5 Hz), 5.84 (1H, H1´, d, J=3.5 Hz), 5.75 (1H, Ar pꢀcym, dd,
J=6.0 Hz, J’=1.0 Hz), 5.17 (1H, Ar pꢀcym, dt, J=6.0 Hz, J’=1.5 Hz), 4.14 (1H, H4, m),
4.07 (3H, H5´+H3´+1H CH2ꢀNH, m), 3.76 (1H, 1H CH2ꢀNH, m), 3.60ꢀ3.54 (2H,
H5+H6, m), 3.49 (1H, H3, m), 3.42ꢀ3.35 (5H, 1H6´+H4´+H2´+CH2ꢀO, m), 3.26ꢀ3.19
(2H, 1H6´+H1, m), 2.50 (1H, iPr, qt, J=7 Hz), 2.39 (1H, 1H2, dt, J=12 Hz, J’=4.0 Hz),
1.99 (1H, 1H2, q, J=12 Hz), 1.76 (3H, CH3, s), 1.66 (4H, NHꢀCH2ꢀCH2ꢀCH2ꢀCH2ꢀCH2ꢀ,
i
m), 1.43 (4H, NHꢀCH2ꢀCH2ꢀCH2ꢀCH2ꢀ, m), 1.19 (3H, CH3 Pr, d, J=7 Hz), 1.16 (3H,
i
CH3 Pr, d, J=7 Hz).
Cytotoxicity assays in cancer cell lines
The cytotoxicity of the conjugates and of the control compounds was determined by the
MTT assay in the breast cancer MCFꢀ7 cell line and the prostate carcinoma DUꢀ145 cell
line, as well as in human embryonic kidney 293 (HEK293) nonꢀtumoral cells. Aliquots
of 4,000 DUꢀ145, 3,500 MCFꢀ7 and 6,000 HEK293 cells were seeded onto flatꢀ
bottomed 96ꢀwell plates. 24 h later, the cells were treated for 72 h with the compounds
at concentrations ranging from 0 ꢁM to 250 ꢁM. After removal of the treatment, the
11
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 12 of 34
1
2
3
4
5
6
7
8
9
cells were washed with PBS and incubated for 3 additional hours with 100 ꢀL of fresh
culture medium together with 10 ꢀL of MTT (SigmaꢀAldrich). The medium was
discarded and DMSO (SigmaꢀAldrich) was added to each well to dissolve the purple
formazan crystals. Plates were shaken at room temperature for 2 minutes and the
absorbance of each well was determined on a Multiscan Plate Reader (ELX800, Biotek,
Winooski, USA) at a wavelength of 570 nm. Three replicates were used in each
experiment. For each treatment, cell viability was determined as a percentage of the
control untreated cells, by dividing the mean absorbance of each treatment by the mean
absorbance of the untreated cells. The concentration that reduces cell viability by 50%
(IC50) was established for each compound.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ruthenium accumulation in cancer cells. For ruthenium cell uptake studies, 1.5 x 106
DUꢀ145 or HEK293 cells were plated in 100 mm Petri dishes and allowed to attach for
48 h. Next, the plates were exposed to compounds 1-3 at a concentration corresponding
to a fifth of their IC50 (DUꢀ145: 1.3 ꢀM for 1, 17 ꢀM for 2 and 1.4 ꢀM for 3; HEK293:
0.5 ꢀM for 1, 10 ꢀM for 2 and 2.4 ꢀM for 3). Additional plates were incubated with
medium alone as negative control. After 24 h incubation, the cells were rinsed three
times with cold PBS and harvested by trypsinization. The number of cells in each
sample was counted manually in a hemocytometer using the Trypan Blue dyeꢀexclusion
test. Then the cells were centrifuged to obtain the whole cell pellet for ICPꢀMS analysis.
All experiments were conducted in triplicate.
ICP-MS analysis. The whole cell pellets were dissolved in 400 ꢂL of concentrated
60% v/v nitric acid, and the samples were then transferred into Wheaton vꢀvials (Sigmaꢀ
Aldrich) and heated in an oven at 373 K for 18 h. The vials were then allowed to cool,
and each cell sample solution was transferred into a volumetric tube and combined with
washings with MilliꢀQ water (1.6 mL). Digested samples were diluted 10 times with
MilliꢀQ to obtain a final HNO3 concentration of approximately 1.2% v/v. Ruthenium
content was analyzed on an ICPꢀMS Perkin Elmer Elan 6000 series machine at the
Centres Científics i Tecnològics of the University of Barcelona. The solvent used for all
ICPꢀMS experiments was MilliꢀQ water with 1% HNO3. The ruthenium standard (Highꢀ
Purity Standards, 1000 ꢂg/mL + 5 ꢂg/mL in 2% HCl) was diluted with MilliꢀQ water to
100 ppb. Ruthenium standards were freshly prepared in MilliꢀQ water with 1% HNO3
before each experiment. The concentrations used for the calibration curve were in all
12
ACS Paragon Plus Environment

Page 13 of 34
Molecular Pharmaceutics
1
2
3
4
5
6
7
cases 0, 1, 2, 5 and 10 ppb. The isotope detected was 101Ru and readings were made in
triplicate. Rhodium was added as an internal standard at a concentration of 10 ppb in all
samples.
8
9
RESULTS
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis and characterization of ruthenium-amino(guanidino)glycoside
conjugates.
