Highly swelling pH-responsive microgels for dual mode near infra-red fluorescence reporting and imaging

Near infra-red (NIR) fluorescence is a desirable property for probe particles because such deeply penetrating light enables remote reporting of the local environment in complex surroundings and imaging. Here, two NIR non-radiative energy transfer (NRET) fluorophores (Cy5 and Cy5.5) are coupled to preformed pH-responsive poly(ethylacrylate-methacrylic acid-divinylbenzene) microgel particles (PEA-MAA-5/5.5 MGs) to obtain new NIR fluorescent probes that are cytocompatible and swell strongly. NIR ratiometric photoluminescence (PL) intensity analysis enables reporting of pH-triggered PEA-MAA-5/5.5 MG particle swelling ratios over a very wide range (from 1–90). The dispersions have greatly improved colloidal stability compared to a reference temperature-responsive NIR MG based on poly(N-isopropylacrylamide) (PNP-5/5.5). We also show that the wavelength of maximum PL intensity (λmax) is a second PL parameter that enables remote reporting of swelling for both PEA-MAA-5/5.5 and PNP-5/5.5 MGs. After internalization the PEA-MAA-5/5.5 MGs are successfully imaged in stem cells using NIR light. They are also imaged after subcutaneous injection into model tissue using NIR light. The new NIR PEA-MAA-5/5.5 MGs have excellent potential for reporting their swelling states (and any changes) within physiological settings as well as very high ionic strength environments (e.g., waste water).


Introduction
Fluorescent reporting and imaging are invaluable tools for remotely studying complex environments such as those within gels and the body. [1][2][3][4][5][6][7] Fluorescence-based technologies provide relatively simple and inexpensive operating procedures whilst delivering real-time imaging and diagnostic data 8,9 with high sensitivity, 10,11 low background noise 12 and potentially low-cost imaging reagents. 13,14 The reporter probes studied to date include organic dyes, 15 uorescent dots, 16 plasmonic nanomaterials, 17 uorescent proteins 18 and upconverting nanoparticles. 16 However, uorescent probes can have limited sensitivity, 19 poor spatial resolution, 20 inadequate stability, 21 poor biocompatibility or even toxicity. 22 They may also suffer from high autouorescence, 23 or require irradiation conditions that result in excitation phototoxicity or thermally damage tissue. 24,25 Consequently, there is a continuing need to develop new uorescent probes with improved reporting characteristics. Here, we introduce a new pH-responsive probe that operates solely within the near infra-red (NIR) range and can report its dimensions using steady-state photoluminescence (PL) spectroscopy.
Responsive microgel particles (MGs) [26][27][28][29][30][31][32] are crosslinked polymer colloid particles that swell in a good solvent or as the pH approaches the pK a of the particle. 33 MGs are appealing as probes due to their fast response, chemical tunability, biocompatibility, suitability for functionalisation, biodegradability and soness. [34][35][36][37][38] MGs containing complementary uorescent non-radiative energy transfer 39,40 (NRET) uorophores have been used for monitoring MG size and swelling. 41,42 Fluorescent MGs can be prepared with high uorescent stability, quantum yield and large Stokes shis. [43][44][45] In addition, the MGs can detect intracellular pH, temperature, and ion concentration changes. [46][47][48] However, the PL intensity of many uorophores can be greatly reduced or completely quenched when assembled into nanostructures due to the intermolecular p-p-stacking. [49][50][51] Consequently, establishing new ratiometric probes with bright and stable emission is desirable. 52,53 We recently investigated uorescent probes that used shortwavelength excitation and emission. 7,41 Short-wavelengths limit applications of such probes as a result of autouorescence, 23 photo-toxicity 54 and poor light penetration depth. 55 The probes were also weakly swelling with particle volume swelling ratios (Q) < 6. To overcome these limitations in this study we report highly swollen pH-responsive MG probes that contain complementary NIR uorophores. NIR light is deeply penetrating in tissue. Moreover, NRET is highly sensitive to distance changes over the 1-10 nm range. 56 NRET requires a pair of uorophores wherein the donor emission overlaps with the acceptor absorption and energy transfer can occur via dipole-induced dipole coupling. 57 The NIR uorophores studied here have been used in temperature-responsive MGs based on poly(N-isopropylacrylamide) (PNP), 42 The new NIR MG probes introduced here (Scheme 1) comprise poly(ethyl acrylate-co-methacrylic acid-co-divinylbenzene) (PEA-MAA-DVB) and the covalently linked, complementary, sulfo-cyanine ourophores Cy5 (donor) and Cy5.5 (acceptor). The MG probes are termed PEA-MAA-5/5.5. We selected the PEA-MAA-DVB MGs to attach the NRET uorophores because previous work showed these MGs have very high pH-triggered swelling. 66 The new PEA-MAA-5/5.5 MG probes introduced here have four major advantages compared to our previous probes: (1) they enable ratiometric PL intensity detection of particle swelling changes over much higher Q ranges (from 1-90). (2) Emission occurs when they are illuminated with NIR irradiation. (3) They emit in the NIR region. (4) The wavelength at maximum PL intensity (l max ) provides a second MG size-dependent parameter for reporting changes in MG swelling. These advantages and also the differences between the present study and our previous work are shown in Fig. S1 (ESI †).
