17
Preparation, structure, cytotoxicity and mechanism of action of ferritin-Pt(II) terpyridine compound nanocomposites Giarita Ferraro 1 , Andrea Pica 2 , Ganna Petruk 1 , Francesca Pane 1 , Angela Amoresano 1 , Agostino Cilibrizzi 3,4 , Ramon Vilar 3 , Daria Maria Monti 1 & Antonello Merlino* 1 1Department of Chemical Sciences, University of Naples Federico II, Napoli, Italy 2EMBL Grenoble, 71 avenue des Martyrs, CS 90181, 38042 Grenoble Cedex 9, France 3Department of Chemistry, Imperial College London, London SW7 2AZ, United Kingdom 4Institute of Pharmaceutical Science, King’s College London, Stamford Street, London SE1 9NH, United Kingdom *Author for correspondence: Tel.: +39 081 674 276; Fax: +39 081 674 090; [email protected] Abstract: Aim: A Pt(II)-terpyridine compound, bearing two piperidine substituents at positions 2 and 2’ of the terpyridine ligand (1), is highly cytotoxic and shows a mechanism of action distinct from cisplatin. 1 has been incorporated within the ferritin nanocage (AFt). Materials & methods: Spectroscopic and crystallographic data of the Pt(II)- AFt nanocomposite have been collected and in vitro anticancer activity has been explored using cancer cells. Results: Pt(II)-containing fragments bind His49, His114 and His132. Pt(II)-AFt nanocomposite is less cytotoxic than 1, but it is more toxic than cisplatin at high concentrations. The Pt(II)-AFt nanocomposite triggers necrosis in cancer cells, as free 1 does. Conclusions: Pt(II)-AFt nanocomposites are promising vehicles to deliver Pt-based drugs to cancer cells. Keywords: protein nanocage; protein metalation; protein-metallodrug interactions; platinum-based drug; Pt(II)- terpyridine compounds There is a growing scientific interest for the study of protein-based self-assembling cages that provide confined chemical environments for the encapsulation of drugs [1]. An important class of biologically relevant protein nanocages is that of ferritins (Fts) [2]. Fts are widely distributed cytosolic proteins that are responsible for iron- storage. They usually have 24 subunits, folded in a four-helix bundle structure, which assemble to form a hollow protein cage (Ft nanocage) [3]. In vertebrates, they are heteropolymeric, i.e. they consist of two types of chains, called H- (~21 kDa) and L-chain (~19 kDa), sharing about 55% of sequence identity. H- and L-chains have distinct physiological roles. Specifically, H-chain is enzymatically active as it contains a dinuclear ferroxidase center that is responsible for Fe uptake and oxidation by O2 or H2O2. L-chain is enzymatically inactive, but it possesses a nucleation site that is needed for iron mineralization. H- and L- chains can be recognized and internalized by the transferrin receptor 1 (TfR1) and Scavenger receptor class A type 5 (SCARA5), which are over-expressed in a variety of malignant cells [4-6]. Therefore, Ft nanocages represent a fully biologically compatible nanomaterial for specific drug-delivery applications, generating limited or no immune response [7]. The protein shows a complete lack of cytotoxicity [8, 9] and allows the precise control of the number of encapsulated molecules, which is a critical feature when defining drug dosage. Different anticancer drugs have been entrapped in the apoferritin (AFt) shell, including widely used anticancer agents like cisplatin (cis-Pt) [10-13], carboplatin [14], oxaliplatin [15] and doxorubicin [7], as well as poorly soluble Ru- and Au-based cytotoxic compounds [16-20]. Cis-Pt and carboplatin were encapsulated in 2007 by

Preparation, structure, cytotoxicity and mechanism of

  • Upload
    others

  • View
    3

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Preparation, structure, cytotoxicity and mechanism of

Preparation, structure, cytotoxicity and mechanism of action

of ferritin-Pt(II) terpyridine compound nanocomposites

Giarita Ferraro1, Andrea Pica2, Ganna Petruk1, Francesca Pane1, Angela Amoresano1, Agostino Cilibrizzi3,4, Ramon Vilar3, Daria Maria Monti1 & Antonello Merlino*1 1Department of Chemical Sciences, University of Naples Federico II, Napoli, Italy 2EMBL Grenoble, 71 avenue des Martyrs, CS 90181, 38042 Grenoble Cedex 9, France 3Department of Chemistry, Imperial College London, London SW7 2AZ, United Kingdom 4Institute of Pharmaceutical Science, King’s College London, Stamford Street, London SE1 9NH, United Kingdom *Author for correspondence: Tel.: +39 081 674 276; Fax: +39 081 674 090; [email protected]

Abstract:

Aim: A Pt(II)-terpyridine compound, bearing two piperidine substituents at positions 2 and 2’ of the terpyridine

ligand (1), is highly cytotoxic and shows a mechanism of action distinct from cisplatin. 1 has been incorporated

within the ferritin nanocage (AFt). Materials & methods: Spectroscopic and crystallographic data of the Pt(II)-

AFt nanocomposite have been collected and in vitro anticancer activity has been explored using cancer cells.

Results: Pt(II)-containing fragments bind His49, His114 and His132. Pt(II)-AFt nanocomposite is less cytotoxic

than 1, but it is more toxic than cisplatin at high concentrations. The Pt(II)-AFt nanocomposite triggers necrosis

in cancer cells, as free 1 does. Conclusions: Pt(II)-AFt nanocomposites are promising vehicles to deliver Pt-based

drugs to cancer cells.