Conjugation of the ruthenium complex to neomycin B was planned through the 5’’ꢀOH,
since this primary hydroxyl group can be regioselectively converted to amino group,
allowing the attachment of the imidazole ligand via an amide bond formation. In
addition, this modification does not alter the number of chargeable groups that are
essential for RNA binding and cell uptake. First, 4ꢀ(1HꢀImidazolꢀ1ꢀyl)benzoic acid was
coupled with the Bocꢀprotected amino derivative of neomycin52 (5, Scheme 2) by using
PyBOP as a coupling reagent in anhydrous DMF for 2 h at rt in the presence of DIPEA.
The metal complex was assembled on the imidazoleꢀderivatized neomycin (6) by
reaction with [(η6ꢀpꢀcym)RuCl(PPh3)2][PF6] (1 mol equiv) in the presence of LiCl (1.2
mol equiv) and NEt3 (10 mol equiv) in a DMF/EtOH 9:1 mixture for 40 min at 60ºC
under microwave irradiation. Reversedꢀphase HPLC analysis of the deprotected crude
(30% TFA in DCM, 25 min rt) showed the presence of a main peak, which was isolated
and characterized by MS as the expected conjugate 2. Hence, as previously found with
imidazoleꢀderivatized peptideꢀbound resins,24 the imidazole ligand incorporated in the
aminoglycoside moiety is able to replace one of the two PPh3 ligands in [(η6ꢀpꢀ
cym)RuCl(PPh3)2]+. However, in this case it was necessary to use microwave irradiation
to assemble the ruthenium complex, since no evidence of the formation of the Bocꢀ
protected intermediate 7 was obtained after prolonged heating (24 h) at 50ºC. After
purification by reversedꢀphase mediumꢀpressure liquid chromatography, the
trifluoroacetate salt of conjugate 2 was obtained as a yellow solid (overall yield from 6:
54%) that was unambiguously characterized by mass spectrometry and NMR
spectroscopy.
13
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 14 of 34
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Synthesis of the rutheniumꢀneomycin and rutheniumꢀguanidinoneomycin
conjugates.
Highꢀresolution ESI MS analysis of the neomycinꢀruthenium conjugate 2 afforded an
m/z value that was consistent with the calculated value of the monocharged species
([M]+) and with the expected isotopic distribution of ruthenium (Figure 1). In addition,
1
2 was fully characterized by H NMR spectroscopy by using 1D and 2D experiments
1
(COSY and TOCSY). The region of the H NMR spectra between 5.0 and 8.5 ppm is
shown in Figure 1, where signals from the ruthenium complex (pꢀcymene, imidazole
and PPh3 ligands) and from the aminoglycoside moiety (anomeric H1’ protons) are
indicated.
14
ACS Paragon Plus Environment

Page 15 of 34
Molecular Pharmaceutics
1
2
3
4
5
A
B
Ar+PPh3
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
H1’
Im
Im
Ar
H1’
p-cym
H1’
p-cym
C
p-cym
8.00
7.50
7.00
6.50
6.00
5.50
ppm (t1)
Figure 1. 1H NMR spectra of conjugate 2 (A) in CD3OD, showing the region between 4.5 and
8.5 ppm. Expanded ESI mass spectrum of the molecular peak of 2 ([M]+), experimental (B) and
calculated (C).
Our next objective was the synthesis of the guanidinylated analogue of the neomycinꢀ
ruthenium conjugate (compound 3 in Scheme 1), which was planned by using N,N’ꢀdiꢀ
BocꢀN”ꢀtriflylguanidine, a powerful guanidinylating reagent usually employed in the
preparation of guanidinoglycosides.55 Conjugate 2 was dissolved in methanol and
reacted with a large excess of N,N’ꢀdiꢀBocꢀN”ꢀtriflylguanidine (200 mol equiv) in the
presence of triethylamine (1000 mol equiv) at RT under an Ar atmosphere (Scheme 2).
The reaction was very slow, especially the incorporation of the sixth guanidinium
group, as followed by ESI MS analysis. Once the reaction reached completion (about 20
days), the Bocꢀprotected guanidino conjugate was treated with a TFA/DCM mixture
(1:1) for 4 h at RT. Reversedꢀphase HPLC analysis revealed the presence of a main
peak, which was isolated and characterized by highꢀresolution MS and NMR as the
expected conjugate 3 (overall yield from 2: 34%). It is interesting to note that the
ruthenium complex remained unaltered during the prolonged reaction time, which
accounts for its high stability in solution (see below).
One of the main problems of aminoglycoside antibiotics is their inherent toxicity,
usually nephrotoxicity or ototoxicity, which are associated with nonꢀspecific
electrostatic binding to RNA.43,44 To minimize these problems, we selected the
pseudodisaccharide known as neamine (Scheme 1), because a reduction in the number
15
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 16 of 34
1
2
3
4
5
6
7
8
9
of amino groups in the aminoglycoside scaffold usually gives less toxic compounds.
