Design of functional cationic microgels as conjugation scaffolds

Abstract

We present the development and detailed characterization of a range of amine functionalized microgels for utilization in conjugation reactions. Cationic N-isopropylmethacrylamide (NIPMAm) based microgels were synthesized through copolymerization with a primary amine containing monomer, N-(3-aminopropyl)methacrylamide hydrochloride (APMA). A range of synthesis conditions and monomer feed ratios generated microgels of diverse architectures, in different size ranges and with varying amounts of incorporated primary amines. The efficiency of amine incorporation was quantified using a fluorescence-based assay in order to determine the potential applicability of these particles for controlled bioconjugation reactions. The pH responsivity of all microgels was studied via dynamic light scattering and their height profile was investigated through atomic force microscopy following deposition on a functionalized flat substrate. Tunable resistive pulse sensing was employed to characterize microgels with respect to their number densities and molecular weights. These microgels were then conjugated to two dyes, malachite green and rose bengal with the purpose of investigating accessibility of primary amine groups for conjugation reactions. The microgel–dye constructs were then analyzed for dye content/microgel. Finally, the viability of NIH 3T3 fibroblasts incubated in the presence of varying concentrations of non-conjugated and dye conjugated microgels was studied. Confocal imaging revealed low cellular toxicity under conditions of incubation with low concentrations of microgel–dye conjugates, which is promising for eventual utilization of these constructs in bioimaging applications.

---

Introduction

The widespread applicability of colloidal bioconjugates has led to the development of a range of systems that can be utilized for controlled bioconjugation. Through different routes, such constructs have been developed for various purposes like the delivery of therapeutics, imaging, and solubilization of active agents.^1–6 A number of these constructs have the advantage of being in the nano- to micro-range, which makes them increasingly attractive for employment in conjunction with biological systems due to their similarity in dimensions to cells and cellular components, and thus increased control over processes happening at that scale.^7–10 Furthermore, the opportunity to tune their size to match specific applications provides additional advantages.¹¹ Microgels constitute a promising sub-class of such materials. Similar to their bulk hydrogels counterparts, microgels are water swollen, yet can be synthesized to have sizes ranging from tens of nanometers to a few microns.^12,13 They have been demonstrated to be beneficial in a wide array of biological applications, including tumor targeting for delivery of therapeutic agents,¹⁴ bioresponsive microlenses,¹⁵ artificial platelets¹⁶ and in non-fouling coatings.¹⁷ The versatility in their synthesis also makes it possible to make them more ‘biofriendly’ through the utilization of biocompatible constituent polymers¹⁸ and degradable crosslinkers,^19,20 further enhancing their suitability for use in biological systems.

A number of species such as single chain antibodies,¹⁶ peptides²¹ and biologically-relevant small molecules²² have been conjugated to microgels in order to make the resulting constructs useful for some of the applications mentioned above. This has primarily been made possible through the introduction of functional groups within the microgel matrices during synthesis, by copolymerization with molecules containing the desirable functionalities. Thus, microgels with versatile functionalities such as carboxylic acid groups, amine groups and others^23,24 have been synthesized in the past. Microgels being porous networks, the presence of these entities throughout their matrix provides the potential for multivalency to an even higher extent than that observed in comparable hard sphere microparticles, which are only capable of displaying functionalities on their surface. This high-density display of functional groups is useful for development of conjugates with very high local concentrations of the relevant biomolecules. One such functional group that can be incorporated in microgel networks is the amine group. The incorporation of amine groups brings with it the opportunity to perform numerous conjugation reactions to a range of functional groups on the biomolecules, such as NHS esters, isocyanates or isothiocyanates, acyl azides, carbodiimides, and aldehydes.²⁵ Thus, keeping the potential of these systems in mind, cationic polymeric particles containing amine groups have been synthesized and characterized by several groups in the past.^26–29 Both N-isopropylacrylamide (NIPAm) and NIPMAm have been used along with comonomers like APMA,³⁰ 2-aminoethylmethacrylate hydrochloride (AEMA),^31–33 and N-vinylformamide.^34,35 A detailed description of the various constructs that have been synthesized from several different constituent monomers has been provided by Ramos et al.³⁶ In spite of their manifold merits though, the extensive utilization of amine functionalized microgels has been hindered due to the inherent limitations of their syntheses as noted by Thaiboonrod et al.³⁵ and Farley et al.³⁷ The main issue relates to the stunted incorporation of amine containing comonomers during synthesis and therefore, an inability to generate stable, monodisperse solutions of large colloids with a high density of amine groups. Microgels with diameters of ∼1 μm particularly, have remarkable potential in being useful for the visualization and manipulation of biological processes occurring at this length scale. However, a simple, reproducible, one step reaction for generation of colloidally stable microgels in the micrometer range, with a high density of amine groups has remained largely elusive. This has necessitated the employment of multi-step, complicated pathways in order to achieve the desirable end products.^37,38

In the work presented here, a series of primary amine functionalized microgels were developed in different size ranges and with varying amounts of amine groups. Having been synthesized primarily from the monomer NIPMAm leading to the formation of the thermoresponsive polymer pNIPMAm, the loosely crosslinked microgels are swollen below 37 °C, which makes them viable candidates for applications requiring colloidally stable particles at the physiological temperature. We demonstrate that simple one step syntheses can be utilized to generate monodisperse solutions of cationic microgels in a reproducible manner, with synthesis conditions being altered to optimize formation of the end product. For comparison, a two-step approach for the synthesis of core/shell microgels is also investigated. All microgels were analyzed in detail for their primary amine content. This analysis is crucial for any system that has the potential to be used for well-controlled bioconjugations. On the basis of these studies, a model microgel system was used for further characterization with respect to molecular weight and number density. Two different dyes, malachite green and rose bengal were then conjugated to this microgel system, in order to demonstrate availability of the primary amine groups for conjugation reactions. We found that the local concentrations of these dyes were nearly three orders of magnitude above their overall concentration in solution, demonstrating the merit of using microgels as the conjugation scaffolds that enable the local concentration of active agents. The viability of NIH 3T3 fibroblasts in the presence of these microgel–dye constructs was also studied. This investigation was carried out particularly to address the toxicity of polycations, which has previously been observed to occur due to their interactions with the cell membrane.²³ Through qualitative analysis via imaging, the microgel–dye constructs were found to be less toxic to cells as compared to the primary amine functionalized microgels containing no dye, presumably due to consumption of amine groups during conjugation, which is a promising outcome for their utilization in conjunction with biological systems.

Experimental

Materials

All materials were purchased from Sigma-Aldrich unless specified otherwise. The primary monomer N-isopropylmethacrylamide (NIPMAm) was purified via recrystallization from n-hexane (J.T. Baker). The comonomer N-(3-aminopropyl)methacrylamide hydrochloride (APMA, Polysciences, Inc.), crosslinker N,N′-methylenebis(acrylamide) (BIS), NaCl, cationic initiator 2,2′-azobis(2-methylpropionamidine)dihydrochloride (V50), anionic initiator ammonium persulfate (APS), buffer preparation materials sodium dihydrogen phosphate, N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid) (HEPES), 3-(cyclohexylamino)-1-propanesulfonic acid (CAPS), 2-(cyclohexylamino)ethanesulfonic acid (CHES) and 4-morpholineethanesulfonic acid (MES), dyes malachite green isothiocyanate (MGITC, Life Technologies) and rose bengal (RB), 3-aminopropyltrimethoxysilane (APTMS) and poly(4-styrenesulfonic acid) sodium salt (PSS) for substrate functionalization for AFM imaging, coupling reagents N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC) and N-hydroxysulfosuccinimide sodium salt (NHSS), ethanol, isopropanol, acetone, calcium nitrate, 1× PBS (phosphate buffered saline, Corning Cellgro), Dulbecco's Modified Eagle Medium (DMEM, Corning Cellgro), methanol, dimethyl sulfoxide (DMSO) and dyes from the live/dead cytotoxicity kit (Life Technologies) calcein AM and ethidium homodimer-1 (EthD-1) were all used as received. The solutions were prepared using distilled water, deionized to a resistance of 18 MΩ (Barnstead E-Pure system). Solutions were filtered through a 0.2 μm Acrodisc syringe filter before use.