Keywords: protein nanocage; protein metalation; protein-metallodrug interactions; platinum-based drug; Pt(II)-

terpyridine compounds

There is a growing scientific interest for the study of protein-based self-assembling cages that provide confined

chemical environments for the encapsulation of drugs [1]. An important class of biologically relevant protein

nanocages is that of ferritins (Fts) [2]. Fts are widely distributed cytosolic proteins that are responsible for iron-

storage. They usually have 24 subunits, folded in a four-helix bundle structure, which assemble to form a hollow

protein cage (Ft nanocage) [3]. In vertebrates, they are heteropolymeric, i.e. they consist of two types of chains,

called H- (~21 kDa) and L-chain (~19 kDa), sharing about 55% of sequence identity. H- and L-chains have distinct

physiological roles. Specifically, H-chain is enzymatically active as it contains a dinuclear ferroxidase center that

is responsible for Fe uptake and oxidation by O2 or H2O2. L-chain is enzymatically inactive, but it possesses a

nucleation site that is needed for iron mineralization. H- and L- chains can be recognized and internalized by the

transferrin receptor 1 (TfR1) and Scavenger receptor class A type 5 (SCARA5), which are over-expressed in a

variety of malignant cells [4-6]. Therefore, Ft nanocages represent a fully biologically compatible nanomaterial

for specific drug-delivery applications, generating limited or no immune response [7]. The protein shows a

complete lack of cytotoxicity [8, 9] and allows the precise control of the number of encapsulated molecules,

which is a critical feature when defining drug dosage.

Different anticancer drugs have been entrapped in the apoferritin (AFt) shell, including widely used anticancer

agents like cisplatin (cis-Pt) [10-13], carboplatin [14], oxaliplatin [15] and doxorubicin [7], as well as poorly

soluble Ru- and Au-based cytotoxic compounds [16-20]. Cis-Pt and carboplatin were encapsulated in 2007 by

Page 2: Preparation, structure, cytotoxicity and mechanism of

Yang et al. [21], but their biological effect was reported only later by Xing, who observed a cellular uptake of

these nanoparticles (43 to 51 cis-Pt molecules per cage) and a cytotoxic effect on rat pheochromocytoma PC12

cells [15]. Huang and colleagues reported that they are able to encapsulate 11.2 molecules of cis-Pt within apo-

pig pancreas ferritin (pAFt) and that the cage induces apoptosis of gastric cancer cells [12]. More recently, Falvo

and coworkers reported an engineered version of human H-chain, able to covalently bind a monoclonal antibody

against the human melanoma-specific antigen CSPG4, with about 50 cis-Pt molecules within the cage. This large

nanocage (about 900 kDa) shows a high specificity in binding melanoma cells, anti-proliferative effect and a

reduction of tumor size in mice bearing melanoma [13]. 20 to 55 molecules of cis-Pt can be trapped in the horse

spleen AFt (hsAFt) [10]. An average value of 17 and 23 atoms of Pt have been found within the cage in the

carboplatin- and oxaliplatin-loaded hsAFts, respectively [15]. The X-ray structure of hsAFt-cis-Pt [10] and hsAFt-

carboplatin [14] nano-composites have shown that Pt atoms are bound to the side chain of His132 or to the side

chains of His132 and His49 [10, 14]. Cytotoxicity assays have also demonstrated that cis-Pt -encapsulated pAFt

induces apoptosis in gastric cancer cells BGC823 (GCC) [12] and that Pt(II)-hsAFts nanocomposites induce the

death of rat pheochromocytoma cells (PC12) and human epithelia cancer cells (HeLa) [15], through a different

mechanism of action in comparison to cis-Pt.

Recently, it has been reported that the Pt(II) terpyridine compound in Figure 1 (compound 1), which is cytotoxic

for proliferating NIH 3T3, as well as for cancerous U2OS and SH-SY5Y cell lines, exhibits a mechanism of action

distinct from cis-Pt [22] and shows an unusual reactivity towards the model protein lysozyme, by producing an

extensive metalation of the protein [23]. These data prompted us to incorporate compound 1 in hsAFt in order

to prepare a Pt(II)-hsAFt nanocomposite, which could have entrapped much more Pt atoms than those found

when cis-Pt, carboplatin and oxaliplatin were loaded in hsAFt.

Figure 1. Structure of compound 1.

Materials and Methods

Sample preparation an analytical characterization

Synthesis of compound 1 has been performed through a previously established procedure [22]. Horse spleen

ferritin (hsFt) was purchased from Sigma (>95% L-chain). Compound 1 was trapped in hsAFt nanocage following

the protocol previously described by Huang et al. [12] and already used to incorporate cis-Pt, carboplatin and

other metallodrugs [9, 10, 14]. Briefly, hsFt was dissociated into individual subunits by adjusting the pH to 13

with NaOH (0.1 M). Then, compound 1 was added to this solution at a concentration of 15 mM, i.e. at a final

hsAFt chain to metal molar ratio of 1:30. After 60 minutes, the solution was slowly neutralized using sodium

phosphate at pH 7.4 (1.0 M) and centrifugated at 5000 rpm/min for 10 minutes at 4 °C to remove the denatured

hsAFt chains that precipitate during the experiment. The recovered supernatant was extensively dialyzed

Page 3: Preparation, structure, cytotoxicity and mechanism of

against sodium phosphate buffered at pH 7.4 using a 10 kDa cutoff Centricon filter to remove the excess of the

metal complex that was unbound and the salts. Pt content was quantified by ICP-MS with a method previously

reported [10, 12, 22]. Briefly, Pt concentration has been measured with three replicates using an Agilent 7700

ICP-MS instrument (Agilent Technologies) equipped with a frequency-matching radio frequency (RF) generator

and 3rd generation Octopole Reaction System (ORS3), operating with helium gas in ORF and the following

parameters: RF power: 1550 W, plasma gas flow: 14 L min−1; carrier gas flow: 0.99 L min−1; He gas flow: 4.3 mL

min−1. 103Rh was used as an internal standard (final concentration: 50 μg L−1). Standard solutions were prepared

in 5% nitric acid at four different concentrations (1, 10, 50, and 100 μg L−1).