This smaller aminoglycoside is still an attractive starting molecule, since it incorporates
rings I and II of most naturallyꢀoccurring aminoglycosides, such as neomycin B, which
are important structural motifs in the recognition of RNA targets, including rRNA, RRE
and TAR RNA. Hence, our next objective was the synthesis of the neamineꢀruthenium
conjugate (4) by the optimized procedure used for the preparation of the neomycin
analogue 2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The attachment of the metal complex was planned through the 5ꢀhydroxyl group of the
2ꢀdeoxystreptamine ring (Scheme 1), since the use of trityl protective groups for the
amino functions allows the regioselective protection of all hydroxyl functions with 4ꢀ
methoxybenzyl groups, except that located at the 5ꢀ position.53 Again, the required
imidazoleꢀderivatized neamine (9) was obtained by reaction between 4ꢀ(1HꢀImidazolꢀ1ꢀ
yl)benzoic acid and the (Trt)4(PMB)3ꢀprotected amino derivative of neamine (8) by
using PyBOP as a coupling reagent (Scheme 3). The metal complex was assembled on
the imidazole ring by reacting 9 with [(η6ꢀpꢀcym)RuCl(PPh3)2][PF6] in the presence of
LiCl and NEt3 in a DMF/EtOH 9:1 mixture for 1 h at 60ºC under microwave irradiation.
In this case, deprotection was achieved by treatment of the crude with a 1:1 mixture of
TFA/anisole under Ar for 15 h at RT. Purification by semiꢀpreparative HPLC afforded
the TFA salt of conjugate 4 as a yellow solid (yield from 9: 46%), which was fully
characterized by highꢀresolution ESI and MALDIꢀTOF MS and NMR.
Scheme 3. Synthesis of the rutheniumꢀneamine conjugate.
16
ACS Paragon Plus Environment

Page 17 of 34
Molecular Pharmaceutics
1
2
Studies on the activation of the ruthenium complex in conjugates 2 and 3
Since the cytotoxic activity of most metallodrugs is intimately related to their hydrolysis
behavior in aqueous media, prior to cytotoxicity studies we wanted to assess whether
hydrolysis of the RuꢀCl bond occurs when the ruthenium complex is conjugated to the
glycoside carrier. This process is known to facilitate the interaction of the metal with the
biological target (e.g. nucleic acids or proteins) through the generation of
monofunctional adducts on guanine nucleobases by the activated aqua species.
Although anticancer activity of ruthenium(II) arene complexes has been attributed in
most cases to DNA ruthenation, some compounds such as 1 do not experience aqueous
hydrolysis of the RuꢀCl bond,24 which hinders the covalent interaction with the
biological target.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
On the basis of these precedents, the stability of conjugates 2 and 3 (Scheme 1) was
investigated in aqueous solution at a chloride concentration mimicking the typical cell
nucleus (4 mM). As previously found with 1 or its peptide conjugate,24 reversedꢀphase
HPLC analysis together with MS revealed that no hydrolysis occurred in the case of the
rutheniumꢀ(guanidino)neomycin conjugates on incubation at 37ºC for 48 h.
Surprisingly, the compounds also remained unaltered upon incubation at 37ºC for 72 h
with a large excess of glutathione (250 mol equiv) under physiologically relevant
conditions (pH 7 phosphate buffer containing 22 mM NaCl, the cytoplasmatic
concentration of chloride).30 These results suggest that the ruthenium moiety in the
amino(guanidino)glycoside conjugates (2ꢀ4) does not follow the typical activation
mechanisms of most organometallic anticancer complexes (e.g. hydrolysis of the RuꢀCl
bond and/or redox activation through the participation of the tripeptide glutathione).14
The presence of a bulky ligand such as triphenylphosphine might explain this behavior,
which is different from other anticancer ruthenium(II) arene complexes such as [(η6ꢀ
biphenyl)Ru(en)Cl)]+.1,14
Cytotoxicity in cancer (MCF-7 and DU-145) and normal (HEK293) cell lines
Our next objective was to evaluate the cytotoxicity of conjugates 2, 3 and 4 to
determine their potential as anticancer agents. Their antiꢀproliferative activity, together
with that of cisplatin as positive control and the unconjugated compounds (complex 1,
neomycin, neamine and guanidinoneomycin), was first determined in two human tumor
cell lines, the breast cancer cell line MCFꢀ7 and the prostate carcinoma cell line DUꢀ
145, by using the MTT test that measures mitochondrial dehydrogenase activity as an
17
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 18 of 34
1
2
3
4
5
6
7
8
9
indication of cell viability (Table 1). The compounds were screened at a wide range of
concentrations (from 0 to 250 ꢀM) to determine the concentration that inhibits cell
growth by 50% (IC50). The compounds that did not inhibit cell growth by more than
50% at 250 ꢀM were considered to be inactive. Interestingly, conjugates 2-4 were
completely soluble in water.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
IC50 (ꢀM)/cell lines
Compound
MCFꢀ7
3.04 + 0.25
3.99 + 1.45
84.67 + 5.13
200.75 + 15.75
> 250
DUꢀ145
3.10 + 0.00
6.48 + 0.54
84.63 + 15.87
221.67 + 3.33
> 250
HEK293
4.63 + 1.60
2.50 + 0.71
49.40 + 17.25
125.00 + 2.02
> 250
Cisplatin
Ruthenium complex 1
Ruꢀneomycin conjugate 2
Ruꢀneamine conjugate 4
Neomycin or neamine
Ruꢀguanidinoneomycin conjugate 3
Guanidinoneomycin
11.33 + 2.56
> 250
7.17 + 0.29
> 250
12.75 + 1.32
> 250
Table 1. Sensitivity of tumor MCFꢀ7 and DUꢀ145 cells and of nonꢀtumorigenic HEK293 cells
to cisplatin, compounds 1-4 and control amino(guanidino)glycosides. The concentration of the
compounds that inhibits cells viability by 50% (IC50) after 72 h was determined by means of the
MTT assay. Each value represents the mean of three independent experiments + standard error.