Here, the new MG probes are rst characterized and then the effects of pH variation on their PL spectra examined. These studies reveal that the maximum wavelength of the PL emission maximum provides a second new PL-based reporting mode for MG swelling. We then compare the PL reporting and colloidal stability of PEA-MAA-5/5.5 to a reference uorescent NIR PNP MG system. This part of the study shows the advantages of our new pH-responsive MG. We also use the pH-responsive MG probes to demonstrate imaging both in stem cells and aer subcutaneous injection in a tissue model. The new PEA-MAA-5/5.5 MGs have excellent future potential for remote NIR uorescent reporting and imaging in physiological settings as well as high ionic strength environments such as waste water.
Synthesis of uorescent NIR PEA-MAA-5/5.5 microgels The precursor PEA-MAA-DVB MG was prepared by seed/starvedfeed emulsion polymerization (see Scheme S1 †). A mixed comonomer solution (250 g) containing EA (164.4 g, 1.64 mol), MAA (82.2 g, 0.95 mol) and DVB (3.4 g, 0.026 mol) was prepared. Seed formation was conducted using a portion of the comonomer mixture (31.5 g) aer addition to water (517.5 g) containing SDS (1.8 g), K 2 HPO 4 (3.15 g of 7.0 wt% solution) and APS (10.0 g of 2.0 wt% solution). The seed was formed at 80 C with stirring under a nitrogen atmosphere over 30 min. The remaining co-monomer solution was added uniformly to the seed at a rate of 2.4 g min À1 . Aer completion of the feed stage the temperature was maintained at 80 C for 2.5 h. The MG dispersion was extensively dialyzed against water.
Synthesis of PEA-MAA-5 and PEA-MAA-5.5 microgels The methods used to synthesize non-NRET PEA-MAA-5 or PEA-MAA-5.5 particles were identical to that described above for PEA-MAA-5/5.5 (Scheme S2 †). The only difference in procedure is that only one uorophore was used for each system.

Synthesis of PNP-5/5.5 microgel probes
The precursor PNP MG dispersion was prepared by precipitation polymerization. 58 NIPAM (568 mg, 5.00 mmol), MBAAm (40.5 mg, 263 mmol), APMA (1.89 mg, 10.6 mmol) and SDS (23.9 mg) were dissolved in water (75 mL). This mixture was then deoxygenated with an N 2 purge for 1 h at 70 C in a 250 mL glass reactor, followed by the addition of aqueous APS solution (0.5 g, 4.0 mmol). The polymerization was conducted at 70 C for 5 h under an N 2 atmosphere. The dispersion was then cooled to room temperature and dialyzed extensively against water.

Physical measurements
Titration measurements were conducted in the presence of aqueous NaCl (0.10 M) at room temperature with aqueous NaOH solution (0.10 M) and a Mettler Toledo DL15 instrument. The z-average particle size (d z ) was determined by dynamic light scattering (DLS) using a Malvern Zetasizer Nano ZS. All measurements were conducted at 25 C unless otherwise stated.