Protein concentration was evaluated by BCA method using bovine serum albumin as a standard, as performed

in previous studies [10]. The amount of compound 1-encapsulated hsAFt recovered after the encapsulation and

purification processes is about 20% of the theoretical value.

Biophysical characterization

UV-Vis absorption spectra were registered using a Jasco V650 UV-visible spectrophotometer at 700-240 nm in

10 mM sodium phosphate pH 7.4 using a compound 1-encapsulated hsAFt concentration of 0.25 mg mL-1.

Experimental conditions: 0.1 cm path-length quartz cuvettes, scan rate: 200 nm min-1, data pitch: 1 nm.

Far-UV circular dichroism (CD) spectra were recorded at 250–190 nm at 20 °C using a V815 CD spectrometer

(Jasco) in 10 mM sodium phosphate pH 7.4 at 20 °C by using a 0.1 cm optical path length cell. Measurements

were acquired in triplicate with a time constant of 4 s, a 2 nm bandwidth, and a scan rate of 50 nm min−1, using

compound 1-encapsulated hsAFt concentration = 0.05 mg mL-1. CD spectra were plotted as molar ellipticity

versus wavelength.

Crystallization, X-ray diffraction data collection, structure resolution and refinement

Compound 1-encapsulated hsAFt was concentrated to 10 mg ml−1 and set up for crystallization at 298 K. The

final optimized crystals of the adduct were obtained from hanging-drop conditions using 0.6 M (NH4)2SO4, 0.1

M Tris HCl pH 7.4, 60 mM CdSO4 as the reservoir solution. The crystals grew to maximum dimensions of 0.5 ×

0.5 × 0.5 mm within a week. Crystallization information is given in Table S1 of the supporting information. For

data collection, the crystals were flash-cooled in liquid nitrogen. X-ray diffraction data were collected at

beamline ID30B of ESRF (Grenoble, France).

Two X-ray diffraction datasets were collected under gaseous nitrogen (100 K) for two different crystals using X-

rays at λ=0.97 Å and 1.54 Å for the first crystal, and λ=1.06 Å and 1.08 Å for the second crystal, respectively.

Best data set was collected to 1.33 Å resolution. All data sets were processed and scaled using AutoProc [24].

Crystal data, data collection parameters and processing information are given in Table S2 of supporting

information.

Molecular replacement was performed using Phaser [25] and the coordinates of the protein extracted from the

PDB file 5ERK as the search model [10]. Model building was carried out using Coot [26] in iterative cycles with

refinement using Refmac5 [27] and autoBuster [28]. Comparison of the anomalous maps obtained using all the

collected datasets allows an unambiguous determination of Pt(II) and Cd(II) binding sites. Refinement statistics

are reported in Table S2. Details of the assignment of Pt versus Cd atoms are reported in Table 1. The best

structure of compound 1-encapsulated hsAFt (Figure 3) comprises 1889 non-H atoms, including 7 Cd2+, 3 Cl-, 1

SO42-, 1 glycerol molecule and five Pt atoms and refines to Rfactor of 0.127 and Rfree of 0.157 with good

stereochemistry. Validation reports, coordinates and structure factors, including anomalous data, have been

provided to reviewers and Editor for review process and deposited in the Protein Data Bank under the accession

Page 4: Preparation, structure, cytotoxicity and mechanism of

codes 6HJT and 6HJU, for the structures obtained using the data collected from the two crystals. Uninterpreted

peaks of electron density in the e.d. maps and validation reports are commented in the Supporting information.

Cell culture and toxicity experiments

Human breast cancer cells MCF-7 and epidermoid carcinoma cells A431 from ATCC were cultured as previously

described [9]. For dose-response experiments, cells were seeded in 96-well plates at a density of 2.5×103 cells

per well. 24 h after seeding, increasing concentrations of compounds were added to the cells. MTT (3-(4,5-

dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assays were carried out after 48 h incubation as

previously described [29]. The IC50 values (i.e. the concentrations of the drug that reduce cell survival to 50%

when compared to control cells treated with identical volumes of buffer) were calculated from Pt concentration-

dependent cellular viability curves. The cytotoxicity results are the mean values of three independent analyses,

each carried out in triplicates.

LDH release

The occurrence of necrosis was determined by measuring the release of the cytosolic enzyme lactate

dehydrogenase (LDH) in the culture medium, using the in vitro toxicology assay kit LDH based (Sigma Aldrich).

The LDH content of the medium from untreated cells was referred to the spontaneous release, whereas the

total content of intracellular LDH was determined upon cell lysis. The percentage of LDH release was calculated

as:

LDH release (%)=((experimental release-spontaneous release))/((total content - spontaneous release) ) x100

Results are the mean values of three independent analyses, each analysis was carried out in triplicates.

Significance was determined by Student’s t-test.

Table 1. Comparison of anomalous peak intensity in the data collected at different wavelengths on the two crystals

of compound 1-encapsulated hsAFt analyzed herein.

compound 1-encapsulated AFt Assignment

Source, =(Å) 0.9795 1.5400 1.06 1.08

Resolution (Å) 1.58 1.96 1.33 1.3

Location of Cd atoms identified in previously solved structures of hsAFt

anomalous peak intensity (x10)

e/Å3

Page 5: Preparation, structure, cytotoxicity and mechanism of

Close to Glu11 0.010 0.018 0.012 0.011 Cd2+

Close to Glu53/Glu56

0.005 0.005 0.007 0.008 Cd2+

Close to Glu60 0.003 0.005 0.007 0.009 Cd2+

At the binary axis. Close to Asp80

0.034 0.071 0.071 0.077 Cd2+

Close to ND1 atom of His132

0.031 0.018 0.049 0.023 Pt2+

Close to NE2 atom of His132

0.009 0.007 0.016 0.007 Pt2+

At the ternary axis. Close to Glu130

0.005 0.011 0.009 0.007 Cd2+

Close to Cys48 0.004 0.006 0.005 0.005 Cd2+

Pt close to NE2 atom of His49

0.023 0.017 0.034 0.014 Pt2+

Pt close to ND1 atom of His49

0.016 0.015 0.027 0.009 Pt2+

Pt close to NE2 atom of His114

0.011 0.009 0.016 0.009 Pt2+

Close to Asp127 0.004 0.012 0.006 0.016 Cd2+

Mitochondrial membrane potential measurements

The mitochondrial membrane potential (Δψm) was measured as described by Monti et al [8]. Cells were plated

at a density of 2 × 104 cells per well and were treated after 24 h as described above. Subsequently, the cells

were incubated with 200 nM cationic lipophilic dyetetramethylrhodamine ethyl ester (TMRE) for 20 min at 37

