Nile Blue-Based Nanosized pH Sensors for Simultaneous Far-Red and Near-Infrared Live Bioimaging

Diblock copolymer vesicles are tagged with pH-responsive Nile Blue-based labels and used as a new type of pH-responsive colorimetric/fluorescent biosensor for far-red and near-infrared imaging of live cells. The diblock copolymer vesicles described herein are based on poly(2-(methacryloyloxy)ethyl phosphorylcholine-block-2-(diisopropylamino)ethyl methacrylate) [PMPC-PDPA]: the biomimetic PMPC block is known to facilitate rapid cell uptake for a wide range of cell lines, while the PDPA block constitutes the pH-responsive component that enables facile vesicle self-assembly in aqueous solution. These biocompatible vesicles can be utilized to detect interstitial hypoxic/acidic regions in a tumor model via a pH-dependent colorimetric shift. In addition, they are also useful for selective intracellular staining of lysosomes and early endosomes via subtle changes in fluorescence emission. Such nanoparticles combine efficient cellular uptake with a pH-responsive Nile Blue dye label to produce a highly versatile dual capability probe. This is in marked contrast to small molecule dyes, which are usually poorly uptaken by cells, frequently exhibit cytotoxicity, and are characterized by intracellular distributions invariably dictated by their hydrophilic/hydrophobic balance.


Figures
Hydrogen abstraction from Nile Blue by a polymeric radical leading to formation of a Nile Blue radical (corresponding to reaction (6) in (A)) and 2) Reaction between a Nile Blue radical and a (different) polymer radical chain (corresponding to reaction (7)

Synthesis of NBM monomer
Nile Blue A (sulfate salt form) (5.032 g, 13.7 mmol) was placed in a 500 mL round-bottom flask fitted with a magnetic stir bar. Dichloromethane (250 mL) and triethylamine (6 mL, 4.356 g, 43.0 mmol) were added and the resulting dark red mixture was placed under nitrogen and cooled on an ice-bath.
Methacrylic anhydride (3 mL, 3.1 g, 20.1 mmol) was added as a solution in dichloromethane (50 mL), followed by DMAP (0.1072 g, 0.877 mmol), and the mixture was allowed to heat to room temperature (~22 °C). The reaction was monitored by TLC (dichloromethane:methanol 19:1 v/v). After 22 h, further methacrylic anhydride (1.10 mL, 1.14 g, 7.38 mmol) and triethylamine (2.00 mL, 1.45 g, 14.3 mmol) was added and the reaction mixture was stirred for a further 26 h. Solvent was evaporated at 35 °C under reduced pressure. The solid was washed with water (4 x 150 mL) and filtered.
The solid was redispersed in diethyl ether (500 mL). Then, 2 M HCl in diethyl ether (10 mL, 20 mmol HCl) was added, which led to the formation of a dark blue precipitate. The flask was placed at -25 °C.
After 15 h, the flask was allowed to heat to room temperature and the resulting crystals were collected on filter paper and washed with diethyl ether (250 mL). The product was dried in a vacuum oven at 25 °C to give 4.3 g (74 %) of product.
The product was further purified on a silica column using a 9:1 dichloromethane:methanol mixture as a mixed eluent. Data below are reported for the non-columned material.

Reverse-phase HPLC analysis of Nile Blue monomers
HPLC chromatograms were acquired using a Shimadzu UFLC XR system consisting of two LC20AD

Dye purification
Due to the relatively low level of dye incorporation the amount of dye which is not incorporated into the polymer becomes significant. Therefore, it is beneficial that the dye is relatively cheap or can be prepared with a minimal amount of purification. For this reason, a technical quality of Nile Blue was used for incorporation into polymers and for preparing NBM and NBC. Similarly, only precipitation was used to purify NBM and NBC used for incorporation into polymers. As shown in Table S1, these dyes each had purities of at least 85 % as measured by analytical HPLC prior to further chromatographic purification. After the final sample was removed, the reaction mixture was diluted with methanol and dialyzed against methanol until the dialyzate was colorless. The solution was then dialyzed against water and freeze-dried.