Microgel synthesis

Microgels were synthesized by free radical precipitation polymerization. In a typical synthesis, NIPMAm, BIS and APMA in the relevant molar compositions (refer to Table 1 for monomer feed ratios) were dissolved in 49 mL distilled, deionized water to a total monomer concentration of 140 mM (except for μgel11 and μgel12) along with sodium chloride, where pertinent, at the indicated concentration. The resulting solution was filtered through a 0.8 μm Acrodisc syringe filter and introduced into a 100 mL three necked, round bottom flask along with a magnetic stirrer. The flask was fitted with a thermometer, condenser and N₂ inlet and introduced into an oil bath, which was heated at 100 °C h⁻¹. Stirring was kept constant at 400 rpm and the solution was purged with N₂. Once the temperature was stable at 70 °C, 1 mL of the initiator (APS or V50) was added after filtration through a 0.2 μm Acrodisc syringe filter. The reaction was allowed to proceed for 4 h under a N₂ blanket, after which it was cooled down to room temperature. The solution was then filtered through glass wool.

Table 1 (a) Reaction conditions and R_H values of microgels synthesized at constant temperature. (b) Reaction conditions and R_H values of microgels synthesized with a temperature ramp. (c) Reaction conditions for microgel cores (μgel10) and shells (μgel11, μgel12) synthesized on μgel10 cores^a

(a)
Microgel identity	[APMA] (mol%)	[NIPMAm] (mol%)	[BIS] (mol%)	[V50] (mM)	[APS] (mM)	[NaCl] (mM)	R H (nm) at pH 7.4	R H (nm) at pH 11
a (a) [Total monomer] = 140 mM, temperature = 70 °C. (b) [Total monomer] = 140 mM, reaction initiation temperature: 50 °C, reaction final temperature: 70 °C. (c) [Total monomer] = 50 mM, temperature = 65 °C, *[total monomer] = 140 mM.
μgel1	1	97	2	0.50	—	25	563 ± 33	559 ± 40
μgel2	5	97	2	0.50	—	50	436 ± 23	419 ± 21
μgel3	7	91	2	0.25	—	85	433 ± 17	380 ± 13
μgel4	7	91	2	—	4	—	425 ± 16	397 ± 17
μgel5	9	89	2	0.25	—	—	94 ± 2	86 ± 3
μgel6	9	89	2	0.25	—	100	311 ± 12	296 ± 9
μgel7	9	89	2	—	3	—	261 ± 8	244 ± 7

---

(b)
Microgel identity	[APMA] (mol%)	[NIPMAm] (mol%)	[BIS] (mol%)	[V50] (mM)	[APS] (mM)	R H (nm) at pH 7.4	R H (nm) at pH 11
μgel8	9	89	2	0.25	—	64 ± 2	58 ± 1
μgel9	9	89	2	—	3	184 ± 5	160 ± 4

---

(c)
Microgel identity	[APMA] (mol%)	[NIPMAm] (mol%)	[BIS] (mol%)	[V50] (mM)
μgel10*	0	98	2	1.0
μgel11	1	97	2	0.5
μgel12	5	93	2	0.5

---

For the syntheses performed on a temperature ramp, a similar procedure was utilized. The monomer solution was heated until it reached a stable starting temperature of 50 °C. Polymerization was initiated at this temperature with the addition of 1 mL of a filtered V50 or APS solution and following initiation, the reaction solution was heated to a final temperature of 70 °C using a ramp rate of 30 °C h⁻¹.

Core/shell synthesis details have been previously reported.^21,40 In a typical synthesis, 39 mL of a filtered monomer solution consisting of NIPMAm, BIS and APMA was introduced into a reaction flask along with 10 mL of the purified core (μgel10) solution. This reaction solution was heated to a stable temperature of 65 °C while being continuously stirred and purged with N₂. Polymerization was initiated at this temperature with the addition of 1 mL of filtered V50 solution.

All microgels were purified by pelleting via ultracentrifugation (Beckman Coulter Optima MAX-XP) at 120 700–214 600 × g for 30–60 min, depending on the microgel type. This was followed by removal of the supernatant and resuspension in DI water. This process was repeated four times, after which the microgels were lyophilized prior to characterization.

Dynamic light scattering

Hydrodynamic radii (R_H) of the microgels were determined via dynamic light scattering (DLS, DynaPro, Protein Solutions) measurements in pH 7.4 HEPES (6 mM ionic strength) and pH 11 CAPS (6 mM ionic strength) buffers in order to observe pH responsivity. Microgel samples in the respective buffers were thermally equilibrated for 20 min, following which scattering intensity fluctuations based on Brownian motion of the microgels were detected at a scattering angle of 90°. Twenty-five (25) acquisitions of 20 seconds each were obtained and used to generate intensity time correlation functions. Cumulants analysis was then used to attain diffusion coefficients and the Stokes–Einstein equation was utilized to calculate the R_H values. The procedure was repeated 3 more times on the same sample in order to generate a total of 100 R_H values for each microgel sample and averages of these R_H values along with standard deviations are presented.

Atomic force microscopy

Atomic Force Microscopy (AFM) was performed using an MFP-3D AFM (Asylum Research). Images were acquired in air and under ambient conditions in the tapping mode, using pyramidal cantilevers (Nanoworld, Force Constant 42 N m⁻¹) made of heavily doped silicon. Image processing was performed using software written in an IgorPro environment (Wavemetrics, Inc.).

Samples were prepared on 22 mm × 22 mm glass coverslips (VWR). The coverslips were first cleaned by sequential sonication in Alconox solution, DI water, acetone, 95% ethanol, and isopropanol for 20 min each. They were then functionalized in a 1% (v/v) APTMS/absolute ethanol solution on a shaker table for 2 h, and subsequently modified with a 0.1 mg mL⁻¹ PSS solution in phosphate buffer (pH 7, 50 mM ionic strength) for 30 min in order to render the glass surface anionic. The coverslips were then washed with DI water. Microgel solutions prepared in the same phosphate buffer described above were then used for submonolayer deposition onto the glass substrates by centrifugation at 2250 × g for 15–20 min (depending on the microgel type) at 25 °C using a plate rotor. The coverslips were rinsed well with DI water and dried with nitrogen before imaging.

CBQCA assay for quantitation of amines

The efficiency of APMA incorporation during polymerization was determined by quantifying the accessible amine groups in the microgels using a fluorescence-based assay (Life Technologies), based on the reaction of ATTO-TAG™ CBQCA with primary amines. A reference standard curve was generated using monomeric APMA solutions (sample volume = 135 μL) prepared in sodium borate buffer (pH 9.3, 100 mM ionic strength). The reagent solutions of KCN (20 mM) and ATTO-TAG™ CBQCA (working solution of 5 mM concentration) were also prepared in the same buffer and following addition of these reagents, all sample solutions (total volume = 150 μL) were incubated at room temperature with shaking for 60–90 min and kept protected from light. Microgel solutions prepared in borate buffer were treated in an identical manner as the standards. The fluorescence of all solutions was then measured using a plate reader (Infinite® 200 PRO NanoQuant™, Tecan Group LTD., San Jose, CA) at excitation and emission wavelengths of 465 nm and 550 nm, respectively. After correcting these values for background from the buffer, the percent efficiencies of APMA incorporation in the microgels were determined.

