Hybrid Hierarchical Heterostructures of Nanoceramic Phosphors as Imaging Agents for Multiplexing and Living Cancer Cells Translocation

Existing fluorescent labels used in life sciences are based on organic compounds with limited lifetime or on quantum dots which are either expensive or toxic and have low kinetic stability in biological environments. To address these challenges, luminescent nanomaterials have been conceived as hierarchical, core–shell structures with spherical morphology and highly controlled dimensions. These tailor-made nanophosphors incorporate Ln:YVO4 nanoparticles (Ln = Eu(III) and Er(III)) as 50 nm cores and display intense and narrow emission maxima centered at ∼565 nm. These cores can be encapsulated in silica shells with highly controlled dimensions as well as functionalized with chitosan or PEG5000 to reduce nonspecific interactions with biomolecules in living cells. Confocal fluorescence microscopy in living prostate cancer cells confirmed the potential of these platforms to overcome the disadvantages of commercial fluorophores and their feasibility as labels for multiplexing, biosensing, and imaging in life science assays.


DLS, TEM, EDS and Fluorescence spectroscopy characterization
The degree of agglomeration of the nanoparticles in solution was followed by dynamic light scattering (DLS) measurements working with aqueous suspensions of 1 mg/ml in concentration. The results reveal the formation of agglomerates of around 1 m in size ( Figure S1), which would point to a strong trend of the synthesized nanos towards agglomeration; as explained in the text, this preliminary parameter will not be a major impediment to the subsequent production of composites with a suitable degree with which to achieve an unprecedented incorporation of the manufactured nanoparticles into the cells. Figure S1. Dynamic light scattering spectra in water of a) Y0.9Er0.1VO4 (concentration = 1 mg/ml in water) and b) Y0.9Eu0.1VO4 (concentration = 1 mg/ml in water).  Figure S3 show the EDS of @SiO2 composites (those for Chitosan are alike, barely showing an increased intensity in the C and O signals). The EDX analyses in both samples indicate a homogeneous composition in agreement with the nominal Y0.9Eu0.1VO4 and Y0.9Er0.1VO4 formulas and the published data on commercial powder Eu:YVO4 samples. The results indicate that the lanthanide dopants, either Er or Eu, are effectively incorporated into the vanadate matrix by this novel hydrothermal method and are maintained after the processes of obtaining the coreshell composites.     Figure S8. Normalized 2D fluorescence contour plots recorded for dispersions in water (0.05 mg/mL) of bulk powder particulate Y0.9Eu0.1VO4. Figure S6 shows the excitation / emission spectra of bulk powder particulate Y0.9Eu0.1VO4 with different concentrations in water. As it can be seen on the spectra samples absorbed at 280 nm. So, the calculated quantum yields were obtained at 280 nm using tryptophan which absorbs at 280 and 290 nm as reference. Figure S9 shows the fluorescence emission spectra of tryptophan and fluorescein. The table S1 shows the quantum yield of fluorescein and tryptophan and the refractive index of both compounds.

Quantum yield
= Qr * * * * ns 2 2 The Table S1 and Table S2 show the reference data for two fluorescence references considered to carry out the quantum yield calculations and the refractive index of common solvents.   Table S3 lists the fluorescence data extracted from the fluorescence spectra of bulk powder particulate Y0.9Eu0.1VO4 ( Figure S6b)  Figure S11. Luminescence spectra of bulk powder particulate Y0.9Eu0.1VO4 (λex = 290 nm) in a) PBS 0.05 mg/mL stored in a glass vial, protected from light at room temperature up to 20 days and b) PBS with glycerol 20% 0.05 mg/mL stored in a glass vial, protected from light at room temperature up to 3 days. Figure S12. Luminescence spectra of bulk powder particulate Y0.9Eu0.1VO4 (λex = 290 nm) in cell medium (MEM) (0.8 mL MEM 10% FBS + 0.05 mg/mL NPs, 0.2 mL Milli-Q water) protected from light at room temperature up to 2 days. Cell studies Scheme S1. Cell uptake experiment. Figure S15. Luminescence spectra of bulk powder particulate Y0.9Eu0.1VO4 nanomaterials and control samples (λex = 290 nm).
As it can be observed in the Figure S16, no incorporation of bulk powder particulate Y0.9Eu0.1VO4 into the cell takes place. This was also confirmed by confocal fluorescence microscopy ( Figure S17). Figure S16. Fluorescence image of live HeLa cells incubated with bulk powder particulate Y0.9Eu0.1VO4 (0.05 mg/mL) at a) and e) t = 0 min (without nanoparticles), b) and f) t = 15 min, c) and g) t = 60 min, and d) and h) t = 360 min. ex = 488 nm. Scale bar: 50 m.
In addition, it has been proven experimentally that, after 6 hours of incubation, there is practically no incorporation of uncoated bulk powder particulate Y0.9Eu0.1VO4 into the cell, and the non-toxicity of the nanoparticles Figure S17.

