Anti-cancer applications of titanocene-functionalised nanostructured systems: An insight into cell death mechanisms

DOI: 10.1002/chem.201400300

Full Paper

&

Drug Delivery

Anti-cancer Applications of Titanocene-Functionalised

Nanostructured Systems: An Insight into Cell Death Mechanisms

Jesﬄs Ceballos-Torres,^[a]Piroska Virag,^[b]Mihai Cenariu,^[c]Sanjiv Prashar,^[a]Mariano Fajardo,^[a]

Eva Fischer-Fodor,*^[b]and Santiago Gómez-Ruiz*^[a]

Abstract: A series of alkenyl-substituted titanocene com-

pounds have been supported on the mesoporous silica-

based material KIT-6. The corresponding functionalised ma-

terials were completely characterised by different techniques

(solid-state multinuclear NMR spectroscopy, IR spectroscopy,

N₂adsorption–desorption isotherms, X-ray fluorescence and

diffraction, SEM and TEM) to observe the incorporation of

the titanocene derivatives on the external surface of the ma-

terial KIT-6. Both the titanocene compounds and the materi-

als were tested in vitro against a wide variety of human

cancer and normal cell lines. A very high cytotoxicity of the

synthesised titanocene derivatives (IC₅₀values in the range

of those described in the literature for the most active cyto-

toxic titanocene compounds), with selectivity towards

cancer cell lines was observed. The cytotoxic activity of the

materials is the highest reported to date for titanocene-func-

tionalised materials. In addition, higher Ti uptake (from 4 to

23% of the initial amount of Ti) of the cells treated with ma-

terials was observed with respect to those treated with

“free” titanocene derivatives (which gave Ti uptake values

from 0.4 to 4.6% of the initial amount of Ti). Additional ex-

periments with the titanocene derivatives and the function-

alised materials revealed that changes to the morphological

and functional dynamics of apoptosis occurred when the

active titanocene species were incorporated into mesopo-

rous materials. In addition, the materials could induce pro-

grammed cell death in tumour cell populations by impairing

the damaged DNA repair mechanisms and by upregulation

of intrinsic and extrinsic apoptotic signalling pathways.

Introduction

Most of the problems associated with the application of

these metal-based drugs in humans are due to the low effec-

tiveness of transport processes of active species to the biologi-

cal target. For this reason, the scientific community has now

focused its efforts on the design of novel therapeutic technolo-

gies mainly based on nanostructured materials, which may act

as vectors for the delivery of the metallodrug or simply as

a protector of the active species for the enhancement of its

anti-cancer action.^[8,9]

Chemotherapeutic treatments based on metallodrugs have tra-

ditionally used platinum complexes due to their excellent cyto-

toxic properties.^[1]However, the relatively high number of side

effects associated with platinum treatments has opened up in-

terest in the use of other metal complexes with similar anti-

neoplastic properties, but with higher potential application in

humans. Thus, in recent years, outstanding results have been

obtained by using alternative metal complexes in preclinical

trials in vitro, although most of them have not been translated

into comparable developments in vivo due to problems in

their administration, which does not permit formulations with

high selectivity and minimum collateral damage.^[2–7]

In this context, many different systems have been used as

scaffolds for loading with metal-based drugs. Macromolecular

systems based on curcubit[n]urils and cyclodextrins have al-

ready been successfully used in preclinical trials, especially

with platinum drugs and metallocene-based therapeutic com-

plexes.^[10–13]Liposomes and lipid nanocapsules have also been

used as vehicles for the delivery of hydrophobic drugs to the

desired biological target, although, again, these systems have

only been loaded with platinum complexes^[14]or ferrocene de-

rivatives.^[15]Other approaches have focused on alternative mac-

romolecular systems, such as proteins, for example, human

serum albumin^[16–18]or apoferritin,^[19]and, recently, even poly-

nuclear organometallic compounds have been used as encap-

sulators of platinum and palladium complexes with cytotoxic

activity.^[20]Some additional studies have been performed by

using nanosystems based on carbon nanotubes^[21–23]and poly-

meric nanoparticles.^[24–26]

[a] J. Ceballos-Torres, Dr. S. Prashar, Prof. M. Fajardo, Dr. S. Gꢀmez-Ruiz

Departamento de Quꢁmica Inorgꢂnica y Analꢁtica, E.S.C.E.T.

Universidad Rey Juan Carlos, 28933 Mꢀstoles, Madrid (Spain)

E-mail: santiago.gomez@urjc.es

[b] Dr. P. Virag, Dr. E. Fischer-Fodor

Research Department, “I.Chiricuta” Oncology Institute

Cluj Napoca 400015 (Romania)

E-mail: fischer.eva@iocn.ro

[c] Dr. M. Cenariu

Biotechnology Research Center

University of Agricultural Science and Veterinary Medicine

Cluj Napoca, 400372 (Romania)

Supporting information for this article is available on the WWW under

http://dx.doi.org/10.1002/chem.201400300.

Chem. Eur. J. 2014, 20, 1 – 19

1

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

However, our interest is focused on the use of ceramic mate-

rials, such as calcium phosphate, hydroxyapatite or mesopo-

rous silicas, because they can be used as components of syn-

thetic hard tissues, such as bone fillers or prostheses. Thus,

functionalisation of these materials with metallodrugs allows

the formation of systems that have the potential to be applied,

after surgery, for the in situ treatment of bone tumours with

the aim of avoiding the reappearance of the tumour in the

same area.^[27–29]With this objective, several groups have fo-

cused their research on the adsorption, desorption and thera-

peutic properties of cisplatin and other platinum compounds

towards two biomimetic synthetic hydroxyapatite-based nano-

crystalline materials with different morphologies and physico-

chemical properties.^[30,31]For mesostructured silica-based mate-

rials, the first study with metal complexes was performed for

the encapsulation of cisplatin with the help of CdS caps,^[32]and

subsequently, only our group has worked with other metal-

based drugs, namely, titanocene derivatives, and observed that

all functionalised surfaces exhibited in vitro cytotoxicity against

a wide variety of human cancer cells.^[33–36]However, these stud-

ies did not clarify the mechanism of action of the titanocene-

functionalised materials. Nevertheless, the results showed that

the incorporation of titanocene derivatives on mesoporous

silica-based particles was a viable alternative for the adminis-

tration of titanocene derivatives and may help in the enhance-

ment of their therapeutic properties by overcoming the prob-

lems of hydrophobicity associated with these compounds.^[37,38]

Herein, we report novel insights into the mechanisms of

action of titanocene-functionalised mesoporous silica-based

materials. We have prepared a series of alkenyl-substituted tita-

nocene compounds that have very high cytotoxic activity

(comparable to that of titanocene-Y, which is one of the most

potent metallodrugs based on titanium). Their corresponding

mesoporous materials present much higher cytotoxic activity

(lower M₅₀or Ti-M₅₀values) and selectivity than similar nano-

structured systems functionalised with titanocene deriva-

tives.^[33–36]

Results and Discussion

Synthesis and characterisation of titanocene complexes

Synthesis and spectroscopic characterisation of titanocene(IV)

complexes

The synthesis of the titanocene(IV) complexes (11–20) was per-

formed according to the synthetic routes described in

Scheme 1, starting from the corresponding titanium precursor

([Ti(h⁵-C₅H₅)Cl₃] or [TiCl₄(thf)₂]) and the stoichiometric ratio of

the corresponding lithium cyclopentadienide derivative (4, 5,

6, 9 or 10; Scheme 1 and the Supporting Information). Com-

plexes 11–20 were characterised by multinuclear NMR spec-

troscopy, FTIR spectroscopy, mass spectrometry and elemental

analysis.

NMR spectroscopy of compounds 14 and 20 revealed the

formation of the expected rac and meso isomers (1:1 ratio),

which arose from the presence of two chiral centres. In addi-

tion, it is important to note the existence of planar chirality

due to the disubstitution of the cyclopentadienyl (Cp) ring in

15 and 16. In the case of compound 16, the planar chirality to-

gether with the chiral centre of the alkenyl substituent leads

to the formation of two isomers. NMR spectroscopy showed

that these isomers were present in a 1:1 ratio. For further infor-

mation about the possible isomers in 14, 16 and 20, see Fig-

ures S1 and S2 in the Supporting Information.

As a general rule for asymmetric titanocene complexes with

1

mixed rings (11, 12, 15, 16, 17 and 18), the H NMR spectra

showed a singlet assigned to the unsubstituted Cp ring be-

tween d=6.1 and 6.6 ppm. In addition, a set of two signals

corresponding to the protons of the substituted Cp ring be-

tween d=6.0 and 6.6 ppm was observed for 11, 12, 17 and

18. In the case of complexes 15 and 16, which have a 1,3-di-

substituted Cp ligand, three signals corresponding to the three

protons of the Cp ring between d=6.0 and 6.7 ppm and a sin-

glet at dꢀ0.2 ppm, assigned to the protons of the trimethyl-

silyl group, were observed.

In addition, we show that titanium-functionalised materials

act as Trojan horses for the incorporation of titanium active

species into the cells, leading to much higher quantities of tita-

nium uptake in cells, than that observed with “free” titanocene

derivatives. These results indicate that transport processes

(either active or passive) of “free” titanocene derivatives to the

cell are less effective than those of the titanocene-functional-

ised materials.

A set of signals associated with the alkenyl fragments was

1

also observed in the H NMR spectra of 11–20. Thus, titano-

cene compounds 11, 13 and 15 showed a singlet at about d=

1.20 ppm, due to two methyl groups substituting the carbon

atom bonded to the Cp ring, and five sets of signals, two cor-

responding to the methylene protons (two multiplets between

d=1.5 and 2.0 ppm), one corresponding to the proton of the

C-d (a multiplet at d=5.6 ppm) and two for the terminal ole-

finic protons (multiplets at dꢀ4.7 and 4.9 ppm). Compounds

12, 14 and 16 showed an additional set of three aromatic

proton signals assigned to the phenyl ring protons.

Finally, we report additional apoptotic experiments, such as

the study of the apoptosis induction capacity, an evaluation of

the modulator effect on intrinsic apoptotic pathway and apop-

tosis regulation at DNA level. We observed that the dynamics

of apoptotic morphological and functional changes were

modified significantly when the active titanocene species were

incorporated onto mesoporous materials. These materials are

also able to program cell death in tumour cell populations by

impairing the damaged DNA repair mechanisms and by up-

regulation of intrinsic and extrinsic apoptotic signalling path-

ways.

For complexes 17 and 19, a singlet due to the two methyl

groups substituting the carbon atom adjacent to the Cp ring

was observed at about d=1.2 ppm. In addition, five sets of

signals, two corresponding to the CH₂methylene protons (two

multiplets between d=1.4 and 1.9 ppm), one corresponding

to the alkenylic proton of the C-d (triplet at dꢀ5.0 ppm) and

two singlets corresponding to the protons of the terminal

methyl groups (between d=1.4 and 1.7 ppm), were observed.

&

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

2

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

Scheme 1. Synthesis of titanocene compounds 11–20.

Complexes 18 and 20 showed an additional set of three aro-

matic proton signals assigned to the phenyl ring protons.

As previously mentioned, in the cases of compounds 14, 16

and 20, due to the existence of two isomers all of the ob-

served signals were twinned.

The ¹³C{¹H} NMR spectra of 11–20 showed the expected sig-

nals. The mass spectra showed the molecular ion peak and

other peaks corresponding to typical fragmentations of the

molecules.

X-ray structure of 17 and 18

Compounds 17 (Figure 1) and 18 (Figure 2) crystallise in the

monoclinic space groups P2₁/n and P2₁/c, respectively, both

with four molecules in the unit cell. Both compounds crystal-

lise in centrosymmetric space groups; compound 18 crystal-

lises as a racemate. Table 1 lists crystallographic details and

Figure 1. X-ray molecular structure and atom labelling scheme of compound

17 with thermal ellipsoids at 50% probability. Hydrogen atoms have been

omitted for clarity.^[80]

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

3

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

Table 2. Selected bond lengths (pm) and angles (^o) for 17 and 18.

17

18

Ti1ÀCent1^[a]

Ti1ÀCent2^[a]

av Ti1ÀC(C1ÀC5)^[b]

av Ti1ÀC(C6ÀC10)^[b]

Ti1ÀC1

207.0(2)

207.4(2)

238.1(2)

239.5(2)

235.7(2)

237.1(2)

239.4(3)

238.9(2)

239.3(2)

245.8(2)

241.8(2)

238.1(2)

233.2(2)

238.5(2)

238.03(8)

235.82(9)

1.500(3)

1.324(3)

1.497(3)

1.501(3)

205.4(8)

208.5(7)

236.4(8)

239.7(7)

238.0(8)

236.7(8)

233.2(8)

234.6(7)

239.5(8)

248.3(7)

242.1(7)

237.6(7)

233.8(7)

236.8(6)

235.3(3)

231.9(3)

1.35(2)

Ti1ÀC2

Ti1ÀC3

Ti1ÀC4

Ti1ÀC5

Ti1ÀC6

Ti1ÀC7

Ti1ÀC8

Ti1ÀC9

Ti1ÀC10

Ti1ÀCl1

Ti1ÀCl2

C15ÀC16

C16ÀC17

C17ÀC18

C17ÀC19

C20ÀC21

C21ÀC22

C22ÀC23

C22ÀC24

1.36(2)

1.40(2)

1.49(1)

1.296(9)

1.51(1)

1.48(2)

Figure 2. X-ray molecular structure and atom labelling scheme of compound

18 with thermal ellipsoids at 50% probability. Hydrogen atoms have been

omitted for clarity.^[80]

Cent1-Ti-Cent2

Cl1-Ti1-Cent1

Cl1-Ti1-Cent2

Cl2-Ti1-Cent1

Cl2-Ti1-Cent2

Cl1-Ti1-Cl2

C15-C16-C17

C16-C17-C18

C16-C17-C19

C18-C17-C19

C20-C21-C22

C21-C22-C23

C21-C22-C24

C23-C22-C24

131.8(2)

105.5(2)

107.1(2)

106.1(1)

106.7(2)

92.82(2)

127.7(2)

124.7(2)

121.8(2)

113.6(2)

132.8(9)

106.3(9)

105.6(9)

105.0(9)

107.6(9)

92.4(1)

Table 1. Crystallographic data of 17 and 18

17

18

121.4(9)

118.4(9)

formula

M_r

C₁₉H₂₆Cl₂Ti

373.20

C₂₄H₂₈Cl₂Ti

435.26

298(2)

monoclinic

P2₁/c

1138.2(13)

1248.1(12)

1606(2)

90

110.49(5)

90

2138(4)

4

T [K]

298(2)

crystal system

space group

a [pm]

b [pm]

c [pm]

a [8]

monoclinic

P2₁/n

653.10(13)

1892.4(4)

1457.1(3)

90

129.5(7)

125.1(8)

120.9(8)

114.0(7)

[a] Cent1 and Cent2 are the centroids of C1ÀC5 and C6ÀC10, respectively.

