New insight into the structural, electrochemical and biological aspects of macroacyclic Cu(II) complexes derived from S-substituted dithiocarbazate schiff bases

Article abstract of DOI:10.1016/j.ejmech.2016.04.027

Copper (II) complexes synthesized from the products of condensation of S-methyl- and S-benzyldithiocarbazate with 2,5-hexanedione (SMHDH2 and SBHDH2 respectively) have been characterized using various physicochemical (elemental analysis, molar conductivity, magnetic susceptibility) and spectroscopic (infrared, electronic) methods. The structures of SMHDH2, its copper (II) complex, CuSMHD, and the related CuSBHD complex as well as a pyrrole byproduct, SBPY, have been determined by single crystal X-ray diffraction. In order to provide more insight into the behaviour of the complexes in solution, electron paramagnetic resonance (EPR) and electrochemical experiments were performed. Antibacterial activity and cytotoxicity were evaluated. The compounds, dissolved in 0.5% and 5% DMSO, showed a wide range of antibacterial activity against 10 strains of Gram-positive and Gram-negative bacteria. Investigations of the effects of efflux pumps and membrane penetration on antibacterial activity are reported herein. Antiproliferation activity was observed to be enhanced by complexation with copper. Preliminary screening showed Cu complexes are strongly active against human breast adenocarcinoma cancer cell lines MDA-MB-231 and MCF-7.

Full text of DOI:10.1016/j.ejmech.2016.04.027

Accepted Manuscript
New insight into the structural, electrochemical and biological aspects of macroacyclic
Cu(II) complexes derived from S-substituted dithiocarbazate schiff bases
May Lee Low, Laure Maigre, Mohamed Ibrahim M. Tahir, Edward R.T. Tiekink,
Pierre Dorlet, Régis Guillot, Thahira Begum Ravoof, Rozita Rosli, Jean-Marie Pagès,
Clotilde Policar, Nicolas Delsuc, Karen A. Crouse
PII:
S0223-5234(16)30320-8
10.1016/j.ejmech.2016.04.027
EJMECH 8542
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 31 August 2015
Revised Date: 25 March 2016
Accepted Date: 9 April 2016
Please cite this article as: M. Lee Low, L. Maigre, M.I.M. Tahir, E.R.T. Tiekink, P. Dorlet, R. Guillot,
T.B. Ravoof, R. Rosli, J.-M. Pagès, C. Policar, N. Delsuc, K.A. Crouse, New insight into the structural,
electrochemical and biological aspects of macroacyclic Cu(II) complexes derived from S-substituted
dithiocarbazate schiff bases, European Journal of Medicinal Chemistry (2016), doi: 10.1016/
j.ejmech.2016.04.027.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
1
2
New Insight into the Structural, Electrochemical and Biological Aspects of Macroacyclic
Cu(II) Complexes Derived from S-Substituted Dithiocarbazate Schiff Bases
3
4
5
6
7
8
May Lee Lowa,b, Laure Maigrec, Mohamed Ibrahim M. Tahira, Edward R. T. Tiekinkd, Pierre
Dorlete, Régis Guillotf, Thahira Begum Ravoofa, Rozita Roslig,h, Jean-Marie Pagèsc, Clotilde
Policarb*, Nicolas Delsucb*and Karen A. Crouseai*
aDepartment of Chemistry, Universiti Putra Malaysia, 43400 Serdang, Selangor (Malaysia), Fax: +6 03
9
89435380; E-mail: kacrouse@gmail.com
10 bÉcole Normale Supérieure-PSL Research University, Département de Chimie, Sorbonne Universités -
11
12
UPMC Univ Paris 06, CNRS UMR 7203 LBM, 24, rue Lhomond, 75005 Paris, (France), Fax: +33 1 44 32
24 02; E-mail: clotilde.policar@ens.fr ; nicolas.delsuc@upmc.fr
13 cUMR-MD1, Aix-Marseille Université, IRBA, Faculté de Médecine, 27 boulevard Jean Moulin, 13385 Marseille
14 (France)
15 dResearch Centre for Crystalline Materials, Faculty of Science and Technology, Sunway University, No. 5 Jalan
16 Universiti, 47500 Bandar Sunway, Selangor Darul Ehsan, (Malaysia), edward.tiekink@gmail.com
17 eInstitute for Integrative Biology of the Cell (I2BC), Université Paris- Saclay, CEA, CNRS, Université Paris-
18 Sud, Bât 532 CEA Saclay, 91191 Gif sur Yvette cedex (France)
19 fInstitut de Chimie Moléculaire et des Matériaux d'Orsay, Bât. 420 Université Paris-Sud, 91405 Orsay (France)
20 gDepartment of Obstetrics and Gynaecology, Universiti Putra Malaysia, 43400 Serdang, Selangor (Malaysia)
21 hUPM-MAKNA Cancer Research Laboratory, Institute of Bioscience, Universiti Putra Malaysia, 43400
22
Serdang, Selangor, (Malaysia)
23 iDepartment of Chemistry, Cape Breton University, Sydney, Nova Scotia B1P 6L2 (Canada)
24
25 Keywords: dithiocarbazate; Schiff base; macroacyclic ligand; tetradendate ligand; NNSS
26 ligand; copper complexes; bioactivity; MDA-MB-231; MCF-7; Gram-positive; Gram-
27 negative
28
29 Abstract
30 Copper(II) complexes synthesized from the products of condensation of S-methyl- and S-
31 benzyldithiocarbazate with 2,5-hexanedione (SMHDH2 and SBHDH2 respectively) have
32 been characterized using various physicochemical (elemental analysis, molar conductivity,
33 magnetic susceptibility) and spectroscopic (infrared, electronic) methods. The structures of
34 SMHDH2, its copper(II) complex, CuSMHD, and the related CuSBHD complex as well as a
35 pyrrole byproduct, SBPY, have been determined by single crystal X-ray diffraction. In order
36 to provide more insight into the behaviour of the complexes in solution, electron
37 paramagnetic resonance (EPR) and electrochemical experiments were performed.
38 Antibacterial activity and cytotoxicity were evaluated. The compounds, dissolved in 0.5%
39 and 5% DMSO, showed a wide range of antibacterial activity against 10 strains of Gram-
40 positive and Gram-negative bacteria. Investigations of the effects of efflux pumps and
41 membrane penetration on antibacterial activity are reported herein. Antiproliferation activity
42 was observed to be enhanced by complexation with copper. Preliminary screening showed Cu
43 complexes are strongly active against human breast adenocarcinoma cancer cell lines MDA-
44 MB-231 and MCF-7.

ACCEPTED MANUSCRIPT
45 TOC diagram
46
47
2

ACCEPTED MANUSCRIPT
48
49
50
1. Introduction
Effective treatment of multi-drug resistant (MDR) bacterial infections has become
51 increasingly challenging as the efficiency of the available antibiotic arsenal is reduced,
52 resulting in increased frequency of therapeutic failure [1, 2]. Over-expression of efflux pumps
53 can contribute to resistance of bacteria by expulsion of structurally unrelated compounds
54 causing a decrease in the intracellular concentration of antibiotics [3, 4]. It is essential to
55 understand efflux-mediated resistance in bacterial pathogens to develop efficient antibacterial
56 agents circumventing this mechanism. In addition, parallel concerns relating to acquired drug
57 resistance as well as the serious side-effects of anticancer drugs in the midst of the increasing
58 rate of cancer diagnoses drives the effort to develop better alternatives [5, 6]. Due to their
59 many tunable functionalities, dithiocarbazate compounds are exciting candidates for
60 exploration and potential development as antimicrobial and cytotoxic agents.
61
62
Sulphur-nitrogen chelating agents derived from S-alkyl/aryl esters of dithiocarbazic acid
63 have been extensively investigated in recent years for their cytotoxicity [7, 8], antibacterial
64 [9], antiamoebic [10], anti-Trypanosoma cruzi [11] and anti-Mycobacterium tuberculosis [12]
65 activities. Considerable attention continues to be given to these and related Schiff bases [13-
66 16], since their properties can be modulated by introducing different substituents through
67 condensation of various S-substituted dithiocarbazate esters with a wide array of aldehydes
68 and ketones. In many cases, the bioactivities of various dithiocarbazate derivatives have been
69 shown to differ widely although there may be only slight modifications in their molecular
70 structures [8]. Since these ligands possess both hard nitrogen and soft sulfur donor atoms they
71 are capable of coordinating with a wide range of transition and non-transition metal ions
72 forming metal complexes with interesting physicochemical and enhanced biological
73 properties [17-19]. The wide diversity of structures displayed by macrocyclic and
74 macroacyclic Schiff bases [20] results in various coordination abilities that could potentially
75 lead to applications ranging from diagnostics to therapeutics [21-22]. As part of our ongoing
76 exploration of these interesting properties, we investigated the synthesis and characterization
77 of some macroacyclic bis(dithiocarbazato) Schiff bases and their Cu(II) complexes. Copper
78 complexes derived from the analogues thiosemicarbazate have also been subjected to
79 intensive research [23-26] and appear to be very efficient antimicrobial [27] and anticancer
80 [28] agents. The copper(II) complexes of quadridentate NNSS donor ligands reported in the
81 literature are also known to be neutral, stable (Kass=1018) compounds that easily cross cellular
3