PEG-ylation Procedures
The amino PEG precursors and the aminooxy H2N-PEG-OC3H6-COOH polymer with Mw ca. 2000 g/mol, 3000 g/mol and 5000 g/mol were purchased from Rapp Polymere or NanoCS and functionalised in-house through the synthetic approaches described below, giving rise to Compounds Type I and II below.
In parallel experiments the surface of the bulk powder material Y0.9Ln0.1VO4 (denoted Eu:YVO4) and of the core-shell Y0.9Ln0.1VO4@SiO2 nanoparticles was functionalized with functional PEGylated silane polymers with Mw of ca. 2000, 3000 and 5000 Da, whereby n were of 40 (C7H16NO4Si(C2H4O)40CH3O), 60 (for C7H17N2O4Si(C2H4O)60C4H7O3), and finally C7H17N2O4Si(C2H4O)100C4H7O3 of ca. 100 PEG mer units), through a number of optimised routes, described below. In case of the bulk powder material Y0.9Ln0.1VO4 (denoted Eu:YVO4) one-pot reaction functionalisation, to prevent the possibility that the -NCO functional reactants (2) may still be present on the NP surface, the quenching reaction with the amino PEG precursors (compounds Type I) were carried out in situ, aiming to attach these directly on the -NCO terminal groups on the NPs. These type of reactions were also carried out for the linker functionalisation optimisation strategies, applied in a stepwise reaching, after the isolation of the molecular compounds of types I and II, characterised by IR, mass spectrometry and GPC.
Scheme S1. General functionalisation of bulk powder materials, in one-pot reaction approach used for the optimization reactions for functionalization with -COOH groups followed by quenching in situ of -NCO residues.