°C. Then, the cells were gently washed with 0.2% BSA in PBS three times and the fluorescence was measured in

a microplate reader with peak Ex/Em = 549/575 nm. Each value is the mean of three independent experiments,

with three measurements for each experiment. Significance was determined by Student’s t-test.

Page 6: Preparation, structure, cytotoxicity and mechanism of

Western blot analysis

To analyze intracellular TNF-α levels, A431 cells were plated at a density of 2 × 104 cells per cm2 for 24 h. After

seeding, cells were treated with 39 µM of cis-Pt, 3 µM of compound 1 and 41 µM of compound 1-encapsulated

hsAFt (corresponding to the IC50 values). After 48 h of incubation, cells were lysed in 0.3 M NaCl, 0.5% NP-40 in

0.1 M Tris-HCl, pH 7.4 containing proteases inhibitors. Then, lysates were analyzed by Western blotting as

reported by Galano et al. [30]. Antibodies anti-TNF-α and anti-GAPDH, as well as chemiluminescence detection

system (SuperSignal® WestPico), were purchased from Thermo Fisher (Rockford, IL, USA).

Results

Platinum-based drug encapsulation within hsAFt

Compound 1 was encapsulated within the hsAFt nanocage following the alkaline disassembly/reassembly

protocol [12]. Both UV–Visible absorption and CD spectroscopy were used to reveal possible structural

differences in the Ft features due to the presence of compound 1 within the protein cage. As shown in Figure

2A, UV-Vis spectrum of compound 1-encapsulated hsAFt overlaps well with that of hsAFt, used as a control.

Similarly, the two CD spectra, i.e. that of compound 1-encapsulated hsAFt and that of control protein, show a

good overlap in the range 190–250 nm (Figure 2B). Deconvolution of CD spectra indicates that 74% of the

residues of the entire cage are involved in the formation of alpha helices (almost 80% considering also the turns),

consistently with the crystal structure of hsAFt (74%) and with a helical percentage even higher that that

observed by CD deconvolution of native hsFt (70% of residues in helices, 72% considering helices and turns).

These results suggest that the molecular features of compound 1-encapsulated hsAFt are similar to those of the

native protein.

To determine the amount of Pt within the cage, ICP-MS data were then collected. Data indicate that the sample

of compound 1-encapsulated hsAFt used in the experiments reported in this paper contained 7.5±0.5 atoms of

Pt per subunit.

Molecular structure of compound 1-encapsulated hsAFt.

For the structural characterization of compound 1-encapsulated hsAFt, the nanocomposite was crystallized in

order to solve its X-ray structure (Figure 3). Preliminary crystallographic analysis of the difference Fourier

electron density maps clearly indicated the presence of several metal binding sites in the structure of compound

1-encapsulated hsAFt. Therefore, to discriminate between Cd(II) ions from the crystallization media and Pt(II)

binding sites, X-ray diffraction data were collected for two crystals, each at two different wavelengths.

In the first experiment, we applied the same protocol that was previously used to reveal the location of Pt(II)

atoms in the structures of cis-Pt- [10] and carboplatin-[12] encapsulated hsAFt. This procedure relies on the

comparison of anomalous difference electron density maps which are calculated using the data collected at 1.54

Å and 0.98 Å. At these wavelengths, the Cd and the Pt f″ signals change in opposite ways: i.e. Pt f″ value

increases, whereas Cd f″ signal decreases. Thus, the comparison of the peaks in the anomalous difference

Fourier electron density map calculated using data collected at the two wavelengths elegantly enables an

unambiguous assignment of these two atoms.

The second experiment discriminates between Pt(II) and Cd(II) by comparing the anomalous electron density

maps which are calculated using the data collected at 1.06 Å and 1.08 Å resolution. These wavelengths are close

Page 7: Preparation, structure, cytotoxicity and mechanism of

to the edge of Pt, thus allowing to identify Pt signal that is high using λ=1.06 Å and low using the latter

wavelength.

Data indicate the existence of five Pt(II) binding sites (Figure 3), close to NE2 and ND1 atoms of the side chains

of His49 (Figure 4A) and His132 (Figure 4B), and close to ND1 atom of the side chain of His114 (Figure 4C).

Indirect evidences to support the proposed assignments were obtained through the observation of the

geometry of metal ligand coordination sphere (i.e. the preference of square planar geometry for Pt(II) and of

octahedral geometry for Cd(II)).

Although Pt ligands are difficult to assign due to the limited occupancy (between 0.20 and 0.40) and resolution

of X-ray diffraction data, as previously discussed by Tanley and Helliwell (2016) [31], an attempt to complete

the metal coordination spheres has been carried out by interpreting the electron density map with solvent

molecules. In particular, it appears that the side chain of His132 binds a [Pt(H2O)3]2+ and a [Pt(H2O)(DMSO)(X)]2+

fragment with X being an undefined ligand (Figure 4B), whereas the side chain of His49 binds a [Pt(H2O)3]2+ and

a [Pt(H2O)2Cl]+ (Figure 4A). Close to the side chain of His114 the electron density map is too poor to assign Pt

ligands (Figure 4C).