ACCEPTED MANUSCRIPT
82 membranes [23, 29]. Thus, copper ion was a logical choice for complexation in our search for
83 effective metallodrugs.
84
85
The main aim of the present work is to explore the biological potential of newly
86 synthesized bis(dithiocarbazato) ligands and their Cu(II) complexes by determining their
87 potencies against different bacterial strains expressing a multi-drug resistance phenotype and
88 the effect of efflux pumps and membrane penetration on their antibacterial activity. In
89 addition cytotoxicity assays against two breast cancer cell lines was carried out to determine
90 the effect of complexation with copper upon the activity of the ligands against these cells.
91 Whereas syntheses of many dithiocarbazate compounds have been reported in the literature,
92 reports on the bioactivities [30], crystallography, EPR and electrochemistry [31, 32] of Cu(II)
93 bis(dithiocarbazate) complexes are limited. To develop such compounds with effective
94 pharmacological activity, it is essential to orient effort towards correlating the biological
95 activities of this class of compounds with their solid and solution structures as well as their
96 physicochemical properties to identify the optimum geometry about the Cu ion. This goal can
97 be achieved through the synthesis of a graduated series of ligands designed to reveal the
98 mode of bioaction.
99
100
101
102
103
104
4

ACCEPTED MANUSCRIPT
105
2. Results and Discussion
106 2.1. Synthesis and characterization
107
The synthesis of S-substituted dithiocarbazates was performed as previously
108 described [33, 34]. Carbon disulfide and hydrazine were reacted in basic ethanol. After
109 workup, the dithiocarbazate produced was directly reacted with methyl iodide or benzyl
110 chloride to afford S-methyldithiocarbazate (SMDTC) and S-benzyldithiocarbazate (SBDTC),
111 respectively. Schiff bases were then prepared by a slight variation of the method described by
112 Ali et al. [35]. The respective S-substituted dithiocarbazates and 2,5-hexanedione were
113 condensed in 2:1 ratio (Scheme 1). The initial attempts to synthesize the ligand SBHDH2
114 with prolonged heating followed by purification using column chromatography were
115 unsuccessful. NMR, ESI, elemental analysis and single crystal X-ray diffraction confirmed
116 cyclization to a pyrrole derivative. We postulate that bis(dithiocarbazate) indeed formed but
117 was then hydrolyzed to mono(dithiocarbazate) and S-benzyldithiocarbazate [36, 37] with
118 subsequent cyclization of the mono(dithiocarbazate) to a pyrrole via the Paal-Knorr reaction.
119 To our knowledge, this is the first pyrrole derived from a dithiocarbazate reported although
120 there are two recent reports of formation of pyrrole byproducts upon reaction of
121 thiosemicarbazone with 2,5-hexanedione [38, 39]. Encouraged by the remarkable
122 pharmacological properties of functionalized pyrroles [40, 41], we tested the compound for
123 its antimicrobial activity, the results of which are discussed below. The Schiff base,
124 SBHDH2, was finally obtained using either of the following two methods: stirring the dione
125 and SBDTC at room temperature for 30 minutes or heating for only 5 minutes after which the
126 white precipitate formed immediately. SMHDH2 was synthesized without the complication
127 of side-reaction occurrence. The precipitate was recrystallized to afford pure SMHDH2 (70%
128 yield).
129
130
Cu(II) complexes with NNSS coordination were obtained from the reaction of
131 copper(II) acetate with an equimolar amount of the respective ligand (in acetonitrile for
132 SBHDH2 and methanol for SMHDH2). The complexes were isolated by filtration with yields
133 of 77% and 73% for CuSMHD and CuSBHD, respectively. Black crystals were grown from
134 acetonitrile.
5

ACCEPTED MANUSCRIPT
135
136 Scheme 1. Synthesis of the copper complexes derived from bis(dithiocarbazate) ligands. a)
137 CS2, KOH, EtOH, 0°C, 1 h; b) CH3I or PhCH2Cl, EtOH, 0°C, 5 h; c) for SMHDH2 (2,5-
138 hexanedione, EtOH, 79°C, 1 h), for SBHDH2 (2,5-hexanedione, EtOH, 79°C, 5 min) and d)
139 for CuSMHD [Cu(OAc)2·H2O, MeOH, 65°C, 1 h], for CuSBHD [Cu(OAc)2, acetonitrile, r.t.,
140 1 h].
141
142
143 2.2. Characterization of the complexes in the solid state
144
The characteristic infrared bands of the S-substituted dithiocarbazate ligand, ν(N-H)
145 at ca. 3129 cm-1 and ν(C=S) at ca. 1050 cm-1 disappeared upon formation of the Cu(II)
146 complexes. In addition, ν(C=N) of the azomethine bond shifted to lower energy (1611 cm-1
147 and 1606 cm-1 for CuSMHD and CuSBHD, respectively) and a second ν(N=C) band in
148 complexes containing anionic dithiocarbazate moieties appeared [42]. The hydrazinic band,
149 ν(N-N), at ca. 828 cm-1 in the free ligand also shifted upon complexation, to higher
150 (CuSBHD) and lower (CuSMHD) wavenumbers. These observations confirm deprotonation
151 of the Schiff bases with coordination through the azomethine nitrogen atom. The ν(CSS)
152 band ca. 985 cm-1 (ligand) splits into two components at 1000-955 cm-1 upon complexation.
153 The presence of this band and the absence of the C=S band in the spectra of the metal
154 complexes provide additional evidence of the coordination of the Schiff base to the metal in
155 its thiolate form [43, 44]. To confirm the 1:1 stoichiometry, the complexes were also
156 characterized by elemental microanalyses for which the analytical data were found to agree
157 with the formulations proposed for the complexes. As expected for paramagnetic 3d9 ions, the
158 magnetic susceptibility values measured at room temperature for the CuSMHD and CuSBHD
159 complexes (1.66 B.M and 1.48 B.M, respectively) suggest a square-planar environment (spin-
160 only value 1.73 B.M) [44, 45]. The slightly low values observed can be attributed to
161 interaction between Cu(II) ion centers [46, 47] or distortion in the Cu(II) environment [48].
162
163
Crystals of SBPY, SMHDH2, CuSMHD and CuSBHD suitable for single crystal X-
164 ray diffraction were obtained; crystallographic data are given in Table 1. As mentioned in
165 Section 2.1, the pyrrolyl derivative, SBPY, is a cyclized product obtained during an attempted
166 synthesis of SBHDH2. The molecular structure of SBPY is shown in Fig. 1a. In SBPY, the
6

ACCEPTED MANUSCRIPT
167 central CN2S2 chromophore is planar (r.m.s. = 0.0490 Å) and forms dihedral angles of
168 88.49(4) and 68.14(4)º with the pyrrolyl and phenyl rings, respectively. As the rings lie to the
169 same side of the molecule, opposite to the thione-S1 atom, the overall conformation is best
170 described as being U-shaped. The dihedral angle between the rings is 60.874(6)º indicating a
171 splayed relationship. The thione-S1 and amine-H atoms are syn which might be expected to
172 lead to an eight-membered {…HNCS}2 synthon in the crystal packing. Nevertheless, the
173 most prominent feature of the crystal packing is the formation of N–H…π (pyrrolyl)
174 interactions, Fig. 1b, which lead to the formation of centrosymmetric dimeric aggregates.
175
176
177 Fig. 1. (a) The structure of SBPY, showing its atom-labelling scheme, and (b) the
178 supramolecular dimer sustained by N–H…π (pyrrolyl) interactions.
179
180
SMHDH2 (Fig. 2) crystallises about a crystallographic centre of inversion located at
181 the mid-point of the C3-C3i bond indicting the molecule has an anti disposition of the
182 dithiocarbazate residues; symmetry operation i: 1-x, 2-y, 1-z. The conformation about the
183 hydrazone bond is E. The entire molecule is planar with the r.m.s. for the 18 non-hydrogen
184 atoms comprising the entire molecule being 0.038 Å, with the maximum deviations being
185 ±0.061 Å for the S2 atom.
186
7