Tunable resistive pulse sensing

The qNano (IZON Science, Oxford, UK) was utilized for performing resistive pulse sensing via examination of translocation events of particles through a tunable polyurethane nanopore (IZON Science Ltd. NZ) NP400, which is suitable for analysis of particles with diameters between 200 and 800 nm. HEPES buffer (pH 7, 150 mM ionic strength) was first introduced in the cells on both sides of the nanopore, in order to generate a stable baseline current. Following this, the buffer in the cell above the nanopore was replaced with 40 μL of the relevant sample solution. All microgel solutions were used at a concentration of 0.2 mg mL⁻¹, and the measurements were standardized against solutions of the reference, carboxylated polystyrene beads at a known number density of 7.5 × 10⁻⁸ particles per mL and a mean diameter of 350 nm, by performing sample and standard measurements under identical conditions of voltage, nanopore stretch and pressure. The proprietary software provided by Izon was utilized for sample and data analysis. Three sets of measurements generated a total of ∼215 values for R_qNano for each sample. The R_qNano distribution from these values is presented. The three sets also provided three values of microgel number density per sample. The averages and standard deviations of these values are presented.

Conjugation of dyes to microgels

Malachite green. The lyophilized microgels were weighed and resuspended overnight in CHES buffer (pH 9.8, 20 mM ionic strength) to a concentration of ∼3–5 mg mL⁻¹. Following this, MGITC dissolved in DMSO was added to the microgel solution in a 4.5× molar excess based on the amount of amine groups in the microgels. The conjugation reaction was allowed to proceed overnight at 4 °C on a shaker after which the solution was dialyzed against DI water for 2–4 weeks at 4 °C in order to get rid of excess dye and buffer.

Rose bengal. For conjugation to RB, the lyophilized microgels were weighed and resuspended in phosphate buffer (pH 7, 50 mM ionic strength) to a concentration of ∼2–4 mg mL⁻¹. The dye was then conjugated to the microgels by dissolving it in the same buffer and adding it to the microgel solution in a 4× molar excess (for μgel1–RB, μgel2–RB and μgel6–RB) or a 10.5× molar excess (for μgel2b–RB) based on the concentration of amines. EDC and NHSS were also used in the same molar excess as RB, for the respective conjugations. The reaction was allowed to proceed overnight at room temperature on a shaker, after which the solution was dialyzed against CAPS buffer (pH 11, 500 mM ionic strength) for 3–5 weeks to remove excess dye and then against DI water for 2–3 weeks to remove buffer components.

The dialyzed dye conjugated microgel solutions were left on a shaker overnight in order to ensure homogeneity after which small portions of the solutions were lyophilized to determine the concentration of the ‘stock’ solutions.

Absorbance measurements

The dye conjugated microgel solutions of known concentrations were analyzed for dye content via absorbance measurements done in a plate reader (Infinite® 200 PRO NanoQuant™, Tecan Group LTD., San Jose, CA). Standard curves were generated by using dye solutions of MGITC (stock solution in DMSO, remaining dilutions in DI water) and RB (in water). Upon determining the λ_max values (625 nm for MGITC and 540 nm for RB), absorbance measurements were performed at the corresponding wavelengths. Absorbance values of the dye conjugated microgels were then determined at these wavelengths, after ensuring that scattering by microgels was not causing perturbation of the obtained results and the moles of dye conjugated and conjugation efficiencies were calculated.

Cell culture

NIH 3T3 cells were grown in DMEM (supplemented with 10% Fetal Bovine Serum and 1% penicillin streptomycin) in T75 flasks until they were 80% confluent. The medium was then removed from the flask, after which the cells were split through treatment with 5 mL trypsin/EDTA (0.25%) followed by a 5 min incubation at 37 °C. The cell count was then determined using a hemocytometer (Reichert Bright Line), following which the cells were pelleted down by centrifugation at 1200 × g for 5 min. The supernatant was aspirated out and the cells were resuspended in medium to a concentration of 1 million cells per mL. The cells were further diluted using medium and plated to a concentration of 10 000 cells per cm² in a 96 well plate and allowed to spread on the surface for 3–4 h. Following this, the medium was removed via aspiration and the microgel solutions diluted in medium were added to the corresponding wells. A change of medium was made for the control wells to maintain uniformity in procedure. The well plate was then incubated for 24 h at 37 °C and the live/dead assay was performed on the cells.

Viability studies

The live/dead assay protocol (Life Technologies) was performed in a sterile tissue culture hood. The medium was aspirated from the sample wells and controls and the wells were washed gently with 200 μL 1× PBS. The cells in the dead control wells were then killed by incubating them with 100 μL of 70% methanol for 30 min. 100 μL 1× PBS was added to the remaining wells for the same period of time. The methanol and 1× PBS were then removed gently with a pipette, following which single or both dye solutions (Calcein AM and EthD-1) were added to the corresponding wells with controls and samples. The 96 well plate was subsequently incubated at room temperature for 45 min and protected from light. The solutions from all wells were then gently removed using a pipette and all wells were washed gently with 200 μL 1× PBS, following which a 4% v/v formaldehyde/1× PBS solution was introduced into the wells for the purpose of fixing the cells. After an incubation of 15 min, the formaldehyde was pipetted out and all wells were gently washed twice with 200 μL 1× PBS. 100 μL 1× PBS was added to all the wells before conducting fluorescence measurements. The excitation and emission wavelengths of Calcein, 528 nm and 645 nm respectively, were utilized for the measurements and the percent viability values of the cells were calculated.

Confocal microscopy

The samples were prepared in a sterile tissue culture hood. Round coverslips (12 mm diameter, 0.17–0.25 mm thickness, Electron Microscopy Sciences) were immersed in 70% ethanol for 30 min, after which they were dried and introduced into a 24 well plate. They were incubated in DMEM for 1 h at 37 °C in order to facilitate cell attachment. The cells were then plated at a concentration of 10 000 cells per cm² in each well and were allowed to spread for 3–4 h. Following this, a similar procedure as the one for the 96 well plate was used for incubation with microgels and staining of the cells, with volume adjustments. The coverslips were then air dried and preserved in ProLong® Gold Antifade reagent (Life Technologies). The samples were sealed with nail polish and kept protected from light, at 4 °C until they were imaged.

Images were obtained via Confocal Laser Scanning Microscopy (CLSM, Zeiss LSM 700) by setting the pinhole diameter to 1 airy unit. Image processing was performed using the Zen 2012 (Carl Zeiss, SP1) software.

Results and discussion

Development and characterization of amine functionalized microgels

Cationic microgels synthesized with varying monomer molar feed ratios of NIPMAm : APMA and under numerous experimental conditions (Tables 1a–c) were purified and analyzed for their hydrodynamic radii (R_H) using DLS.

The pH responsivity observed in low ionic strength pH 7.4 HEPES and pH 11 CAPS buffers (pK_a of APMA ≈ 10) demonstrated successful incorporation of the cationic comonomer APMA. Buffers of low, 6 mM ionic strength were used in order to reduce the effects of charge screening, thus facilitating observation of pH responsivity to a greater extent. The motivation behind the development of a large range of microgels was two-fold. Firstly, modifications in synthesis conditions generated microgels of different dimensions. This versatility with respect to size is a desirable quality, particularly for particles that have the potential to be used in different environments in biological systems. Secondly, the variation in APMA incorporation within the particles during synthesis provides tunability with regards to degree of functionality and thus further bioconjugation. Since it has been shown that cationic polymers can potentially be toxic to cells,³⁹ a definitive control on the density of amine groups present in the microgels can also be leveraged for utilization of these constructs in biological environments. The interplay between the aforementioned microgel features was quantified by measuring the R_H and amine content of the microgels. The ‘knobs’ that were tuned to achieve differences in microgel characteristics included monomer ratios, concentration of NaCl, concentration and identity of the initiator, initiation temperature and microgel architecture.