A

Page 8: Preparation, structure, cytotoxicity and mechanism of

B

Figure 2. The structure of compound 1-encapsulated hsAFt has been studied by UV-Vis spectroscopy (A) and circular

dichroism (B) in 10 mM sodium phosphate at pH 7.4. hsAFt was used as a control. Protein concentration is 0.25 mg

mL-1 for UV-Vis absorption experiments and 0.05 mg mL-1 for CD measurements.

Figure 3. Overall structure of the four-helix bundle of compound 1-encapsulated hsAFt with highlighted the

location of Pt atoms. The occupancy of Pt centers is 0.40, 0.30, 0.35, 0.25 and 0.20 for the atoms bound to ND1

and NE2 atoms of His49, ND1 and NE2 atoms of His132, and ND1 of His114, respectively. The B-factors for the Pt

atoms are within the range 34.9- 64.5 Å2. Anomalous difference electron density map is reported at 4.0 σ.

Cytotoxicity

Page 9: Preparation, structure, cytotoxicity and mechanism of

The cytotoxicity of compound 1 and compound 1-encapsulated hsAFt was determined by the colorimetric MTT

assay. Two human cancer cell lines, the epidermoid carcinoma A431 and breast adenocarcinoma MCF-7 cells

were exposed to increasing amounts of compound 1 and of the nanocomposite for 48 h. cis-Pt was used for

comparison. In Figure 5, cell survival is reported as a function of the amount of Pt provided to the cells.

Compound 1 resulted to be the most toxic compound on both cell lines, followed by cis-Pt and then by

compound 1-encapsulated hsAFt. A complete lack of toxicity was found for hsAFt, even at very high

concentrations (> 1000 µM, data not shown). The corresponding IC50 values are reported in Table 2 and indicate

that compound 1 has IC50 values that are 10 to 13 times lower than those obtained for compound 1-

encapsulated hsAFt. Compound 1-encapsulated hsAFt has an IC50 value comparable to that of cis-Pt for the A431

cells, whereas it is less active than cis-Pt on MCF-7 cells. However, the nanocomposite shows more toxic than

cis-Pt at high Pt concentrations (80 μM) for both cell lines (Figure 5).

A B

C

Figure 4. Pt binding sites in the structure of compound 1-encapsulated hsAFt. (A) Pt binding sites close to His49. (B)

Pt binding sites close to His132. (C) Pt center close to His114. 2Fo-Fc electron density map is reported at 0.8 σ level

(grey).

Page 10: Preparation, structure, cytotoxicity and mechanism of

Table 2. IC50 values (μM) obtained for compound 1 and compound 1-encapsulated hsAFt on A431 and MCF-7

cell lines after 48 h incubation. cis-Pt is reported for comparison.

Drug (µM) Cell line

A431 MCF-7

Pt in compound 1-encapsulated hsAFt 41.4 ± 5.1 55.9 ± 4.2

compound 1 3.05 ± 0.29 5.48 ± 1.44

cis-Pt 39.2 ± 12.6 18.3 ± 1.7

Page 11: Preparation, structure, cytotoxicity and mechanism of

Figure 5. Dose-effect curves of compound 1, compound 1-encapsulated hsAFt and cis-Pt determined by the MTT

assay on A431 (A) and MCF-7 (B) cell lines. Cells were incubated in the presence of increasing amount of each

drug for 48 h.

Cell death mechanism

Subsequently, we have analyzed the cell death mechanism induced by compound 1-encapsulated hsAFt. First,

we measured the loss of mitochondrial membrane potential (Δψm), which is known to be generally involved in

cell death due to various insults. Indeed, it is well known that dissipation of Δψm precedes shrinkage and

fragmentation of cells, contributing to cell death. In this regard, A431 cells were incubated for 48 h with

increasing amount of each drug, as this cell line was the most sensitive to compound 1-encapsulated hsAFt

treatment. As shown in Figure 6A-C, a significant depolarization of cell membrane was observed in the presence

of each drug. Then, since it has been demonstrated that compound 1 induces necrosis [22], we have measured

the release of the cytosolic enzyme lactate dehydrogenase (LDH), a known marker for necrosis. Indeed,

permeabilization of the plasma membrane during necrosis allows the release of this intracellular enzyme into

the extracellular media. Following a 48h incubation of A431 cells with increasing amount of each drug, the

release of the enzyme was measured and related to the physiological LDH release, as described in Materials and

Methods section. Interestingly, LDH release was not observed in A431 cells after cis-Pt exposure (data not

shown), whereas both compound 1 and compound 1-encapsulated hsAFt induced an apparent membrane

rupture when compared to untreated cells (Figure 6D). As necrosis results in the secretion of cytokines, we have

also evaluated by Western blotting the intracellular protein tumor necrosis factor-α (TNF-α), as a further marker

of necrosis. Figure 7 shows that TNF-α release was not observed in untreated cells or cells exposed to cis-Pt,

whereas no signal associated with intracellular TNF-α was observed in cells incubated with either compound 1-

encapsulated hsAFt, or compound 1. Overall, these results clearly indicate that compound 1-encapsulated hsAFt

induces necrosis in A431 cells, similarly to the free drug.

Page 12: Preparation, structure, cytotoxicity and mechanism of

Figure 6. Analysis of cell death induced by compound 1-encapsulated hsAFt, compound 1 and cis-Pt on A431

cells. Cells were incubated with increasing amounts of each drug for 48 h. At the end of the experiment, changes

in the mitochondrial membrane potential (Δψm) were determined using TMRE staining and expressed as Δψm

over control (%) (A-C). D, LDH release of compound 1-encapsulated hsAFt (black bars) and compound 1 (white

bars). The release is expressed as described in Material and Methods section. * indicates p<0.05; ** indicates

p<0.01; *** indicates p<0.001; **** indicates p<0.0001 with respect to untreated cells. § indicates p<0.05; §§

indicates p<0.01; §§§ indicates p<0.001 with respect to 1 µM of compound 1 or 20 µM of compound 1-

encapsulated hsAFt. # indicates p<0.05 with respect to 4 µM of compound 1 or 40 µM of compound 1-

encapsulated hsAFt.