ACCEPTED MANUSCRIPT
187
188 Fig. 2. (a) The structure of SMHDH2, showing its atom-labelling scheme. Unlabelled atoms
189 are related by the symmetry operation 1-x, 2-y, 1-z, and (b) supramolecular chain mediated
190 by N–H…S hydrogen bonds via centrosymmetric eight-membered {…HNCS}2 synthons.
191 [N1–H1n…S1 = 2.64(2) Å, N1…S1 = 3.455(2) Å, and angle at H1n = 156(3)º; symmetry
192 operation i: 2-x, 2-y, 2-z]
193
194
The doubly deprotonated SMHD species functions as a tetradentate N2S2 donor in its
195 complex with copper(II). The molecular structure of CuSMHD is shown in Fig. 3 and
196 selected geometric bond lengths (Å) and angles (°) for this, CuSBHD, and for SMHDH2 are
197 given in Table 2. To a first approximation, the seven-membered ring may be described as
198 having a half-chair conformation where the C4 atom lies 0.9317(16) Å above the plane
199 defined by the Cu, N2, N3, C2, C3 and C5 atoms; r.m.s. = 0.1373 Å with maximum
200 deviations 0.1590(6) Å for Cu and -0.1595(10) Å for C2. The two five-membered chelate
201 rings have similar conformations. The S1-containing ring is an envelope with the flap atom,
202 Cu, lying 0.7857(16) Å above the least-squares plane defined by the remaining four atoms
203 (r.m.s. = 0.0119 Å). The S3-containing ring is considerably more planar but can still be
204 described as having an envelope conformation with Cu being the flap atom. In this
205 description, the Cu atom lies 0.2103(18) Å out of the plane defined by the four remaining
206 atoms which has a r.m.s. of 0.0043 Å. There is a clear distortion away from the ideal square
207 planar geometry as is commonly observed in seven-membered rings having two hydrazone
208 bonds [49]. In CuSMHD, the angle between the two five-membered chelate rings is 46.10(2)º
209 and the range of angles subtended at the Cu atom is 84.89(3)º for S1–Cu–N2 to 164.90(4)º
210 for S1–Cu–N3.
211
8

ACCEPTED MANUSCRIPT
212
213 Fig. 3. The structure of CuSMHD, showing its atom-labelling scheme.
214
215
The availability of the crystal structure of SMHDH2 enables a comparison of the
216 geometric parameters in the free molecule and in its coordinated dianion in CuSMHD. Two
217 quite distinct differences are noted in the bond lengths collected in Table 2. First and
218 foremost, there has been a significant elongation, 0.08 Å, of the formally C1=S2 thione bond
219 in SMHDH2 once this atom is complexed to Cu. Secondly, there has been a notable
220 reduction, 0.05 Å, of the amine C1–N2 bond in SMHDH2, consistent with the formation of
221 an imine bond in the complex. The reorganisation of electron density around the NCS2
222 residue results in contraction of the S–C–S angle with concomitant expansion in the angles
223 involving the double bonded nitrogen atom, Table 2. The crystal packing diagram reveals that
224 the molecules stack in columns aligned along the b-axis with no directional interactions
225 between them.
226
The molecular structure of CuSBHD (Fig. 4a and Table 2) shows the same features as
227 CuSMHD, consistent with the notion that the nature of the S-bound substituent, methyl or
228 benzyl, does not exert a significant effect upon the structure. This observation is highlighted
229 in the overlay diagram shown in Fig. 4b. The seven-membered ring in CuSBHD has a half-
230 chair conformation with the C4 atom lying 0.9970(19) Å above the plane defined by the Cu,
231 N2, N3, C2, C3 and C5 atoms; r.m.s. = 0.2078 Å with maximum deviations 0.2432(7) Å, for
232 Cu, and -0.3012(10) Å, for N3 indicating that this chelate ring is more distorted than in
233 CuSMHD. The S1-containing five-membered chelate ring is an envelope with the flap atom
234 being the Cu atom which lies 0.7703(19) Å above the least-squares plane defined by the
235 remaining four atoms (r.m.s. = 0.0004 Å), as is the S3-chelate ring with the Cu atom lying
236 0.423(2) Å above the plane of the four remaining atoms (r.m.s. = 0.0027 Å). In CuSBHD, the
237 angle between the two five-membered chelate rings is 48.93(4)º, and the range of angles
238 subtended at the Cu atom is 84.35(4)º for S3–Cu–N3 to 175.08(4)º for S1–Cu–N3, marginally
239 broader than observed in CuSMHD, Table 2.
9

ACCEPTED MANUSCRIPT
240
241
242 Fig. 4. (a) The structure of CuSBHD, showing its atom-labelling scheme, and (b) overlay
243 diagram of CuSMHD (red image) and CuSBHD (blue). The complex molecules are
244 overlapped so that the S1, Cu and S3 atoms are coincident.
245
246 The molecular packing of CuSBHD also resembles that of CuSMHD in that columns of
247 molecules are evident, aligned along the a-axis. However, the CuSBHD molecules are linked
248 by a combination of C–H…S and C–H…π (phenyl) interactions. Geometric parameters
249 characterising the intermolecular interactions operating in the crystal structure of CuSBHD:
250 C4–H4a...S4i = 2.83 Å, C4...S4i = 3.7647(16) Å, and angle at H4a = 158º for i: -x, -½+y, -½-
251 z; C4–H4b...S3ii = 2.84 Å, C4...S3ii = 3.6946(19) Å, and angle at H4b = 145º for ii: -x, 1-y, -
252 z; C18–H18...S4iii = 2.86 Å, C18...S4iii = 3.6811(17) Å, and angle at H18 = 145º for iii: x,
253 1½-y, -½+z; C3–H3a...Cg(C17–C22)i = 2.89 Å, C3...Cg(C17–C22)i = 3.6538(17) Å, and
254 angle at H3a = 135º.
255
256
The four-coordinate structures described here for CuSBHD and CuSMHD, with the
257 dianions in the iminothiolate form, are consistent with literature precedents [15, 16, 49-52].
258
10

ACCEPTED MANUSCRIPT
259 Table 1
260 Crystallographic and refinement details for SBPY, SMHDH2, CuSMHD and CuSBHD
261
262 Compound
263 Formula
264 Formula weight
265 Crystal colour/habit
266 Crystal dimensions/mm
267 Crystal system
268 Space group
269 a/Å
270 b/Å
271 c/Å
272 α/°
273 β/°
SBPY
C14H16N2S2
276.41
SMHDH2
C10H18N4S4
322.52
Yellow needle
0.03 x 0.06 x 0.24
triclinic
CuSMHD
C10H16CuN4S4
384.05
Black prism
0.04 x 0.12 x 0.18
monoclinic
C2/c
24.6441(8)
7.9100(2)
16.8972(6)
90
111.167(4)
90
CuSBHD
C22H24CuN4S4
536.23
Black prism
0.11 x 0.22 x 0.28
monoclinic
P21/c
10.7937(1)
18.8337(2)
11.8412(2)
90
103.410(1)
90
Colourless plate
0.04 x 0.20 x 0.21
monoclinic
P21/c
9.2991(4)
15.9635(8)
9.4848(5)
90
96.155(1)
90
1399.87(12)
4
¯
P1
5.1646(5)
7.2792(8)
10.7840(12)
100.652(9)
90.751(9)
107.305(10)
379.39(7)
1
274 γ/°
275 V/Å3
3071.62(18)
8
1.661
2341.51(5)
4
1.521
276
277 Dc/g cm-3
Z
1.312
1.412
278 F(000)
584
0.364
170
5.662
1576
1.956
1108
1.308
279 ꢀ/mm-1
280 Measured data
281 Radiation
19199
MoKα
4869
CuKα
19072
MoKα
58909
MoKα
282 θ range/°
283 Unique data
2.5–27.5
3207
4.2–71.6
1455
2.6–27.5
3490
2.2–27.5
5353
284 Observed data (I ≥ 2.0σ(I))
285 R, obs. data; all data
286 a, b in weighting scheme
287 Rw, obs. data; all data
288 Residual electron density
289 peaks/e Å3
2722
1197
3291
4827
0.032; 0.041
0.030, 0.621
0.071; 0.075
0.046; 0.055
0.078, 0.048
0.121; 0.130
0.019, 0.021
0.032, 2.432
0.054, 0.055
0.028, 0.032
0.050, 1.156
0.077, 0.080
0.30, -0.26
0.43, -0.28
0.37, -0.31
0.67, -0.47
11