[b] The average distance between Ti1 and the carbon atoms of the C₅

ring of the corresponding Cp moiety.

b [8]

100.37(3)

90

1771.5(6)

4

0.200.150.10

1.399

784

g [8]

V [³]

Z

crystal size [mm³]

0.100.060.04

1.352

912

1_calcd[mgm^À3

F(000)

]

carbon atoms of titanocene complexes 17 and 18 vary from

about 233 to 248 pm. The longest TiÀC bond lengths are ob-

served for the substituted carbon atom of the Cp ring (C6).

The TiÀCent distances in 17 and 18 are between 205 and

209 pm and the Cent-Ti-Cent angles are about 1328; these

values are in the expected range for titanocene deriva-

tives.^[39–41]In addition, the Cl1-Ti1-Cl2 angles of about 928 are

similar, and comparable with those recorded for the X-ray crys-

tal structures of other related titanocene complexes.^[39–41]

The C17ÀC18 and C21ÀC22 bond lengths for the double

bonds of the alkenyl chain of 17 and 18, respectively, have

values of about 130 pm and are in agreement with reported

structures of other titanocene complexes with alkenyl substitu-

ents.^[42–44]

absorption coefficient [mm^À1

]

0.779

0.657

hkl range

À7ꢁhꢁ7,

À21ꢁkꢁ21,

À16ꢁlꢁ16

46.60

À12ꢁhꢁ12,

À14ꢁkꢁ14,

À18ꢁlꢁ18

47.42

28862/3166

0.1845

3166/0/247

1.040

0.0763/0.1737

0.1197/0.1946

0.561/À0.343

2q_max[8]

reflns collected/unique

R(int)

21730/2567

0.0413

data/restraints/parameters

GoF on F²

2567/0/203

1.031

R₁/wR₂[I>2s(I)]

R₁/wR₂(all data)

largest diff. peak/hole [e

0.0232/0.0527

0.0277/0.0547

0.253/À0.183

À3

]

Table 2 gives selected bond lengths and angles of 17 and 18.

The molecular structures reveal a tetrahedral geometry about

the titanium atom with coordination of the Cp rings in an h⁵

manner. The bond lengths between titanium and the Cp

&

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

4

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

Synthesis and characterisation of materials incorporating ti-

tanocene complexes (K11–K20)

Finally, characterisation of this material by ²⁹Si MAS NMR

spectroscopy showed typical Q⁴(Si(OSi)₄), Q³(Si(OSi)₃(OH)) and

Q²(Si(OSi)₂(OH)₂) signals at d=À112, À104 and À92 ppm, re-

spectively.

Synthesis and characterisation of KIT-6

Mesoporous material KIT-6 was prepared according to the syn-

thetic method described by Ryoo and co-workers (see the Sup-

porting Information).^[45]The physical parameters of nitrogen

adsorption–desorption isotherms (Figure 3) of this material

Synthesis and characterisation of titanocene-functionalised

materials

The synthesis of the titanocene-functionalised materials was

performed by grafting of the different titanocene compounds

synthesised in this study (11–20) on dehydrated (at 5008C) KIT-

6 mesoporous silica. This reaction is very simple and consists

of the treatment of the dehydrated material overnight with

a solution of the titanocene compounds in toluene under

reflux (Scheme 2).

Because the non-functionalised KIT-6 has a low concentra-

tion of hydroxyl groups from silanol moieties due to its treat-

ment at high temperature, the only possible reactions between

titanocene derivatives and KIT-6 mesoporous silica are those

involving the chlorine atoms of the titanocene complexes,

either with the acidic silanol groups or with the reactive silox-

ane bridges, to give m-oxo surface species, as observed in pre-

vious reports (Scheme 2).^[33–36]Notably, adsorbed titanocene

complexes may also be located in the pore of the correspond-

ing materials.

Following our previous work on the functionalisation reac-

tions of mesoporous silicas with titanocene derivatives, graft-

Figure 3. Nitrogen adsorption/desorption isotherms of KIT-6 and K11.

showed a typical type IV iso-

therm (according to the IUPAC

classification)^[46]with an H2 hys-

teresis loop. In addition, synthe-

sised KIT-6 had a BET surface

area (S_BET) of 694 m²g^À1, an aver-

age pore diameter of 63.7 and

an average pore volume of

0.76 cm³g^À1

; these values are

typical for mesostructured mate-

rials. The low-angle XRD pattern

of KIT-6 (which has a three-di-

mensional

cubic

symmetry

(Ia3d)) gave an intense peak at

2q=0.958 (assigned to the (211)

plane) and two very weak peaks

at 2q=1.65 and 1.908 (indexed

as (220) and (320);^[45,47]see Fig-

ure S3 in the Supporting Infor-

mation).

Additionally, analysis of this

Scheme 2. Synthesis of titanocene-functionalised materials K11–K20.

material by FTIR spectroscopy

gave a spectrum with a broad

band between 3400 and 3200 cm^À1, which was attributed to

OÀH stretching of the silanol groups of the surface and the re-

maining adsorbed water molecules. The stretching band of the

SiÀO bonds was observed at 805 cm^À1and a broad strong

band between 1100 and 1250 cm^À1, corresponding to the si-

loxane groups (SiÀOÀSi), was also recorded.

ing reactions were performed by using a theoretical 5 wt% of

Ti.^[33–36]From data obtained by X-ray fluorescence, values of Ti

wt% for functionalised surfaces K11–K20 are between 1.20

and 4.85%, with most of the values falling between 2 and 3%.

It is worth noting the unusually high functionalisation rate of

K20 with a Ti wt% of 4.85 (Table 3). The relatively low Ti

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

5

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

face of the mesoporous scaffold. The difference in the location

of the titanocene derivative in K11–K20 with KIT-6 as the sup-

port, compared with MCM-41, SBA-15, HMS and MSU-2 materi-

als,^[33–36]may be associated either with the higher pore diame-

ter of free KIT-6 (of almost 70 ) or with a higher number of si-

lanol groups on the external surface, which is typical in KIT-6.

These two textural properties of KIT-6 may minimise the inter-

actions between the silanol groups of the pores on direct com-

parison with the silanol groups of the external surface, forcing

functionalisation onto the external surface of the material.

In addition, all materials were characterised by powder X-ray

diffraction, which gave low-angle patterns of KIT-6. The intensi-

ty of the peaks of materials K11–K20 is lower than that of free

KIT-6; this is probably due to blocking of the dispersion centres

by the titanocene derivatives (see Figure S3 in the Supporting

Information). Thus, the diffractograms displayed one intense

peak at 2qꢀ0.958, which can be indexed as (211). However,

the two weak reflections at 2qꢀ1.65 and 1.98 (corresponding

to (220) and (320), respectively) were barely observed due to

lowering of their intensity by functionalisation with the metal

complex.

Table 3. Physical parameters of unmodified KIT-6 and of the modified

materials K11–K20 measured by N₂adsorption–desorption isotherms and

percentage of Ti grafted on the silica support in surfaces K11–K20, as

measured by X-ray fluorescence.

BET surface area

Total pore volume

Pore diameter

(BJH)^[a][]

Ti

[wt%]

[a]

[m²g^À1

]

[cm³g^À1

]

KIT-6

K11

K12

K13

K14

K15

K16

K17

K18

K19

K20

694

575

592

610

469

561

532

492

442

365

348

0.76

0.63

0.67

0.68

0.42

0.44

0.59

0.44

0.41

0.37

0.33

63.7

63.6

63.7

68.7

63.8

64.3

63.9

64.2

–

2.13

1.20

2.17

2.26

3.06

2.43

2.23

2.00

2.43

4.85

[a] BJH=Barrett, Joyner and Halenda.

amount in the functionalised materials may be due either to

the saturation of the surface or to the weak basicity of both

chloro ligands of the titanocene complexes and the weak acid-

ity of SiÀOH groups of the KIT-6 mesoporous material.

Incorporation of the titanocene derivatives in the mesopo-

rous scaffold was also confirmed by FTIR spectroscopy. All

functionalised materials K11–K20 showed, in addition to the

bands of KIT-6, characteristic weak bands for CÀH stretching vi-

brations between 3000 and 2800 cm^À1and bending vibration

bands at about 1400 cm^À1due to the Cp ligands (for further

information, see Figures S6 and S7 in the Supporting Informa-

tion).

For K11–K20, the functionalisation rate (Ti wt%) is some-

what higher than those obtained in previous studies of the

functionalisation of MCM-41, SBA-15, HMS and MSU-2 with ti-

tanocene derivatives, for which a maximum loading of the

metallocene complex on the mesoporous surfaces was always

lower than 2.0% of the metal, even when starting from high

metal/SiO₂ratios.^[33–36]

In addition, characterisation by multinuclear solid-state NMR

spectroscopy confirmed functionalisation of the materials with

the corresponding titanocene derivatives. The ¹H MAS NMR

spectra of all materials showed expected signals corresponding

to the different protons of the grafted titanocene complex. In

each case, a broad signal assigned to protons situated in the

sp³-carbon atoms of the alkenyl fragment was observed be-

tween d=1.5 and 4 ppm, whereas the protons of the sp²-

carbon atoms (Cp ligands and those of the C=C double bond

of the alkenyl fragment) were recorded between d=4 and

7 ppm. Moreover, in the special case of the materials K12, K14,

K16, K18 and K20, which contain phenyl groups, an additional

signal above d=7 ppm was recorded due to the resonance of

the aromatic protons of the phenyl groups. An additional

signal between d=0 and 1 ppm, corresponding to the protons

of the trimethylsilyl group, was observed for the materials K15

The physical parameters of nitrogen adsorption–desorption

isotherms, such as BET surface area (S_BET), average pore diame-

ter, average pore volume, and wall thickness, for unmodified

KIT-6 and the modified materials K11–K20 were measured

(Figure 3).

Typical type IV isotherms (according to the IUPAC classifica-

tion)^[46]with an H2 hysteresis loop were observed; this indicat-

ed the mesoporous nature of all of the studied materials. Ca-

pillary condensation of nitrogen within the uniform mesopore

structure, as indicated by the increase in the adsorbed volume

at a relative pressure (P/P₀) of about 0.4, was observed for all

studied surfaces. The inflection position shifted slightly to-

wards lower relative pressures after functionalisation. Interest-

ingly, a decrease in the surface area and pore volume of mate-

rials K11–K20 was observed on direct comparison with the

corresponding non-functionalised material (KIT-6; Table 3). The

pore size distribution in all materials is very narrow, both

before and after functionalisation with the titanocene deriva-

tives (see Figures S4 and S5 in the Supporting Information).

However, the pore diameter seems to be unaffected after the

functionalisation reactions for KIT-6-based materials, contrary

to that reported for the functionalisation of MCM-41, SBA-15,

HMS and MSU-2, which showed a significant decrease of the

pore diameter (Table 3). Thus, it seems plausible that for K11–

K20 only a very low amount of the titanocene derivative is lo-

cated in the pore of the materials and the major part of the

grafted titanocene compound is located on the external sur-

1

and K16. Figure 4 shows the H MAS NMR spectrum of K16.

The ¹³C CP MAS NMR spectra showed different sets of sig-

nals, depending on the carbon atoms present in the grafted ti-

tanocene complex. Thus, the sp³-carbon atoms of the alkenyl

fragment gave signals between d=10 and 40 ppm for all ma-

terials; the quaternary carbon atom was generally observed

above d=40 ppm. In addition to these signals, resonances cor-

responding to the sp²-carbon atoms of the Cp, phenyl rings

and alkenyl fragments were normally observed between d=

110 and 140 ppm (Figure 5). Finally, for materials K15 and K16,

the resonance due to the carbon atom of the trimethylsilyl

group was recorded at d=À2.6 and À3.4 ppm, respectively.

&

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

6

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

shape. However, the aggregates were of a similar diameter of

approximately 5 mm. TEM images of all materials demonstrated

a clear arrangement of the porous material with a uniform

size. Figure 7a (SEM) and b (TEM) shows the micrographs of

K13.

Figure 4. ¹H MAS NMR spectrum of material K16.

Figure 5. ¹³C CP MAS NMR spectrum of material K11.

Figure 7. a) SEM and b) TEM micrographs of K13.

In addition, the ²⁹Si MAS NMR spectra of the functionalised

materials (K11–K20) showed signals corresponding to the dif-

ferent Qⁿsites of the mesoporous structure, which were also

observed for the non-functionalised material. However, a drastic

decrease in the intensity of these signals was observed in all

materials relative to those of free KIT-6, as a consequence of

functionalisation with the titanocene derivative (Figure 6).

Finally, SEM and TEM were used to explore the morphology

of the studied materials. Results revealed that K11–K20 were

aggregates of mesoporous small particles without a defined

In vitro studies

Compounds and materials with an in vitro anti-proliferative

effect against human cells

The capacity of the newly synthesised titanocene compounds

to inhibit human cell proliferation in vitro was studied by eval-

uating their cytotoxicity on three human tumour cell lines:

A2780 ovary carcinoma, MCF-7 breast adenocarcinoma and

DLD-1 colon carcinoma. To establish their selectivity, the toxici-

ty of the titanocene derivatives to HEK-293 human normal em-

bryonic kidney cells was also tested.