Scheme S2.
Stepwise functionalization protocol for optimization reactions: (i) Formation of functional linker molecules followed by (ii) Direct NPs surface functionalization and quenching, adopted hereby as the functionalisation approach for direct functionalisation of Y0.9Eu0.1VO4 bulk powder nanomaterial used for kinetic stability investigations. Box: Structures of linkers precursors and quenchers isolated, Types I and II, whereby compounds type I (n = 40, 60 or 100-111 range) were typically used in the quenching reactions. The functionalised linkers (with PEGs 2 kDa, 3 kDa and 5kDa) denoted compounds type II and derived from the aminooxy H2N-PEG-O-C3H6-COOH (with Mw ca. 2 kDa, 3 kDa and 5 kDa) were isolated and characterised and used directly as molecular linkers for the nanophosphors conjugation and introduction of -COOH units. Trimethoxy(3-ureapropyl)polyethylene glycol-silane (black line).
A band at ca. 2300 cm -1 would be expected for an isocyanate group, which is clearly not present. However, a band and 1095 cm -1 were detected which are indicative of Si-O, showing that coating has occurred ( Figure S19).
Therefore, it appears that the -NCO group has hydrolysed, leaving an amine group. Bands at 1648 cm -1 and 1554 cm -1 (N-H bend region) and at 1198 cm -1 (C-N stretch region) are confirmative of the presence of a -NH2 group. There is also a band at 2940 cm -1 not visible in the uncoated, which could be indicative of one of the two expected the N-H stretch (despite being a little low). The other N-H stretch may be overlapping with the broad band occurring 1198 cm -1 and 1095 cm -1 1648 cm -1 and 1554 cm -1 2940 cm -1 S13 in the uncoated nanophosphors and is therefore not easily visible. To avoid the observed hydrolysis a one-pot approach and pre-reacting approach were carried out.
A Schlenk was pre-silylated using dimethyldichloride silane in vacuo using a desiccator for 2 hours and was then moved to an oven for 2h at 200°C. 10 mL dry, THF and 10 µL aqueous ammonia (34%) was added followed by 100 mg Eu:YVO4 nanophosphor under a flow of nitrogen and stirred for 1h. 1 mL Trimethoxy(3-isocyanotopropyl)silane was added and the reaction mixture was stirred overnight at room temperature. 100 mg of the 5,000 g/mol (NH2-PEG-COOH)HCl was added followed by 1 eq. of triethylamine under a flow of nitrogen. After 20h 250 mg of Poly(ethylene glycol) monomethyl ether (5,000 g/mol) was added under a flow of nitrogen (Scheme S3). This was allowed to react for 5h and then was centrifuged at 3000 rpm for 5 minutes and subsequently washed 3x with dry THF. After removal of THF the nanophosphors, the supernatant was analyzed by fluorimetry in Milli-Q water ( Figure S19). The characteristic spectrum can be visualized here (λex = 290 nm), where there was an absence of the emission peak observed in poorly dispersed coated nanoparticle (a broad peak ca. 350 nm). This therefore indicates that there is a good dispersability in THF and in water.
Both the liquid which remained within the dialysis membrane and the dialysate were isolated in vacuo ( Figure S23). The IR spectra of the un-purified product, the dialysate and that, which stayed within the membrane, are compared below ( Figure S24). Importantly the FTIR spectrum of the purified product is notably different from the unreacted nanophosphor and contains a significant band at 2890 cm -1 ( Figure S24) which is attributable to the carboxylic acid. Furthermore, the intense luminescence of which stayed within the membrane when irradiated with short UV light (ca. 254 nm) indicates that this reaction was successful and yielded 335 mg of coated Y0.9Eu0.1VO4 nanoparticles.

Approach 2: Test reaction approach for alternative PEG linkers functionalisation
Scheme S4. A model reaction with a short amine-functionalised PEG, leading to a mono-substituted siloxane compound.
Another approach was attempted whereby reacting the silane isocyanate with the NH2-PEG-tBoc protected amine prior to coating to the nanophosphor. Trimethoxy(3-isocyanotopropyl)silane, 1.084 g (0.0053 mmols) was diluted in 100 mL anhydrous DCM and added dropwise (over 8 h) to 0.783 g (0.0053 mmols) and allowed to stir over weekend (Scheme S4). DCM was removed in vacuo and an oily product was analysed by ESMS and 1 H and 13 C NMR. Mass spectrometry confirmed that there was a mixture of the mono(substituted) and di(substituted) silane (peaks ionised in positive mode as in the reaction scheme above). Importantly, there was no significant peak according to MS corresponding to the silane starting material (or its hydrolysed form) ( Figure S26) as was there no peak in the region of 120 ppm in the 13 C NMR (75.5 MHz).