Page 13: Preparation, structure, cytotoxicity and mechanism of

Figure 7. Effects of drugs on TNF-α intracellular levels in A431 cells. Cells were treated with cis-Pt, compound 1

and compound 1-encapsulated hsAFt by using the concentrations reported in Materials and Methods section.

After 48 h incubation, the levels of TNF-α were detected by Western blot analysis using human TNF-α antibody.

GAPDH was used as the loading control.

Discussion

Compound 1 has been previously reported to target nuclear DNA and induce cell death via necrosis, i.e. with a

different mechanism of action when compared to cis-Pt and its derivatives [22]. Herein we have encapsulated

compound 1 within hsAFt expecting that this nanocomposite could act on cancer cells in a different way when

compared to the free compound [10-15,21]. hsAFt has been chosen as a suitable nanocarrier since it has the

advantage of being commercially available and is highly biocompatible. Recently, it has been shown that hsFt is

less toxic for HeLa cells than the human H-chain, when loaded with iron ions [32].

Compound 1-encapsulated hsAFt has been characterized by both biophysical and structural techniques.

Crystallographic data are in agreement with ICP-MS/BCA analysis indicating that the compound produces a

metalated hsAFt with a higher number of Pt centres within the cage when compared to various literature

records [10-15,21]. Three Pt binding sites have been identified in the structure of the Ft-Pt(II) nanocomposite:

the side chains of His49, His114 and His132. The overall structure of the protein in the adduct is superimposable

to that of cis-Pt- [10] and carboplatin-encapsulated hsAFt [14]. The root mean square deviations between the

Cα atoms of these structures is within the range 0.37-1.14 Å (171 atoms). His132 was previously identified as

Pt binding site in the structure of cis-Pt-encapsulated hsAFt [10], while His49 was found as a Pt binding site in

the structure of carboplatin-encapsulated hsAFt [14]. In the structure of 4-PF6-encapsulated hsAFt, a different

Ft-metallodrug adduct containing Pt and Au, Pt centres have not been found directly bound to the protein,

although this metal was present in the cage (PDB code 5ERJ) [18]. The different structures of Ft-Pt(II)

nanocomposites formed upon encapsulation of Pt(II) complexes within hsAFt cage indicate that metal ligands

have a role in determining the reactivity of metallodrugs with the protein, although they do not appear in the

final adducts. This conclusion is in agreement with what has been already found when Pt-[33], and Ru-based

drugs react with model proteins [34-35].

The electron density map close to the Pt centre is not well defined, thus precluding the possibility to assign Pt

ligands. These findings could indicate that at least an undefined amount of compound 1 could be degraded

during the encapsulation process. This would not be surprising, since the metal complex is not stable at the high

alkaline pH used for the encapsulation and degradation of compound 1 upon interaction with proteins has been

already observed [23]. Thus, it is likely that compound 1 and/or compound 1 fragments react with hsAFt during

the encapsulation process. However, it is also possible that Pt ligands are simply not visible in the electron

density map because of mobility and disorder in the crystal. Furthermore, it should be noted that comparing

the results of ICP MS/BCA data analysis with the crystallographic model of compound 1-encapsulated hsAFt, it

appears that there is a lot of Pt (from compound 1 or fragments) that is trapped within the cage but is not

covalently bound to hsAFt.

Page 14: Preparation, structure, cytotoxicity and mechanism of

Compound-1-encapsulated hsAFt is cytotoxic for the epidermoid carcinoma A431 and breast adenocarcinoma

MCF-7 cells and appear more toxic than cis-Pt at high Pt concentrations. A direct comparison between

cytotoxicity of different Ft-Pt(II) nanocomposites reported in literature is not trivial, since the biological

properties of these molecules have been tested using different protocols, cell lines and incubation time. Cis-Pt-

loaded hsAFt is 10-fold less toxic than the free drug for PC12 cells after 72 h of incubation (IC50= 20 and 2 μM,

respectively), even at 100 μM [15]. Carboplatin and oxaliplatin-loaded hsAFt have a low toxicity towards these

cell lines, even at very high concentrations, contrarily to free oxaliplatin that is highly toxic (IC50 = 4 μM) [15].

cis-Pt-encapsulated H-chain induces death of breast carcinoma and melanoma cells, with 25 % 3H-thymidine

incorporation within cells treated with the nanocomposite, when compared to the control, after 96 h incubation

at concentration 100 μM [13].

We have also shown that compound-1-encapsulated hsAFt induces necrosis of A431 cells, as the free compound

1 does. This is the first evidence of necrosis induced by a compound encapsulated within the ferritin nanocage.

These findings suggest that the biological properties of the compound 1-hsAFt nanocomposite depend on the

features of compound 1 rather than on the structure of the Pt-bound hsAFt. Previous studies have indicated

that cis-Pt-encapsulated pAFt induces apoptosis of gastric cancer cells BGC823 (GCC) and HeLa cells upon

releasing of the cis-Pt [12] and that apoptosis rate of GCC induced by cis-Pt.-encapsulated pAFt is lower than

that of the free drug [12]. A gold-based compound encapsulated within hsAFt also leads to the cell death by

apoptosis [8]. Altogether these data indicate that the Ft-metallodrug nanocomposites can act as powerful

nanovectors and that the nature of the delivered drug is what determines the induced mechanism of cell death.

Although the exact mechanism of action of compound 1-encapsulated hsAFt has not been studied in detail,

based on our results and literature data [1, 7, 32, 36, 37] the following scenario can be drawn: drug-loaded hsAFt

administered to the cell can be internalized via SCARA5 leading to endosomal internalization. Endosomal-

mediated import of compound 1-encapsulated hsAFt leads to drug release induced by the low pH in endosome.