Fig. 1 presents images of microgels obtained using atomic force microscopy. Owing to their deformability, microgels deposited on functionalized glass substrates are often observed to be flattened (∼hemispherical), as illustrated by these images. Within the group of microgels synthesized using V50 as an initiator, an increase in NaCl concentration during synthesis led to an increase in particle diameter. This is indicated by the R_H values of μgel5 and μgel6. This trend can be attributed to the screening of charges caused by NaCl, leading to a decrease in coulombic repulsion between APMA-containing polymer chains. The electrostatic repulsion when uninhibited, prevents chain collapse due to positioning of like positive charges along the polymer backbone. This leads to free polymer chains in solution and thus smaller microgel size resulting from less polymer incorporation in the particles.³¹ Additionally, the charged monomer in solution acts as a polymerizable surfactant and displays behavior similar to that of an ionic surfactant, thus further limiting the growth of particle size.^41,42 The R_H values of microgels μgel1, μgel2, μgel3 and μgel6 are demonstrative of this effect of NaCl concentration on particle size. However, above a threshold concentration, aspherical deformation of microgels was observed (Fig. S1A, ESI†). This threshold varied with monomer feed concentrations and its occurrence can likely be attributed to the formation of block copolymers at the periphery of the microgels in a charge screening environment, which leads to the development of particles that are not perfectly spherical.^30,43 These artifacts in structure, while well demonstrated in the AFM images, were also revealed by an inconsistency in R_H values obtained using DLS. They were found to manifest in the form of high SOS errors (>200), which commonly indicate polydispersity or, in this case, irregularity in the shapes of the particles. This is because the Stokes–Einstein equation relates the diffusion coefficient D of a spherical ‘Stokes’ particle to its R_H. Upon increasing the NaCl concentration further, macroscopic aggregation of the polymer was observed, demonstrating disappearance of electrostatic stabilization in the system due to increased charge screening.

Fig. 1 AFM height traces of μgel1 to μgel12 (A to L respectively) deposited on pre-functionalized glass by centrifugal deposition at 2250 × g for 15–20 min at 25 °C, depending on the microgel type. Images qualitatively demonstrated monodispersity of individual microgel suspensions and size differences in particles depending on changes in synthesis conditions.

For syntheses initiated using the anionic initiator APS, microgel size increased with increasing concentration of the initiator. At first glance this is counterintuitive, as ordinarily an increase in initiator concentration would be expected to lead to a higher rate of nucleation than propagation, which leads to a large number of smaller particles. This expected trend is observed in the case of microgels synthesized using the cationic initiator V50 (Table S1, ESI†). However, in the case of the ones synthesized using APS, there are two possible explanations for the observation of a reverse trend. Firstly, proton abstraction from the carbon adjacent to the primary amine group in APMA, causes the generation of secondary reactive sites. This occurs due to the nucleophilicity of the radical center oxygen in the persulfate radical. Self crosslinking and larger particle size result from this phenomenon due to an increase in the efficiency of polymer incorporation. Hu et al. observed larger yields in cases of reactions initiated using APS when compared to those initiated using V50.³⁰ These observations were also demonstrative of better polymer incorporation in the microgels. Secondly, the process of stabilization of microgels by positive charges contributed by APMA is hindered by the presence of negatively charged sulfate groups from APS. This necessitates the accumulation of a larger number of primary particles to generate a microgel stabilizing net positive charge, causing build-up of particle size. This reasoning is supported by the observations of Still et al. concerning positive zeta potential values obtained in syntheses of p(NIPAm-co-AEMA) and negative values obtained in a similar synthesis conducted in the absence of AEMA.³² Above a threshold concentration of APS, this ‘neutralization’ of charges leads to an effect similar to that observed with higher NaCl concentrations in syntheses described previously, i.e. the formation of misshapen microgels (Fig. S1B, ESI†). Microgels were also synthesized using a different crosslinker, poly(ethylene glycol) (200) diacrylate (Polysciences, Inc.) or PEG-DA, and were designated μgelPEG. This synthesis was performed with molar feed ratios of 93 : 2 : 5 corresponding to NIPMAm : PEG-DA : APMA and in the presence of 50 mM NaCl. These microgels were also analyzed for their R_Hvia DLS and their hemispherical profile via AFM (Fig. S2, ESI†), in order to demonstrate the versatility of the approach.

Microgels synthesized by means of a temperature ramp commencing after initiation (μgel8 and μgel9) were found to be smaller than their counterparts obtained via traditional syntheses (μgel5 and μgel7 respectively). pNIPAm and pNIPMAm based microgels have been synthesized in a similar manner before,^13,44 with initiation at a temperature lower than that used for conventional syntheses, followed by a gradual, steady increase in solution temperature as the reaction progresses. In these cases, large microgels were obtained due to a bias created towards particle growth through chain propagation vs. the generation of nucleation sites. A continuous increase in temperature during the synthesis ensures that the propagation rate is not sacrificed due to consumption of the monomer, by increasing the number of radicals generated. It is noteworthy that these ramp syntheses performed in the past have involved the use of an anionic comonomer. The observation of a somewhat reverse trend in the case of p(NIPMAm-co-APMA) microgels synthesized on a temperature ramp (50–70 °C) via initiation with the cationic initiator V50 (μgel5 and μgel8) or anionic initiator APS (μgel7 and μgel9) may be attributed to the heavy influence of cationic comonomers on the formation of water soluble polymers during the synthesis. A similar phenomenon was observed by Mai-ngam et al. when analyzing the phase separation of p(NIPAm-co-APMA) polymers.⁴³ In their experiment, conformational collapse was not observed at temperatures up to 45 °C in pure water, in spite of the LCST of pNIPAm being 31 °C. This was due to strong electrostatic repulsion between the positively charged amine groups along the polymer chain. The addition of salt however, facilitated phase separation through charge screening. The same argument may be extended to microgel synthesis of μgel8 and μgel9, since these were conducted in the absence of any salt. At the temperature of initiation, coulombic repulsion between the positively charged amine groups along the polymer chain dominates over the collapse of polymer chains. Thus, a majority of the monomer is utilized in the formation of polymer chains in solution. This can be observed as a temperature driven amplified effect of the same phenomenon discussed earlier, that leads to the generation of small microgels when a high feed ratio of amine containing comonomer is used. Despite the fact that the syntheses discussed here were initiated at a temperature higher than the ones conducted by Mai-ngam et al., the same reasoning may be considered applicable because of the higher LCST of pNIPMAm (≈44 °C). As the temperature gradually increases, the coil to globule transition of the polymer begins to occur. However, by the time this stage of the synthesis is reached, another factor, namely the incorporation of the crosslinker BIS into the particles comes into play. It has been postulated by Meunier et al. that the depletion of free BIS through incorporation into the polymer, acts as a deterrent towards any further capture of free polymers by the microgels from solution.³¹ Thus particle growth is further impeded. All these factors contribute to the formation of microgels smaller than those generated through a conventional synthesis, under similar conditions. It is worthwhile to note that the reactivity ratios of NIPAm and BIS are different from those of NIPMAm and BIS. However, the results obtained here indicate that these differences may not be a major contributing factor due to the overpowering effect of the comonomer APMA on the syntheses.