ACCEPTED MANUSCRIPT
290 Table 2
291 Selected geometric parameters (Å, º) for SMHDH2, CuSMHD and CuSBHD
292 ___________________________________________________________________________
293
294 Compound
SMHDH2
CuSMHD
CuSBHD
295 ___________________________________________________________________________
296 Parameter
297 Cu–S1
298 Cu–S3
299 Cu–N2
300 Cu–N3
–
–
–
–
2.2480(4)
2.2523(4)
2.0555(12)
1.9792(12)
2.2458(4)
2.2659(4)
2.0704(13)
1.9927(14)
301 C1–S1, S2
302 C6–S3, S4
303 N1–C1, N2–C2
304 N1–N2
1.655(3), 1.763(3) 1.7373(14), 1.7579(14) 1.7354(16), 1.7573(17)
1.7380(14), 1.7533(14) 1.7404(16), 1.7560(16)
1.339(3), 1.281(3) 1.2892(19), 1.2914(18) 1.286(2), 1.292(2)
–
1.391(3)
1.4182(16)
1.4181(18)
305 C5–N3, C6–N4
306 N3–N4
–
–
–
–
–
–
–
–
1.2872(18), 1.2887(18) 1.285(2), 1.286(2)
1.4073(16)
92.795(14)
84.89(3)
1.4019(18)
91.655(15)
85.26(4)
307 S1–Cu–S3
308 S1–Cu–N2
309 S1–Cu–N3
310 S3–Cu–N2
311 S3–Cu–N3
312 N2–Cu–N3
313 C1–N1–N2
314 C2–N2–N1
315 C5–N3–N4
316 C6–N4–N3
317
164.90(4)
148.04(3)
85.76(4)
175.08(4)
148.89(4)
84.35(4)
104.21(5)
113.22(11)
112.01(11)
115.20(11)
112.83(11)
99.65(5)
119.2(2)
117.0(2)
–
–
113.38(12)
112.92(13)
115.88(13)
112.70(12)
318 S1–C1–S2
319 N1–C1–S1
320 N1–C1–S2
321 S3–C6–S4
322 N4–C6–S3
323 N4–C6–S4
123.83(16)
122.7(2)
113.51(18)
–
–
–
113.87(8)
127.49(11)
118.63(11)
113.94(8)
127.17(11)
118.89(10)
111.63(9)
128.34(13)
120.01(12)
113.16(9)
126.59(12)
120.23(12)
324 ___________________________________________________________________________
325
326
12

ACCEPTED MANUSCRIPT
327 2.3. Solution characterization of the complexes
328 The UV-Vis absorption spectra of the compounds in DMSO (25 ꢀM and 1 mM) are
329 given in Figure 5. Both complexes showed π→π* and n→π* intra-ligand transitions at 272
330 nm, 295 nm and 338 nm and a d-d band at approximately 600 nm that can be attributed to
331 Jahn-Teller distortion from square planar geometry [47]. The presence of the S→Cu(II)
332 LMCT band at ~400 nm in the spectra of both metal complexes is strong evidence that the
333 metal ion is coordinated to sulphur [43, 53].
334
335
336 Fig. 5. UV-Vis spectra for all compounds (25 ꢀM, DMSO). The insert shows the d-d band of
337 the Cu complexes (1 mM solutions).
338
339
That the ligands and their corresponding complexes be stable at physiological pH is
340 an important prerequisite for evaluation of their biological activity. The molar conductance
341 values for the complexes in DMSO were in the range 12-13 ꢁ−1 cm2 mol−1, indicating that
342 there is essentially no dissociation in that solvent [54]. To more precisely evaluate their
343 stability, reverse phase HPLC experiments were performed. The ligands and their complexes
344 were eluted on a C18-column with a gradient increase in concentration of CH3CN in H2O
345 (from 5% to 100% over 30 min), containing 0.1 % TFA to maintain pH. The compounds were
346 detected at 220 and 280 nm. The chromatograms of the purified ligands showed three peaks
347 that could correspond to the expected ligand, the hydrolyzed hydrazone and the pyrrole
348 byproduct whereas the complexes showed only the single peak of the copper complexes (see
349 Supplementary Data). It is noteworthy that the hydrazone bond stability significantly
350 increased upon metal-complexation under acidic conditions suggesting that complexation
13

ACCEPTED MANUSCRIPT
351 could be used as a means to protect the ligand from degradation that might occur in biological
352 systems before free ligand could reach its target.
353
354 2.4. Electron Paramagnetic Resonance (EPR)
355
The EPR spectra recorded in DMF (Figure 6) are typical of distorted square planar
356 Cu(II) complexes having axial symmetry with the unpaired electron mainly in the dx
2-y2
357 orbital. The spectra also exhibit partially resolved superhyperfine features. The g∥ values for
358 all the complexes are similar to those previously reported for analogous Cu(II)N2S2
359 complexes [16, 24, 55]. Kivelson and Nieman [56, 57] suggested that g|| values higher than
360 2.3 are indicative of a predominantly ionic character for metal-ligand bonds, whereas g||
361 values smaller than 2.3 reveal metal-ligand bonds of predominantly covalent character, as is
362 the case here (see Table 3). In addition, the relatively small g value (~2.20) suggests a strong
∥
363 nitrogen character in the singly occupied molecular orbital. The equation below [42, 58] was
364 used to calculate the molecular orbital coefficients, α2 (in-plane σ-bonding):
365
α2 = (A / 0.036) + (g − 2.0036) + 3 / 7 (g − 2.0036) + 0.04
∥ ∥
366 An α2 value of 0.5 indicates complete covalent bonding, while 1.0 suggests complete ionic
367 bonding. The observed value of 0.64 for the present complexes is evidence that these copper
368 complexes have some covalent character, as suggested above. EPR spectroscopy is sensitive
369 to angular distortions at the Cu(II) centre, particularly those involving distortions from planar
370 to tetrahedral geometry which generally result in a decrease in A and an increase in g [53].
∥
∥
371 The empirical factor f (= g||/Α||) [59, 60] is a measure of deviation from idealized geometry. Its
372 value ranges between 105 and 135 cm for square planar complexes, depending on the nature
373 of the coordinated atoms, while, for tetrahedral structures, values from 160 to 242 cm suggest
374 a moderate to considerable tetrahedral distortion. In solution as well as in the solid, CuSBHD
375 displays a slightly higher degree of tetrahedral distortion than CuSMHD. Both are slightly
376 more distorted than analogues, probably due to their extended carbon backbones [16, 24, 55].
377
14

ACCEPTED MANUSCRIPT
378
379 Fig. 6. EPR spectra of CuSBHD and CuSMHD recorded at a microwave frequency 9.50
380 GHz, power 0.25 mW, modulation amplitude 0.2 mT, modulation frequency 100 kHz, and
381 time constant 164 ms, at 50 K. Samples were prepared in DMF (1 mM).
382
383 Table 3
384 EPR parameters measured from the spectra of CuSBHD and CuSMHD in DMF
385
f[b]
α2
[a]
g||
Α
||
g
CuSMHD 2.15 2.06
CuSBHD 2.16 2.05
460 (153) 141 0.64
451 (150) 143 0.64
386
387
[a] MHz (x104 cm-1) [b] cm
388 2.5. Electrochemistry
389 As redox properties have been linked to SOD and anticancer properties of metal
390 complexes [61, 62], we describe herein the electrochemical properties of Cu(II) bis(dithio
391 carbazate). Figure 7 shows the profile of the Cu(II) complexes obtained with SMHDH2 and
392 SBHDH2 at scan rate 100 mV s−1. Both complexes undergo an irreversible one-electron
393 reduction at Epc = -0.328 and -0.285 V/(AgCl/Ag and Fc+/Fc = 0.563 V), respectively,
394 coupled with oxidation at Epa = 0.069 and 0.129 V/(AgCl/Ag). These waves can be assigned
395 to the irreversible oxidation/reduction of Cu(II)/Cu(I) [52]. The ligands were found to be
396 redox innocent. The irreversible nature of the copper-centered redox waves contrasts with the
397 quasi-reversible reduction previously reported for CuATSM and CuAATSM analogues [50,
15