Synthesis of PMPC-PDPA-NBM with Nile Blue dye present throughout
MEBr (0.0392 g, 140 µmol, 1 eq.) and MPC (1.0363 g, 3.5 mmol, 25 eq.) was dissolved in 1.5 mL ethanol and the solution was purged with nitrogen for 20 minutes. Then, bpy (0.0445 g, 285 µmol, 2 eq.) and CuBr (0.0274 g, 191 µmol, 1.4 eq.) was added. The dark brown solution was left at 30 °C for 1.5 h. Then a nitrogen-purged solution of DPA (2.0310 g, 9.52 mmol, 68 eq.) and NBM (0.0057 g, 14.7 µmol, 0.1 eq.) in ethanol (3 mL) was added. The reaction mixture was stirred for 18 h, whereupon 1 H NMR indicated that all vinylic groups had reacted. Then the viscous solution was exposed to the atmosphere and diluted with ethanol (50 mL). The solution was passed through a silica column using ethanol as eluent to remove spent catalyst. Then the ethanolic solution was transferred to a dialysis bag (MWCO 1,000) and dialyzed against ethanol, then methanol and finally water. The opaque solution was freeze-dried to give the polymer as a beige powder.
The reaction mixture was stirred at 30 °C for 16 h. Then Nile Blue or Nile Blue monomer (265 µmol, 1 eq.) was added as a solid and the reaction mixture was stirred for a further 24 h. The viscous solution was exposed to the atmosphere, diluted with methanol (50 mL) and dialyzed (MWCO 1,000) against methanol until the dialyzate was colorless, followed by dialysis against water. Finally the polymer was freeze-dried overnight.

S8
The general protocol described in the literature 4   After cooling, the solution was transferred to a dialysis bag (MWCO 1,000 Da) and dialyzed against methanol until the dialyzate was colorless, followed by water. Finally the polymer was freeze-dried overnight.

Determination of molar absorption coefficients and integrated absorption coefficients for dyes in ethanol
A PC-controlled CARY 50 PROBE UV/Visible spectrophotometer was used to record spectra from 300 nm to 800 nm at a scan rate of 4800 nm min -1 . All measurements were performed using disposable UV-grade cuvettes. Solutions for measuring the molar absorption coefficient of Nile Blue and Nile

Spectroscopic determination of polymer dye contents in ethanol
Between 10.0 and 20.0 mg polymer was weighed into a 10 mL volumetric flask. Three solutions were prepared for each polymer. In general, the absorption maximum of these solutions was between 0.2 and 1.2. For all solutions, the maximum absorbance and integrated absorbance (determined as described above) was used to assess the dye content in the solution. Knowledge of the polymer molecular weight from the target degree of polymerization and monomer conversion (as judged by 1 H NMR) was used to obtain the value given in Table 2, see main text.

Absorption of Polymers in PBS and determination of relative quantum yield
Relative quantum yields were determined using the procedure of Fery-Forgues and Lavabre. 6

pH-dependent emission of Nile Blue-containing PMPC homopolymers in PBS
To 10.0 mL of a solution of the homopolymer in PBS used for evaluating quantum yield (see above), was added 0.5 mL 1 M HCl. This lowered the pH to below 2. A 3.0 mL aliquot was removed after measuring the solution pH using a calibrated pH meter (Hanna Instruments) and the emission spectrum was measured, exciting at 550 nm and 650 nm. The aliquot was then recombined with the original solution. The settings of the spectrofluorimeter was adjusted to give a maximum emission of less than 50 % of the maximum range. In general, settings were those used for obtaining the quantum yield (see above) but in some cases it was necessary to adjust the slit widths when the dye content was very low.
These settings were maintained throughout the experiment. The pH was then slowly increased using 1 M NaOH added via a syringe pump (World Precision Instruments). At the desired pH, the emission spectrum was obtained by removing 3.0 mL of solution, followed by replacing it as described above. 176.751-QS, 3 mm light path) placed within a Fluoromax-3 fluorimeter. The fluorescence spectra were recorded at predetermined pH values. The spectrofluorimeter settings were adjusted to give a maximum emission of less than 50 % of the maximum range. In general, settings were those used for obtaining the quantum yield (see above) but in some cases it was necessary to adjust the slit widths when the dye content was very low. These settings were maintained throughout the experiment.