The concentrationꢀresponse curves, plotted in Figure 2, revealed that the cytotoxic
activity of conjugates 2-4 was very dependent on the nature of the glycoside moiety in
these cell lines. As previously reported,24 the effectivity of the complex [(η6ꢀpꢀ
cym)RuCl(ImꢀBzCOOMe)(PPh3)]+ (1) was comparable to that of cisplatin and slightly
higher in MCFꢀ7 cells than in DUꢀ145 cells, showing IC50 values of 3.99 + 1.45 ꢀM and
6.48 + 0.54 ꢀM, respectively (Table 1). Neomycinꢀruthenium conjugate 2 had moderate
antiꢀproliferative activity in both cell lines (IC50≈ 80), but its guanidinylated analogue
(3) was highly cytotoxic (IC50= 11.33 + 2.56 in MCFꢀ7 cells and IC50= 7.17 + 0.29 in
DUꢀ145 cells), with an IC50 value similar to that of complex 1 in the prostate cancer cell
line. However, conjugation of the ruthenium complex to the small neamine
aminoglycoside led to a strong decrease in cytotoxic activity (IC50≈ 200). Interestingly,
18
ACS Paragon Plus Environment

Page 19 of 34
Molecular Pharmaceutics
1
2
3
4
5
neither the aminoglycosides alone (neomycin or neamine) nor guanidinoneomycin were
found to be cytotoxic (IC50 > 250).
6
7
8
9
100
100
90
90
DU-145
MCF-7
80
80
70
60
50
70
60
50
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
40
30
20
10
0
40
30
20
10
0
0
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
0
10 20 30 40 50 60 70
80 90 100 110 120 130 140 150
Concentration (Micromolar)
Concentration (Micromolar)
120
110
100
90
80
70
60
50
40
30
20
10
0
HEK293
Ruthenium complex (1)
Ruthenium-neomycin conjugate (2)
Ruthenium-guanidinoneomycin conjugate (3)
0
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Concentration (Micromolar)
Figure 2. Cytotoxic effect of compounds 1-3 in the DUꢀ145, MCFꢀ7 and HEK293 cell lines.
Cells were treated for 72 h with the indicated concentrations of each compound. Cell viability
was determined by the MTT assay. Each point in the graphs represents the mean of three
independent experiments ± SD.
Despite the promising anticancer activities of the rutheniumꢀglycoside conjugates,
particularly those of guanidinoneomycin conjugate 3, an efficient targeted strategy
requires that the potential anticancer drug must be less toxic to normal cells than to
tumor cells. Otherwise, such anticancer agents generate undesired sideꢀeffects, as
current platinum drugs in clinical use do.
To assess the conjugates’ selectivity for tumor cells rather than normal cells, we
determined the cell viability of a nonꢀtumorigenic HEK293 cell line, in the presence of
conjugates 2-4. To our surprise, ruthenium complex 1 (IC50= 2.50 ꢀM) was found more
cytotoxic in the normal cell line than in the two cancer cell lines, whereas cisplatin
cytotoxicity was similar (IC50= 4.63 ꢀM) (Table 1). Again, control aminoglycosides and
guanidinoneomycin were found to be nonꢀcytotoxic (IC50 > 250). As shown in Figure 2,
the cytotoxicity of the rutheniumꢀglycoside conjugates against HEK293 cells can be
ranked in the following order: 3 >> 2 > 4, which reproduces the tendency found in
cancer cell lines. Interestingly, although control complex 1 and aminoglycosideꢀ
ruthenium conjugates (2 and 4) were more cytotoxic in the normal cell line than in both
19
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 20 of 34
1
2
3
cancer cell lines, this tendency was reversed for the guanidinylated conjugate (3), which
4
5
was much less active in HEK293 than in DUꢀ145 cells.
6
7
Cell uptake in DU-145 and HEK293 cell lines
8
9
To gain insight into the involvement of the glycoside moiety in the cytotoxic activity of
the compounds, the cell uptake of the control ruthenium complex (1) and of the
neomycinꢀruthenium (2) and guanidinoneomycinꢀruthenium (3) conjugates was
investigated in the prostate cancer cell line and in the normal cell line. Ruthenium
accumulation (determined here as the net effect of influx and efflux of Ru) in both cell
lines was quantified by inductivelyꢀcoupled plasma mass spectrometry (ICPꢀ
MS)24,56,57,58 after a 24 h exposure to the compounds at equicytotoxic concentrations,
which were in all cases a fifth of their IC50 values in each cell line.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As shown in Figure 3, the intracellular level of ruthenium after exposure to neomycin
conjugate 2 (107.14 ± 8.93 pmol Ru/106 cells) was substantially higher in the DUꢀ145
cell line than guanidinylated analogue 3 (24.10 ± 0.72 pmol Ru/106 cells) was, whereas
a similar level was obtained in the HEK293 cell line for both compounds (63.75 ± 8.46
pmol Ru/106 cells for 2 and 58.02 ± 3.95 pmol Ru/106 cells for 3).
350
Cellular uptake
300
DU-145 cells
250
HEK293 cells
200
150
100
50
0
(1)
(2)
(3)
Compound
Figure 3. Ruthenium uptake in DUꢀ145 and HEK293 cell lines after 24 h exposure to
equicytotoxic concentrations of compounds 1-3. The ruthenium content is related to the cell
number. Results are means of three independent samples and are expressed as mean + SD.