ACCEPTED MANUSCRIPT
398 51]. The loss of reversibility observed in this work is most likely related to differences in the
399 geometric rearrangement about the Cu(II)/Cu(I) ions in this ligand system with two carbon
400 atoms between the two hydrazone functions. The Cu(II)/Cu(I) redox potentials of CuSMHD
401 and CuSBHD are also more positive than those reported for previous examples. The ease of
402 deformation seems to favour reduction. The difference in redox potential between CuSMHD
403 and CuSBHD may be rationalized by induction due to the stronger electron-donating effect of
404 the methyl group compared to benzyl [63].
405
406
As mentioned above, the oxidation proceeding at higher positive potential has
407 previously been assigned to the copper(III/II) redox couple. It is interesting to note the
408 occurrence of an additional peak, which can be attributed to the reduction of a species
409 produced by the second oxidation. However, the nature of this oxidized complex has not been
410 determined.
411
412
413 Fig. 7. Cyclic voltammograms of the Cu complexes (1.7 mM) in anhydrous deoxygenated
414 DMF containing 0.1 M tetrabutylammonium hexafluorophosphate as the supporting
415 electrolyte. Working electrode: glassy carbon; counter electrode: Pt wire; reference electrode:
416 AgCl/Ag, scan rate: 100 mV/s. All sweeps were initiated in the direction of the arrow.
16

ACCEPTED MANUSCRIPT
417 Table 4
418 Electrochemical data for CuSMHD and CuSBHD vs AgCl/Ag.
419
Cu(II)/Cu(I) Cu(III)/Cu(II)
Epc/V Epa/V Epc/V Epa/V
CuSMHD -0.328 0.069 0.195 0.899
CuSBHD -0.285 0.129 0.357 0.870
420
34.21 Biological evaluation
422 3.1. Antibacterial activity
423
The free Schiff bases and their metal complexes were tested for their ability to inhibit
424 the growth of ten strains of Gram-negative and Gram-positive bacteria (Table 5). The effects
425 of a membrane permeabilizing agent and efflux pumps were investigated in an attempt to
426 correlate the activity of the compounds with their penetration into the bacteria and the
427 resistance mechanisms of the bacteria.
428
429
One of the limitations of this class of compounds is their poor solubility in aqueous
430 solution. The universal solvent DMSO has been used in many studies to pre-dissolve the
431 compounds for biological assays. However, it has been shown that DMSO solutions (1% to
432 10%) considerably affect the growth of fungi and cancerous cells, and, at 15%, DMSO
433 effectively eliminates the growth of certain bacteria [64-66]. DMSO has also been reported to
434 decrease membrane rigidity, thus facilitating membrane diffusion of exogenous species [67-
435 70]. As DMSO is used in this work to encourage dissolution of the compounds and since
436 there is no rule of thumb for the amount of DMSO to be used for antibacterial assay, it was
437 essential to examine the influence of DMSO concentrations on the growth curve of the
438 selected bacterial strains. The minimum inhibititory concentration (MIC) values were
439 determined at 0.5% and 5% (v:v) DMSO. We found that the growth of A. baumannii and P.
440 aeroginosa is inhibited by DMSO at a concentration of only 5% thus preventing
441 determination of MIC at this concentration. The growth of E. coli and E. aerogenes (see
442 Supplementary Data) was also affected by DMSO at 5%. Differences were observed between
443 MIC values against the mutated strains E. coli AcrAB- and E. aerogenes 298 TolC- obtained
444 in the presence of 0.5% or 5% DMSO for certain molecules, in particular, CuSMHD.
445 Additional MIC values determined for CuSMHD using DMSO 50%, 30% and 20% (2.5%,
17

ACCEPTED MANUSCRIPT
446 1.5% and 1% final v:v DMSO) were all higher than 128 µM while with 5% of DMSO, the
447 MIC values were in the range of 1-2 and 0.5-1 µM against E. coli AcrAB- and E. aerogenes
448 298 TolC-, respectively. Because of the effect of DMSO on bacterial growth, we are unable to
449 confirm whether the value truly reflects the specific antimicrobial activity of the tested
450 compound and not a synergetic effect involving the compound and DMSO. MIC values
451 recorded using 0.5% DMSO are used for discussion of the role of membrane permeabilizing
452 agents and efflux pumps since DMSO at this concentration was shown not to interfere with
453 bacterial growth.
454
455
Since it has been reported that low permeability of the outer membrane and the
456 efficiency of efflux pumps [3, 4] are prime factors limiting intracellular activity of potential
457 antimicrobial compounds, it is expected that the presence of a substance known to increase
458 membrane permeability, such as polymyxin B nonapeptide (PMBN) [71], would act
459 synergistically to improve uptake of the compounds under study and consequently would
460 affect their antimicrobial efficiency in a positive manner. The compounds were tested in the
461 presence and absence of sub-inhibitory concentrations (1/5 of the MIC value) of PMBN. In
462 the absence of PMBN only SMHDH2 was active against the strains tested (MIC ≥ 64 µM).
463 SMHDH2 showed moderate activity against S. aureus. However, up to 3-fold improvement
464 in activity (MIC values) was observed for the organic compounds SMHDH2, SBHDH2 and
465 SBDP in the presence of PMBN against both Gram-negative and Gram-positive bacteria.
466 These results imply that apparent lack of activity was due to the inability of the compounds to
467 efficiently permeate the bacterial membrane. SMHDH2 showed a broad range of moderate
468 activity against various strains. It was most effective against E. coli AcrAB-, A. baumannii, P.
469 aeruginosa and S. aureus (MIC values ~16 µM), thus making it a potential antimicrobial
470 agent in the presence of PMBN. These results are not unexpected since many previous
471 reports have shown that the biological activity of dithiocarbazato compounds can be greatly
472 modified by the presence of different substituents. The enhanced activity observed for
473 SMHDH2 compared to its S-benzyl analog is consistent with the observation that the Schiff
474 base prepared from 2-benzoylpyridine with S-methyldithiocarbazate (SMDTC) was a highly
475 effective inhibiter of E. coli and S. aureus whereas that prepared with the S-
476 benzyldithiocarbazate (SBDTC) analog showed no activity [72].
477
18

ACCEPTED MANUSCRIPT
478
479
480
Table 5
Antibacterial activity.
Minimum Inhibitory Concentration (MIC) (µM)
Gram-
Compound
Gram+
A.
K.
P.
S.
E. coli
E. aerogenes
S. aureus
baumannii pneumoniae
aeruginosa
enterica
AG100
WT
5
AG100A
AcrAB-
5
EA289
AcrAB+
0.5
EA294
AcrAB-
0.5
EA298
TolC-
0.5
>128-128 >128
ATCC
19606
0.5
ATCC
11296
0.5
PA01
0.5
SL696
SA1199
% DMSO
SMHDH2
0.5
0.5
5
5
5
5
0.5
>128 >128
64 32
5
0.5
5
>128
>128
>128
>128
16 >128-128 64
>128 128-64
64
128
64
128-64
32
64-32
>128
64
+PMBN
32
32
16
128
32
16
64
32-16
>128
>128
128
16-8
>128
>128
>128
32-16 64-32
128-32
64
>128
>128
>128
>128
>128
128
>128
>128
>128
>128
1-2
>128
>128
>128 >128
>128
>128
>128
>128
>128
>128
>128
>128
>128
>128 >128 >128
>128 >128 >128
>128
>128
CuSMHD
+PMBN
>128
>128
>128
>128-128 0.5-1
>128-128 64
SBHDH2
+PMBN
128
>128 128-64
>128 >128 >128 64-32
>128
64
>128
64
128-32 32-16 >128-128 64
>128-64 16-4
128-64
128-64 32-16
64-32
>128
64
16
128
>128-64
64
>128
>128
>64
64
>128
>128
>128
>128
>64
>128
>128
>64
>128
>128
>64
>128
>128
>64
>128 >128
>128 >128
>128
>128
>64
>128
>128
>64
>128
>128
>64
>128
>128
>128
>128
>128
>128
>128
>128
>128
>128
>128
>128
64
>128
>128
>128
>128
>128
>128 >128 >128
>128 >128 >128
>128
>128
CuSBHD
+PMBN
SBPY
+PMBN
Cu(Ac)2
+PMBN
>64
64
>64
32
>128 >128
128 128-64
64
>128
>128
32
>128
>128
16
>128
>128
>64
>128
>128
4
32
>128
>128
4
>128-128
>128
>128
>128-128
>128
>128
64 64-32 128-64 128
>128 >128 >128
>128 >128 >128
>128 >128
>128 >128
>128
>128
>128
>128
>128
>128
481
482
483
Colour code: MIC values or average MIC values ≥ 64 µM = red, ≤10 µM = green, in between 64 µM and 10 µM = colourless. MIC values
higher than 64 µM indicate poor activity.
19