The particle volume swelling ratio (Q) is calculated from the ratio of the swollen to collapsed particle volumes using the respective d z values. Zeta potential measurements were also obtained using the Malvern Zetasizer Nano-ZS instrument. TEM images were obtained using a FEI Tecnai 12 BioTwin instrument operating at an accelerating voltage of 110 kV. The particles were stained using 1% uranyl acetate solution. Confocal laser scanning microscopy images (CLSM) images were obtained using a broadband confocal Leica TCS SP5 microscope. UV-visible absorption spectra were recorded with a Hitachi U-1800 UV spectrophotometer with a scan rate of 480 nm min À1 . PL spectra were obtained using an Edinburgh Instruments FLS980 spectrometer. The position of the wavelength at maximum intensity (l max ) was determined by tting the main peak to a 6 th order polynomial. Unless otherwise stated the excitation wavelength (l ex ) was 650 nm. Details concerning the reversibility studies, MG uptake by the cells and cytotoxicity assays are given in the ESI. †

NIR imaging
NIR imaging was conducted using a laser with an incident wavelength of 650 nm. The sample investigated consisted of chicken breast (lean chicken breast purchased from Sainsbury Ltd., U.K.). A NIR camera (DCC 1240C from Thorlabs Ltd) tted with a lter which blocked light below 715 nm was used to obtain digital NIR photographs of the samples. A PEA-MAA-5/ 5.5 dispersion (0.50 mL, 0.050 wt%) was injected below the surface of the chicken breast at room temperature using a 21gauge needle.

Results and discussion
PEA-MAA-5/5.5 microgel probe characterization PEA-MAA-5/5.5 construction was performed in two steps. Firstly, PEA-MAA-DVB MG particles were synthesized by emulsion polymerization (Scheme S1, ESI †). The composition for the latter was PEA 0.58 -MAA 0.41 -DVB 0.01 based on titration data (for MAA content, Fig. S2A, ESI †) and the assumption that all of the DVB added was incorporated. Then the uorophores Cy5-NH 2 and Cy5.5-NH 2 (ref. 67) were covalently linked to the MG COO À groups at pH 7.5 using DMTMM activation 68 (Scheme S2, ESI †). The MAA content and apparent pK a for PEA-MAA-5/5.5 estimated from potentiometric titration data (Table S1 (ESI †)) are 41.1 mol% and 6.5, respectively. The PEA-MAA-5/5.5 diameter from TEM is 57 nm (Fig. 1A). DLS data ( Fig. 1B and S2B (ESI †)) gave a z-average diameter (d z ) of 66 nm in the collapsed state (pH 4.5). At pH 7.4 the particles have a d z value of 287 nm and are highly swollen (with Q ¼ 82) due to deprotonation of the RCOOH groups. (Size polydispersity index data appear in Fig. S2D and DLS diameter distributions are shown in Fig. S3, ESI †). Fig. S1A (ESI †) shows that the Q values for PEA-MAA-5/5.5 are much higher than those for our earlier (non-NIR) probe particles. 7,41 The zeta potential (z) was pH-responsive (Fig. S2C, ESI †) and decreased from À48 mV (at pH 7.4) to À29 mV (at pH 4.5). Hence, the anionic MGs have a high volume charge density in the swollen state, which is attributed to COO À groups.  Table S1 (ESI †). We consider further the data relating to the pHresponsiveness of PEA-MAA-5/5.5. Comparison of the TEM image for the particles deposited at pH $ 5 ( Fig. 1A) with the CLSM image of the particles swollen at pH 7.4 ( Fig. 1D) provides clear evidence that a large pH-triggered size increase occurred. Moreover, the d z and z values strongly increase with increasing pH and, interestingly, follow each other (see Fig. S2C, ESI †). Both increases begin aer pH 4.5 and are complete at pH 8.0 (highlighted in Fig. S2C †). The d z value originates from the whole particle; 69 whereas, z is determined by the electrophoretic mobility which, in turn, is governed by the charge density at the particle periphery. 70 Because d z and z follow each other closely it follows that the pH-triggered swelling/de-swelling transitions for PEA-MAA-5/5.5 MG probes occur uniformly throughout the particles, i.e., affine swelling occurs.