Anti-proliferative activity was quantified by using IC₅₀values.

The IC₅₀value is the concentration necessary to reduce the cell

population’s growth by 50%. Consequently, large IC₅₀values

indicate nontoxic compounds and a decrease in the IC₅₀value

indicates anti-proliferative behaviour. For 11–20, these parame-

ters were calculated from the dose-response sigmoid curves

provided by the GraphPad Prism biostatistics program. The

studied compounds exhibit different grades of toxicity against

the tested cell lines (Figure 8 and Table 4). Most of them are

less toxic against the normal cell line (HEK-293) than to the

human cancer cell lines. In particular, A2780 cells and MCF-7

cells were more affected by the titanocene complexes, which

indicated some degree of selectivity of these compounds

against the studied cancer cell lines.

Figure 6. ²⁹Si MAS NMR spectrum of material K13.

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

7

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

stituents at the Cp ring,^[49–52]which are considered to be highly

cytotoxic, comparable to titanium(IV)–salen complexes^[53]and

mixed titanocene–gold(I) derivatives.^[54]

On the other hand, in human breast carcinoma (MCF-7

cells), the differences are more pronounced between the

group of the less cytotoxic compounds formed by 11, 12, 13,

14 and 16 (with IC₅₀values from ca. 60 to 170 mm) versus the

more cytotoxic 15, 17, 18, 19 and 20 (with IC₅₀values from ca.

5 to 17 mm). For the last five compounds, the IC₅₀values dis-

play a significant drop in comparison with the rest of the tita-

nocenes (one-way ANOVA test, Bonferroni post-test, p<0.05).

The most cytotoxic compounds of the series reported herein

present higher cytotoxicity than other titanocene deriva-

tives^[55–58]and comparable cytotoxicity in this cell line with the

famous and highly cytotoxic compound titanocene-Y reported

by Tacke and co-workers.^[59,60]

Proliferation of the aggressive K-ras mutant DLD-1 colon

cells was also reduced by using the titanocene compounds

11–20. For this cell line IC₅₀values were somewhat larger com-

pared with the other two cell lines. Compounds 15,

Figure 8. Histograms showing the IC₅₀values in mm for 11–20 and M₅₀in

mgmL^À1for KIT-6 and K11–K20 on A2780, MCF-7, DLD-1 and HEK-293 at

24 h of action.

17, 18 and 20 maintained their relatively high cyto-

toxicity against DLD-1 cells, although with higher IC₅₀

Table 4. Cytotoxicity values (IC₅₀in mm for 11–20 and M₅₀in mgmL^À1for KIT-6 and

K11–K20) on A2780, MCF-7, DLD-1 and HEK-293 at 24 h of action.^[a]

values than those in the other cell lines (from ca. 30

to 47 mm, which are the lowest IC₅₀values described

in the literature for DLD-1 cells). The remaining tita-

nocene derivatives 11–14, 16 and 19 showed rela-

tively low cytotoxic activity against DLD-1 with IC₅₀

values ranging from about 70 to 250 mm.

A correlation between the toxicity of the com-

pounds (nonparametric Spearman two-tailed correla-

tion, non-significant p value summary r>0.3, p>0.2)

and the different cell types was not found.

IC₅₀[mmÆSD]/M₅₀values [mgmL^À1ÆSD]

A2780

MCF-7

DLD-1

HEK-293

11

12

13

14

15

16

17

18

100.07Æ7.85

7.49Æ1.06

65.67Æ12.25

56.23Æ11.20

7.83Æ1.54

26.57Æ1.12

28.46Æ4.66

25.72Æ5.47

41.42Æ6.86

16.91Æ1.66

2404Æ323

1281Æ156

193Æ29

143.80Æ16.39

103.13Æ19.42

62.15Æ28.40

90.19Æ10.95

12.55Æ2.79

170.33Æ32.61

8.84Æ1.09

16.45Æ3.64

10.80Æ2.44

5.92Æ1.00

>2500

250.32Æ17.20

96.67Æ6.89

152.75Æ33.48

97.65Æ10.45

29.67Æ3.55

136.33Æ13.97

46.60Æ5.11

37.44Æ5.65

69.62Æ8.54

41.06Æ1.94

2212Æ204

2057Æ136

446Æ38

293.90Æ8.42

232.67Æ21.75

183.70Æ15.76

157.41Æ28.98

326.50Æ23.22

134.43Æ10.86

174.30Æ41.78

79.92Æ8.49

95.39Æ15.20

70.77Æ10.96

>2500

Analysis of the toxicity of 11–20 towards normal

cell lines (HEK-293) shows, in general, much higher

IC₅₀values (lower toxicity) for all complexes in com-

parison with the IC₅₀values obtained for cancer cell

lines. In particular, compounds 15 and 20 show

a very high degree of selectivity (much greater than

that observed for the other compounds). This is due

to the ability of 15 and 20 to strongly inhibit the

three studied tumour cell lines, which results in IC₅₀

value ratios between tumoural and normal cells su-

perior to those of the other compounds (for further

information, see Table S1 in the Supporting Informa-

tion).

19

20

KIT-6

K11

K12

K13

K14

K15

K16

K17

K18

K19

K20

1971Æ120

2040Æ242

857Æ62

717Æ102

1081Æ214

1086Æ172

445Æ48

766Æ152

299Æ58

1834Æ138

504Æ57

486Æ67

111 Æ18

163Æ24

236Æ31

1762Æ183

590Æ148

225Æ34

1260Æ131

278Æ25

840Æ59

996Æ92

1079Æ86

1254Æ216

704Æ104

489Æ24

450Æ70

714Æ27

424Æ50

124Æ11

508Æ46

650Æ47

85Æ12

36Æ5

149Æ27

256Æ42

[a] SD=standard deviation.

Additionally, the anti-proliferative activity of the ti-

tanocene-functionalised materials K11–K20 and the

For the A2780 cell line, compounds 11, 13, 14 and 19

showed the highest IC₅₀values (from ca. 41 to 100 mm), where-

as 16, 17 and 18 were slightly more cytotoxic (IC₅₀values of

ca. 26 mm). However, compounds 12, 15 and 20 showed the

highest cytotoxic activities against this cancer cell line with IC₅₀

values from (7.49Æ1.06) to (16.91Æ1.66) mm. The cytotoxicity

of the last compounds is much higher than that of titanocene

dichloride in this cell line.^[48]In addition, the cytotoxic activity

of 12, 15, 16 and 20 is also higher than that reported for other

active titanocene compounds with alkyl or arylammonium sub-

unmodified mesoporous material KIT-6 was tested against

cancer cell lines A2780, MCF-7, DLD-1 and normal cells HEK-

293. The M₅₀values (which is the concentration of material

necessary to reduce the cell population’s growth by 50%)

were determined and are given in Table 4 and Figure 9.

The non-functionalised mesoporous matrix KIT-6 does not

show any significant activity against the studied cell lines,

whereas materials K11–K20 exhibited, in general, high cytotox-

ic activity against cancer cell lines, but only moderate activity

against normal cell line HEK-293.

&

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

8

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

Table 5. Cytotoxicity values (Ti M₅₀in mgmL^À1of Ti anchored on the sur-

face) on A2780, MCF-7, DLD-1 and HEK-293 of the titanocene-functional-

ised materials K11–K20 at 24 h of action.

Material

Ti M₅₀values (mgmL^À1ÆSD)

A2780

MCF-7

DLD-1

HEK-293

K11

K12

K13

K14

K15

K16

K17

K18

K19

K20

27.3Æ3.3

2.3Æ0.3

16.6Æ3.3

6.8Æ1.3

3.4Æ0.6

5.5Æ0.8

22.2Æ2.1

9.8Æ0.5

10.3Æ1.2

4.1Æ0.6

42.0Æ2.6

24.5Æ2.9

18.6Æ1.3

11.0Æ1.5

5.0Æ0.7

30.6Æ3.2

6.2Æ0.6

9.0Æ1.4

3.0Æ0.3

1.7Æ0.2

43.8Æ2.9

5.4Æ0.5

39.8Æ3.0

11.4Æ1.3

7.2Æ0.9

20.4Æ1.4

24.1Æ1.9

14.3Æ0.5

12.3Æ1.1

7.2Æ1.3

15.3Æ2.2

13.0Æ2.6

23.6Æ3.7

10.1Æ1.1

53.9Æ5.6

14.3Æ3.6

28.0Æ4.8

14.1Æ2.1

15.8Æ1.1

12.4Æ2.0

Figure 9. Histograms showing the M₅₀values in mgmL^À1for KIT-6 and K11–

K20 on A2780, MCF-7, DLD-1 and HEK-293 at 24 h of action.

and show again remarkable selectivity towards cancer cell lines

(for further information, see Table S2 in the Supporting Infor-

mation).

In general, the materials reported herein, especially K15 and

K20, present much higher cytotoxic activity (lower M₅₀or Ti

M₅₀values) and selectivity than other similar nanostructured

systems functionalised with titanocene derivatives reported by

us.^[33–36]

In particular, materials K11, K13, K16 and K17 display lower

toxicity in tumour cell lines than K15, which exhibits a statisti-

cally significant anti-proliferative effect against the three

tumour cell lines. Furthermore, material K12 shows a high cy-

totoxic effect against ovary and colon carcinoma cells and K18

against ovary and breast tumoural cells (one-way ANOVA test,

Dunnet post-test, 95% confidence interval), but their selectivity

is low because the M₅₀values against the normal cell line HEK-

293 are in the same concentration range.

Due to their superior cytotoxic properties in terms of activity

and selectivity, 15, 20, K15 and K20 were selected to study the

cell death mechanisms and signalling pathway interferences

(see the section on apoptotic studies).

Again, as in the case of the titanocene derivatives 15 and

20, materials K15 and K20 are the most potent cytotoxic

agents in all the three tumour cell populations, with a pro-

nounced statistical significance (Bonferroni post-test, p<

0.001). This confirms the fact that these titanocene structures

give the best activity for the studied compounds. The propor-

tion between the titanocene IC₅₀and material M₅₀values is

maintained in all tumour cell lines. In A2780 cells, the correla-

tion is significant (nonparametric correlation between two data

sets, Spearman r value of 0.7697, two-tailed p value of 0.0126);

the same is observed in DLD-1 cells (r=0.66, p=0.0438). How-

ever, in MCF-7 cells, the association is even more pronounced

(r=0.90, p=0.0008). For example, ratios of 11.67Æ3.06 in K15/

15 activity, and 4.90Æ1.23 in K20/20 activity in tumour cell

lines are observed. For this reason, material K15 appears to

have a higher selectivity than K20 when compared with the

M₅₀value of the studied normal cell line. This phenomenon

occurs even if there is no correlation between the absolute

values of titanocene toxicity versus material toxicity for most

of the compounds, mainly because the relative proportion of

titanocene is not identical in all nanostructured materials.

Therefore, we have calculated the cytotoxic activity of all of

the functionalised materials in terms of Ti M₅₀index (which

represents the quantity of anchored titanium needed to inhibit

normal cell growth by 50%; Table 5). The results show a similar

tendency to that found for the M₅₀values; materials K15 and

K20 are the most cytotoxic compounds for the three cancer

cell lines and present moderate toxicity to the HEK-293 cells,

Apoptosis induction capacity of titanocene complexes and ti-

tanocene-functionalised mesoporous materials

To establish the type of damage induced by the novel synthe-

sised compounds, the mechanism of cell death was analysed

by flow cytometric methods. The behaviour of compounds 15

and 20, along with their analogous nanomaterials (K15 and

K20), was analysed in A2780 ovary carcinoma, MCF-7 breast

adenocarcinoma and DLD-1 colon carcinoma cell lines.

For these studies, the fluorescein isothiocyanate (FITC)

stained, AnnexinV marked cells corresponded to the early

apoptotic stages, whereas double colouration indicated a dis-

tinct population in early/mid apoptotic stages in which the

translocation of membrane proteins was added to DNA frag-

mentation and loss of nuclear DNA content, marked by propid-

ium iodide (PI) fluorochrom, which was capable of binding and

labelling DNA (Figure 10).^[61]PI colouration only indicates late

apoptosis; the magnitude of the PI-marked DNA damage in

the context of cell cycle arrest is also reflected (Figure 11).

For A2780 ovary cells, a strong anti-proliferative activity of

complex 15 was confirmed and in all compounds the shift to

the mid and late stages of apoptosis was observed within 8 h.

Translation to the late stages of apoptosis with DNA damage is

more pronounced for K15 and K20 materials (for further de-

tails, see Figure S8 in the Supporting Information). Regarding

MCF-7 breast carcinoma, all studied compounds and materials

had a better capacity to induce apoptosis, in comparison with

the standard chemotherapy drug oxaliplatin. The apoptotic ac-

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

9

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

Figure 11. Flow cytometric evaluation of apoptosis induction by titanocenes

and nanostructured materials: a) untreated control MCF-7 cells, b) oxalipla-

tin-treated cells, c) cells treated with 15, d) cells treated with 20, e) cells

treated with K15, f) cells treated with K20. Viable cells are found in the Q3

box, the Q1 box corresponds to FITC-labelled early apoptosis cells, double

colouration is located in the Q2 quadrant, and in Q4 PI indicates late apop-

tosis.

In general, despite the very low quantity of titanium in func-

tionalised materials K15 and K20, compared with that of the

“free” titanocene derivatives, the materials present a higher

apoptotic activity, which indicates that the grafting of titano-

cene derivatives on silica-based mesoporous materials has

a positive influence on the cytotoxic properties of the titano-

cene complexes.

Figure 10. Apoptosis induction by titanocene derivatives in a) A2780,

b) MCF-7 and c) DLD-1 cancer cells. Flow cytometry provides data about the

resulting cell populations after 8 h of treatment with titanocene compounds

15 and 20 and nanostructured materials K15 and K20; colourless cells are

living cells, and the rest of the cells are non-viable and divided into three

different populations with specific attributes of cell death.