Following analysis of the physical characteristics of all microgels, the ATTO-TAG™ CBQCA reagent was used to determine the content of accessible primary amines on the microgels via a fluorescence-based assay.⁴⁵ This reagent is non-fluorescent by itself in aqueous solutions. However, it reacts with primary amines in the presence of cyanide to yield a fluorescent derivative, and is capable of detecting proteins in nanogram quantities though primary amine derivatization. Historically, the incorporation of primary amine functional groups during microgel synthesis has not been very efficient, as described above. Moreover, the low magnitude of pH responsivity, when observed for the microgel solutions (Table 1) indicated that the degree of APMA incorporation during the synthesis of almost all microgels was fairly low. Hence, the CBQCA assay was the analysis method of choice, due to its low detection limit, which should permit the quantitative analysis of amines even if incorporation degrees were low. A standard curve was generated using APMA solutions in borate buffer and used for analysis.

Taking into consideration the monomer feed ratios during the syntheses, the efficiencies of incorporation of APMA during polymerization were calculated and are presented in Fig. 2. Analysis on μgel10 solutions demonstrated fluorescence close to that of the background, thus indicating the absence of any interference from non-primary amine groups, at the concentrations under consideration. A general upward trend in the moles of APMA incorporated within the microgels is observed, based on the monomer feed ratio of APMA, in microgels μgel1 (1 mol% feed APMA), μgel2 (5 mol%) and microgels μgel3 and μgel4 (7 mol%). For μgel5, μgel6, μgel7, μgel8 and μgel9, the absence of a clear trend can be attributed to the stark difference in R_H values of these microgels compared to the others, and thus dissimilar number densities. With a low feed ratio of APMA in μgel1, a low amount of amine groups are successfully assimilated within the microgel network, but a high incorporation efficiency of 33% is observed. This result is in agreement with the argument put forth by Meunier et al., described earlier. At lower APMA feed ratios, the electrostatic repulsion between positively charged amine groups randomly positioned along the forming polymer chains is minimal and polymer collapse is not impeded. This results in higher APMA incorporation efficiency and larger microgels. Therefore, in spite of a lower concentration of NaCl being used in the synthesis, microgels μgel1 have larger R_H values (in comparison to μgel2, μgel3 and μgel6). As the APMA feed ratio is increased, the percent efficiency of amine incorporation within the microgels remains within a range from 16–23%, despite differences in the microgel sizes, hinting at the fact that in the synthesis, a compromise of microgel size is favored over that of the APMA incorporation efficiency.

Fig. 2 Incorporation of APMA in moles, in 135 μL of 1 mg mL⁻¹ solutions of the respective microgels (grey bars) and percent efficiency of incorporation of APMA (red circles) in conventional microgel populations with varying APMA feed ratios (A) and core/shell microgel architectures (B), analyzed via CBQCA assay (values presented as averages of three trials, error bars represent standard deviations of these three values).

In order to analyze variants of microgel architecture, two core/shell structures were synthesized using a ‘seed and feed’ method. Microgels with such architectures have been demonstrated to be useful for applications requiring dual functionality, such as loading of therapeutics in the core, and cell specific targeting through conjugation of relevant molecules to the shell,¹⁴ and for ratiometric imaging.⁴⁶ The cores used for both syntheses were the same and were synthesized with no APMA (μgel10). They were found to have an R_H value of 459 ± 24 nm in pH 7.4 HEPES buffer of 6 mM ionic strength. Core/shell architectures μgel11 and μgel12 were generated by depositing two different shells on μgel10 with APMA feed ratios of 1 mol% and 5 mol%, respectively.

The AFM images of the core/shell structures revealed well-defined, hemispherically-deformed particles (Fig. 1K and L). Reliable R_H values could not be obtained for these microgels, which may be attributed to their larger size leading to a combination of non-random movement, i.e. sedimentation and number fluctuations due to low particle concentration in the DLS measurement volume. Analysis of the primary amine content of these microgels via the CBQCA assay revealed the moles of APMA incorporated in μgel11 and μgel12 to be approximately an order of magnitude below their conventional microgel counterparts, μgel1 and μgel2 (Fig. 2A and B). The inclusion of APMA only in microgel shells is a possible explanation for this observation. However, the trend observed was similar, i.e. a lower feed ratio of APMA resulted in a lower number of moles being incorporated. The percent efficiency of APMA incorporation was found to be 8 ± 0.31% and 4 ± 0.08% in μgel11 and μgel12 respectively. This low efficiency may be due to a reasoning similar to that proposed by Meunier et al. as explained above, relating to the formation of polymer chains in solution. Additionally, the formation of secondary ‘core’ microgels in solution (detected via DLS on supernatants following ultracentrifugation) may be responsible for a further decrease in incorporation efficiency of APMA within the shells formed on pre-existing cores.

Based on the microgel sizes, profiles and primary amine contents, μgel2 was selected as a model system for further characterization and conjugations. The reasons behind this selection were multi-fold. Firstly, the content of primary amine groups in these microgels was intermediate in the range of microgels studied. This content measured via the CBQCA assay was corroborated with the help of a second technique, pH titrations (Fig. S3, ESI†). The comparative results from both these analysis methods are presented in Table 2. The agreement in these values demonstrated the quantitative accuracy of the CBQCA assay. Finally, μgel2 has a convenient size for analysis and imaging, is synthetically reproducible, and possesses good amine incorporation.

Table 2 Comparative values of percentage efficiency of primary amine incorporation during microgel synthesis using two different techniques (data from μgel2)^a

Method used for primary amine quantitation	Calculated % efficiency of primary amine incorporation
a Values presented as averages of three trials.
CBQCA assay	22 ± 2
pH titrations	20 ± 2

---

Synthesis and detailed characterization of μgel2a and μgel2b

Two new batches of the model system μgel2 were synthesized, μgel2a and μgel2b (feed ratios of NIPMAm : BIS : APMA::93 : 2 : 5). Their R_H values in pH 7.4 HEPES and pH 11 CAPS buffers were found to be 413 ± 29 nm and 377 ± 24 nm for μgel2a and 409 ± 19 nm and 387 ± 15 nm for μgel2b respectively. The incorporation of APMA during synthesis is qualitatively evident from these values. The microgels were also observed via AFM imaging following centrifugal deposition on pre-functionalized glass. They were found to be monodisperse with a hemispherical profile (Fig. 3A and C).

Fig. 3 AFM height traces of μgel2a (A), μgel2a–MGITC (B), μgel2b (C) and μgel2b–RB (D) deposited on pre-functionalized glass by centrifugal deposition at 2250 × g for 15 min at 25 °C. Images qualitatively demonstrated monodispersity of individual microgel suspensions. Solutions of μgel2a–MGITC and μgel2b–RB and dried microgel samples are shown adjacent to the corresponding images.

To determine the physical radii of individual microgels and number densities of microgel solutions, dispersions of both μgel2a and μgel2b were subjected to tunable resistive pulse sensing in a high ionic strength buffer. This technique based on the Coulter principle detects passage of single particles through a nanopore. A decrease in the intensity of current whenever a particle passes through the pore, detected as a blockade event, provides information about the population density of the dispersion and properties of individual particles. The distributions of microgel radii as obtained by this method are presented in Fig. 4. All measurements were calibrated against standard solutions of polystyrene beads of relevant dimensions and values were obtained as averages from three sets of measurements on the qNano. When compared to the hydrodynamic diameters (D_H) of the microgels in similar conditions of ionic strength, the diameters obtained via the qNano (D_qNano) were seen to be smaller by ∼35% for μgel2a and ∼38% for μgel2b (Table 3). This difference in measured microgel dimensions can be attributed to the fact that the R_H measured by DLS takes into account the dangling polymer chains around the microgels and hence tends to be an overestimation of the microgel radius when compared to the physical size. Furthermore, since resistive pulse analysis involves standardization against hard spheres, it does not take into account the volume of buffer within the porous microgels, which is responsible for making them conductive. Thus, when examined in parallel to the resistive pulse due to a hard sphere, that observed for a comparable microgel is of lower magnitude due to its buffer content and therefore, its volume is underestimated, leading to a further discrepancy in size. Based on the number density obtained via the qNano and the weight percent of the starting microgel solutions, the molecular weights of μgel2a and μgel2b were found to be (8.3 ± 3.0) × 10¹⁰ g mol⁻¹ and (5.4 ± 2.2) × 10¹⁰ g mol⁻¹ respectively (Table 3).