ACCEPTED MANUSCRIPT
484
The role of efflux pumps was investigated using pump-deleted strains of Gram-negative
485 E. coli and E. aerogenes. Both SMHDH2 and SBPY seemed more active (16 µM and 32 µM,
486 respectively) towards the isogenic strain in which the AcrAB pump was deleted as compared
487 to wild-type E. coli. No significant activity was observed for SMHDH2 in the presence or
488 absence of efflux pump for E. aerogenes. SBPY, on the other hand, showed differences in the
489 MDR clinical isolate EA289 over expressing AcrAB efflux pump and on its AcrAB- and
490 TolC- derivatives, EA294 and EA298, with improvement in MIC from > 64 µM to 32 µM.
491 These results confirmed that SBPY and SMHDH2 are recognized by the efflux pumps and
492 expelled from the bacteria, thus limiting their bioactivity. Both SMHDH2 and SBHDH2
493 showed activity towards Gram-positive S. aureus. Typically, antibacterial molecules are more
494 active toward Gram-positive than Gram-negative bacteria [73, 74], because the additional
495 outer membrane of the latter organisms impairs or slows down the drug uptake. It has often
496 been reported in the literature that bioactivity of a ligand is enhanced by metal complexation
497 [49, 60], but in our case, the formation of the copper complexes induces a loss of antibacterial
498 potency of the compounds. A similar loss in activity was previously reported for
499 palladium(II) and platinum(II) complexes of acetone Schiff bases [7]. This can be explained
500 by the lower solubility of the metal complexes or by the lower stability of the hydrazone
501 moiety in the case of the free ligands. As noted above, hydrolysis of the ligands occurring in
502 aqueous solution at specific pH, may lead to reactive products that can be responsible for the
503 toxicity. Since these observations strongly suggest that enhancement of antibacterial activity
504 can be expected as a result of increasing solubility and membrane diffusion, efforts are
505 currently being made to significantly improve the aqueous solubility of these compounds.
506
507 3.2. Cytotoxic assay
508
Cytotoxicity was evaluated in vitro against two breast cancer cell lines MDA-MB-231
509 (human breast carcinoma cells not expressing nuclear estrogen receptors) and MCF-7 (human
510 breast carcinoma cells expressing nuclear estrogen receptors). Measurements were carried out
511 using MTT assay [75] which is based on the metabolic reduction of tetrazolium salt to form
512 water insoluble formazan crystals. DMSO was used as a negative control in the assay. There
513 was no perceptible precipitation of the compounds. The concentrations required to inhibit the
514 growth of cancer cells by 50% (IC50) are given in Table 6.
515
516
20

ACCEPTED MANUSCRIPT
517 Table 6
518 Cytotoxic assay results.
519
IC50 (ꢀM)
MCF-7
MDA-MB-231
SMHDH2
SBHDH2
CuSMHD
CuSBHD
Tamoxifen
138.90
9.69
2.60
1.49
11.20
9.61
1.05
2.34
0.71
13.40
520
521
Both ligands displayed at least 9-fold better toxicity towards MDA-MB-231, indicating
522 that ligand toxicity is not only mediated by nuclear estrogen receptors. The stronger toxicity
523 exhibited by SBHDH2 may be related to facilitated diffusion into cells resulting from its
524 comparatively higher lipophilicity [12]. Complexation of the Schiff base ligands with
525 copper(II) has been found to produce synergistic effects on the antiproliferative activities of
526 some parent ligands [76] and here the complexes showed a marked cytotoxicity with IC50
527 values < 5.0 µM towards both cell lines. Like the ligands, the complexes are also more active
528 towards MDA-MB-231 cells, suggesting that their toxicity does not involve estrogen
529 receptors. For both cell lines, the benzyl substituted complex CuSBHD showed slightly
530 higher IC50 values. Although a definitive structure-activity relationship cannot be deduced
531 since only a limited number of compounds was tested, observations indicate that the
532 enhanced activity of CuSBHD may be linked to higher cellular uptake resulting from
533 increased lipophilicity as suggested above for the ligand. Redox potential may also be a
534 discriminating factor since its higher redox potential means that reduction of Cu(II) is easier,
535 and consequently a higher content of Cu(I) could be generated. Cu(I) is prone to participate in
536 Fenton-type reactions that produce reactive oxygen species (ROS), which can damage
537 biomolecules within cells [62].
538
539
4. Conclusions
540
This study provides new insight into the structural, electrochemical and biological aspects
541 of macroacyclic Cu(II) complexes derived from S-substituted dithiocarbazate. All the
542 compounds exhibited good cytotoxicity toward MDA-MB-231 and MCF-7 breast cancer cell
543 lines. Their low antibacterial activity can be related to poor bacterial penetration and limited
544 solubility, both of which should be amenable to improvement by further functionalization of
21

ACCEPTED MANUSCRIPT
545 the ligands. The anticancer activity of the ligands was enhanced by complexation with Cu(II).
546 Expanding the carbon backbone between the hydrazone moieties resulted in further distortion
547 from square planar geometry in both the solid state and in solution as well as a positive shift
548 in the Cu(II)/Cu(I) reduction potential. A higher reduction potential could be related to the
549 production of reactive oxygen species resulting in the enhanced bioactivity observed in this
550 present work. Taking into consideration the serious side effects and the poor efficacy of
551 clinical reference drugs, as well as the appearance of resistance during treatment, these
552 complexes are potentially useful lead candidates for the development of new therapeutic
553 agents to treat cancer and bacterial infections. In addition, this work underlines the need to
554 consider the concentration of DMSO used to dissolve compounds, since DMSO may pre-
555 sensitize the bacterial cells to the tested compounds. Reduced uptake of compounds due to
556 low permeability of the outer cell membrane and the efficiency of efflux pumps were also
557 shown to be issues to be addressed in subsequent studies. With these considerations in mind,
558 our group is attempting to improve antimicrobial and anticancer activities of compounds in
559 this family by exploring the design and synthesis of a new generation of S-substituted
560 dithiocarbazate derivatives and their metal complexes.
561
5. Experimental
562 5.1. Materials-instrumentation-physical measurements
563 All chemicals and solvents were of analytical grade and were used as received.
564 Chemicals: Potassium hydroxide (Merck), hydrazinium hydroxide (Merck), carbon disulfide
565 (Sigma Aldrich), 2,5-hexanedione (Merck), and copper(II) acetate monohydrate (Analar). The
566 IR spectra were recorded in the range of 550-4000 cm−1 on a Perkin-Elmer 100 series FT-IR
567 spectrophotometer in ATR mode. Magnetic susceptibility was measured with a Sherwood
568 MSB-AUTO instrument at room temperature. All susceptibilities were corrected for the
569 diamagnetic contribution using Pascal’s constant. Microanalyses were carried out using either
570 a LECO CHNS-932 analyzer or performed at the CNRS (Gif-sur-Yvette and Vernaison,
571 France). The molar conductance of a 10-3 M solution of each metal complex in DMSO was
572 measured at 29°C using a Jenway 4310 conductivity meter and a dip-type cell with platinized
573 electrode. The UV-Vis spectra were recorded on a Cary 300 Bio spectrophotometer (200-800
1
13
574 nm) or Perkin Elmer Lambda 45 with a 1 cm optical path quartz cuvette. H NMR and C
575 NMR spectra were recorded using Bruker DRX300 spectrometers. The chemical shifts
576 (δ/ppm) were calibrated relative to residual solvent signals. Electrospray-ionization mass
22