Mounting evidence in the literature shows that tumour cell

evasion of programmed cell death is critical for cancer prog-

ress and unremitting growth and the success of a chemothera-

peutic drug is dependent on its capacity to promote apopto-

sis.^[62]Therefore, not only the toxicity, but the apoptosis induc-

ing capacity of the compounds are important issues associated

with the analysed agents, especially for K15 and K20.

tivity of K15 and K20 was comparable or superior to that of

the analogous titanocene complexes. In addition, the propor-

tion of cells in mid- and late-stage apoptosis was superior to

that of the “free” titanocene derivatives 15 and 20 (Figure 11).

In DLD-1 cells, treatment with the titanocene-functionalised

nanostructures not only conveyed cell death, but also acceler-

ated the transition of cells to late apoptotic stages and in-

duced massive DNA damage (for further details, see Figure S9

in the Supporting Information).

Evaluation of the modulator effect on the intrinsic apoptotic

pathway

We have measured the expression of two antagonist proteins,

members of the Bcl-2 family, which have a very important

function in the intrinsic apoptotic pathway. The Bax-a proapop-

&

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

10

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

totic and Bcl-xL anti-apoptotic molecule expressions vary be-

tween different cells lines because they are tumours derived

from different organs. This observation is supported by numer-

ous reports in the literature.^[62–67]

Bearing in mind that the Bax-a/Bcl-xL ratio is a mathematical

parameter with relevance for promotion of the apoptotic pro-

cess, we can infer that when this ratio is higher than one,

apoptotic cell death is induced and when this ratio is much

lower than one and close to zero the cell is resistant to apop-

totic stimuli.^[64]For K15 and K20, this ratio is higher than one

in all the cell lines, with a single exception (material K15 in

MCF-7 breast cancer), and is much higher in comparison with

the untreated cell basal value. Thus, in A2780, the ratios are

4.897 for K20, 2.596 for K15; 1.366 for K20 in MCF-7 cells, and

1.056 for K20 and 2.887 for K15 in DLD-1 cells. The Bax-a/Bcl-

xL ratio induced by the materials is superior to the ratio corre-

sponding to the titanocene compounds in DLD-1 cells, for

which the ratio is 0.407 for 20 and 0.598 for 15. The same

trend is observed for titanocene 20 in the cell lines MCF-7

(ratio value 1.074) and A2780 (ratio value 3.58). In both cases,

the “free” titanocene action is inferior to that of the analogous

materials. This tendency is not followed by K15 in A2780 cells,

for which its ratio is lower than that observed for 15. Although

both K15 and K20 induce a large amount of apoptotic events,

the intrinsic pathway seems to be triggered more efficiently by

K20, whereas K15 appears to take other pathways.

In A2780 cells, treatment with 15 shows evident Bax-a over-

expression, whereas in the other cases the molecule is sup-

pressed (Figure 12). In MCF-7 breast carcinoma cells, 15, 20

and K20 lead to proapoptotic Bax-a upregulation in a statisti-

cally relevant proportion; the concentration of the molecule is

much higher than that observed upon treatment with oxalipla-

tin (Figure 12). In colorectal carcinoma DLD-1, Bax-a expression

is prominent for K15; for K20, suppression is less than that

provoked by oxaliplatin (all data compared with column statis-

tics, one-way analysis of variance, p<0.001; Figure 12).

Apoptotic cell death is regulated by complex interactions

between pro-survival members and two subgroups of pro-

apoptotic members of the B-cell lymphoma-2 (Bcl-2) protein

family. In the intrinsic apoptotic pathway, two crucial cell death

switch molecules have important regulatory functions: Bax-

a and Bcl-xL. The Bax-a protein promotes the apoptotic pro-

cesses, for which caspases activation is involved, whereas anti-

apoptotic Bcl-xL, which is a protein located in the mitochondri-

al membrane, acts in the direction of cell survival by blocking

the apoptotic signals.^[63]The two molecule expression is used

as a molecular marker in tumours and has an important prog-

nostic value in cancer diagnostics and chemotherapy follow

up.^[68]Modern molecular therapies use innovative drugs to

reach these molecular targets in breast cancer,^[69]ovarian carci-

noma, and other malignant diseases;^[70]consequently the ca-

pacity of K15 and K20 to modulate production of the two pro-

teins is very important in defining their biological potential,

and thus, their possible application in chemotherapy.

Apoptosis regulation at the DNA level: Inhibition of PARP-1

Figure 12. Modulation of Bax-a expression in untreated malignant cells and

malignant cells treated with oxaliplatin, 15, 20, K15 and K20 (top). Modula-

tion of Bcl-xL expression in untreated malignant cells and malignant cells

treated with oxaliplatin, 15, 20, K15 and K20 (bottom).

When cells are exposed to external factors, such as apoptosis

trigger molecules or drugs, and DNA damage occurs, cellular

stress responses are determined by specific regulatory process-

es. In the nucleus, the poly(ADP-ribose) polymerase-1 enzyme

(PARP-1) functions at the centre of cellular stress responses

and facilitates DNA repair by catalysing the ADP-ribosylation of

proteins.^[71]This is why PARP-1 is a target for anti-cancer drugs,

and its inhibition may be a way to enhance the apoptotic sig-

nalling pathways in tumour cells. The capacity of 15, 20, K15

and K20 to cleave PARP-1 was assessed after incubation of

a histone-coated plate with different dilutions of the com-

pounds/materials of interest for 1 h.

The anti-apoptotic Bcl-xL protein expression is downregulat-

ed in A2780 cells for the studied compounds and materials: 15

is strongly suppressed and for 20 and K20 downregulation is

more important than that caused by the platinum drug

(Figure 12). In MCF-7 cells, material K15 exhibits the most im-

portant suppression (higher than that caused by oxaliplatin),

whereas in DLD-1 cells 15, K15 and K20 downregulate the

anti-apoptotic molecule similarly to oxaliplatin (all data ana-

lysed by column statistics, two-tailed ANOVA test, Bonferroni

post-test, p=0.001).

All analysed agents display moderate PARP-1 inhibitory ac-

tivity, below that of the reference compound, 3-aminobenz-

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

11

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

amide (Figure 13). The maximum PARP-1 inhibitory activity is

above the IC₅₀concentration, especially for the titanocenes.

Thus, the activity of K15 and K20 is somewhat better than

that of 15 and 20. By taking into account that there is a re-

duced amount of active titanocene compounds in the meso-

porous materials, a plausible conclusion is that the titanocene

compounds anchored onto nanostructured materials are more

efficient than the “free” titanocene derivatives in counteracting

DNA damage repair.

that the expression is weaker compared with that of the other

tumour cells, and a reduction in the PARP-1 activity caused by

K20 and K15 is observed.

Bearing in mind that PARP-1 is implicated in the Fas-induced

extrinsic apoptosis pathway, in the caspases cascade, in

tumour growth factor b (TGF-b) receptor signalling and is of

utmost importance in cancer cell growth regulation, its cleav-

age provides evidence of DNA damage caused by the titano-

cene-functionalised materials K15 and K20. Thus, in breast car-

cinoma, data in the literature show that there is a significant

correlation between PARP expression and members of the Bcl-

2 family.^[69]In MCF-7 breast cancer cells,^[71]PARP-1 cleavage by

caspase-3 is an important step in the apoptotic signal pathway,

although, even in the absence of caspase-3, Bax-a overexpres-

sion induces DNA fragmentation and apoptosis. On the other

hand, apoptotic events occur in treated DLD-1 cells^[73]by PARP

fragmentation from its 113 kDa intact form to a 85 kDa cleaved

protein, following metal-based drug action. In addition, after

certain treatments, the 82 kDa band can appear, which is de-

tected by the anti-cleaved PARP (Asp214) antibody available.

Thus, PARP-1 can recognise DNA lesions caused by plati-

num-based drugs through PARP-1 binding to the damaged in-

trastrand cross-links.^[74]Titanocene compounds are also capa-

ble of causing DNA damage,^[50]and we expect a similar interac-

tion with PARP-1 and the material-damaged DNA.

The interconnections between phosphatidylserine transloca-

tion, Bax-a and Bcl-xL expression, and the PARP enzyme activi-

ty may be one reason why a decrease in PARP-1 cleavage and

DNA repair is observed for the three cell lines studied, even

though the compounds themselves are modest enzyme inhibi-

tors.

In 15 or K15 action against tumour cells, for which proapop-

totic Bax-a is overexpressed, significant poly-ADP-ribose cleav-

age was recorded, despite the weak PARP-1 inhibitory capacity

of the compounds.

Figure 13. a) Capacity of etoposide, 15, 20, K15 and K20 to inhibit the

PARP-1 enzyme and counteract DNA damage repair and b) DNA damage

caused by treatment with etoposide, 15, 20, K15 and K20 quantified by the

decrease of PARP-1 activity. In the treated malignant cells, the PARP-1 activity

decreases due to cleavage of the molecule during apoptosis.

Titanium release and cellular uptake

In this context, DNA damage caused by titanocene-function-

alised nanostructured materials K15 and K20 was quantified

by the magnitude of PARP-1 cleavage. The PARP-1 activity was

evaluated in material-treated cell lysates to establish the

amount of fragmented PARP-1. Etoposide is a recognised

apoptosis inducer drug and it was used as a reference com-

pound provided by the PARP-1 measuring kit.

In view of the interesting results obtained in all biological ex-

periments associated with 15, 20, K15 and K20, we decided to

explore the titanium uptake of the treated cells in an attempt

to correlate this with the cytotoxicity and other effects ob-

served previously. The first question that we had to try to

answer was whether there were some differences in the quan-

tities and/or way of titanium uptake by cells treated with the

titanium derivatives (15 and 20) or with the titanium-function-

alised materials (K15 and K20). Thus, cells were treated with

the studied compounds or materials at concentrations of

500 mm (15 and 20) and 500 mgmL^À1(K15 and K20); the final

concentration of Ti in the cell culture media was 23.91 mgmL^À1

for 15, 23.90 mgmL^À1for 20, and, according to the Ti-function-

alisation rates, 15.3 mgmL^À1in K15 and 24.25 mgmL^À1in K20.

After 24 h of treatment, surprisingly, the quantity of titanium

in populations of 10⁶cells was higher in the case of titanium-

functionalised materials (K15 and K20); incorporation ranging

from about 4 to 23% of the total amount of titanium was ob-

served, whereas 15 and 20 showed incorporation that ranged

In the A2780 cell population, the PARP-1 molecule is strong-

ly expressed and it has been reported that is cleaved by apop-

tosis-inducing drugs, such as platinum drugs.^[72]The analysed

materials K15 and K20 are also able to cleave PARP-1, but

their capacity is below the standard etoposide. In DLD-1 cells,

where PARP-1 is also overexpressed (higher levels outside the

limits of our assessment), the two nanostructured materials

K15 and K20 cause a higher level of DNA damage in the cells,

although the PARP-1 activity is decreased compared with the

etoposide treatment. However, this is an important aspect be-

cause standard drugs, such as cisplatin, are unable to cleave

this enzyme in DLD-1 cells.^[73]The results in MCF-7 cells show

&

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

12

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

Table 6. Titanium concentration in a population of 10⁶treated cancer cells and titanium incorporation efficiency calculated as titanium incorporated in

10⁶cells relative to the amount of titanium in the cell culture media.

A2780 Ti

[mg per 10⁶cells]

A2780 Ti

incorporation rate [%]

MCF-7 Ti

[mg per 10⁶cells]

MCF-7 Ti

incorporation rate [%]

DLD-1Ti

[mg per 10⁶cells]

DLD-1 Ti

incorporation rate [%]

15

20

K15

K20

110 Æ2

57Æ1

4.6

2.4

10.8

10.2

9Æ0.2

25Æ1

0.4

1.0

23.0

19.0

14Æ0.2

10Æ0.4

168Æ1.4

114Æ4

0.6

0.4

11.0

4.7

165Æ0.2

248Æ2

352Æ4

460Æ2

from about 0.4 to 4.6% titanium (Table 6; for further details,

see Tables S3–S5 in the Supporting Information). These results

indicate an interesting property of the titanium-functionalised

materials, which appear to be acting as a Trojan horse for the

incorporation of the titanium active species into the cells. This

leads to much higher quantities of titanium uptake in cells,

probably because permeation or adherence of the materials to

the cells is relatively high. What is clear from these experi-

ments is that the transport processes (either active or passive)

of “free” titanocene derivatives to the cell are less effective

than in the case of titanocene-functionalised materials; this

may be due to the tendency towards hydrolysis and/or decom-

position of titanocene compounds in physiological medium,

which leads to a decrease in the cytotoxicity due to the impos-

sibility of titanium reaching the cell. It seems, therefore, that

the most important steps associated with the in vitro proper-

ties of these compounds are the transport of active species to

the cell.

Thus, we have determined that the titanium uptake data in-

dicate that the Ti cellular uptake amount delineates the capaci-

ty of 15 and 20 to trigger apoptosis, whereas in the case of

the materials the Ti uptake amount is correlated with their cy-

totoxicity and PARP cleavage.

To be certain that this high quantity of titanium uptake in

cells treated with materials K15 and K20 is due to protection

of the titanium compounds by the mesoporous system KIT-6,

release studies were performed. Thus, taking into account that

the titanocene compounds supported on the KIT-6 matrix are

normally bound to the material through TiÀOÀSi bonds and

that these are normally stable in aqueous solutions at physio-

logical pH (as previously observed by our group),^[33–36]no re-

lease, or a very low amount, of titanium-containing water-solu-

ble species is expected for these materials. In this context, sus-

pensions of materials K15 and K20 in simulated body fluid

were incubated at 378C in a water bath for 72 h and the sus-

pensions subsequently filtered. When the filtrate was analysed

by inductively coupled plasma atomic emission spectroscopy

(ICP-AES), a very low titanium amount (lower than the 0.01%

of the titanium in each material) was observed, which indicat-

ed that no release of the titanocene derivative occurred in

these experiments.