On the basis of the amount of intracellular ruthenium after exposure to compounds 1-3,
the molar intracellular concentration was calculated by considering the mean cellular
20
ACS Paragon Plus Environment

Page 21 of 34
Molecular Pharmaceutics
1
2
3
4
5
6
7
8
9
volume, as previously reported by Osella et al.59 The accumulation ratio was obtained
from the ratio between the intracellular concentration and the concentration of the
compounds in the extracellular medium at the beginning of the incubation period
(IC50/5). As shown in Table 2, the accumulation ratio of the neomycinꢀruthenium
conjugate (2) was similar in both cell lines. However, the accumulation ratio of the
guanidinylated analogue (3) was much higher than that of 2 both in the cancer DUꢀ145
cell line (about 3ꢀfold) and in the nonꢀtumorigenic HEK293 cell line (about 4ꢀfold).
Hence, these results are consistent with the generally accepted idea that the
incorporation of guanidinium groups in a molecule facilitates its internalization through
cell membrane,60ꢀ62 such as in the case of cellꢀpenetrating peptides. As previously
mentioned, cell uptake studies with aminoꢀ and guanidinoglycosides had demonstrated
an approximately 20ꢀfold internalization enhancement of neomycin upon
guanidinylation.45 In fact, guanidinoneomycin shows similar or even better cell uptake
efficiency than some polyarginineꢀcontaining peptides,45,60,61,62 which has been
attributed to the semiꢀrigid preꢀorganization of the guanidinium groups on the glycoside
core.45 Accordingly, ICPꢀMS accumulation studies with conjugates 2 and 3 showed the
same tendency in both cell lines, since guanidinylation of the neomycin moiety leads to
a compound (conjugate 3) with greater accumulation than its amino precursor
(conjugate 2). Moreover, it should be noted that accumulation of the guanidinylated
analogue (3) is about 1.4ꢀfold higher in the normal cell than in the tumor cell, whereas 2
is accumulated at the same ratio in both cell lines.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Regarding the parent complex (1), as shown in Figure 3, the intracellular level of
ruthenium in DUꢀ145 cells (240.86 ± 84.40 pmol Ru/106 cells) was much higher than in
HEK293 cells (27.64 ± 2.29 pmol Ru/106 cells). Similarly, the accumulation ratio was
about 3ꢀfold higher in DUꢀ145 than in HEK293 cells (Table 2), which indicates higher
accumulation in the prostate cancer cell line than in the normal cell line. In
consequence, we can conclude that conjugation of the ruthenium complex through the
imidazole ligand to a hydrophilic molecule, either neomycin or guanidinoneomycin,
leads to compounds (2 and 3) with reduced accumulation in both cell lines, particularly
in the case of conjugate 2, since the incorporation of the guanidinium groups in the
glycoside moiety seems to ameliorate this reduction.
21
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 22 of 34
1
2
3
4
5
6
7
8
Finally, the fact that in all cases the intracellular concentrations were greater than the
extracellular concentrations indicates an active cell uptake process for all the
compounds, particularly in the case of complex 1 and, to a lesser extent, its
guanidinoneomycin conjugate 3.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Intracellular concentration
(ꢀM)
Accumulation
ratio
Compound / Cell line
Ruthenium complex 1
Ruꢀneomycin 2
DUꢀ145
120.40
53.57
HEK293
13.82
DUꢀ145
80.29
3.15
HEK293
27.64
3.19
31.87
Ruꢀguanidinoneomycin 3
12.05
29.01
8.61
12.09
Table 2. Intracellular ruthenium concentrations determined in DUꢀ145 and HEK293 cells after
exposure to compounds 1-3 for 24 h at a concentration that was a fifth of their IC50 value. The
volume of a single cell was considered to be about 2 pL.63 Accumulation ratio59 represents the
ratio between the intracellular Ru concentration and the Ru concentration in the extracellular
medium at the beginning of the incubation period (DUꢀ145: 1: 1.5 ꢀM, 2: 17 ꢀM, 3: 1.4 ꢀM;
HEK293: 1: 0.5 ꢀM, 2: 10 ꢀM, 3: 2.4 ꢀM).