Fig. 4 Particle size distribution of dispersions of μgel2a (A) and μgel2b (B) as determined through tunable resistive pulse sensing (each graph represents a distribution of ∼215 R_qNano values obtained through a total of three trials) and a schematic demonstrating working of the qNano (C).

Table 3 Population characteristics of μgel2a and μgel2b with comparative microgel sizes obtained via tunable resistive pulse sensing using the qNano and hydrodynamic diameter obtained using DLS^a

Microgel identity	Number density (particles per mL)	Molecular weight (g mol^−1)	D qNano (nm)	D H (nm)
a Values presented as averages of three sets of trials. All measurements performed in HEPES buffer (pH 7, 150 mM ionic strength).
μgel2a	(7.8 ± 2.4) × 10^9	(8.3 ± 3.0) × 10^10	499 ± 114	764 ± 42
μgel2b	(1.3 ± 6.9) × 10^10	(5.4 ± 2.2) × 10^10	461 ± 103	746 ± 32

---

Further analysis of microgels μgel2a and μgel2b with respect to their primary amine content was undertaken using the ATTO-TAG™ CBQCA reagent. A standard curve was generated using APMA solutions in pH 9.3 borate buffer and the efficiency of APMA incorporation in the microgels was found to be 25 ± 3% for μgel2a and 20 ± 2% for μgel2b. The average number of moles of APMA in the microgels were found to be (1.3 ± 0.2) × 10⁻⁸ and (1.1 ± 0.1) × 10⁻⁸ respectively.

Conjugation of dyes to microgels and characterization of microgel–dye constructs

Following characterization of the cationic microgels, two dyes, malachite green and rose bengal were conjugated to μgel2a and μgel2b respectively. The conjugations were performed via the amine groups in the polymer matrices, using different techniques. This study was performed to determine the applicability of these microgels as agents for conjugation through a demonstration of the synthetic accessibility of their primary amines. Malachite green (MG) was conjugated in the form of its derivative MGITC (4.5× molar excess) in a manner similar to that utilized by Nayak et al.⁴⁷ RB was conjugated in its native form by carbodiimide coupling using EDC in the presence of NHSS (10.5× molar excess of reagents used) (Scheme S2, ESI†). The two conjugates thus obtained, μgel2a–MGITC and μgel2b–RB have potential applications in bioimaging through techniques like photoacoustic and fluorescence imaging.^48,49 The conjugations were followed by purification through dialysis. μgel2a–MGITC was dialyzed against DI water, while μgel2b–RB was subjected to a two step dialysis method due to the non-specific interactions between free dye and the microgels. The first step involved dialyzing the microgel–dye conjugate solution against pH 11 CAPS buffer of very high ionic strength (500 mM). This buffer served the purpose of decreasing coulombic interactions between the two species through the deprotonation of the amine groups on the microgels and by decreasing the Debye screening length of the charged species. Following this first step of dialysis, the second step involved dialysis against DI water to facilitate replacement of the buffer.

After purification, μgel2a–MGITC and μgel2b–RB were analyzed for their R_H values by DLS in low ionic strength buffers. The results obtained are summarized in Table 4. The pH responsivity of μgel2a–MGITC can be attributed partially to the free amine groups remaining post conjugation. Additionally, it may also be ascribed to the pK_a of malachite green being 6.9,⁵⁰ leading to the presence of the charged dye species R = N(CH₃)₂⁺ at pH 7.4. R_H measurements performed on μgel2b–RB in pH 11 CAPS buffer did not yield reliable results, which may be due to aggregate formation through charge screening at the higher pH, leading to enhanced hydrophobic interactions. AFM on μgel2a–MGITC and μgel2b–RB following deposition on pre-functionalized glass demonstrated no distortion in the hemispherical profile of the microgels post dye conjugation (Fig. 3B and D). NMR spectra for RB, μgel2b and μgel2b–RB were also obtained (Fig. S5, ESI†).

Table 4 Concentrations of conjugated dyes (MGITC and RB) per microgel in comparison to dye concentration in solution and efficiency of conjugation reactions expressed as a percentage

Microgel identity	R H (nm) at pH 7.4	R H (nm) at pH 11	‘Overall’ concentration of dye in solution^a (mM)	Concentration of dye per microgel^a (mM)	% efficiency of dye conjugation^a
a Values presented as averages of measurements on two different sample dilutions, error bars represent standard deviations of these two values.
μgel2a–MGITC	424 ± 27	404 ± 22	0.035 ± 0.010	20.9 ± 7.7	43 ± 12
μgel2b–RB	379 ± 18	—	0.034 ± 0.001	14.0 ± 5.7	46 ± 3

---

In order to determine the conjugation efficiencies of MGITC and RB, absorbance measurements were performed. For these, the concentrations of microgel–dye constructs used were low enough to prevent interference due to scattering from the colloidal microgel particles. Measurements were done on μgel2a–MGITC and μgel2b–RB solutions alongside standard absorbance curves prepared using MGITC and RB solutions, at λ_max values of 625 nm and 540 nm respectively (Fig. S4, ESI†). These revealed that the dye concentrations of the stock 1 mg mL⁻¹ microgel solutions were 0.035 ± 0.010 mM and 0.034 ± 0.001 mM respectively. For comparison, μgel1, μgel2 and μgel6 from the original series were conjugated to RB, but with a lower molar excess of the conjugation agents (4×) (AFM height traces of microgel–dye conjugates in Fig. S6, ESI†). The motivation behind the synthesis of these microgels was to demonstrate the trend in primary amine group availability for reactions in two ways. First, through the conjugation of RB to microgels with different starting amounts of amines (μgel1–RB, μgel2–RB and μgel6–RB) and second, through the analysis of microgels with the same starting density of amines but varying amounts of reagents used for conjugations (μgel2–RB and μgel2b–RB). The results obtained from absorbance measurements are presented in Fig. 5. A similar trend as that observed for APMA incorporation was seen for μgel1–RB, μgel2–RB and μgel6–RB. The percent efficiency of conjugation was found to decrease with an increase in the number of APMA units in the microgels, which may be a result of differences in accessibility, particularly due to differences in microgel R_H values and thus surface presentation of functional groups. Additionally, μgel2–RB and μgel2b–RB showed differences in RB conjugation amounts as expected, due to differences in reaction equilibria. Consequently, the efficiency of conjugation was also seen to be higher in μgel2b–RB, synthesized with a larger starting concentration of the reagents.

Fig. 5 Microgel–RB conjugates analyzed for dye content via absorbance measurements at λ_max = 540 nm. Bars with dashed pattern represent conjugates μgel1–RB, μgel2–RB and μgel6–RB generated using low ratios of conjugation reagents. The brick pattern represents μgel2b–RB with primary amine content identical to μgel2–RB, but conjugated to RB using higher concentrations of conjugation reagents. Red diamonds represent percent efficiencies of RB conjugations to microgels (values presented as averages of measurements on two different dilutions, error bars represent standard deviations of these two values).