Alternative functionalisation of linkers -reaction approach 3
Scheme S5. A model reaction with a short t-boc protected-amine-functionalised PEG, with the purpose of synthesising a monosubstituted siloxane compound.
277.3 mg (1.351 mmols) Trimethoxy(3-isocyanotopropyl)silane, was added to 335.4 mg (1.351 mmols) of a short t-Butoxycarbonyl-Amine-PEG-Amine, with vast excess of triethylamine (136.8 mg) in anhydrous dichloromethane (Scheme S5). A preliminary ES-MS demonstrated that a mixture of the desired product and undesirable bis(substituted) product were formed ( Figure S27). Figure S27. ES-MS of the product from reaction approach 3, showing a mixture of the desired mono(substituted) product (blue) and the by-product (red).
Alternative reaction approach 4 31.8 mg (29.4 µL, 0.1551 mmols) trimethoxy(3-isocyanotopropyl)silane was added to 0.1 g (0.1551 mmols) of t-Butoxycarbonyl-Amine-PEG-Amine (Mr = 644.7 g/mols) and to 3 equivalents of triethylamine. As was observed with model reaction 3 a preliminary ES-MS demonstrated that a mixture of the desired product and undesirable bis(substituted) product were formed ( Figure S28).  Both t-Boc protected amine reaction approaches 3 and 4 resulted in deprotection, despite being carried out under basic conditions. It can be speculated that improved results may be achieved, by rinsing glassware with base prior to the start of the reaction and addition of the t-boc protected amine as the last component.

Reactions using protocols involving a pre-reacting to silane step
To allow both more control over the reaction and analysis via additional methods an approach whereby the silane was reacted to the PEG prior to coating was explored.

S20
Scheme S6. Reaction to enable a controlled ratio of non-reacting PEG and to hetero-labelled PEG-amine.

Reaction A:
A Schlenk was pre-silylated using dimethyldichloride silane in vacuo using a desiccator for 2 hours and was then moved to an oven for 2h at 200°C. 10 mL anhydrous dioxane was added under a flow of nitrogen to which, 150 mg (0.73067 mmols) of Trimethoxy(3-isocyanotopropyl)silane and 1.607 g (ca 0.8 mmols) of PEG methyl ether (2,000 g/mol) were added, with the reaction mixture stirred for 24h at room temperature, and dried in vacuo yielding 1.7221 g, 96 % (Scheme S6). The reactants and the product were analysed by FT-IR ( Figure S32), TGA and GPC. TGA of the compound 1 possessed a first derivative at 251.94 °C, and plateaued at ca. 600 °C, with 7.4 % remaining at 900 °C. PEG methyl ether (2,000 g/mol) contained two Mn(GPC) of = 2423 and 667, with the Mn (theoretical) = 2014 (see 4.6 GPC section, Figure S47). GPC showed that compound 1 has three Mn of 5073, 2374 and 674, Mn(theoretical) = 2234 (see 4.6 GPC section, Figure S48). The major peak of the Mn(GPC) for PEG methyl ether (2,000 g/mol) was slightly smaller than that of compound 1, which is expected suggesting that the reaction took place. The disappearance of the isocyanate peak in the FT-IR spectrum at 2265 cm -1 and the presence of 1663 cm -1 and 1556 cm -1 , which indicate O-C and C=O respectively and would likewise suggest that the reaction happened. Preliminary experiments have been carried out using NMR on compound 1. Without suppression, the CH2 groups of the polyethylene glycol chain it was not possible to observe the other peaks within this polymer (see 4.4 NMR section Figure S43). With suppression, however in the region of CH2 groups of the polyethylene glycol chain, it was possible to presence of shifts at δ = 0.49 (SiCH2CH2CH2, 2H, m), 1.47 (SiCH2CH2CH2, 2H, m) (see NMR section Figure S44).

Thermogravimetric analysis
Thermogravimetric analysis is a technique whereby the mass of a sample is measured according to an increase in temperature. This approach was designed whereby a decrease in mass of coated nanophosphor can be correlated to the mass of the coated material. If the molecular weight of the coating material is known, the number of moles required to coat a given mass of nanoparticle can therefore be calculated.
The mass of the coating material can be calculated by: Where %C is the estimated percentage mass of coating material and % is the remaining percentage mass of sample Interestingly, the calculated molar ratio is 1:4.6, interestingly the ratio added was 1:8.17.