In addition, from the data obtained in the uptake experi-

ments, we were able to correlate the titanium uptake with

some other properties discussed previously. Thus, in cancer

cells treated with materials K15 and K20, the Ti incorporation

rate is significantly correlated with the cytotoxicity, expressed

as M₅₀values (two-tailed correlation, Pearson r=0.8925 in the

95% confidence interval, p=0.0167; R²=0.966), whereas in

compounds 15 and 20 no such association was found.

However, the cellular uptake in cells treated with 15 and 20

is related to the number of apoptotic cells in the treated popu-

lation (the sum of early and late apoptotic cells, those with

Alexa Fluor stain and double PI/Alexa Fluor stain; Pearson r

value=0.8315, significant p=0.0402, and R²⁼0.6914). Addition-

ally, cellular inclusion of the compounds correlates significantly

with the Bax-a expression (two-tailed correlation, Pearson r

value 0.8267 in the 95% confidence interval, p=0.0425; R²=

0.6834). The Ti incorporation efficiency is also strongly related

with Bax-a expression (r=0.8755, p=0.0223; R²=0.7665). Fur-

thermore, Ti uptake is related with Bcl-xL expression when

tumour cells are treated with 15 and 20, showing a negative

correlation with the anti-apoptotic molecule, (r=À0.8597, p=

0.0281, 0.7392). The Ti incorporation rate is also negatively

connected to Bcl-xL with statistical significance (r=À0.9133,

p=0.0109, R²=0.8341).

The lack of release of titanium-soluble species confirms that

the cytotoxic properties of the corresponding titanocene-func-

tionalised materials is probably due to particle action and not

to the release of the corresponding titanocene compound or

some other titanium-containing active species. These results

show again that the mesoporous silica-based material there-

fore promotes higher titanium uptake by the cells.

It also seems plausible that the different TiCp₂moieties at-

tached to the surfaces, together with the titanium content of

the material, may have a real influence on the cytotoxicity and

that is why different materials with similar titanium contents,

but different supported compounds have completely different

cytotoxic properties. Therefore, the release of the titanocene

compound does not seem to be necessary for cytotoxicity.

However, functionalisation of materials with titanocene is cru-

cial because non-functionalised KIT-6 does not present cytotox-

ic properties.

Conclusion

On the other hand, in tumour cells treated with materials

K15 and K20, Ti uptake can be significantly correlated with the

PARP value (Pearson r=0.9289 in the 95% confidence interval,

two-tailed p value=0.0074, R²=0.8629).

A series of alkenyl-substituted titanocene compounds have

been synthesised and characterised. These compounds have

been subsequently grafted onto a nanostructured silica-based

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

13

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

d=1.79, 1.80 (s, 6H each; CMePh), 1.80–2.00 (m, 8H; CH₂CH₂-

CH=CH₂), 2.16 (m, 8H; CH₂CH₂CH=CH₂), 4.91 (m, 8H; CH₂CH₂CH=

CH₂), 5.22, 5.60, 6.11, 6.15, 6.29, 6.43, 6.51, 6.68 (m, 2H each; C₅H₄),

5.73 ppm (m, 4H; CH₂CH₂CH=CH₂); ¹³C{¹H} NMR (100 MHz, CDCl₃,

258C, for the mixture of isomers): d=23.3, 24.3 (CH₂CH₂CH=CH₂),

28.8, 28.9 (CH₂CH₂CH=CH₂), 41.2, 42.6 (CMePh), 44.4, 44.5 (CpC),

145.7, 145.8 (C₅H₄), 146.8, 147.0 (C₆H₅), 138.7, 138.8 ppm

(CH₂CH₂CH=CH₂); FTIR (KBr): n˜ =1636 (C=C), 2870, 2934 (C_al-H),

3060 cm^À1(C_ar-H); EI-MS: m/z (%): 529 (10) [M⁺ÀCl], 341 (15)

[M⁺ÀC₅H₄{CMePhCH₂CH₂CH=CH₂}], 250 (13) [M⁺ÀC₅H₄{CMe-

PhCH₂CH₂CH=CH₂}ÀCH₂CH₂CH=CH₂ÀClÀH], 185 (75) [M⁺ÀC₅H₄-

{CMePhCH₂CH₂CH=CH₂}ÀCMePhCH₂CH₂CH=CH₂+3H], 143 (77)

[CMePhCH₂CH₂CH⁺À2H] 117 (96) [M⁺À2C₅H₄{CMePhCH₂CH₂CH=

CH₂}], 105 (38) [CMePh⁺+H], 91 (100) [C₇H₇⁺]; elemental analysis

calcd (%) for C₃₄H₃₈Cl₂Ti (565.44 gmol^À1): C 72.22, H 6.77; found: C

72.61, H 6.63.

porous material, KIT-6, to give titanocene-functionalised mate-

rials, which have been synthesised and characterised by vari-

ous techniques. It has been observed that the titanocene de-

rivatives are normally bound, through m-oxo surface species, to

the external surface of the mesoporous matrix.

All of the synthesised compounds and materials have been

tested in vitro against different human cancer cell lines and

a normal human cell line. A very high activity of the studied ti-

tanocene derivatives has been observed and is much higher

than the majority of the titanocene derivatives reported in the

literature, and comparable to that of famous titanocene-Y,

which is one of the most potent metallodrugs based on titani-

um.

In addition, the titanocene-functionalised materials reported

herein, present much higher cytotoxic activity (lower M₅₀or Ti

M₅₀values) and selectivity than other similar nanostructured

systems functionalised with titanocene derivatives.

Synthesis of 17: This compound was synthesised by using the

same procedure as that described for 14, with 5 (20.00 mL, 0.75 g,

3.83 mmol, 0.19m in THF) and [Ti(h⁵-C₅H₅)Cl₃] (0.80 g, 3.65 mmol).

1

From studies of the apoptosis induction capacity, evaluation

of the modulator effect on the intrinsic apoptotic pathway,

apoptosis regulation at the DNA level and titanium uptake by

cells, one can conclude that the dynamics of apoptotic mor-

phological and functional changes is modified when the active

titanocenes are incorporated onto mesoporous materials. 15,

20, K15 and K20 are able to induce programmed cell death in

tumour cell populations, by impairing the damaged DNA

repair mechanisms and by up-regulation of intrinsic and extrin-

sic apoptotic signalling pathways. In addition, it seems that ti-

tanium-functionalised materials are acting as Trojan horses for

the incorporation of the titanium active species into the cells,

leading to much higher quantities of titanium uptake in cells,

so that transport processes (either active or passive) of “free”

titanocene derivatives to the cell are less effective than in the

case of titanocene-functionalised materials.

Yield: 0.36 g, 26%; H NMR (400 MHz, CDCl₃, 258C): d=1.34 (s, 6H;

CpCMe₂), 1.44, 1.65 (m, 2H each; CH₂CH₂CH=CMe₂), 1.49, 1.63 (s,

3H each; CH₂CH₂CH=CMe₂), 4.96 (m, 1H; CH₂CH₂CH=CMe₂), 6.49,

6.60 (m, 2H each; C₅H₄), 6.56 ppm (s, 5H; C₅H₅); ¹³C{¹H} NMR

(100 MHz, CDCl₃, 258C): d=17.8, 23.4 (CH₂CH₂CH=CMe₂), 25.9, 26.9

(CH₂CH₂CH=CMe₂), 37.6 (CMe₂), 47.3 (CpC), 118.8, 119.7, 148.9

(C₅H₄), 120.3 (C₅H₅), 124.3 (CH₂CH₂CH=CMe₂), 131.8 ppm

(CH₂CH₂CH=CMe₂); FTIR (KBr): n˜ =1637 (C=C), 2854, 2924, 2963

(C_al-H), 3108 cm^À1(C_ar-H); EI-MS: m/z (%): 373 (11) [M⁺+H], 337

(26) [M⁺ÀCl], 307 (28) [M⁺ÀCp], 300 (36) [M⁺À2Cl+2H], 253 (63)

[M⁺ÀTiÀ2Cl+H], 189 (29) [C₅H₄{CMe₂CH₂CH₂CH=CMe₂}⁺], 99 (65)

[Cp+ClÀH], 69 (43) [2ClÀH]; elemental analysis calcd (%) for

C₁₉H₂₆Cl₂Ti (373.18 gmol^À1): C 61.15, H 7.02; found: C 60.92, H 7.03.

Synthesis of 18: This compound was synthesised by using the

same procedure as that described for 14, with 6 (24.00 mL, 0.74 g,

2.88 mmol, 0.12m in THF) and [Ti(h⁵-C₅H₅)Cl₃] (0.61 g, 2.77 mmol).

Yield: 0.74 g, 58%; ¹H NMR (400 MHz, CDCl₃, 258C): d=1.46, 1.64

(s, 3H each; CH₂CH₂CH=CMe₂), 1.81 (s, 3H; CMePh), 1.98, 2.18 (m,

2H each; CH₂CH₂CH=CMe₂), 5.05 (m, 1H; CH₂CH₂CH=CMe₂), 6.04,

6.43 (m, 2H each; C₅H₄), 6.13 (s, 5H; C₅H₅), 6.87 (t, 1H; H_paraof Ph

ring), 7.27 (m, 2H; H_metaof Ph ring), 7.38 ppm (m, 2H; H_orthoof Ph

ring); ¹³C{¹H} NMR (100 MHz, CDCl₃, 258C): d=17.8, 23.4

(CH₂CH₂CH=CMe₂), 23.8, 25.9 (CH₂CH₂CH=CMe₂), 43.8 (CMePh), 44.4

(CpC), 113.8 (CH₂CH₂CH=CMe₂), 117.2, 122.4, 146.2 (C₅H₄), 120.8

(C₅H₅), 124.1, 129.9, 127.7, 128.6 (C₆H₅), 145.3 ppm (CH₂CH₂

CH=CMe₂); FTIR (KBr): n˜ =1672 (C=C), 2863, 2923, 2965 (C_al-H),

3054, 3104 cm^À1(C_ar-H); EI-MS: m/z (%): 435 (20) [M⁺], 399 (91)

[M⁺ÀCl], 369 (41) [M⁺ÀCp], 251 (22) [C₅H₄{CMePhCH₂CH₂CH=

CMe₂}⁺], 181 (17) [M⁺ÀC₅H₄{CMePhCH₂CH₂CH=CMe₂}ÀH], 99 (100)

[Cp+Cl+H], 91 (11) [C₇H₇⁺], 69 (74) [2ClÀH]; elemental analysis

calcd (%) for C₂₄H₂₈Cl₂Ti (435.25 gmol^À1): C 66.23, H 6.48; found: C

66.05, H 6.36.

Therefore, the functionalised nanostructured materials K15

and K20 are good candidates for targeted anti-cancer therapy,

not only because of their interesting biological properties,

which have been reported herein, but also because of the

higher permeation of the membranes of cancer cells in com-

parison with healthy cells. This may be advantageous for the

specific use of these materials in chemotherapy, especially

against bone tumours due to the physical properties of silica-

based materials.

Experimental Section

Synthesis of titanocene complexes

Synthesis of 19: This compound was synthesised by using the

same procedure as that described for 14, with 5 (30 mL, 1.12 g,

5.70 mmol, 0.19m in THF) and [TiCl₄(thf)₂] (0.95 g, 2.85 mmol).

Complexes 11,^[44]12,^[75]13,^[75]15,^[75]and 16^[75]were prepared ac-

cording to synthetic protocols described by us previously.

1

Synthesis of 14: A solution of 4 (1.79 g, 7.79 mmol) in THF (50 mL)

was added dropwise over 10 min to a solution of [TiCl₄(thf)₂]

(1.30 g, 3.89 mmol) in THF (50 mL) at 08C. The reaction mixture

was allowed to warm to room temperature and stirred for 2–3 h.

The solvent was then removed in vacuo and a 9:1 mixture of tolu-

ene/hexane (50 mL) was added to the resulting solid. The suspen-

sion was filtered to remove LiCl and the filtrate was concentrated

(20 mL) and cooled to À308C to give 14 as a red solid (1.39 g,

Yield: 0.16 g, 6%; H NMR (400 MHz, CDCl₃, 258C): d=1.35 (s, 12H;

CpCMe₂), 1.42, 1.62 (m, 4H each; CH₂CH₂CH=CMe₂), 1.47, 1.61 (s,

4H each; CH₂CH₂CH=CMe₂), 4.94 (m, 2H; CH₂CH₂CH=CMe₂), 6.43,

6.50 ppm (m, 4H each; C₅H₄); ¹³C{¹H} NMR (100 MHz, CDCl₃, 258C):

d=17.8, 23.4 (CH₂CH₂CH=CMe₂), 25.9, 26.8 (CH₂CH₂CH=CMe₂), 37.6

(CMe₂), 47.5 (CpC), 117.6, 120.3, 148.6 (C₅H₄), 124.6 (CH₂CH₂

CH=CMe₂), 131.7 ppm (CH₂CH₂CH=CMe₂); FTIR (KBr): n˜ =1672

(C=C), 2864, 2921, 2965 (C_al-H), 3086, 3116 cm^À1(C_ar-H); EI-MS: m/z

(%): 461 (68) [M⁺ÀCl], 377 (7) [M⁺ÀTiÀ2ClÀH], 189 (73)

1

63%). H NMR (400 MHz, CDCl₃, 258C, for the mixture of isomers):

&

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

14

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

[C₅H₄{CMe₂CH₂CH₂CH=CMe₂}⁺], 69 (100) [CH₂CH=CMe₂⁺]; elemen-

tal analysis calcd (%) for C₂₈H₄₂Cl₂Ti (497.41 gmol^À1): C 67.61, H

8.51; found: C 67.22, H 8.43.