DISCUSSION
Current research efforts in metalꢀbased anticancer complexes have focused on the
development of new compounds, with the aim of overcoming the high toxicity and
sensitivity to resistance of platinumꢀbased drugs in clinical use, such as cisplatin and its
analogues.1ꢀ8 Organometallic complexes, in particular ruthenium(II) arene complexes,
have attracted attention because of their promising cytotoxic activities in several tumour
cell lines, including cisplatinꢀresistant cells.6ꢀ12 In medicinal chemistry terms, their
“pianoꢀstool” structure allows the optimization of lead compounds through structureꢀ
activity relationship studies by modifying ligands around the metallic center.15ꢀ16
Among such modifications, derivatization of nonꢀleaving ligands with a carrier
molecule has tremendous potential to generate selective, less toxic metallodrugs. The
potential of this targeted strategy is particularly promising when metal complexes,
either organometallic or classical coordination compounds, are attached to
biomolecules17ꢀ24 whose receptors are overexpressed on the membrane of tumoral
cells25ꢀ28 or to organic molecules with high affinity with a given biological target
22
ACS Paragon Plus Environment

Page 23 of 34
Molecular Pharmaceutics
1
2
3
4
5
6
7
8
(proteins or nucleic acids).29 Besides improving selectivity against cancer cells or
affinity towards the ultimate target, such modified ligands are also expected to generate
hybrid compounds with better pharmacological properties than the original metal
complex, such as aqueous solubility or cellular uptake.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
This paper reports the effect of conjugating an anticancer ruthenium(II) arene complex
(1) to aminoꢀ and guanidinoglycosides on the cytotoxic activity and cell uptake of the
resulting hybrid compounds in cancer and normal cells. The ultimate aim of this project
is to develop novel metalꢀbased anticancer compounds with reduced toxicity and side
effects that might have RNA as a final biological target. The fact that aminoꢀ and
guanidinoglycosides are known to be selective RNA binders, might facilitate the
interaction of the ruthenium moiety with potential RNA targets involved in the
pathogenesis of cancer, such as miRNAs or their precursors (preꢀmiRNAs). For the
synthesis of neomycinꢀ (2) and neamineꢀruthenium conjugates (4), the imidazole ligand
incorporated on the suitablyꢀprotected aminoglycoside moieties was reacted under
microwave irradiation with the complex [(η6ꢀpꢀcym)RuCl(PPh3)2]+ (Scheme 1). For the
guanidinylated analogue of the neomycin conjugate (3), the use of N,N’ꢀdiꢀBocꢀN”ꢀ
triflylguanidine was found compatible with the integrity of the ruthenium moiety. After
TFA deprotection, all conjugates were isolated by semipreparative HPLC (overall
yields: 45ꢀ55%) and fully characterized by high resolution MS and NMR spectroscopy.
The cytotoxicity of conjugates 2-4 together with that of cisplatin, complex 1 and control
aminoꢀ and guanidinoglycosides was tested in two human cancer cell lines, MCFꢀ7
(breast) and DUꢀ145 (prostate), as well as in a normal cell line, HEK293 (human
embryonic kidney). As shown in Table 1, the cytotoxic activity of the rutheniumꢀ
glycoside conjugates against cancer and normal cells can be ranked in the following
order: 3 >> 2 > 4. The fact that both natural aminoglycoside antibiotics, neamine and
neomycin, and guanidinoneomycin were found to be nonꢀcytotoxic in all cell lines (IC50
> 250) indicates that the anticancer activity of compounds 2-4 is provided by the
ruthenium moiety. However, an accurate analysis of the IC50 values of the compounds
reveals that, as well as the ruthenium or the glycoside moieties, the nature of the cell
(tumoral or healthy) is another key factor to be considered. Two important observations
can be drawn when comparing IC50 values of these compounds in the normal cell line
with those in cancer cells. On the one hand, rutheniumꢀaminoglycoside conjugates (2
23
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 24 of 34
1
2
3
4
5
6
7
8
9
and 4) were more cytotoxic in the nonꢀtumorigenic cell line than in cancer cells.
Effectively, their antiꢀproliferative activity in HEK293 cells was increased by about
twoꢀfold in both cases (e.g. IC50 ≈ 80 ꢀM for conjugate 2 in both cancer cell lines vs
IC50= 49.4 ꢀM for 2 in the normal cell line). On the other hand, the guanidinylated
conjugate 3 was about twoꢀfold less cytotoxic in the healthy cell line than in the DUꢀ
145 cell line (e.g. IC50 = 12.75 ꢀM vs IC50 = 7.17 ꢀM, respectively), whereas a similar
IC50 value was obtained when compared with the MCFꢀ7 cell line (IC50 = 11.33 ꢀM),
revealing that the response to 3 is also related to the nature of the cancer cells. These
results indicate that aminoglycosideꢀcontaining conjugates 2 and 4 show the same
behavior as their parent ruthenium complex 1, since their cytotoxicity in normal cells
was always greater than in cancer cells. However, this tendency was inverted with the
guanidinylated analogue of the neomycinꢀruthenium conjugate, since the antiꢀ
proliferative activity of 3 in cancer cells was similar to or even higher than in a normal
cell line.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Such differences in cytotoxic activity for each compound in a particular cell line, either
normal or tumoral, could be interpreted in terms of cell uptake and accumulation
efficiency. First, the nature of the aminoglycoside moiety seems to be an important
factor, since the cytotoxicity of the neomycinꢀruthenium conjugate (2) in the three cell
lines is about 2.5ꢀfold higher than that of the neamineꢀcontaining conjugate (4).
Although both aminoglycosides are polycations at physiological pH, the greater number
of ammonium groups in neomycin than in neamine (six vs four) might favor
accumulation of conjugate 2, resulting in a slightly greater antitumoral activity. Second,
the fact that the cytotoxicity of the guanidinylated conjugate (compound 3) in the three
cell lines is substantially higher than that of the parent aminoglycoside conjugate
(compound 2) (7ꢀfold in MCFꢀ7 cells, 12ꢀfold in DUꢀ145 cells and 4ꢀfold in HEK293
cells) could be attributed to greater permeability through cell membranes, a property
that might facilitate its accumulation in the cell and, for instance, higher antiꢀ
proliferative activity. Hence, an apparently good correlation between the expected
glycoside uptake efficiency in eukaryotic cells (guanidinoneomycin >> neomycin >
neamine) and cytotoxic activity (3 >> 2 > 4) can be established in all cell lines, although
this does not explain the differences between normal and cancer cells in antiꢀ
proliferative activity of the conjugates.