Cellular uptake of Nile Blue vesicles: Live imaging via confocal fluorescence microscopy.
Cells were seeded at a density of 5 × 10 3 cells/well in BD Falcon 96-well imaging plates and grown until 50% confluence. Cells were treated overnight (typically 16 h) with 1.0 mg/ml vesicles. The cells were washed three times with PBS and imaging medium (culture medium without phenol red) was added to each well in preparation for live imaging with a Zeiss LSM510 Meta system (40X). The pHdependent subcellular localization was monitored using a commercial early endosomes marker for Rab-5 (CellLight® Early Endosomes-GFP, Invitrogen) and a commercial lysosomal marker (Lysotracker®, Invitrogen) following the manufacturer's instructions.

3D multicell tumour spheroid (MCTS) culture and image analysis
S12 3D MCTS cultures of the MDA-MB-231 breast cancer cell line were produced using ultra-low attachment (ULA) 96-well round bottomed plates, as previously described. 9 For optimal 3D growth and high reproducibility, 1.5 x 10 4 cells per well were seeded in 200 µl of cell media (RPMI-1640 supplemented with 10 % (v/v) foetal calf serum, 2 mM L-glutamine, 100 IU/ml penicillin, 100 mg/ml streptomycin). Spheroids were fed on the third day after seeding and thereafter every other day until the sixth day, whereupon they were fed daily thereafter.
MCTS were treated with Nile Blue vesicles (1.0 mg/ml in cell medium) for approximately 30-36 h.
Afterwards, live spheroids were washed three times with PBS and imaged with a digital camera (5x, full spheroid per image ( Figure 2B) or under an optical microscope (20x magnification, approximately 25 % of a spheroid per image, see Figure S13) in order to identify the boundary for the color change from physiological pH (pink) to hypoxic/low pH (blue). Bright-field optical images were recorded using a Leica DMI4000B equipped with a Nuance Multispectral imaging system. Untreated control spheroids were used to subtract the background noise from the images. The image analysis was performed using a double-blind protocol. Four independent observers were asked to score the center of the tumor (marked with a white star in Figure S13), the edge (red dotted line in Figure S13) and identify the color change, the acidic/blue color margin line (black dotted line in Figure S13). Following this analysis, acidic regions (blue colored) of the spheroids were compared to the full size of the MCTS and the ratio of the hypoxic (acidic) radius to the total radius of the MCTS was calculated over time.
Hypoxic (low pH regions) of the spheroids were compared to a control marker for hypoxia (Glut-1).
Briefly, control spheroids were grown under the same conditions as described above and fixed with 4% buffered paraformaldehyde at specific time points. Control spheroids were then incubated for 1 h in a solution of 5% BSA to block unreactive sites. Immunolabeling was performed using rabbit-anti human Glut-1 1:100 (Abcam, UK) for 30 mins followed by goat-anti rabbit alkaline phosphatase conjugated Mean ratios were calculated for N = 3 independent experiments and analysis of differences between the groups was performed using the Student's paired t-test.
Scheme S1: Electronic structure of Nile Blue dye dissolved in polar and non-polar solvents.   [a] GPC studies were conducted in a 3:1 chloroform/methanol mixed eluent containing 2.0 mM LiBr using a refractive index detector and a series of near-monodisperse poly(methyl methacrylate) calibration standards.
[b] Based on integrated absorption coefficient data.