When considering utility in an imaging application like photoacoustics, the dye content in a specific, concentrated volume is a more relevant parameter than the overall dye concentration in solution. This is because in photoacoustic imaging, signals are generated by irradiation of a region of the sample with very small dimensions (∼1 μm diameter), defined either by a point source such as a highly focused laser beam or by excitation of a small ∼1 μm sized contrast agent. Thus, the concentration of dye per microgel was determined from the absorbance measurements using the Beer–Lambert law, the molecular weights of the microgels determined previously using the qNano (Table 3) and the volume of the microgels, which was calculated from their R_H values. The dye concentrations per microgel were found to be 20.9 ± 7.7 mM for μgel2a–MGITC and 14.0 ± 5.7 mM for μgel2b–RB (Table 4), which were almost three orders of magnitude above the ‘overall’ solution concentrations for the microgel dispersions. Based on these values and the efficiency of APMA incorporation obtained via the CBQCA assay, the efficiency of dye conjugation was found to be 43 ± 12% for μgel2a–MGITC and 46 ± 3% for μgel2b–RB. These efficiencies observed were similar in spite of different conjugative approaches and molar excesses of reagents being used for both dyes. This may be partially due to the inherent reaction efficiencies of the conjugation reactions utilized, but the accessibility of the amine groups in the microgels may also be a major contributing factor in determining the number of dye molecules attached per microgel.

Viability studies with NIH 3T3 fibroblasts

The development of new constructs for use in conjunction with biological systems brings with it the necessity to examine their potential toxic effects on cells. A viability assay was performed using NIH 3T3 fibroblasts. These particular cells were utilized because they have been studied extensively as model systems for analysis of cell function in regards to adhesion, movement, and proliferation. They have also often been used to examine the toxicity of different kinds of nanoparticles.^51,52

In the current study, NIH 3T3 fibroblasts were grown on a multi-well plate at a uniform density of 10 000 cells per cm² and were allowed to adhere and spread, after which they were incubated with the relevant microgels for 24 h. The microgel solutions were subsequently removed from the wells and the cells were stained with Calcein AM and EthD-1 homodimer to determine viability. The measurements performed at λ_ex = 528 nm and λ_em = 645 nm for Calcein AM were normalized against ‘all live’ and ‘all dead’ cell controls and the percent viabilities were obtained (Fig. S7, ESI†). Images of cells obtained via confocal microscopy following staining with the dyes, are presented in Fig. 6. These images demonstrated red fluorescence from all microgels that had been incubated with μgel2b–RB (Fig. 6D–F). A closer examination of the images revealed that this red fluorescence was emanating from the cytoplasm of the cell and not from the nucleus. Since EthD-1 is known to stain the nucleus of dead cells (Fig. 6K), this observation along with the high Calcein fluorescence exhibited by these cells suggested that the cells were ‘live’ and that the red fluorescence was emanating from the RB in μgel2b–RB particles within the cells, rather than the membrane penetrating EthD-1. Therefore, we concluded that the microgels had been engulfed by the cells during the incubation period. This observation could be validated only with the μgel2b–RB particles, because the unconjugated cationic microgels μgel2b are non-fluorescent and the μgel2a–MGITC particles are conjugated to MG, which has a very low quantum yield and is hence virtually non-fluorescent. We hypothesize that both these species of microgels had been engulfed by the fibroblasts as well. In the interest of avoiding any interference from engulfed μgel2b–RB particles, all viability calculations were performed based only on fluorescence intensities obtained from Calcein AM.

Fig. 6 CLSM images of NIH 3T3 fibroblasts incubated with varying microgel conjugates and concentrations for 24 h, and then stained using Calcein AM (green, live cell stain) and EthD-1 (red, dead cell stain) for determination of viability. Cells grown at a concentration of 10 000 cells per cm² were incubated with 0.125 mg mL⁻¹ μgel2a–MGITC (A), 0.25 mg mL⁻¹ μgel2a–MGITC (B), 0.5 mg mL⁻¹ μgel2a–MGITC (C), 0.125 mg mL⁻¹ μgel2b–RB (D), 0.25 mg mL⁻¹ μgel2b–RB (E), 0.5 mg mL⁻¹ μgel2b–RB (F), 0.125 mg mL⁻¹ μgel2b (G), 0.25 mg mL⁻¹ μgel2b (H) and 0.5 mg mL⁻¹ μgel2b (I). ‘All live’ and ‘all dead’ controls (J and K) shown for comparison.

A statistical examination via one way ANOVA analysis rendered the quantitative viability results inconclusive, presumably due to variability in cell cultures (Fig. S7, ESI†). However, the assay revealed low viability when the cells were incubated with the unconjugated, cationic microgels. This is expected since polycations are known to be cytotoxic, as mentioned earlier. Qualitatively, via imaging, this cytotoxicity was seen to escalate with increasing concentrations of these microgel solutions incubated with the cells. However, when incubated with dye-conjugated microgels, the images of the fibroblasts obtained hinted at the comparative lower toxicity of these constructs, especially at lower concentrations. At higher concentrations of the microgel–dye conjugates, the results indicated that the microgels might be toxic to the cells, which may be due to the ability of the dyes to generate biologically harmful reactive oxygen species (ROS) in solution.^53,54 However, in spite of its toxicity, RB has been used in biological systems in the past.⁵⁵ At the lowest concentration of the microgel–dye conjugates (0.125 mg mL⁻¹) used for these viability studies, low cytotoxicity was observed. This was encouraging, since the local concentrations of the dyes were very high as described in the earlier section. Thus, decreasing weight percent during introduction into a biological system may be feasible, in order to improve cell viability further. If the microgel–dye constructs need to be utilized at concentrations higher than 0.125 mg mL⁻¹, microgels synthesized from biocompatible materials like polyethylene glycol (PEG)^18,56 may present lower toxicity to the cells and thus may be the material of choice moving forward. Additionally, dyes that do not generate ROS can be conjugated to the microgels to serve the same purpose.

Conclusions

We have successfully synthesized and characterized a large range of cationic microgels with varying amounts of primary amine groups incorporated within the polymer networks. These microgels have potential applications in bioconjugation reactions. The syntheses were optimized to generate monodisperse, stable microgels and the particles were analyzed in detail with the goal of generating information about the incorporation efficiency of the amine containing comonomer. Furthermore, the effects of various synthesis conditions on microgel size and APMA incorporation efficiency were also investigated. A model system was characterized further and conjugated to two different dyes in order to demonstrate availability of functional groups for conjugation. Local concentrations of the dyes calculated through analysis of the molecular weights of the microgels were found to be almost three orders of magnitude higher than the concentration of dye in solution. This was very promising for applications like bioimaging, for the generation of high intensity signals. This high local concentration of the conjugated entity may also be useful for phenomena requiring multivalency for efficient targeting in biological systems. Additionally, it may be beneficial for heightened interactions with biological materials leading to events such as clot contraction, such as that observed by Brown et al. when they developed artificial platelets.¹⁶ Viability studies conducted with NIH 3T3 fibroblasts indicated that microgel–dye conjugates were not toxic to cells, especially when concentrations of 0.125 mg mL⁻¹ were used for incubation with the cells. All these studies have helped in the establishment of a toolbox of amine functionalized microgels, with constructs being suited to a wide range of applications. In particular, these investigations have revealed previously unknown aspects of these systems, with extensive studies providing corroboration through trends observed in microgel behaviors. Straightforward conjugation reactions can be performed using these microgels, as demonstrated through coupling to the two different dyes utilized in this paper. Due to the large number of possible reactions with primary amines, the opportunities for bioconjugation to these microgels are extensive, with ample flexibility available for a vast array of systems. These particles thus have prospective utility in a number of significant biomedical applications such as the delivery of therapeutics and bioimaging.