From the molar extinction coefficients measured for the uorophores (Fig. S6, ESI †) and the UV-visible spectra for PEA-MAA-5/5.5 (Fig. 1C) the Cy5 and Cy5.5 contents in the MG were calculated as 0.010 mol% and 0.002 mol%, respectively. The relatively high concentration of Cy5 is attributed to differences in MG-uorophore electrostatic interactions. The MG particles were negatively charged at pH 7.5; whereas, Cy5-NH 2 and Cy5.5-NH 2 had net charges of 0 and À2, respectively. (The structures of Cy5-NH 2 and Cy5.5-NH 2 are shown in Scheme S2 (ESI †).) Furthermore, the uorophore contents for PEA-MAA-Cy5 and PEA-MAA-Cy5.5 were found to be 0.010 and 0.005 mol%, respectively. The relatively high concentration of Cy5 in the PEA-MAA-5/5.5 MGs together with the small separation of the two PL peaks (25 nm, Fig. S5A (ESI †)), resulted in one main PL peak in the PL spectra for PEA-MAA-Cy5/5.5 (Fig. 1C). We show below that the wavelength of this peak is dependent on d z . The PEA-MAA-5/5.5 dispersion emitted bright red light when illuminated at 600 nm (inset of Fig. 1C). A CLSM image of the MGs at pH 7.4 (Fig. 1D) conrms emission occurred in the red region of the electromagnetic spectrum.

Reporting of pH-triggered microgel swelling
NRET is possible for PEA-MAA-5/5.5 MGs because there is substantial overlap of the main emission peak for Cy5 (667 nm) with the main absorption band of Cy5.5 (674 nm) (see Fig, S7, ESI †). We used the PL maxima for Cy5 in PEA-MAA-5 (667 nm) and Cy5.5 in PEA-MAA-5.5 (692 nm) (Fig. S5A, ESI †) for the donor intensity (I D ) and acceptor intensity (I A ), respectively, to analyze the PEA-MAA-5/5.5 MG spectra. Fig. 2A shows the effect of pH on the PL spectra. The wavelength of maximum intensity (l max ) blue-shis as the pH increased. Both the (I D /I A ) ratio and l max are very sensitive to pH-triggered swelling (see Fig. 2B). The I D /I A ratio increased from 1.4 to 2.4 with increasing pH. Simultaneously, l max decreased from 672 to 665 nm. Both the excitation and emission wavelengths for PEA-MAA-5/5.5 Multiple-run experiments for I D /I A , l max and d z of PEA-MAA-5/5.5 using pH cycling between 4.5 and 8.0 were conducted to investigate reversibility (Fig. 2C). The results show negligible dri and, hence, good reversibility for all the parameters. This is attributable to the high zeta potentials for PEA-MAA-5/5.5 (Fig. S2C, ESI †) which help prevent aggregation. Dispersion stability during storage is also important for applications. The I D /I A ratio (Fig. 2D), l max data (Fig. S9A, ESI †) and d z data (Fig. S9B, ESI †) for PEA-MAA-5/5.5 MGs were measured at pH values in the range of 4.5-9.8 for 22 days. These values changed by an average of less than 5% and so PEA-MAA-5/5/5 MG had good stability.
Comparing reporting from pH-responsive and temperatureresponsive NIR microgels We next compare the reporting properties of PEA-MAA-5/5.5 to an established temperature-responsive poly(Nisopropylacrylamide)-based NIR MG containing Cy5 and Cy5.5. 58 The latter is denoted as PNP-5/5.5 and was prepared by precipitation polymerization (Scheme S3, ESI †). The particles are spherical with an average TEM diameter of 40 nm (Fig. 3A). The PNP-5/5.5 MG particles contained $0.020 mol% Cy5 and 0.050 mol% Cy5.5 using the UV-visible spectrum (Fig. S10, ESI †) and the molar extinction coefficients for the NHSfunctionalized uorophores (Fig. S11, ESI †). The PL maximum for PNP-5/5.5 (Fig. 3B) is dominated by Cy5.5. The relatively high Cy5.5 concentration in the PNP-5/5.5 MGs may be due to greater electrostatic attraction during coupling to the -NH 3 + groups of the MGs for Cy5.5-NHS (charge ¼ À3) compared to Cy5-NHS (charge ¼ À1). The coupling reaction and Cy5.5-NHS and Cy5-NHS structures are shown in Scheme S3 (ESI †). Fig. 3B shows the PL spectra change considerably as the temperature approaches the volume phase transition temperature (VPTT). The latter is $32 C based on the temperaturedependence for d z (see Fig. 3C, top), which agrees with the literature. 