Compound 1 (or compound 1 fragments) is thus internalized and can act as free compound, inducing necrosis,

as it does when it is protein-free. It is expected that the use of protein cage would make the Pt(II) compound

more biocompatible and selective in vivo. The cage should protect it from the attack of other

biomacromolecules, but further studies are needed to verify this hypothesis.

Conclusions

The combined use of X-ray crystallography and mass spectrometry, which is a valuable tool for the structural

characterization of protein-metal adduct [38-39], has indicated that we have prepared the ferritin

nanocomposite with the highest amount of Pt atoms within the cage reported so far. Compound 1-encapsulated

hsAFt is cytotoxic towards MCF-7 and A431 human cancer cell lines and is more toxic than cis-Pt at high

concentrations; it induces necrosis in cancer cells, similarly to the free drug. This suggests that the delivered

drug is responsible for the biological properties of the Ft-Pt(II) nanocomposites and dictates the cell death

mechanism. These findings and literature data [1, 7, 32, 36, 37] suggest that hsAFt can be used as an efficient,

biocompatible delivery system, since it can selectively release drugs into the cancer cells and protect them in

the cellular environment.

Future perspectives

hsAFt is a good drug delivery system also because it has a long circulation half-life and is available at low cost

and biodegradable. Our data could be useful for the design of new Ft-Pt(II) nanocomposites. However, several

open questions should be addressed in future research. The mechanisms of encapsulation of drugs into the Ft

cage and their release are not known and other in vitro and in vivo studies are needed to clarify the mechanism

of action of these molecules, to investigate their fate after systemic injection and to evaluate possible use in

clinics.

Page 15: Preparation, structure, cytotoxicity and mechanism of

References:

1 Abe S, Maity B, Ueno T. Design of a confined environment using protein cages and crystals for the development

of biohybrid materials. Chem. Commun. 52, 6496-6512 (2016).

•• of considerable interest. A feature article summarizing the development of ferritin as the most promising

molecular template protein cage and in vivo and in vitro engineering of protein crystals as solid protein

materials.

2 Uchida M, Kang S, Reichhardt C, Harlen K, Douglas T. The ferritin superfamily: Supramolecular templates for

materials synthesis., BBA-Gen. Subjects. 1800(8), 834-845 (2010).

3 Andrews SC, Arosio P, Bottke W et al. Structure, Function, and Evolution of Ferritins. J. Inorg. Biochem. 47(3-

4), 161-174 (1992).

4 Hogemann-Savellano D, Bos E, Blondet C et al. The transferrin receptor: A potential molecular Imaging marker

for human cancer. Neoplasia. 5(6), 495-506 (2003).

5 Mendes-Jorge L, Ramos D, Valenca A et al. L-Ferritin Binding to Scara5: A New Iron Traffic Pathway Potentially

Implicated in Retinopathy. Plos One, 9(9), e106974 (2014).

6 Li JY, Paragas N, Ned RM et al. Scara5 Is a Ferritin Receptor Mediating Non-Transferrin Iron Delivery. Dev. Cell.

16(1), 35-46 (2009).

7 Liang MM, Fan KL, Zhou M et al. H-ferritin-nanocaged doxorubicin nanoparticles specifically target and kill

tumors with a single-dose injection. Proc. Natl. Acad. Sci. USA. 111(41), 14900-14905 (2014).

** of considerable interest. A very interesting paper describing the efficacy of H-ferritin nanoparticles as

effective vehicle for doxorubicin.

8 Monti DM, Ferraro G, Petruk G et al. Ferritin nanocages loaded with gold ions induce oxidative stress and

apoptosis in MCF-7 human breast cancer cells. Dalton Trans. 46, 15354-15362 (2017).

9 Ferraro G, Monti DM, Amoresano A et al. Gold-based drug encapsulation within a ferritin nanocage: X-ray

structure and biological evaluation as a potential anticancer agent of the Auoxo3-loaded protein. Chem.

Commun. 52(61), 9518-9521 (2016).

* of interest. An interesting paper describing the encapsulation of gold-based drugs within the ferritin

nanocage.

10 Pontillo N, Pane F, Messori L, Amoresano A, Merlino A. Cisplatin encapsulation within a ferritin nanocage: a

high-resolution crystallographic study. Chem Commun, 52, 4136-4139 (2016).

* of interest. The first complete structural characterization of a ferritin nanocage loaded with a metal-based

drug.

11 Ferraro G, Ciambellotti S, Messori L, Merlino A. Cisplatin Binding Sites in Human H-Chain Ferritin. Inorg. Chem.

56(15), 9064-9070 (2017).

12 Ji XT, Huang L, Huang HQ. Construction of nanometer cisplatin core-ferritin (NCC-F) and proteomic analysis

of gastric cancer cell apoptosis induced with cisplatin released from the NCC-F. J. Proteomics. 75(11), 3145-3157

(2012).

13 Falvo E, Tremante E, Fraioli R et al. Antibody-drug conjugates: targeting melanoma with cisplatin

encapsulated in protein-cage nanoparticles based on human ferritin. Nanoscale. 5(24), 12278-12285 (2013).

14 Pontillo N, Ferraro G, Helliwell JR, Amoresano A, Merlino A. X-ray Structure of the Carboplatin-Loaded Apo-

Ferritin Nanocage. ACS Med. Chem. Lett. 8(4), 433-437 (2017).

15 Xing RM, Wang XY, Zhang CL et al. Characterization and cellular uptake of platinum anticancer drugs

encapsulated in apoferritin. J. Inorg. Biochem. 103(7), 1039-1044 (2009).