ACCEPTED MANUSCRIPT
577 spectra (ESI-MS) were recorded with a Finnigan Mat 95S in the BE configuration at low
578 resolution. Electron paramagnetic resonance (EPR) spectra were recorded on an X-band
579 Bruker Elexsys 500 spectrometer equipped with a continuous flow helium cryostat (Oxford
580 Instruments) and a temperature control system. The field modulation frequency was 100 kHz.
581 The spectra were all recorded under nonsaturating conditions. Cyclic voltammetry (CV)
582 measurements were recorded under argon using a 620C electrochemical analyzer (CH
583 Instruments, Inc). The working electrode was a glassy carbon disk; a Pt wire was used as
584 counter electrode and the reference electrode was AgCl/Ag. Immediately before
585 measurements were taken, the working electrode was carefully polished with alumina
586 suspensions (1, 0.3 and 0.05 ꢀm, successively), sonicated in an ethanol bath and then
587 carefully washed with ethanol. Test solutions were prepared using 100 µL of the complexes
588 in anhydrous deoxygenated DMF (0.01 M) with 0.5 mL of tetrabutylammonium
589 hexafluorophosphate (0.1 M) as the supporting electrolyte (total volume 0.6 mL). Ferrocene
590 was used as internal reference: the ferrocinium/ferrocene one-electron redox process occurs
591 at E1/2 = 0.508 V (DMF) vs AgCl/Ag with scan rate = 0.1 V/s. RP-HPLC analysis was carried
592 out using a Waters HPLC system that consisted of a combination of a dual wavelength UV-
593 Vis absorbance detector (Waters 2487) and a binary pump (Waters 1525) equipped with an
594 analytical cell for reaction monitoring or purity checking connected to Breeze software.
595 Analytical HPLC measurements were performed using an ACE C18 column (250 × 4.5mm)
596 packed with spherical 5 µm particles of 300 Å pore size. Experiments were carried out at a
597 flow rate of 1 mL min-1 at room temperature. Injection volume was 50 µL. Sample
598 concentration was approximately 1 mg mL-1.
599
600 5.2. Preparation of ligands and metal complexes
601 5.2.1 Synthesis of SBHDH2
602
A modification of the method described by Ali et al. [35] was used to synthesize
603 SBHDH2. 2,5-Hexanedione (0.587 mL, 0.005 mol) was added to a hot solution of S-
604 benzyldithiocarbazate (1.983 g, 0.01 mol) in absolute ethanol (150 mL) and the mixture was
605 further heated for 5 min. The white precipitate formed was immediately filtered off, washed
606 with cold ethanol and dried in vacuo over silica gel to yield the expected Schiff base (yield
607 0.997 g, 42%). Elemental analysis for C22H26N4S4: Calc. C 55.66, H 5.52, N 11.80; Found C
608 54.79, H 5.59, N 11.75. 1H NMR (300 MHz, DMSO-d6) δ 12.18 (s, 2H), 7.39 -7.20 (m,
13
609 10H), 4.40 (s, 4H), 1.96 (s, 6H). C NMR (75 MHz, DMSO-d6) δ 197.16, 158.26, 137.15,
610 129.15, 128.41, 127.05, 37.56, 34.05, 17.74. IR: ν (cm-1) = 3147 (m, b), 1640 (w), 1054 (s),
23

ACCEPTED MANUSCRIPT
611 981 (m), 828 (m). UV-Vis in DMSO: λmax nm (log
ε
in Lꢂmol–1ꢂcm–1) = 276 (4.32), 308 (4.41),
612 ≈360 (3.32, sh). RP-HPLC: RT (min) = 15.3, 18.3, 22.4. Molar conductivity: Λ (ohm-1
613 cm2mol-1) = 6.86.
614
615 5.2.2 Synthesis of SMHDH2
616
2,5-Hexanedione (0.587 mL, 0.005 mol) was added to a solution of S-methyldithio
617 carbazate, SMDTC (1.222 g, 0.01 mol) dissolved in hot ethanol (150 mL). The mixture was
618 heated while being stirred to reduce the volume by half. The white precipitate formed from
619 the mixture kept at 4°C overnight was filtered off, washed with cold ethanol and dried in
620 vacuo over silica gel (yield 1.129 g, 70%). The compound was recrystallized from methanol
621 and crystals suitable for X-ray diffraction analysis were obtained from the same solvent.
622 Elemental analysis for C10H18N4S4: Calc. C 37.24, H 5.63, N 17.37; Found C 37.86, H 4.87,
623 N 17.84. 1H NMR (300 MHz, DMSO-d6) δ 12.13 (s, 2H), 2.57 (s, 4H), 2.43 (s, 6H), 2.00 (s,
13
624 6H). C NMR (75 MHz, DMSO-d6) δ 198.95, 157.63, 33.97, 17.77, 16.94. IR: ν (cm–1) =
625 3111 (m, b), 1628 (m), 1046 (s), 988 (m), 827 (m). UV-Vis in DMSO: λmax nm (log ε in
626 Lꢂmol–1ꢂcm–1) = 276 (4.25), 305 (4.37), ≈360 (2.75, sh). RP-HPLC: RT (min) = 6.4, 11.1, 18.7.
627 Molar conductivity: Λ (ohm-1 cm2 mol-1) = 3.58
628
629 5.2.3. Synthesis of CuSBHD
630
CuSBHD was prepared by adding copper(II) acetate monohydrate (0.020 g, 0.0001
631 mol) in acetonitrile (20 mL) to a solution of SBHDH2 (0.047 g, 0.0001 mol) in acetonitrile
632 (150 mL) at room temperature. The solution was stirred for an hour and then concentrated to
633 reduce the volume before being placed at 4°C overnight. The product was filtered off and
634 recrystallized from acetonitrile (yield 0.039 g, 73%). Black crystals of diffraction quality
635 were crystallized from acetonitrile after several days through slow evaporation at 4°C.
636 Elemental analysis for C22H25CuN4S4: Calc. C 49.27, H 4.51, N 11.85; Found C 49.40, H
637 4.63, N 10.46. ESI-MS: m/z = [M+H]+ Calc. 536.04, Found 536.02; [M+Na]+ Calc. 558.02,
638 Found 558.01; [M+K]+ Calc. 573.99, Found 573.98; [2M+3H]+ Calcd. 1073.08, Found
639 1073.04. IR: ν (cm–1) = 1629 (m), 1606 (w), 992 (s), 955 (s), 857 (m). UV-Vis in DMSO: λmax
640 nm (log
ε
in Lꢂmol–1ꢂcm–1) = 275 (4.37), ≈294 (4.26, sh), ≈340 (4.01, sh), ≈400 (3.55, sh),
641 ≈600 (2.45, sh). RP-HPLC: RT (min) = 28.5. Magnetic moment: ꢀ (B.M.) = 1.48. Molar
642 conductivity: Λ (ohm-1 cm2 mol-1) = 13.01.
643
24

ACCEPTED MANUSCRIPT
644 5.2.4. Synthesis of CuSMHD
645 CuSMHD was prepared by adding copper(II) acetate monohydrate (0.200 g, 0.001
646 mol) in methanol (20 mL) to a hot solution of SMHDH2 (0.322 g, 0.001 mol) in methanol
647 (100 mL). The reaction was heated until the volume reduced to half and then placed at 4°C
648 overnight. The product was filtered off and recrystallized from acetonitrile to afford 0.296 g
649 of CuSMHD (yield 77%). Black crystals of diffraction quality crystallized from acetonitrile
650 after several weeks through slow evaporation at room temperature. Elemental analysis for:
651 C10H17CuN4S4: Calc. C 31.27, H 4.20, N 14.59; Found C 31.35, H 4.24, N 14.64. ESI-MS:
652 m/z = [M + H]+ Calc. 383.97, Found 383.96; [M+Na]+ Calc. 405.96, Found 405.94; [M+K]+
653 Calc. 421.93, Found 421.92. IR: ν (cm–1) = 1628 (m), 1611(w), 1000 (s), 964 (s), 821 (m).
654 UV-Vis in DMSO: λmax nm (log ε
in Lꢂmol–1ꢂcm–1) = 273 (4.34), ≈294 (4.24, sh), ≈340 (3.99,
655 sh), ≈400 (3.49, sh) ≈600 (2.43, sh). RP-HPLC: RT (min) = 23.3. Magnetic moment: ꢀ (B.M.)
656 = 1.66. Molar conductivity: Λ (ohm-1 cm2 mol-1) = 12.80.
657
658 5.2.5. Synthesis of SBPY
659
SBPY was a side product from the initial attempt to synthesize SBHDH2 using
660 prolonged heating. Single crystals of diffraction quality were obtained from DMSO and
661 analyzed by single crystal X-ray diffraction. ESI-MS: m/z = [M + H]+ Calc. 277.08, Found
1
662 277.08; [M + Na]+ Calc. 299.07, Found 299.06. H NMR (300 MHz, DMSO-d6): δ (ppm) =
13
663 12.29 (s, 1H), 7.45 – 7.20 (m, 5H), 5.69 (s, 2H), 4.45 (s, 2H), 2.00 (s, 6H). C NMR (75
664 MHz, DMSO-d6): δ (ppm) = 204.09, 136.30, 129.02, 128.55, 127.43, 126.50, 104.34, 38.17,
665 10.99. IR: ν (cm–1) = 3264 (m), 2917 (w), 1055 (s), 972 (w), 828 (w). UV-Vis in DMSO: λmax
666 nm (log ε
in Lꢂmol–1ꢂcm–1) = 282 (4.02). RP-HPLC: RT (min) = 22.3.
667
668 5.3. Biological studies
669 5.3.1. Antimicrobial assay
670 5.3.1.1. Bacterial strains, culture media and chemicals
671
The bacteria used in this study are listed in Table 7. The microorganisms studied
672 included reference (from the American Type Culture Collection) and clinical (Laboratory
673 collection) strains of Gram-negative bacteria Escherichia coli, Enterobacter aerogenes,
674 Acinetobacter baumannii, Klebsiella pneumoniae, Pseudomonas aeruginosa and Salmonella
675 enterica serotype Typhimurium as well as Gram-positive strain Staphylococcus aureus.
676 EA289 is an E. aerogenes KANS (susceptible to kanamycin, MDR isolate that exhibits active
25