72 The PNP-5/5.5 MG de-swelling is due to temperature-triggered disruption of the hydrogen bonding of water with the amide groups. The I D /I A ratio for PNP-5/5.5 decreases from 0.8 to 0.4 when the temperature is increased from 9 to 40 C (Fig. 3C, middle) due to increased NRET and mirrors the d z changes. These data conrm that PNP-5/5.5 is also able to report changes of d z in pure water. 58 The l max value can also be used to report temperature-triggered de-swelling as shown in the bottom graph of Fig. 3C. Based on these results the ability to use both I D /I A and l max to report particle swelling appears to be general for MGs containing Cy5 and Cy5.5. To test the ability of the MG probes to report swelling in the presence of electrolyte a series of PEA-MAA-5/5.5 and PNP-5/5.5 dispersions were prepared in pH 7.4 buffer (0.10 M) at 37 C in the presence of various added NaCl concentrations. Visual inspection ( Fig. 3D and S12, ESI †) showed that the PEA-MAA-5/ 5.5 dispersion did not have any aggregates in the presence of added NaCleven at concentrations as high as 0.80 M! In contrast, at all NaCl concentrations (and even in the absence of added NaCl) a sedimented layer of aggregates was present for PNP-5/5.5. The d z data (Fig. 3E) show that the PEA-MAA-5/5.5 particles slightly de-swelled with increasing NaCl concentration, which is due to electrostatic screening. 73 In contrast, the d z values for PNP-5/5.5 increased strongly with increasing NaCl concentration at 37 C due to aggregation.
The different colloidal stabilities of the PEA-MAA-5/5.5 and PNP-5/5.5 dispersions to electrolyte are due to differences in the dispersion stabilisation mechanisms. Zeta potential (z) data measured for PEA-MAA-5/5.5 and PNP-5/5.5 (Fig. S13, ESI †) show that the former has much larger z values in the presence of NaCl than the latter. This difference is due to the high content of COO À groups in PEA-MAA-5/5.5 at pH 7.4. Because the PNP-5/5.5 MGs collapsed at 37 C, which is above their VPTT of 32 C (above), they relied exclusively on electrostatic stabilization for colloidal stability. However, this was compromised in the presence of electrolyte concentrations greater than or equal to 0.10 M. In contrast the PEA-MAA-5/5.5 particles remained mostly swollen based on the d z data (Fig. 3E). Hence, PEA-MAA-5/5.5 dispersions are stabilized by electrostatic and steric interactions, i.e., they are electrosterically stabilized. 74 Such strong stabilization imparted superior colloidal stability under all conditions studied. Indeed, the PEA-MAA-5/5.5 probe has colloidal stability at electrolyte concentrations at least 5 times higher than physiological ionic strength. The superior stability of PEA-MAA-5/5 originates from the very high MAA content (41 mol%) which provides high charge densities and robust particle swelling in electrolyte solutions.
The PL spectra for the two probes ( Fig. S14A and B, ESI †), I D / I A (Fig. 3F) and l max values (Fig. S14C, ESI †) show major differences. The I D /I A and l max values for PEA-MAA-5/5.5 are close to their fully swollen values (from Fig. 2B) and decrease slightly with increasing NaCl concentration ( Fig. 3F and S14C (ESI †)). These decreases are due to electrolyte induced deswelling of the PEA-MAA-5/5.5 probe particles (Fig. 3E). In contrast all of the I D /I A (Fig. 3F) and l max values (Fig. S14C, ESI †) for PNP-5/5.5 correspond to the collapsed state (from Fig. 3C). Whilst the PNP-5/5.5 PL data have correctly reported the swelling state, the system is compromised in terms of further reporting because its swelling state cannot change at physiological temperature. It is also aggregated. Hence, the PEA-MAA-5/5.5 NIR probe MG has the potential advantages over PNP-5/5.5 of being colloidally stable and in a swollen state (and able to report swelling changes) under physiological conditions and in high salt concentration environments. PNP-5/5.5 is best suited to very low electrolyte concentrations (<0.10 M) where it remains colloidally stable.