K14: ¹H MAS NMR: d=1.5–3 (CMePhCH₂CH₂CH=CH₂), 4–6

(CMePhCH₂CH₂CH=CH₂, and C₅H₄), 7–8 ppm (CMePhCH₂CH₂

CH=CH₂); ¹³C CP MAS NMR: d=17, 30 (CMePhCH₂CH₂CH=CH₂), 34,

40

(CMePhCH₂CH₂CH=CH₂),

110–118,

125–140 ppm

Synthesis of 20: This compound was synthesised by using the

same procedure as that described for 14, with 6 (26.00 mL, 0.80 g,

3.12 mmol, 0.12m in THF) and [TiCl₄(thf)₂] (0.52 g, 1.56 mmol).

(CMePhCH₂CH₂CH=CH₂and C₅H₄); ²⁹Si MAS NMR: d=À111.9 (Q⁴),

À103.4 (Q³), À91.8 ppm (Q²); FTIR (KBr): n˜ =454 (physisorbed

water), 796 (Si-OH), 1079 (Si-O-Si), 1383 (C-H bending), 1636 (physi-

1

Yield: 0.43 g, 22%; H NMR (400 MHz, CDCl₃, 258C, for the mixture

sorbed water), 2923, 2973, 2986 (C_al-H), 3431 cm^À1(O-H); Ti_exp

2.26%; S_BET=469 m²g^À1; V_p=0.42 cm³g^À1; D_p=68.7 .

=

of isomers): d=1.43, 1.44, 1.61, 1.62 (s, 6H each; CH₂CH₂CH=CMe₂),

1.79 (s, 12H; CMePh), 1.70–2.20 (m, 16H; CH₂CH₂CH=CMe₂), 5.03

(m, 4H; CH₂CH₂CH=CMe₂), 5.16, 5.58, 6.07, 6.14, 6.27, 6.43, 6.51,

6.69 (m, 2H each; C₅H₄), 7.00–7.50 ppm (m, 20H; C₆H₅);

¹³C{¹H} NMR (100 MHz, CDCl₃, 258C): d=17.7, 17.7, 23.1, 23.2

(CH₂CH₂CH=CMe₂), 23.0, 24.0, 25.8, 25.8 (CH₂CH₂CH=CMe₂), 42.3,

43.7 (CMePh), 44.4, 44.5 (CpC), 112.2, 115.9 (CH₂CH₂CH=CMe₂),

117.8, 120.8, 120.8, 121.4, 122.9, 145.7, 145.9, 146.8, 147.0 (C₅H₄),

124.1, 124.3, 126.4, 126.5, 126.8, 127.5, 128.3, 128.3 (C₆H₅), 131.8,

131.8 ppm (CH₂CH₂CH=CMe₂); FTIR (KBr): n˜ =1628 (C=C), 2874,

2928, 2966 (C_al-H), 3064, 3101 cm^À1(C_ar-H); EI-MS: m/z (%): 585 (28)

[M⁺ÀCl], 369 (100) [M⁺ÀC₅H₄{CMePhCH₂CH₂CH=CMe₂}⁺], 334 (17)

[M⁺ÀC₅H₄{CMePhCH₂CH₂CH=CMe₂}ÀCl], 250 (47) [C₅H₄{CMe-

PhCH₂CH₂CH=CMe₂}⁺], 91 (68) [C₇H₇⁺], 69 (89) [2ClÀH]; elemental

analysis calcd (%) for C₃₈H₄₆Cl₂Ti (621.54 gmol^À1): C 73.43, H 7.46;

found: C 72.64, H 7.22.

K15: ¹H MAS NMR: d=0–1 (SiMe₃), 2.5–7.5 ppm (CMe₂CH₂CH₂

CH=CH₂, C₅H₃and C₅H₅); ¹³C CP MAS NMR: d=À2.6 (SiMe₃), 24, 27,

31 (CMe₂CH₂CH₂CH=CH₂), 40 (CMe₂CH₂CH₂CH=CH₂), 117, 126, 131,

144 ppm (CMe₂CH₂CH₂CH=CH₂, C₅H₃and C₅H₅); ²⁹Si MAS NMR: d=

À112.4 (Q⁴), À103.7 (Q³), À91.9 ppm (Q²); FTIR (KBr): n˜ =457 (physi-

sorbed water), 796 (Si-OH), 1082 (Si-O-Si), 1383 (C-H bending), 1621

(physisorbed water), 2925, 2933, 2970 (C_al-H), 3434 cm^À1(O-H);

Ti_exp=3.06%; S_BET=561 m²g^À1; V_p=0.44 cm³g^À1; D_p=63.8 .

1

K16: H MAS NMR: d=0–1 (SiMe₃), 2–4 (CMePhCH₂CH₂CH=CH₂), 6–

8 ppm (CMePhCH₂CH₂CH=CH₂, C₅H₃and C₅H₅); ¹³C CP MAS NMR:

d=À3.4 (SiMe₃), 26, 37, 42, 47 (CMePhCH₂CH₂CH=CH₂), 117,

127 ppm (CMePhCH₂CH₂CH=CH₂, C₅H₃and C₅H₅); ²⁹Si MAS NMR:

d=À112.4 (Q⁴), À103.4 (Q³), À92.1 ppm (Q²); FTIR (KBr): n˜ =456

(physisorbed water), 805 (Si-OH), 1079 (Si-O-Si), 1383 (C-H bend-

ing), 1636 (physisorbed water), 2924, 2963 (C_al-H), 3431 cm^À1(O-H);

Ti_exp=2.43%; S_BET=532 m²g^À1; V_p=0.59 cm³g^À1; D_p=64.3 .

Synthesis of materials incorporating the titanocene com-

plexes (K11–K20)

K17: ¹H MAS NMR: d=1–4.5 (CMe₂CH₂CH₂CH=CMe₂), 4.5–6.5 ppm

(CMe₂CH₂CH₂CH=CMe₂, C₅H₄and C₅H₅); ¹³C CP MAS NMR: d=14,

21, 22 (CMe₂CH₂CH₂CH=CMe₂), 34 (CMe₂CH₂CH₂CH=CMe₂), 60

A solution of the titanocene derivative (the quantity to obtain

a theoretical level of 5% Ti/SiO₂) in toluene (100 mL) was added to

dehydrated KIT-6 (1.00 g) and the mixture was stirred overnight at

808C. The slurry was filtered through fritted discs and the solid res-

idue was washed with toluene (5200 mL). Finally, the resultant

solid was dried under vacuum at room temperature for 24 h to

give a pale-red, free-flowing powder.

(CMe₂CH₂CH₂CH=CMe₂),

112,

128,

135,

156,

165 ppm

(CMe₂CH₂CH₂CH=CMe₂, C₅H₄and C₅H₅); ²⁹Si MAS NMR: d=À112.5

(Q⁴), À103.0 (Q³), À92.4 ppm (Q²); FTIR (KBr): n˜ =456 (physisorbed

water), 800 (Si-OH), 1081 (Si-O-Si), 1383 (C-H bending), 1636 (physi-

sorbed water), 2854, 2922, 2971 (C_al-H), 3432 cm^À1(O-H); Ti_exp

=

2.23%; S_BET=492 m²g^À1; V_p=0.44 cm³g^À1; D_p=64.3 .

K11: ¹H MAS NMR: d=2–3 (CMe₂CH₂CH₂CH=CH₂), 5–8 ppm

(CMe₂CH₂CH₂CH=CH₂, C₅H₄and C₅H₅); ¹³C CP MAS NMR: d=23, 27

(CMe₂CH₂CH₂CH=CH₂), 35 (CMe₂CH₂CH₂CH=CH₂), 112 (CMe₂-

CH₂CH₂CH=CH₂), 118, 124, 127 (C₅H₄and C₅H₅), 138 ppm (CMe₂-

CH₂CH₂CH=CH₂); ²⁹Si MAS NMR: d=À111.5 (Q⁴), À102.9 (Q³),

À93.9 ppm (Q²); FTIR (KBr): n˜ =801 (Si-OH), 1094 (Si-O-Si), 1383

(C-H bending), 1629 (physisorbed water), 2854, 2922, 2962 (C_al-H

K18: ¹H MAS NMR: d=1–3.5 (CMePhCH₂CH₂CH=CMe₂), 4.5–7

(CMePhCH₂CH₂CH=CMe₂, C₅H₄and C₅H₅), 7–8.5 ppm

(CMePhCH₂CH₂CH=CMe₂); ¹³C CP MAS NMR: d=11, 23, 27, 33

(CMePhCH₂CH₂CH=CMe₂), 57 (CMePhCH₂CH₂CH=CMe₂), 117, 124,

141 ppm (CMePhCH₂CH₂CH=CMe₂, C₅H₄and C₅H₅); ²⁹Si MAS NMR:

d=À112.1 (Q⁴), À102.6 (Q³), À92.1 ppm (Q²); FTIR (KBr): n˜ =456

(physisorbed water), 801 (Si-OH), 1081 (Si-O-Si), 1419 (C-H bend-

ing), 1636 (physisorbed water), 2854, 2922, 2971 (C_al-H), 3431 cm^À1

(O-H); Ti_exp=2.00%; S_BET=442 m²g^À1; V_p=0.41 cm³g^À1; D_p=63.9 .

stretching), and 3440 cm^À1(O-H); Ti_exp=2.13%; S_BET=575 m²g^À1

V_p=0.63 cm³g^À1; D_p=63.6 .

;

K12: ¹H MAS NMR: d=2–3 (CMePhCH₂CH₂CH=CH₂), 4–6.5

(CMePhCH₂CH₂CH=CH₂, C₅H₄and C₅H₅), 7–8 ppm (CMe-

1

K19: H MAS NMR: d=1.5–3.5 (CMe₂CH₂CH₂CH=CMe₂), 4.5–6.5 ppm

PhCH₂CH₂CH=CH₂);

¹³C

CP

MAS

NMR:

d=18,

26

(CMe₂CH₂CH₂CH=CMe₂and C₅H₄); ¹³C CP MAS NMR: d=15, 19, 23

(CMe₂CH₂CH₂CH=CMe₂), 34 (CMe₂CH₂CH₂CH=CMe₂), 63 (CMe₂-

CH₂CH₂CH=CMe₂), 114, 127, 134, 162, 168 ppm (CMe₂CH₂CH₂

CH=CMe₂and C₅H₄); ²⁹Si MAS NMR: d=À112.0 (Q⁴), À102.1 (Q³),

À92.6 ppm (Q²); FTIR (KBr): n˜ =456 (physisorbed water), 798 (Si-

OH), 1082 (Si-O-Si), 1387 (C-H bending), 1636 (physisorbed water),

(CMePhCH₂CH₂CH=CH₂), 38, 42 (CMePhCH₂CH₂CH=CH₂), 115–120,

123–132 ppm (CMePhCH₂CH₂CH=CH₂, C₅H₄and C₅H₅); ²⁹Si MAS

NMR: d=À112.6 (Q⁴), À104.9 (Q³), À92.3 ppm (Q²); FTIR (KBr): n˜ =

457 (physisorbed water), 805 (Si-OH), 1080 (Si-O-Si), 1384 (C-H

bending), 1629 (physisorbed water), 2857, 2926, 2963 (C_al-H

stretching), 3441 cm^À1(O-H); Ti_exp=1.20%; S_BET=592 m²g^À1; V_p=

0.67 cm³g^À1; D_p=63.7 .

2925, 2972 (C_al-H), 3444 cm^À1(O-H); Ti_exp=2.43%; S_BET=365 m²g^À1

V_p=0.37 cm³g^À1; D_p=63.9 .

;

K13: ¹H MAS NMR: d=2.5–6.5 (CMe₂CH₂CH₂CH=CH₂), 6–7 ppm

(C₅H₄); ¹³C CP MAS NMR: d=23, 27 (CMe₂CH₂CH₂CH=CH₂), 34

(CMe₂CH₂CH₂CH=CH₂), 44 (CMe₂CH₂CH₂CH=CH₂), 110, 115, 127

(CMe₂CH₂CH₂CH=CH₂and C₅H₄), 137 ppm (CMe₂CH₂CH₂CH=CH₂);

²⁹Si MAS NMR: d=À112.1 (Q⁴), À103.7 (Q³), À92.0 ppm (Q²); FTIR

(KBr): n˜ =458 (physisorbed water), 802 (Si-OH), 1077 (Si-O-Si), 1383

(C-H bending), 1635 (physisorbed water), 2859, 2926, 2965 (C_al-H),

K20: ¹H MAS NMR: d=1–4 (CMePhCH₂CH₂CH=CMe₂), 4–6.5

(CMePhCH₂CH₂CH=CMe₂and C₅H₄), 7–8 ppm (CMePhCH₂CH₂CH=

CMe₂); ¹³C CP MAS NMR: d=11–57 (CMePhCH₂CH₂CH=CMe₂), 117,

124, 141 ppm (CMePhCH₂CH₂CH=CMe₂and C₅H₄); ²⁹Si MAS NMR:

d=À112.3 (Q⁴), À102.1 (Q³), À92.9 ppm (Q²); FTIR (KBr): n˜ =456

(physisorbed water), 801 (Si-OH), 1082 (Si-O-Si), 1383 (C-H bend-

ing), 1636 (physisorbed water), 2934, 2978 (C_al-H), 3431 cm^À1(O-H);

Ti_exp=4.85%; S_BET=348 m²g^À1; V_p=0.33 cm³g^À1; D_p=64.2 .

3431 cm^À1(O-H); Ti_exp=2.17%; S_BET=610 m²g^À1; V_p=0.68 cm³g^À1

D_p=63.7 .

;

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

15

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

(from Nunc, Thermo Scientific) at

a

concentration of 5

In vitro studies

10⁵cellsmL^À1media, and incubated for 24 h to obtain a well-at-

tached 80–90% confluent population. They were treated for 8 h

with a final concentration above their IC₅₀values of the com-

pounds: 50 mm titanocenes in the cell culture media, 500 mgmL^À1

titanocene-containing nanomaterials and 25 mm oxaliplatin; con-

centrations were proportional to their IC₅₀(or M₅₀) values and they

were chosen to exhibit an effect after just 8 h of treatment. After

treatment, the cells were detached from the plate, washed and re-

suspended in PBS.