16 Cornell TA, Orner BP. Medium throughput cage state stability screen of conditions for the generation of gold

nanoparticles encapsulated within a mini-ferritin. Bioorg. Med. Chem. S0968-0896(17)32341-6 (2018).

Page 16: Preparation, structure, cytotoxicity and mechanism of

17 Maity B, Abe S, Ueno T. Observation of gold sub-nanocluster nucleation within a crystalline protein cage. Nat

Commun. 16, 8,14820 (2017).

18 Ferraro G, Petruk G, Maiore L et al. Caged noble metals: Encapsulation of a cytotoxic platinum(II)-gold(I)

compound within the ferritin nanocage. Int. J. Biol. Macromol. 115, 1116-1121 (2018).

19 Ciambellotti S, Pratesi A, Severi M et al. The NAMI A – human ferritin system: a biophysical characterization.

Dalton Trans. 47(33), 11429-11437 (2018).

20 Belletti D, Pederzoli F, Forni F, Vandelli M, Tosi G, Ruozi B. Protein cage nanostructure as drug delivery

system: magnifying glass on apoferritin. Expert Opin. Drug Deliv. 14(7), 825-840 (2017).

21 Yang Z, Wang XY, Diao HJ et al. Encapsulation of platinum anticancer drugs by apoferritin. Chem. Commun.

33, 3453-3455 (2007).

22 Suntharalingam K, Mendoza O, Duarte AA, Mann DJ, Vilar R. A platinum complex that binds non-covalently

to DNA and induces cell death via a different mechanism than cisplatin. Metallomics. 5, 514-523 (2013).

23 Ferraro G, Marzo T, Infrasca T et al. A case of Extensive Protein Platination: the Reaction of Lysozyme with a

Pt(II) terpyridine complex. Dalton Trans. 47(26), 8716-8723 (2018).

24 Vonrhein C, Flensburg C, Keller Pet al. Data processing and analysis with the autoPROC toolbox. Acta Cryst.

D Biol. Crystallogr. 67(4), 293-302 (2011).

25 McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software.

J. Appl. Crystallogr. 40(4), 658-674 (2007).

26 Emsley P, Lohkamp B, Scott WG, Cowtan K. Features and development of Coot. Acta Cryst. D. Biol. Crystallogr.

66, 486-501 (2010).

27 Murshudov GN, Skubak P, Lebedev AA et al. REFMAC5 for the refinement of macromolecular crystal

structures. Acta Cryst. D Biol. Crystallogr. 67, 355-367 (2011).

28 Bricogne G, Blanc E, Brandl M, Flensburg C, Keller P, Paciorek W, Roversi P, Sharff A, Smart O, Vonrhein C,

Womack T. BUSTER version X.Y.Z. Cambridge, United Kingdom: Global Phasing Ltd, in, 2017.

29 Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and

cytotoxicity assays. J. Immunol. Methods. 65(1-2), 55-63 (1983).

30 Galano E, Arciello A, Piccoli R, Monti DM, Amoresano A. A proteomic approach to investigate the effects of

cadmium and lead on human primary renal cells. Metallomics. 6, 587-597 (2014).

31 Tanley SW, Schreurs AM, Kroon-Batenburg LM, Helliwell JR.Re-refinement of 4g4a: room-temperature X-ray

diffraction study of cisplatin and its binding to His15 of HEWL after 14 months chemical exposure in the presence

of DMSO. Acta Cryst. F. Struct. Biol. 72, 253-254 (2016).

32 Cutrin JC, Alberti D, Bernacchioni C et al. Cancer cell death induced by ferritins and the peculiar role of their

labile iron pool. Oncotarget. 9, 27974-27984 (2018).

33 Messori L., Marzo T, Merlino A. Interactions of Carboplatin and Oxaliplatin with Proteins:insights from X-ray

structures and mass spectrometry studies of their Ribonuclease A adducts. J. Inorg. Biochem., 153, 136-142

(2015).

34 Messori L, Merlino A. Ruthenium metalation of proteins: the X-ray structure of the complex formed between

NAMI-A and hen egg white lysozyme Dalton Trans, 43, 6128-31 (2014).

35 Merlino A. Interactions between proteins and Ru compounds of medicinal interest: a structural perspective.

Coord. Chem. Rev., 326, 111–134 (2016).

36 Zhang L, Li L, Di Penta A, Carmona U, Yang F, Schops R, Brandsch M, Zugaza JL, Knez M. H-chain ferritin: a

natural nuclei targetin and bioactive delivery nanovector. Adv. Healthcare Mater. 4, 1305-1310 (2015).

37 Zhen Z, Tang W, Tod T, Xie J. Ferritins as nanoplatforms for imaging and drug delivery. Expert Opin. Drug Deliv

11(12), 1913-1921 (2014).

38 Messori L, Merlino A. Protein Metalation by Metal-Based Drugs:X-ray crystallography and Mass Spectrometry

studies. Chem. Commun (Camb) 53, 11622-11633 (2017).

Page 17: Preparation, structure, cytotoxicity and mechanism of

39 Merlino A, Marzo T, Messori L. Protein Metalation by Anticancer Metallodrugs: A Joint ESI MS and XRD

Investigative Strategy. Chemistry. 23(29), 6942-6947 (2017).

Summary points:

*A cytotoxic Pt(II)-terpyridine compound has been encapsulated within a ferritin nanocage.

*Inductively plasma coupled mass spectrometry data and X-ray crystallography indicate that Ft remains intact

upon the encapsulation process, while the Pt compound could be at least in part degraded.

*The in vitro cytotoxicity studies of the Pt(II)-AFt nanocomposite show that it is cytotoxic for cancer cells.

*The Pt(II)-AFt nanocomposite induces necrosis in cancer cells.

* Pt(II)-AFt nanocomposites could be powerful tools for cancer therapy.

* The encapsulation of a metallodrug within Ft is a promising strategy to prepare new anticancer agents.

Graphical abstract