The loading and position of attachment of the uorophores within these MG probes will affect NRET. If the loading of either uorophore is too low then the average separation between the donor and acceptor will be much greater than the Förster distance (6.9 nm (ref. 75)) and NRET will no longer be observed. We used reaction solutions containing both uorophores to ensure close attachment of each in the MG network. In this study, the mole ratios of Cy5 to Cy5.5 decreased from 5.0 for PEA-MAA-5/5.5 to 0.40 for PNP-5/5.5. The changes in I D /I A and l max were successfully used to report diameter changes for both MG probe systems. Hence, we conclude that NRET will allow dual mode PL reporting for similar MG-5/5.5 probes provided they are prepared using the methods described this work. Accordingly, the mole ratio of Cy5 to Cy5.5 within the MG probes should be in the range 0.40 to 5.0 and their total (summed) concentration should be between 0.012 and 0.070 mol% (Table S1, ESI †).
Imaging PEA-MAA-5/5.5 within stem cells and aer subcutaneous injection We investigated the ability to image the PEA-MAA-5/5.5 probes within adipose-derived stem cells. The pH values of 6.4 (Fig. 4A-C) and 7.4 (Fig. 4D-F) were investigated. The blue stain shows the nucleus, and the green stain highlights the actin, while the red color is from the PEA-MAA-5/5.5 MGs. These MG probes were able to cross the cell membrane 76,77 of stem cells aer only 4 h incubation. There was no need to use cationic 78 or lipophilic 79 transfection agents. This property is potentially useful because stem cells can be difficult to transfect. 80 The regions of locally high MG concentration in Fig. 4B, C, E and F and are indicated with blue arrows. Additional images recorded using 0, 5, 10, 20 and 40 mg mL À1 are shown for each color channel and also bright eld white light at pH 6.4 and 7.4, respectively, in Fig. S15 and S16 (ESI †). Cell imaging using white light showed normal cell growth morphology indicating that PEA-MAA-5/5.5 produced little toxicity inside the cells (see bright eld images in Fig. S15 and S16 (ESI †)). We measured the uorescence intensity of a series of MG dispersions with the same concentrations as used for stem cell uptake in vitro using PL spectroscopy (see Fig. S17, ESI †). The I D and I A values are linear with PEA-MAA-5/5.5 concentration implying facile tuning of the PL intensity (Fig. S17B, ESI †). The I D /I A ratios and l max values were not affected by probe concentration (Fig. S17C, ESI †), conrming good probe stability. In vitro cytotoxicity of the PEA-MAA-5/5.5 probes was measured via the Alarmar blue™ assay using adipose-derived stem cells. The cells were incubated with PEA-MAA-5/5.5 at concentrations ranging from 5 to 40 mg mL À1 for 1, 3, and 7 days. The cells proliferated in the rst three days, and then the cell viability stayed the same or increased for the next four days. The nal viability aer 7 days was over 95% (see Fig. 4G). There was not any signicant cytotoxicity of PEA-MAA-5/5.5 for the stem cells under the conditions employed.
We investigated the ability to use PEA-MAA-5/5.5 as an injectable MG for NIR imaging. Fig. 5A and B show, respectively, images for PEA-MAA-5/5.5 probe and the parent non-uorescent PEA-MAA-DVB MG control in syringe barrels illuminated by NIR and white light. The PEA-MAA-5/5.5 dispersion appears white when illuminated with NIR light (650 nm) and imaged with an NIR camera (Fig. 5A, le hand side). In contrast PEA-MAA-DVB does not emit NIR light and the dispersion is dark when illuminated with NIR light (Fig. 5B, LHS). Also, both PEA-MAA-5/5.5 and PEA-MAA-DVB do not emit NIR light when illuminated with white light (Fig. 5A and B, RHS). These experiments demonstrate that only PEA-MAA-5/5.5 can be imaged using NIR when illuminated with NIR light. We next investigated the imaging ability for subcutaneous injection of PEA-MAA-5/5.5 into chicken tissue (see Fig. 5C and D). The PEA-MAA-5/5.5 probe was injected at a depth of 5.0 mm within the chicken breast. (That latter tissue is dark under NIR irradiation.) NIR emission appeared (Fig. 5C) and gradually spread (Fig. 5D) as the injection proceeded. Hence, PEA-MAA-5/5.5 can be imaged by NIR light when subcutaneously injected in such tissue.