Cell cultures: Biological evaluations were performed in vitro with

the following cell lines: A2780 human ovary carcinoma, MCF-7

human breast adenocarcinoma, DLD-1 colon carcinoma and HEK-

293 human normal epithelial kidney cell lines; these were acquired

from the European Collection of Cell Cultures through Sigma Al-

drich Company (St Louis, Missouri, USA).

A2780 and DLD-1 cell lines were cultivated in RPMI-1640 culture

medium supplemented with 2 mm glutamine and 10% foetal

bovine serum (FCS); MCF-7 cells required minimum essential

medium eagle (MEM) with 2 mm glutamine, 1% non-essential

amino acids (NEA) and 10% FCS, and for 293 cell line cultivation

we used Dulbecco’s modified eagle medium with 2 mm glutamine,

1% NEA and 10% FCS; all culture media and nutrients were ob-

tained from Sigma Aldrich.

We employed the Alexa Fluor 488 Annexin V/PI Apoptosis Kit for

flow cytometry, a Molecular Probes product (Eugene, OR, USA), fol-

lowing the manufacturer’s indications. This method distinguished

the apoptotic cells within a population, based on characteristic

morphological and biochemical changes, one of which was the

translocation of phosphatidylserine (PS) on the cell membrane

outer leaflet. The Alexa Fluor labelled Annexin V binds to PS locat-

ed on the cell outer surface, and this confers a photostable green

colouration to the apoptotic cells, which is detectable at l=

488 nm by the flow cytometer. In addition, propidium iodide (PI),

a nucleic acid binding dye, penetrated inside the dead cells and in-

dicated the viability of the population. After treatment with the

two dyes, cells were subjected to flow cytometry measurement by

using a method described earlier.^[77]The two colourations were

quantified and depicted by using the histograms of the FACS

Canto equipment (Beckton, Dickinson and Company, Franklin

Lakes, NJ, USA).

Cells were cultivated under sterile conditions by using a CO₂Series

incubator from Uniequip GmbH (Planegg, Germany) at a constant

temperature of 378C and 5% CO₂in cell culture flasks. The cell

passages were made at subconfluency with enzymatic methods by

using 0.25% trypsin ethylenediaminetetraacetic acid (EDTA; from

Sigma). Cell viability was evaluated microscopically through Trypan

Blue staining.

General instrumentation used for cell cultures were as follows:

R320 centrifuge with a swing-out rotor from Hettich (Tuttlingen,

Germany),

a Sartorius analytical scale (Goettingen, Germany),

a Heto Ultra Freeze À808C freezer (Heto Holten, Allerod, Denmark),

an MVE Cryosystem 2000 liquid nitrogen tank (MVE Cryogenics,

Ball Ground, GA, USA), a CKX41 inverted-phase fluorescence micro-

scope (Olympus, Hamburg, Germany), and a Titramax 1000 shaker

with incubator (Heidolph, Schwabach, Germany).

Cell signalling: The expression of apoptosis-related molecules Bcl-

xL and Bax-a were evaluated quantitatively with the enzyme-

linked immunosorbent assay (ELISA). For this purpose, cells were

treated in the same manner as those for the apoptosis assay; as

a reference we used untreated control cells. We used a final con-

centration derived from IC₅₀(or M₅₀) values of the compounds:

20 mm titanocenes in the cell culture media, 200 mgmL^À1titano-

cene-containing materials and 10 mm oxaliplatin. Samples were

subjected to lysis in FNN0011 cell extraction buffer (from Molecular

Probes, Eugene, OR, USA) on ice. Cell lysates were centrifuged at

2000 rpm and aliquots of supernatant were prepared for the tests.

Cytotoxicity: Compounds were dissolved in DMSO (from Fluka

Chemie, Bichs, Switzerland). The dilutions of solvent used were

tested on the four cell lines for their toxicity and the results were

negative. The final concentrations of the compounds were in the

range of 0.3–1500 mm and for the materials we used concentra-

tions between 1 and 2500 mgmL^À1

.

To assess the cytotoxicity of the compounds, we used the colouri-

metric MTT cell viability assay, which is an optimised method of de-

termining viable cell number in proliferation and cytotoxicity stud-

ies.^[76]This test is based on the cleavage of the yellow thiazolyl

blue tetrazolium bromide salt (MTT) to form a water-soluble blue

formazan product by mitochondrial enzymes of the living cells.

Thus, the amount of formazan produced is directly proportional to

the number of viable cells present during MTT exposure.

The protein concentration in cell lysates was measured by the

Bradford method,^[78]using a UV/Vis spectrophotometer (Jasco Ana-

lytical Instruments, Easton, MD, USA). All samples were adjusted to

a protein concentration of 50 mgmL^À1by dilution in PBS.

The DuoSet intracellular human total Bcl-xL ELISA kit and the

DuoSet intracellular human total Bax-a ELISA kit were from R&D

Systems (Minneapolis, MN, USA); reagents were employed accord-

ing to the manufacturer’s indications. All experiments were per-

formed in duplicate. Colourimetric measurements were performed

on a Sunrise Elisa reader and an automated plate washer (both

from Tecan, Männedorf, Germany). The standard values provided

by the pure recombinant human Bax-a and Bcl-xL proteins allowed

the Magellan software to draw the calibration curve and to com-

pute the concentration of relevant proteins from each well.

The cells were plated on 96-well Nunclon TC cell culture micro-

plates (Nalgene Nunc, through Thermo Scientific Company, Wal-

tham, MA, USA), at a 210⁴cell/well density in 200 mL of media.

After 24 h, incubation cells were treated with the serial dilution of

compounds in triplicate. For every compound, colour controls

were used, by treating wells containing cell culture media only

(without cells). As a viability control, we used the untreated cells,

and oxaliplatin (Oxaliplatin Accord, Accord Healthcare Ltd, North

Harrow, Great Britain) was the cytostatic reference drug. Oxaliplatin

was diluted in phosphate buffered saline solution (PBS, Sigma Al-

drich); the concentrations used were in the 0.01–100 mm range.

PARP inhibitory capacity: The PARP inhibitory capacity of new

compounds was assessed by using the HT Universal Colourimetric

PARP assay kit with histone-coated strip wells from Trevigen Incor-

poration (Gaithersburg, MD, USA). For this purpose, the studied

compounds were diluted in PARP buffer; we used a range of 10–

150 mgmL^À1final concentration/well interval for 15, 20, K15 and

K20, and we used two different wells for each concentration. The

capacity of the synthesised compounds to inhibit PARP-HSA

enzyme activity was compared with the reference, 3-aminobenz-

amide, provided by the kit. PARP-1 activity modulation of the new

compounds was tested by measuring the biotinylated poly-ADP-

The spectrophotometric measurements were made by using a Syn-

ergy 2 microplate reader (BioTek Instruments, Winooski, VT, USA) at

l=492 nm. The cells were subjected to titanocene derivate and

material action for 24 h.

Apoptosis evaluation: For apoptosis assay, a flow cytometric eval-

uation was performed. Cells were plated on six-well multi-dishes

&

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

16

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

ribose on histone proteins from the microplate surface. This accu-

mulation is proportional with the magnitude of PARP-1 cleavage,

and can be measured colourimetrically.

search (grant no. PN-II-ID-PCE-2011-3-1057) supported the eval-

uation of the biological activity of the compounds and materi-

als.

PARP-1 cleavage in treated cells: The PARP activity in cell extracts

during apoptosis was assessed by using the HT Colourimetric PARP

Apoptosis Assay Kit (Trevigen), which also included histone-coated

well strips. We used the cell lysates obtained from treated and un-

treated cell, as described earlier, in the amount of 25 mL/well. In

this experiment, lysates from K15 and K20 were utilised; all meas-

urements were made in duplicate. The baseline sample, before

apoptosis, for each cell line was considered as the untreated cell

batch. As a reference, we used etoposide, which was a cytostatic

drug provided by the kit. Cells were treated with etoposide as fol-

lows: 20 mm for A2780 and MCF-7 cell line, whereas for DLD-1 we

used a final concentration of 50 mm. The PARP inhibitory activity of

each compound was determined semi-quantitatively based on

poly-ADP-ribose deposited on the plate and visualised by anti-

PARP monoclonal antibody binding.

Keywords: antitumour agents · apoptosis · cytotoxicity ·

mesoporous materials · titanium

[1] See, for example: I. Ali, W. A. Wani, K. Saleem, A. Haque, Anticancer

Agents Med. Chem. 2013, 13, 296–306.

[2] C. G. Hartinger, P. J. Dyson, Chem. Soc. Rev. 2009 , 38 , 39 1 –4 01.

[3] P. C. A. Bruijnincx, P. J. Sadler, Adv. Inorg. Chem. 2009 , 61 , 1 –6 2.

[4] C.-H. Zhou, Y.-Y. Zhang, C.-Y. Yan, K. Wan, L.-L. Gan, Y. Shi, Anticancer

Agents Med. Chem. 2010, 10, 371–395.

[5] S. Komeda, A. Casini, Curr. Top. Med. Chem. 2012, 12, 219–235.

[6] E. MelØndez, Inorg. Chim. Acta 2012 , 393 , 3 6 –5 2.

[7] C. Nardon, G. Boscutti, D. Fregona, Anticancer Res. 2014, 34, 487–492.

[8] C. Sanchez-Cano, M. J. Hannon, Dalton Trans. 2009, 48, 10702–10711.

[9] S. Gómez-Ruiz, New Directions in the Fight against Cancer: From Metal

Complexes to Nanostructured Materials, in Molecules at Work: Self-Assem-

bly, Nanomaterials, Molecular Machinery (Ed.: B. Pignataro), Wiley-VCH,

Weinheim, 2012.

In both methods the enzyme concentration increase was moni-

tored by TACS-Sapphire colourimetric substrate and measured with

the Synergy 2 microplate reader.

Cellular titanium uptake: For cellular uptake of titanocene com-

pounds and materials, the malignant cells (A2780, MCF-7 and DLD-

1) were cultivated on 25 cm²Nunclon delta surface culture flasks

for 24 h and after total attachment of the cells to surface, at 70–

80% subconfluency, they were treated with the studied com-

pounds, at the concentrations of 500 mm compounds and

500 mgmL^À1materials. The final concentration of Ti in the cell cul-

ture media was 23.91 mgmL^À1for 15 and 23.90 mgmL^À1for 20,

whereas for the materials the Ti concentration in the cell environ-

ment was 15.3 mgmL^À1in K15 and 24.25 mgmL^À1in K20. After

24 h, treatment cells were collected by trypsinisation in centrifuge

tubes; supernatant was also collected to recover the detached

cells. After gentle centrifugation, cell counting and two washes

with PBS, cell pellets were counted immediately and frozen in cryo-

tubes at À708C; every sample contained 10⁶cells suspended in

PBS (200 mL). The compounds and materials had the ability to

attach themselves to the cell surface, with very strong bonds, and

even after repeated washing the pellets were visibly coloured be-

cause of the presence of the compounds not only on their outer

membrane, but inside the cell.

[10] A. R. Kennedy, A. J. Florence, F. J. McInnes, N. J. Wheate, Dalton Trans.

2009, 37, 7695–7700.

[11] C. C. L. Pereira, C. V. Diogo, A. Burgeiro, P. J. Oliveira, M. P. M. Marques,

S. S. Braga, F. A. A. Paz, M. Pillinger, I. S. GonÅalves, Organometallics

2008 , 27 , 494 8 –4 956.

[12] A. M. Krause-Heuer, N. J. Wheate, M. J. Tilby, D. G. Pearson, C. J. Ottley,

J. R. Aldrich-Wright, Inorg. Chem. 2008 , 47 , 688 0 –6 888.

[13] D. Plazuk, A. Vessires, E. A. Hillard, O. Buriez, E. LabbØ, P. Pigeon, M.-A.

Plamont, C. Amatore, J. Zakrzewski, G. Jaouen, J. Med. Chem. 2009, 52,

4964–4967.

[14] M. Galanski, B. K. Keppler, Anticancer Agents Med. Chem. 2007, 7, 55–73.

[15] E. Allard, C. Passirani, E. Garcion, P. Pigeon, A. Vessires, G. Jaouen, J. P.

Benoit, J. Control. Release 2008 , 130 , 14 6 –1 53.

[16] S. Dhar, L. Zhuang, J. Thomale, H. Dai, S. J. Lippard, J. Am. Chem. Soc.

2008 , 130 , 1146 7 –1 1476.

[17] A. A. Bhirde, V. Patel, J. Gavard, G. Zhang, A. A. Sousa, A. Masedunskas,

R. D. Leapman, R. Weigert, J. S. Gutkind, J. F. Rusling, ACS Nano 2009 , 3 ,

30 7 –3 16.

[18] S. Dhar, F. X. Gu, R. Langer, O. C. Farokhzad, S. J. Lippard, Proc. Natl.

Acad. Sci. USA 2008 , 105 , 1735 6 –1 7361.

[19] J. M. Dominguez-Vera, J. Inorg. Biochem. 2004, 98, 469–472.

[20] See, for example: B. Therrien, G. Suess-Fink, P. Govindaswamy, A. K. Ren-

frew, P. J. Dyson, Angew. Chem. 2008 , 120 , 383 3 –3 836; Angew. Chem.

Int. Ed. 2008 , 47 , 377 3 –3 778.

[21] K. Ajima, T. Murakami, Y. Mizoguchi, K. Tsuchida, T. Ichihashi, S. Iijima, M.

Yudasaka, ACS Nano 2008 , 2 , 205 7 –2 064.

[22] R. P. Feazell, N. Nakayama-Ratchford, H. Dai, S. J. Lippard, J. Am. Chem.

Soc. 2007 , 129 , 843 8 –8 439.

[23] B. W. Harper, A. M. Krause-Heuer, M. P. Grant, M. Manohar, K. B. Gar-

butcheon-Singh, J. R. Aldrich-Wright, Chem. Eur. J. 2010 , 16 , 706 4 –7 077.