ACCEPTED MANUSCRIPT
677 efflux of norfloxacin and AcrAB-TolC pump overproduction), EA294 and EA298 derived
678 from EA289 and deleted of AcrA and TolC, respectively [77]. AG100 is an E. coli Wild Type
679 (WT) and AG100A is its KANR (resistant to kanamycin) derivative, deleted of AcrAB and
680 hypersensitive to chloramphenicol, tetracycline, ampicillin and nalidixic acid [78]. Strains
681 were grown at 37°C on Mueller-Hinton medium 24 h prior to any assay. Mueller-Hinton
682 broth (MHB) was used for the susceptibility test. Polymyxin B nonapeptide (PMBN) was
683 obtained from Sigma-Aldrich and the culture medium was purchased from Becton Dickinson.
684 Table 7: Bacterial strains.
685
Bacteria
strains
686
Features
References
687
688
689
690
Escherichia coli
AG100
Wild-type E. coli K-12
AG100 ꢃAcrAB::KANR [78]
[78]
AG100A
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
Enterobacter aerogenes
KAN sensitive
EA289
EA294
EA298
derivative of EA27
EA289 AcrA::KANR
EA 289 TolC::KANR
[77]
[77]
[77]
Acinetobacter baumannii
ATCC19606 Reference strain
Klebsiella pneumonia
-
-
-
ATCC12296 Reference strain
Pseudomonas aeruginosa
PA 01
Reference strain
Salmonella enterica serotype Typhimurium
SL696
Wild-type, metA22,
trpB2, strAi20
[79]
[80]
711
712
713
714
715
Staphylococcus aureus
SA1199 Wild-type clinical,
methicillin-susceptible
716 KANR, resistance to kanamycin
26

ACCEPTED MANUSCRIPT
717 5.3.1.2. Determination of bacterial susceptibility
718 The respective minimum inhibitory concentrations (MIC) of the samples against
719 targeted bacteria were determined using the microdilution method (CLSI) [81].
720 Susceptibilities were determined in 96-well microplates with an inoculum of 2 × 105 cfu in
721 200 µL of MHB containing two-fold serial dilutions of samples. MICs were determined in
722 the presence of 5% or 0.5% of DMSO. In the first case, a 20× concentration range of each
723 compound was prepared in DMSO 100%. In the second case, a 200× concentration range of
724 each compound was prepared in DMSO 100% and then diluted with H2O to obtain a 20×
725 concentration range in DMSO 10%. 10 µl was added to 190 µl of inoculum reducing the
726 DMSO concentration to 0.5%. The MICs of samples were determined after 18 h incubation at
727 37°C, following addition (50 ꢀl) of 0.2 mg/mL iodonitrotetrazolium (INT) and incubation at
728 37°C for 30 minutes. MIC is defined as the lowest sample concentration that prevented the
729 color change of the medium and exhibited complete inhibition of microbial growth. The
730 sample dilution range was from 0-128 µM. Samples were tested alone or in the presence of
731 PMBN at 1/5 of its direct MIC (51.2 mg/L or 102.4 mg/L final concentration). All assays
732 were performed in duplicate or triplicate.
733
734 5.3.2. In vitro cytotoxicity testing
735
MCF-7 (human breast cancer cells possessing nuclear estrogen receptor) and MDA-
736 MB-231 (human breast cancer cells without nuclear estrogen receptor) were obtained from
737 the National Cancer Institute, U.S.A. Both cell lines were cultured in RPMI-1640 / DMEM
738 (High glucose) (Sigma) medium supplemented with 10% fetal calf serum. The cells were
739 plated into 96-well plates at a cell density of 6000 cells/well. After incubation for 24 hours,
740 the medium was discarded and the cells were rinsed with PBS solution. 200 µL of a series of
741 concentrations (50.0, 25.0, 10.0, 5.0, 1.0 and 0.5 µM) for each sample was added to each
742 well. The plate was incubated for another 72 hours. Cytotoxicity was determined using the
743 microtitration of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay
744 (Sigma, USA) as reported by Mosmann [75]. MTT solution (20µL, 5 mg/mL) was added to
745 each well. The plate was wrapped with aluminium foil and incubated for 4 h after which the
746 MTT solution was discarded leaving formazan crystals. 200 µL of DMSO was added to each
747 well to dissolve the formazan. The effect of the compound on cell line viability was measured
748 on an automated spectrophotometric plate reader (model MRX II Elisa microplate reader) at a
27

ACCEPTED MANUSCRIPT
749 test wavelength of 570 nm. Cytotoxicity was expressed as IC50, i.e. the concentration that
750 reduced the absorbance of treated cells by 50% with reference to the control (untreated cells).
751 IC50 values were determined from the plotted absorbance data for the dose-response curves.
752 Controls that contained only cells were included for each sample. Tamoxifen was used as the
753 cytotoxic standard.
754
755 5.4. X-ray crystallography
756
X-ray diffraction measurements for SBPY were performed at 100 K on a Bruker
757 Kappa X8 APEXII CD diffractometer with graphite monochromatised MoKα radiation (λ =
758 0.71073 Å). Correction for absorption was based on a multi-scan technique [82]. Intensity
759 data for SMHDH2, CuSMHD and CuSBHD were measured at 150 K on an Oxford
760 Diffraction Gemini CCD diffractometer employing either CuKα (SMHDH2), λ = 1.54184 Å,
761 or MoKα radiation (CuSMHD and CuSBHD). Corrections for absorption were based on a
762 multi-scan technique [83]. The structures were solved by direct methods and refined using
763 anisotropic displacement parameters, H atoms in the riding model approximation and a
2
2
764 weighting scheme of the form w = 1/[σ2(Fo ) + aP2 + bP] where P = (Fo + 2Fc2)/3) F2 using
765 SHELX programs [84] through the WinGX interface [85]. Crystal data and refinement details
766 are collected in Table 1. The molecular structures shown in Figs 1-4 were drawn with 70%
767 displacement ellipsoids using ORTEP-3 for Windows [85]. The overlay diagram, Fig. 4b, was
768 drawn with QMol [86] and the crystal packing diagrams with DIAMOND [87].
769
770 Acknowledgements
771 Support for the project came from Universiti Putra Malaysia (UPM), the Ministry of Higher
772 Education (Malaysia), French ANR Blanc 2010, METABACT grant and the French
773 Infrastructure for Integrated Structural Biology (FRISBI) ANR-10-INSB-05-01. M. L. Low is
774 grateful for the award of an Erasmus Mundus: Maheva Scholarship and a UPM Graduate
775 Research Fellowship (GRF).
776
777
28

ACCEPTED MANUSCRIPT
778 List of abbreviations
779
A. baumannii
CV
Acinetobacter baumannii
Cyclic voltammetry
DMEM
DMF
Dulbecco's modified Eagle's medium
Dimethylformamide
DMSO
DMSO-d6
E. aerogenes
E. coli
Dimethyl sulfoxide
Deuterated dimethyl sulfoxide
Enterobacter aerogenes
Escherichia coli
EPR
ESI-MS
FT-IR
Electron paramagnetic resonance
Electrospray ionization-mass spectrometry
Fourier transform-infrared spectroscopy
Iodonitrotetrazolium
Resistance to kanamycin
Sensitive to kanamycin
INT
KANR
KANS
K. pneumonia
LMCT
MCF-7
Klebsiella pneumonia
Ligand-to-metal charge transfer
Human breast carcinoma cells expressing
nuclear estrogen receptors
Human breast carcinoma cells not expressing
nuclear estrogen receptors
Multidrug resistance
MDA-MB-231
MDR
MHB
MIC
Mueller-Hinton broth
Minimum inhibitory concentration
3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide
Nuclear magnetic resonance
Oak Ridge thermal ellipsoid plot
Phosphate buffered saline
MTT
NMR
ORTEP
PBS
PMBN
Polymyxin B nonapeptide
Pseudomonas aeruginosa
Reactive oxygen species
P. aeruginosa
ROS
RP-HPLC
Reversed phase-high performance liquid
chromatography
r. t.
S. enterica
SBDTC
SMDTC
SOD
S. aureus
UV-Vis
WT
Room temperature
Salmonella enterica
S-benzyldithiocarbazate
S-methyldithiocarbazate
Superoxide dismutase
Staphylococcus aureus
Ultraviolet-visible
Wild type
780
781
29