24
ACS Paragon Plus Environment

Page 25 of 34
Molecular Pharmaceutics
1
2
3
4
5
6
7
8
9
Ruthenium accumulation studies performed by ICPꢀMS with neomycin (2) and
guanidinoneomycin (3) conjugates in DUꢀ145 and HEK293 cells clearly support the
above conclusions. As shown in Table 2, guanidinylation of the aminoglycoside moiety
leads to a compound (3) with higher accumulation than that of its amino precursor 2.
Indeed, the accumulation ratio of 3 in the cancer cell line was about 3ꢀfold greater than
that of 2, whereas this tendency increased even more in the normal cell line (about 4ꢀ
fold). These results indicate that guanidinylation of the amino functions at the
aminoglycoside moiety in conjugate 2 had a positive effect on cell uptake, thus
improving intracellular accumulation. Moreover, it should be noted that the
accumulation ratio of conjugate 3 in the normal cell line was 1.4ꢀfold higher than in the
cancer cell line, thus revealing behavior the opposite of its amino precursor 2, which
was equally accumulated in both cell lines, or that of control complex 1, which
accumulated at a higher proportion (about 3ꢀfold) in DUꢀ145 than in HEK293.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
On the one hand, recent studies have revealed that guanidinoneomycin uptake in CHO
cells is mediated by cellꢀsurface heparinꢀsulfate proteoglycans, which has been used to
transport large bioactive cargo into cells at low concentration (nM order) in a selective
proteoglycanꢀdependent manner.47ꢀ49 Thus, we may speculate that differences in the
expression level and/or in the composition of proteoglycan receptors on the cell
membrane surface between cancer cells and normal cells would be responsible for such
differences in the accumulation ratio of 3 between DUꢀ145 and HEK293 cells.64
On the other hand, the fact that neomycinꢀ and guanidinoneomycinꢀruthenium
conjugate’s accumulations were lower than that of the parent ruthenium complex 1 also
indicates that conjugation to polar, polycationic glycosides results in reduced cell
uptake in both cell lines, although this reduction was more dramatic in the cancer cell
line. It is well known that lipophilic compounds can cross cell membranes more readily
than hydrophilic compounds, which results in increased intracellular accumulation. This
is also true for ruthenium(II) arene complexes since an increase in their lipophilicity, for
example by increasing the size of the arene ligand, correlates with greater cytotoxicity.65
Accordingly, the decrease in accumulation of complex 1 when conjugated to neomycin
or guanidinoneomycin could be attributed to an overall decrease in the lipophilicity of
the compound induced by the highly hydrophilic glycoside moiety, which would
25
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 26 of 34
1
2
3
4
5
6
7
8
9
therefore diminish cell uptake. Although the lipophilicity provided by the ruthenium
complex (e.g. pꢀcymene and PPh3 ligands) may modulate the ability of
amino(guanidino)glycosides to cross cell membranes when conjugated together, these
results seem to indicate that the glycoside moiety, particularly in the case of
guanidinoneomycin, has a fundamental weight in the cell uptake of the conjugates.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Overall, on the basis of cytotoxicity and cell uptake studies, in DUꢀ145 cells there is a
correlation between accumulation (1 >> 3 > 2) and antiꢀproliferative activity (1 ≈ 3 >>
2). The same tendency was found in the case of the normal cell line, since the cytotoxic
activity ranking (1 > 3 >> 2) follows that of cellular accumulation data (1 > 3 >> 2).
Interestingly, despite the fact that neomycinꢀruthenium conjugate (2) accumulation was
similar in both cell lines, its antiꢀproliferative activity was higher in the normal cell than
in the tumor cell (about 1.7ꢀfold). An opposite tendency was found for the
guanidinylated analogue (3), since the cytotoxic activity was higher in the tumor cell
line than in the normal cell line (about 1.8ꢀfold), although the accumulation ratio in the
normal cell was slightly higher (about 1.4ꢀfold). These results suggest that the glycoside
moiety cannot be seen as a simple carrier that modulates the lipophilicity of the
anticancer ruthenium(II) arene complex and, for instance, cell uptake, but rather as a
dynamic moiety that also modulates the antiꢀproliferative activity of the metal fragment,
depending on the cell type. Indeed, the fact that the antiꢀproliferative activity of 3 in the
DUꢀ145 cancer cells was higher than in the HEK293 normal cells, with a lower
accumulation ratio, suggests that guanidinoneomycin provides some kind of selectivity
against cancer cells. This hypothesis is supported by the greater cytotoxicity of
ruthenium complex 1 or its neomycin conjugate 2 in normal cells than in cancer cells,
despite the fact that their accumulation in normal cells is much lower than (1) or similar
to (2) that found in cancer cells. Hence, conjugation to guanidinoneomycin leads to a
compound (3) with reduced cytotoxicity against normal cells but with a similar antiꢀ
proliferative activity to that of the parent ruthenium complex 1 in DUꢀ145 cells.