[24] W. J. Rieter, K. M. Pott, K. M. L. Taylor, W. Lin, J. Am. Chem. Soc. 2008 ,

130 , 1158 4 –1 1585.

[25] A. Nguyen, V. Marsaud, C. Bouclier, S. Top, A. Vessires, P. Pigeon, R.

Gref, P. Legrand, G. Jaouen, J.-M. Renoir, Int. J. Pharm. 2008 , 347 , 12 8 –

135.

Quantification of the titanium content was performed by ICP-AES

by using a Varian Vista AX Pro Varian 720-ES instrument.

Study of titanium release: The titanium release of the materials

under biological conditions was performed in a blood simulating

fluid. This consisted of a pH 7.4 buffer prepared according to previ-

ously reported procedures.^[79]In duplicate, 1 mL of simulated blood

fluid was added to 10 mg of the studied materials (this ratio was

used to guarantee that the quantity associated with the release of

1% Ti was within the detection limits of the inductively coupled

plasma atomic emission spectrometer). The suspensions were incu-

bated at 378C in a water bath for different time periods up to

72 h. Afterwards, the suspension was subsequently filtered and the

filtrate was analysed by ICP-AES by using a Varian Vista AX Pro

Varian 720-ES instrument.

[26] W. Zhu, Y. L. Li, L. X. Liu, W. L. Zhang, Y. M. Chen, F. Xi, J. Biomed. Mater.

Res. 2011 , 96A , 33 0 –3 40.

[27] A. Salinas, P. Esbrit, M. Vallet-Regí, Biomater. Sci. 2013 , 1 , 4 0 –5 1.

[28] M. Colilla, B. Gonzµlez, M. Vallet-Regí, Biomater. Sci. 2013 , 1 , 11 4 –1 34.

[29] A. J. Salinas, M. Vallet-Regi, RSC Adv. 2013 , 3 , 11116 – 11131.

[30] B. Palazzo, M. Iafisco, M. Laforgia, N. Margiotta, G. Natile, C. L. Bianchi,

D. Walsh, S. Mann, N. Roveri, Adv. Funct. Mater. 2007 , 17 , 218 0 –2 188.

[31] M. Iafisco, B. Palazzo, M. Marchetti, N. Margiotta, R. Ostuni, G. Natile, M.

Morpurgo, V. Gandin, C. Marzano, N. Roveri, J. Mater. Chem. 2009 , 19 ,

838 5 –8 392.

Acknowledgements

We gratefully acknowledge financial support from the Minister-

io de Economía y Competitividad, Spain (grant no. CTQ2012-

30762); the Romanian National Authority for Scientific Re-

[32] C. Y. Lai, B. G. Trewyn, D. M. Jeftinija, K. Jeftinija, S. Xu, S. Jeftinija, V. S.

Lin, J. Am. Chem. Soc. 2003 , 125 , 445 1 –4 459.

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

17

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

ˇ

´

[33] D. PØrez-Quintanilla, S. Gómez-Ruiz, Z. Zizak, Z. D. Juranic, I. Sierra, S.

[56] L. M. Gao, E. MelØndez, Met. Based Drugs 2010, 2010, 286–298.

[57] L. M. Gao, J. L. Vera, J. Matta, E. MelØndez, J. Biol. Inorg. Chem. 2010 , 15 ,

85 1 –8 59.

´

derovic, Chem. Eur. J. 2009 ,

Prashar, I. del Hierro, M. Fajardo, G. N. Kalu

¯

15 , 558 8 –5 597.

´

derovic, D. PØrez-Quintanilla, I. Sierra, S. Prashar, I. del Hierro,

[34] G. N. Kalu

[58] M. Napoli, C. Saturnino, E. Sirignano, A. Popolo, A. Pinto, P. Longo, Eur.

J. Med. Chem. 2011 , 46 , 12 2 –1 28.

[59] P. Beckova, O. Oberschmidt, A. R. Hanouscke, C. Pampillón, V. Schirr-

macher, N. J. Sweeney, K. Strohfeld, M. Tacke, Anticancer Drugs 2007, 18,

311–315.

¯

ˇ

´

Z. Zizak, Z. D. Juranic, M. Fajardo, S. Gómez-Ruiz, J. Mater. Chem. 2010,

20, 806–814.

ˇ

´

ˇ

´

derovic, D. PØrez-Quintanilla, Z. Zizak, Z. D. Juranic, S. Gómez-

[35] G. N. Kalu

¯

Ruiz, Dalton Trans. 2010, 39, 2597–2608.

[36] A. García-PeÇas, S. Gómez-Ruiz, D. PØrez-Quintanilla, R. Paschke, I.

[60] J. Schur, C. M. Manna, A. Deally, R. W. Kçster, M. Tacke, E. Y. Tshuva, I.

Ott, Chem. Commun. 2013 , 49 , 478 5 –4 787.

´

derovic, J. Inorg. Biochem. 2012 ,

Sierra, S. Prashar, I. del Hierro, G. N. Kalu

¯

106 , 10 0 –1 10.

[61] C. Riccardi, I. Nicoletti, Nat. Protoc. 2006 , 1 , 145 8 –1 461.

[62] P. N. Kelly, A. Strasser, Cell Death Differ. 2011 , 18 , 141 4 –1 424.

[63] C. Gao, A. Y. Wang, J. Histochem. Cytochem. 2009 , 57 , 113 9 –1 148.

[64] A. C. Tsamandas, D. Kardamakis, T. Petsas, V. Zolota, V. Vassiliou, T.

Matatsoris, H. Kalofonos, C. E. Vagianos, C. D. Scopa, In Vivo 2007, 21,

113 – 118.

[37] K. Strohfeldt, M. Tacke, Chem. Soc. Rev. 2008 , 37 , 117 4 –1 187.

[38] S. Gómez-Ruiz, D. Maksimovic-Ivanic, S. Mijatovic, G. N. Kalu

´

derovic, Bio-

´

¯

inorg. Chem. Appl. 2012, 140284.

´

derovic, D. Polo-Cerón, S. Prashar, M. Fajardo,

[39] S. Gómez-Ruiz, G. N. Kalu

¯

ˇ

´

Z. Zizak, Z. D. Juranic, T. J. Sabo, Inorg. Chem. Commun. 2007 , 10 , 74 8 –

752.

[65] S. Kagawa, J. Gu, T. Honda, T. J. McDonnell, S. G. Swisher, J. A. Roth, B.

Fang, Clin. Cancer Res. 2001, 7, 1474–1480.

´

derovic, M. Fajardo, R.

[40] J. Ceballos-Torres, S. Gómez-Ruiz, G. N. Kalu

¯

Paschke, S. Prashar, J. Organomet. Chem. 2012 , 700 , 18 8 –1 93.

[66] A. Aranovich, Q. Liu, T. Collins, F. Geng, S. Dixit, B. Leber, D. W. Andrews,

Mol. Cell 2012 , 45 , 75 4 –7 63.

[41] J. Ceballos-Torres, M. J. Caballero-Rodríguez, S. Prashar, R. Paschke, D.

´

derovic, S. Gómez-Ruiz, J. Organomet. Chem. 2012 ,

Steinborn, G. N. Kalu

716 , 20 1 –2 07.

[67] P. J. Beale, P. Rogers, F. Boxall, S. Y. Sharp, L. R. Kelland, Br. J. Cancer

2000, 82, 436–440.

¯

[42] S. Gómez-Ruiz, D. Polo-Cerón, S. Prashar, M. Fajardo, A. AntiÇolo, A.

[68] Y. Edden, S. D. Wexner, M. Berho, Colorectal Dis. 2012 , 14 , 55 5 –5 61.

[69] H. M. Amm, T. Zhou, A. D. Steg, H. Kuo, Y. Li, D. J. Buchsbaum, Mol.

Cancer Res. 2011 , 9 , 40 3 –4 17.

Otero, Eur. J. Inorg. Chem. 2007 , 444 5 –4 455.

´

derovic, D. Polo-Cerón, V. Tayurskaya, S. Pra-

[43] S. Gómez-Ruiz, G. N. Kalu

¯

shar, M. Fajardo, R. Paschke, J. Organomet. Chem. 2009 , 694 , 303 2 –

3038.

[70] T. V. Bagnyukova, I. G. Serebriiskii, Y. Zhou, E. A. Hopper-Borge, E. A. Go-

lemis, I. Astsaturov, Cancer Biol. Ther. 2010 , 10 , 83 9 –8 53.

[71] X. Luo, W. L. Kraus, Genes Dev. 2012 , 26 , 41 7 –4 32.

[72] J. Sand-Dejmek, G. Adelmant, B. Sobhian, A. S. Calkins, J. Marto, D. J.

Iglehart, J. B. Lazaro, Mol. Cancer 2011 , 10 , 74.

[73] S. Aizawa, K. Ookawa, T. Kudo, J. Asano, M. Hayakari, S. Tsuchida, Cancer

Sci. 2003 , 94 , 88 6 –8 93.

[74] G. Zhu, P. Chang, S. J. Lippard, Biochemistry 2010 , 49 , 617 7 –6 183.

ˇ

´

derovic, Z. Zizak, I. Besu, Z. D. Juranic, S. Pra-

[44] S. Gómez-Ruiz, G. N. Kalu

¯

shar, M. Fajardo, J. Organomet. Chem. 2009 , 694 , 198 1 –1 987.

[45] F. Kleitz, S. H. Choi, R. Ryoo, Chem. Commun. 2003 , 213 6 –2 137.

[46] K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pierotti, J. Rou-

quØrol, T. Siemieniewska, Pure Appl. Chem. 1985, 57, 603–620.

[47] T. W. Kim, F. Kleitz, B. Paul, R. Ryoo, J. Am. Chem. Soc. 2005 , 127 , 760 1 –

7610.

´

derovic, S. Prashar, D. Polo-Cerón, M. Fajardo,

[75] S. Gómez-Ruiz, G. N. Kalu

¯

ˇ

´

Z. Zizˇak, T. J. Sabo, Z. D. Juranic, J. Inorg. Biochem. 2008 , 102 , 155 8 –

1570.

[48] C. V. Christodolou, A. G. Eliopoulos, L. S. Young, L. Hodgkins, D. R. Ferry,

D. J. Kerr, Br. J. Cancer 1996, 77, 2088–2097.

[49] P. W. Causey, M. C. Baird, S. P. C. Cole, Organometallics 2004 , 23 , 448 6 –

4494.

[50] O. R. Allen, A. L. Gott, J. A. Hartley, J. M. Hartley, R. J. Knox, P. C. McGo-

wan, Dalton Trans. 2007 , 508 2 –5 090.

[76] P. W. Sylvester, Methods Mol. Biol. 2011 , 716 , 15 7 –1 68.

[77] A. D. Goodrich, A. Ersek, N. M. Varain, D. Groza, M. Cenariu, D. S. Thain,

G. Almeida-Porada, C. D. Porada, F. D. Zanjani, Exp. Hematol. 2010 , 38 ,

51 6 –5 25.

[51] M. Tacke, L. T. Allen, L. Cuffe, W. M. Gallagher, L. Ying, O. Mendoza, H.

Muller-Bunz, F. J. K. Rehmann, N. Sweeney, J. Organomet. Chem. 2004 ,

689 , 224 2 –2 249.

[52] G. D. Potter, M. C. Baird, S. P. C. Cole, J. Organomet. Chem. 2007 , 692 ,

350 8 –3 518.

[53] C. M. Manna, O. Braitbard, E. Weiss, J. Hochman, E. Y. Tshuva, ChemMed-

Chem 2012 , 7 , 70 3 –7 08.

[54] M. Wenzel, B. Bertrand, M. J. Eymin, V. Comte, J. A. Harvey, P. Richard, M.

Groessl, O. Zava, H. Amrouche, P. D. Harvey, P. Le Gendre, M. Picquet, A.

Casini, Inorg. Chem. 2011 , 50 , 947 2 –9 480.

[78] G. A. Filip, I. D. Postescu, C. Tatomir, A. Muresan, S. Clichici, J. Physiol.

Pharmacol. 2012, 63, 423–432.

[79] P. Horcajada, A. Rµmila, J. PØrez-Pariente, M. Vallet-Regí, Microporous

Mesoporous Mater. 2004 , 68 , 10 5 –1 09.

[80] CCDC-983057 (17) and 983058 (18) contain the supplementary crystal-

lographic data for this paper. These data can be obtained free of

charge from The Cambridge Crystallographic Data Centre via

www.ccdc.cam.ac.uk/data_request/cif.

[55] X. Stachtea, N. Karamanos, N. Klouras, Anticancer Res. 2009, 29, 3227–

3231.

Received: January 24, 2014

Published online on && &&, 0000

&

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

18

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

FULL PAPER

Concealed activity: A series of titano-

cene-functionalized silica-based materi-

als have been synthesized, character-

ized, and studied in vitro against human

cell lines (see figure). They act as Trojan

horses for the incorporation of the tita-

nium active species into cells, leading to

much higher quantities of titanium

uptake in cells, changing the dynamics

of apoptotic morphological and func-

tional changes, and inducing pro-

grammed cell death in tumor cell popu-

lations.

&

Drug Delivery

J. Ceballos-Torres, P. Virag, M. Cenariu,

S. Prashar, M. Fajardo, E. Fischer-Fodor,*

S. Gꢀmez-Ruiz*

&& – &&

Anti-cancer Applications of

Titanocene-Functionalised

Nanostructured Systems: An Insight

into Cell Death Mechanisms

Chem. Eur. J. 2014, 20, 1 – 19

www.chemeurj.org

19

ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Article Doi

Anti-cancer applications of titanocene-functionalised nanostructured systems: An insight into cell death mechanisms

DOI: 10.1002/chem.201400300

Source and publish data:

Authors:

Article abstract of DOI:10.1002/chem.201400300

Full text of DOI:10.1002/chem.201400300

Products guided by the article

R&D Labs maybe for 1628028-45-4

Relevant to this article

Hot Product