Dual-fluorophore ratiometric pH nanosensor with tuneable p K a and extended dynamic range †

Ratiometric pH nanosensors with tuneable pKa were prepared by entrapping combinations of two pH-sensitive fluorophores (fluorescein isothiocyanate dextran (FITC-D) and Oregon Green dextran (OG-D)) and a reference fluorophore (5-(and-6)-carboxytetramethylrhodamine dextran (TAMRA-D)), in a biocompatible polymer matrix. Dual-fluorophore pH nanosensors permit the measurement of an extended dynamic range, from pH 4.0 to 7.5.

can be quantified with the use of analytical techniques such as fluorescence spectroscopy and widefield or confocal microscopy.Due to their small size, nanosensors can also provide high spatial resolution when compared to pulled optical fibres. 15H-sensitive fluorescent nanosensors have previously been reported, 12,[16][17][18] however, they are restricted when using commercially available fluorophores to a defined pK a and a limited pH range of measurement.19 For pH-sensitive fluorophores the pK a is generally the pH at which the fluorophore shows half its maximal response. More imprtantly, the pK a is the pH at which the fluorophore exhibits its greatest sensitivity to changes in pH.Therefore, a nanosensor which contains a fluorophore that has a pK a that can be tuned to a defined pH is highly desirable.For example, being able to derive maximal sensitivity from a sensor at selected pH would be of great value when monitoring cellular processes, such as apoptosis where small changes in pH are thought to be key.[20][21][22] Typically a pH-sensitive fluorophore responds in a sigmoidal manner to changes in pH, allowing the dynamic range and detection limit of the fluorophore to be determined.For a pH-sensitive fluorophore the dynamic range can be considered as the pH range between its minimal and maximal response. Whreas, the detection limit is the pH between the intersection of the lower and upper asymptote and the linear portion of the calibration curve.23 The detection limit has predominantly been used to characterise the measurement range for ion-selective electrodes and in some cases optodes.23,24 It has previously been reported that chemical sensors using a combination of absorbance based pH-sensitive dyes give optodes with an extended range of pH measurement [25][26][27] and indeed this is the principle behind Universal Indicator Solution.28 Using an analogous approach we have applied this theory to fluorophore based pH measurements where we have immobilised two pH-sensitive fluorophores within a polyacrylamide nanoparticle to produce a pH-sensitive nanosensor with extended measurement range and tuneable pK a .
Combinations of pH-sensitive fluorophores, fluorescein isothiocyanate dextran (FITC-D) and Oregon Green Ò dextran (OG-D), and a reference fluorophore 5-(and-6)-carboxytetramethylrhodamine dextran (TAMRA-D) were immobilised in an acrylamide based nanosensor matrix.OG-D and TAMRA-D were synthesised through conjugation to amino dextran 10 000 M w with succinimidyl ester forms of the fluorophores prior to entrapment. 29Dynamic light scattering of the nanoparticles showed a single distribution centred at a diameter of 40 nm, in agreement with previously published data (see the ESI † for experimental details). 8,15,30he tuning of pK a was demonstrated by varying the initial concentrations of FITC-D and OG-D during the preparation of nanosensors.Fig. 1 shows the change in pH calibration curves as the OG-D:FITC-D percentage is varied through successive increases in FITC-D concentration.This change is accompanied by an increase in the pK a for the nanosensors, from 4.8 to 6.4, corresponding to a change from OG-D to FITC-D.These values are comparable to previously reported pK a solution values for Oregon Green Ò and fluorescein of 4.7 and 6.4, respectively. 19aking the second derivative of the sigmoidal pH response curve allows improved visualisation of the dynamic range of pH measurement, Fig. 2a.The effective dynamic range, the region where the pH response is most sensitive to changes in pH, is the pH range between the maxima and minima points in the second derivative curve.Using this approach, nanosensors consisting of a single pHsensitive fluorophore, only OG-D (0% FITC-D) or FITC-D (100% FITC-D), have an effective dynamic range of approximately 1.15 pH units, Fig. 2b.Whereas, for nanosensors containing both FITC-D and OG-D fluorophores combined in a 1 : 1 ratio the effective   dynamic range is maximised, to 2.01 pH units.Therefore, through careful selection of the ratios of FITC-D to OG-D, the pK a can be tuned to create a nanosensor with maximal sensitivity at a desired pH, Fig. 1, and an extended pH measurement range, Fig. 2a.
Nanosensors containing both FITC-D and OG-D, in a 1 : 1 ratio, have overlapping emission curves, which peak at approximately 520 nm and increase in intensity when the pH is increased from 3.5 to 8.0, Fig. 3a.Using TAMRA-D as a reference dye, ratiometric measurements can be made over the range of pH to 7.5, Fig. 3b.
In conclusion, fluorescent nanosensors with tuneable pK a and size of approximately 40 nm have been demonstrated.When FITC-D and OG-D were combined in a 1 : 1 ratio and immobilised within the nanosensors the effective dynamic range was extended to 2.01 pH units as opposed to 1.15 pH units for sensors containing the individual fluorophores.We believe tuneable nanosensors will be of great value when applied to biological systems where nanosensors can be designed to have a specific pK a , so that sensitivity in a narrow pH range can be maximised.Additionally, the development of sensors with an extended dynamic range will enable simultaneous measurement of both cytoplasmic and endosomal pH, potentially eliminating the need to perform multiple experiments with more than one type of nanosensor containing different pH-sensitive fluorophores.
The authors gratefully acknowledge Professor Phil Williams for helpful discussions.This research is supported by the Biotechnology and Biosciences Research Council (BBSRC) and an industrial CASE award from GlaxoSmithKline (GSK) Consumer Health (grant number BBG0176381).

Fig. 1
Fig. 1 Normalised ratiometric calibration curves for nanosensors containing varying OG-D:FITC-D percentages and incorporating TAMRA-D as a reference.Percentage of FITC-D, with respect to OG-D, is increasing from the left.The calibration curves are fitted to a sigmoidal function each with R 2 values greater than 0.99 (see the ESI † for further details).

Fig. 2
Fig. 2 (a) Normalised second derivatives of calibration curves for nanosensors containing varying OG-D:FITC-D and incorporating TAMRA-D as a reference.(b) Graph illustrating the change in pK a and effective dynamic range with increasing proportions of FITC-D, with respect to OG-D.The effective dynamic range is fitted to a linear plot with a R 2 value of 0.98.Error bars represent standard error mean (n ¼ 3).

Fig. 3
Fig. 3 Emission (a) and calibration curves (b) for FITC-D, OG-D (1 : 1 ratio) and TAMRA-D nanosensors.Excitation/emission of FITC-D and OG-D at 488 nm/520 nm and TAMRA-D at 540 nm/577 nm.The calibration curve is fitted to a sigmoidal function with a R 2 value of 0.99 and pK a of 5.5 (see the ESI † for further details).Error bars representative of the standard error mean fall within the data points.(c) Residual errors of sigmoidal fit.