Probing the stability of the luminescence with time in aqueous media
The luminescence of coated and uncoated bulk powder particulate Y0.9Eu0.1VO4 nanophosphors, 0.1 mg/mL in Milli-Q water was tested using excitation at 290 nm ( Figures S10, S35-S37). This was carried out in glass vials and in Eppendorfs tubes to assess the suitability of these coated materials under biological testing conditions, whereby plastic Eppendorf tubes were used as standard. These were also assessed at room temperature and at 4 °C. Uncoated materials were investigated as a control at the same w/v concentration ( Figure S10 and S32). A luminescence emission was detectable up to 4 weeks under all conditions, with data gathered comparable at room temperature and at 4 °C as it was in glass vials and in plastic Eppendorf tubes. The luminescence of these samples was monitored over the two weeks.

General Cell culturing methods for fluorescence imaging
HeLa and PC-3 cell lines were purchased from American type culture collection (ATCC). Cells were grown as monolayers in T75 tissue culture flasks and cultured in Eagle's Minimum Essential Medium (EMEM) for HeLa and Roswell Park Memorial Institute medium (RPMI) for PC-3, 1% L-glutamine (200 mM), 0.5% penicillin/streptomycin (10 000 IU mL-1/10 000 mg mL-1). Cells were cultured at 37 °C in a humidified atmosphere of 5% CO2 in air and split once 70% confluence had been reached, using the corresponding cell medium. All steps were performed in absence of phenol red. Once cells reached more than 70% confluence, the supernatant containing dead cell matter and excess protein was aspirated. The live adherent cells were then washed with 10 mL of phosphate buffer saline solution twice to remove any remaining media. Cells were incubated in 3 mL of trypsin solution (0.25% trypsin) for 5 to 7 min at 37 °C.
After trypsinisation, 6 mL of medium containing 10% serum medium was added to inactivate the trypsin and the solution was centrifuged for 5 min (1000 rpm, 25 °C). The supernatant liquid was aspirated and 5 mL of serum medium was added to the cell matter left behind. Cells were counted using a haemocytometer and then seeded as appropriate.
For microscopy, cells were seeded into glass-bottomed Petri dishes and incubated for 12 h for HeLa and 24 h for PC-3 to ensure adhesion. Cells were plated in 35 mm uncoated 1.5 mm thick glass-bottomed dishes as 3 × 10 5 cells per dish and incubated for at least 24 h prior to imaging experiment. Once cells attached firmly, cells were washed with 990 µL Hank's Balanced Salt Solution (HBSS) five times and refilled with 990 µL of serum-free medium (SFM), then in each case, an aliquot of 10 µL of the nanoparticulate material (generally as 1 mg/mL stock dispersion in DMSO) was added. Cells were incubated with compounds for 15 minutes, 1 h, 6h or over-night at 37 °C, or longer, as required. Afterwards, cells were washed with 990 µL Hank's Balanced Salt Solution (HBSS) three times to rinse any remai ning probe traces from the medium and 990 µL of SFM was added.
Once the cell dish was ready for the single photon confocal fluorescence imaging, cells were excited at 405 nm, 488 nm and 561 nm wavelength, then at each wavelength, there were five images c aptured namely a merged image, image between 420 and 480 nm wavelength, image between 516 and 530 nm wavelength, image between 615 and 650 nm wavelength and Differential Interference Contrast (DIC) image.

Cellular Viability Tests
Standard MTT assays of HeLa cells treated with Y0.9Eu0.1VO4@SiO2 composite were performed in order to investigate the effect of silica encapsulation on the cellular viability. The results demonstrate that encapsulation of Y0.9Eu0.1VO4 NPs within a silica shell improves the in vitro biocompatibility. Figure S85. Dose response curve for Y0.9Eu0.1VO4@SiO2 composite. Error bars stand for standard error calculated from the twelve repeats.

S56
DMSO control Y0.9Eu0.1VO4@SiO2 Figure S89. Normalised cell viability in PC3 cells treated with Y0.9Eu0.1VO4@SiO2 composite for 48 hours. Significance in difference between two groups were tested by Student t test. The asterisk marks a significant difference at the level of p <0.05.