We compare the performance of PEA-MAA-5/5.5 in terms of particle swelling ratio (Q) and l ex with other reported stimuliresponsive ratiometric NRET-based uorescent nanoscale probes in Fig. 6. Q is the ratio of swollen to collapsed particle volume and was calculated using the published DLS data. (The data used are shown in Table S2, ESI †). The values for l ex were given in the publications. The probes were responsive to pH, metal cations, pH, sugar or light and are indicated. PEA-MAA-5/ 5.5 (red star) has by far the highest Q value of all of these systems. Furthermore, PEA-MAA-5/5.5 can be excited using a very high l ex . It also emits at a relatively high wavelength (l em ) compared to the other systems. The main competitor in terms of l ex is the temperature-responsive PNP-based system (L in Fig. 6). However, that Q value is more than a factor of 10 lower than that for PEA-MAA-5/5.5. That system was synthesized in this study (PNP-5/5.5) and was unstable to electrolyte (cf. PEA-MAA-5/5.5) as discussed above (Fig. 3). Unlike most systems in Fig. 6, PEA-MAA-5/5.5 is pH-responsive (Fig. 2B) and, due to electrosteric stabilization, is only weakly affected by ionic strength (Fig. 3E and F). Hence, PEA-MAA-5/5.5 has these unique advantages, as well as dual-mode PL reporting, offering potential as a pH-responsive NIR probe for reporting particle swelling.
We envisage potential use for PEA-MAA-5/5.5 as strain reporting probes for gel-based implants. 81 For example injectable gels have been used for increasing the height of degenerated intervertebral discs. 82 Inclusion of our NIR PEA-MAA-5/ 5.5 probes into such gels would, in principle, enable the d z value (and strain) of the MGs to be reported externally via NIR emission using I D /I A and/or l max . Such reporting should also be possible if the PEA-MAA-5/5.5 probes were internalized in cells within so tissue. To test the ability to estimate d z for our probes using I D /I A and/or l max we constructed calibration curves ( Fig. S18A and B, ESI †) using these parameters and the respective measured d z values from Fig. 1B and 2B. We then used the calibration curves to calculate d z values for PEA-MAA-5/ 5.5 from I D /I A and/or l max data in the later parts of this study, i.e., Fig. 2C, D, S9 (ESI †) and 3E. These calculated d z values are plotted against the measured values in Fig. S18C (ESI †). Good agreement between the calculated and measured values is evident using I D /I A , or l max as well as the average of the d z values calculated from I D /I A and l max . Whilst this analysis shows that all our results are self-consistent it also paves the way for future work using such NIR probes in cells and implants to monitor internal strain remotely.   6 Comparison of the properties of PEA-MAA-5/5.5 MGs with other ratiometric fluorescent stimuli-responsive swollen probes that use NRET. The particle volume swelling ratios (Q) and excitation wavelengths (l ex ) used appear in Table S2 (ESI †).

Conclusions
In this work we have studied new PEA-MAA-5/5.5 pH-responsive NIR MG probes. The PEA-MAA-5/5.5 MGs can report pHtriggered swelling using NIR uorescence reversibly in a range of conditions using both ratiometric intensity as well as l max . The PEA-MAA-5/5.5 probes have excellent colloidal and uorescent signal stability at physiological pH, ionic strength and temperature. These new probes, which benet from electrosteric stabilization, have better colloidal stability than the reference PNP-5/5.5 MGs. We have shown that l max provides a second PL addressable mode that is able to report swelling changes for both PEA-MAA-5/5.5 and PNP-5/5.5 MGs. This unexpected and general result is attributed to the incomplete separation of the PL maxima due to the acceptor and donor uorophores used. PEA-MAA-5/5.5 MGs are not cytotoxic to adipose stem cells and can potentially be used for NIR cell imaging. The results from subcutaneous injection study indicate that the PEA-MAA-5/5.5 also has potential for NIR imaging in tissue. Because the PL signal is very sensitive to MG swelling and NIR is deeply penetrating the PEA-MAA-5/5.5 MGs could be used to report pH or swelling changes inside cells or other complex environments such as tissue or synthetic gels. Furthermore, the total concentration of Cy5 and Cy5.5 used to obtain NIR reporting and imaging in this study was only $5 Â 10 À7 M to 5 Â 10 À6 M and PEA-MAA-5/5.5 had good stability (Fig. 2D). Hence, these new probes potentially provide a versatile and cost-effective alternative for NIR reporting and imaging. The PEA-MAA-5/5.5 MG system also has excellent potential to be used for environmental monitoring of aqueous solutions with high electrolyte concentration (e.g., waste water).

Conflicts of interest
There are no conicts of interest to declare.