Finally, we cannot rule out the possibility that amino(guanidino)glycoside conjugation
may also modify not only cell uptake in cancer and normal cells, but also the
mechanism of action of the ruthenium complex or its biological target. As previously
mentioned, guanidinoglycosides have higher binding affinity with RNA sequences than
their parent aminoglycosides do. Among aminoglycosides, neomycin derivatives bind
26
ACS Paragon Plus Environment

Page 27 of 34
Molecular Pharmaceutics
1
2
3
4
5
6
7
8
9
stronger RNA structures than neamineꢀcontaining compounds. Hence, given the
significant differences in cytotoxicity of the conjugates in both cancer cell lines (IC50 ≈
7ꢀ11 ꢀM for 3, IC50 ≈ 80 ꢀM for 2 vs IC50 ≈ 200 ꢀM for 4) and in the normal cell line
(IC50 ≈ 12 ꢀM for 3, IC50 ≈ 50 ꢀM for 2 vs IC50 ≈ 125 ꢀM for 4), we may well speculate
about the involvement of RNA as a target for these conjugates, since there is close
correlation between their antiꢀproliferative activity (3 >> 2 > 4) and the RNA binding
affinity of the glycoside moiety (guanidinoneomycin >> neomycin > neamine). The fact
that the ruthenium moiety in conjugates 2 and 3 does not follow the typical activation
mechanisms of most anticancer organometallic ruthenium complexes (aqueous
hydrolysis of the RuꢀCl bond and/or glutathione mediated redox activation) suggests
that ruthenation at RNA nucleobases would not occur. However, binding to this
biological target could be based on nonꢀcovalent interactions such as electrostatic forces
and/or hydrogen bonds between the negatively charged skeleton of RNA and the
polycationic glycoside moiety. Moreover, aromatic ligands from the ruthenium complex
(phenyl rings in PPh3 and pꢀcymene) could interact with RNA through intercalation or
stacking with nucleobases.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In conclusion, the overall results demonstrate the potential of conjugating anticancer
metal complexes, in particular organometallic ruthenium(II) complexes, to
aminoglycosides, especially to their guanidinylated derivatives, to generate anticancer
metalꢀbased drugs with new modes of action. As far as we are aware, such differences
between
cancer
and
healthy
cells
in
the
cytotoxic
activity
of
amino(guanidino)glycosides conjugated to a ruthenium complex are unprecedented and
open the way to the use of guanidinoglycosides to deliver metalꢀbased anticancer agents
selectively into tumor cells, with the aim of developing new drugs with reduced toxicity
and sideꢀeffects. Modifications on the aminoglycoside scaffold, on the metal complex or
its properties, or in the number of guanidinium groups might lead to increasing such
differences in cytotoxic activity between cancer and normal cells. Moreover, the use of
cleavable linkers in the biological media or metal complexes that might be selectively
activated (e.g. via irradiation) is expected to improve their potential as anticancer drugs.
Since aminoglycosides and their guanidinylated derivatives might be accumulated in
some specific cellular compartments such as lysosomes, this strategy would facilitate
the detachment of the metallodrug from the glycoside carrier, thus increasing the
27
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 28 of 34
1
2
3
4
5
6
7
8
effective concentration of the anticancer active species. Further work is in progress to
investigate the anticancer activity of these compounds in a wide panel of tumoral cells,
as well as to establish their mechanism of action and if RNA is involved as a drug
target.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
28
ACS Paragon Plus Environment

Page 29 of 34
Molecular Pharmaceutics
1
2
ASSOCIATED CONTENT
Supporting Information
3
4
5
6
7
8
9
Highꢀresolution ESI MS and NMR spectra of conjugates 2-4 and of intermediates 6 and
9. Reversedꢀphase HPLC traces of pure conjugates. This information is available free of
charge via the Internet at http://pubs.acs.org.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Author
*Vicente Marchán: Phone +34 934021249, Fax +34933397878, Eꢀmail
vmarchan@ub.edu
ACKNOWLEDGEMENTS
This research was supported by funds from Spain’s Ministerio de Ciencia e Innovación
(grant CTQ2010ꢀ21567ꢀC02ꢀ01), the Generalitat de Catalunya (2009SGR208), the
Instituto de Salud Carlos III (grants RD06/0020/0041) and the Programa
d’Intensificació de la Recerca (UB).
REFERENCES
(1) Peacock, A. F. A.; Sadler, P. J. Medicinal organometallic chemistry: Designing
metal arene complexes as anticancer agents. Chem. Asian J. 2008, 3, 1890ꢀ1899.
(2) Hartinger, C. G; Dyson, P. J. Bioorganometallic chemistryꢀfrom teaching paradigms
to medicinal applications. Chem. Soc. Rev. 2009, 38, 391ꢀ401.
(3) Casini, A.; Hartinger, C. G.; Nazarov, A.; Dyson, P. J. Organometallic antitumour
agents with alternative modes of action. Top. Organomet. Chem. 2010, 32, 57ꢀ80.
(4) Gasser, G.; Ott, I.; MetzlerꢀNolte, N. Organometallic anticancer compounds. J. Med.
Chem. 2011, 54, 3ꢀ25.
(5) Noffke, A. L.; Habtemariam, A.; Pizarro, A. M.; Sadler, P. J. Designing
organometallic compounds for catalysis and therapy. Chem. Commun. 2012, 48, 5219ꢀ
5246.
(6) Levina, A.; Mitra, A; Lay, P. A. Recent developments in ruthenium anticancer
drugs. Metallomics 2009, 1, 458ꢀ470.
(7) SüssꢀFink, G. Arene ruthenium complexes as anticancer agents. Dalton Trans. 2010,
39, 1673ꢀ1688.
29
ACS Paragon Plus Environment