Acknowledgements

We would like to thank Prof. Thomas Barker for granting access to the Barker laboratory's tissue culture facilities at Georgia Tech, where the cell viability studies were performed, and Dr Ashley Carson Brown for providing training with same. We also appreciate the inputs provided by Prof. Robert Dickson and Aida Demissie with regards to the utility of microgels for photoacoustic signal generation. Lastly, we would like to express our gratitude towards Dr Nicole Welsch for her guidance and training with experiments related to tunable resistive pulse sensing.

Notes and references

1. J. P. Magnusson, S. Bersani, S. Salmaso, C. Alexander and P. Caliceti, Bioconjugate Chem., 2010, 21, 671–678.
2. F. Meng, W. E. Hennink and Z. Zhong, Biomaterials, 2009, 30, 2180–2198.
3. J.-F. Lutz and Z. Zarafshani, Adv. Drug Delivery Rev., 2008, 60, 958–970.
4. V. Escriou, M. Carriere, D. Scherman and P. Wils, Adv. Drug Delivery Rev., 2003, 55, 295–306.
5. Y. Chau, F. E. Tan and R. Langer, Bioconjugate Chem., 2004, 15, 931–941.
6. Y. D. Jin, C. X. Jia, S. W. Huang, M. O'Donnell and X. H. Gao, Nat. Commun., 2010, 1, 41.
7. V. Biju, Chem. Soc. Rev., 2014, 43, 744–764.
8. L. Cheng, H. Gong, W. W. Zhu, J. J. Liu, X. Y. Wang, G. Liu and Z. Liu, Biomaterials, 2014, 35, 9844–9852.
9. N. Erathodiyil and J. Y. Ying, Acc. Chem. Res., 2011, 44, 925–935.
10. O. C. Farokhzad, J. Cheng, B. A. Teply, I. Sherifi, S. Jon, P. W. Kantoff, J. P. Richie and R. Langer, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 6315–6320.
11. Y. Matsumura and H. Maeda, Cancer Res., 1986, 46, 6387–6392.
12. L. Li, A. Chang, Y. Hu, L. Zhang and W. Wu, Chem. Commun., 2013, 49, 6534–6536.
13. Z. Y. Meng, M. H. Smith and L. A. Lyon, Colloid Polym. Sci., 2009, 287, 277–285.
14. E. B. Dickerson, W. H. Blackburn, M. H. Smith, L. B. Kapa, L. A. Lyon and J. F. McDonald, BMC Cancer, 2010, 10, 10.
15. G. R. Hendrickson and L. A. Lyon, Soft Matter, 2009, 5, 29–35.
16. A. C. Brown, S. E. Stabenfeldt, B. Ahn, R. T. Hannan, K. S. Dhada, E. S. Herman, V. Stefanelli, N. Guzzetta, A. Alexeev, W. A. Lam, L. A. Lyon and T. H. Barker, Nat. Mater., 2014, 13, 1108–1114.
17. A. W. Bridges, N. Singh, K. L. Burns, J. E. Babensee, L. A. Lyon and A. J. Garcia, Biomaterials, 2008, 29, 4605–4615.
18. J. F. Lutz, Adv. Mater., 2011, 23, 2237–2243.
19. M. H. Smith, E. S. Herman and L. A. Lyon, J. Phys. Chem. B, 2011, 115, 3761–3764.
20. A. B. South and L. A. Lyon, Chem. Mater., 2010, 22, 3300–3306.
21. W. H. Blackburn, E. B. Dickerson, M. H. Smith, J. F. McDonald and L. A. Lyon, Bioconjugate Chem., 2009, 20, 960–968.
22. S. Nayak, H. Lee, J. Chmielewski and L. A. Lyon, J. Am. Chem. Soc., 2004, 126, 10258–10259.
23. T. Hoare and R. Pelton, Langmuir, 2004, 20, 2123–2133.
24. N. Sahiner, S. Demirci, M. Sahiner and H. Al-Lohedan, Appl. Surf. Sci., 2015, 380–387.
25. G. T. Hermanson, Bioconjugate Techniques, Academic Press, San Diego, CA, 2nd edn, 2008.
26. L. Antonietti, C. Aymonier, U. Schlotterbeck, V. M. Garamus, T. Maksimova, W. Richtering and S. Mecking, Macromolecules, 2005, 38, 5914–5920.
27. M. Hahn, E. Gornitz and H. Dautzenberg, Macromolecules, 1998, 31, 5616–5623.
28. L. Nabzar, D. Duracher, A. Elaissari, G. Chauveteau and C. Pichot, Langmuir, 1998, 14, 5062–5069.
29. R. K. Pinschmidt, J. Polym. Sci., Part A: Polym. Chem., 2010, 48, 2257–2283.
30. X. B. Hu, Z. Tong and L. A. Lyon, Colloid Polym. Sci., 2011, 289, 333–339.
31. F. Meunier, A. Elaissari and C. Pichot, Polym. Adv. Technol., 1995, 6, 489–496.
32. T. Still, K. Chen, A. M. Alsayed, K. B. Aptowicz and A. G. Yodh, J. Colloid Interface Sci., 2013, 405, 96–102.
33. D. Duracher, F. Sauzedde, A. Elaissari, A. Perrin and C. Pichot, Colloid Polym. Sci., 1998, 276, 219–231.
34. J. J. Xu and R. Pelton, J. Colloid Interface Sci., 2004, 276, 113–117.
35. S. Thaiboonrod, C. Berkland, A. H. Milani, R. Ulijn and B. R. Saunders, Soft Matter, 2013, 9, 3920–3930.
36. J. Ramos, J. Forcada and R. Hidalgo-Alvarez, Chem. Rev., 2014, 114, 367–428.
37. R. Farley and B. R. Saunders, Polymer, 2014, 55, 471–480.
38. I. Miraballes-Martinez, A. Martin-Molina, F. Galisteo-Gonzalez and J. Forcada, J. Polym. Sci., Part A: Polym. Chem., 2001, 39, 2929–2936.
39. D. Fischer, Y. X. Li, B. Ahlemeyer, J. Krieglstein and T. Kissel, Biomaterials, 2003, 24, 1121–1131.
40. D. J. Gan and L. A. Lyon, J. Am. Chem. Soc., 2001, 123, 7511–7517.
41. W. McPhee, K. C. Tam and R. Pelton, J. Colloid Interface Sci., 1993, 156, 24–30.
42. J. Gao and B. J. Frisken, Langmuir, 2005, 21, 545–551.
43. K. Mai-ngam, K. Boonkitpattarakul, M. Sakulsombat, P. Chumningan and B. Mai-ngam, Eur. Polym. J., 2009, 45, 1260–1269.
44. H. Bachman, A. C. Brown, K. C. Clarke, K. S. Dhada, A. Douglas, C. E. Hansen, E. Herman, J. S. Hyatt, P. Kodlekere, Z. Y. Meng, S. Saxena, M. W. Spears, N. Welsch and L. A. Lyon, Soft Matter, 2015, 11, 2018–2028.
45. W. W. You, R. P. Haugland, D. K. Ryan and N. P. Haugland, Anal. Biochem., 1997, 244, 277–282.
46. X. F. Zhou, F. Y. Su, Y. Q. Tian and D. R. Meldrum, PLoS One, 2014, 9, e88185.
47. S. Nayak and L. A. Lyon, Chem. Mater., 2004, 16, 2623–2627.
48. G. L. Indig, D. G. Jay and J. J. Grabowski, Biophys. J., 1992, 61, 631–638.
49. A. N. Serafini, W. M. Smoak, H. B. Hupf, J. E. Beaver, J. Holder and A. J. Gilson, J. Nucl. Med., 1975, 16, 629–632.
50. M. C. Nguyen, I. Shoji and K. Noboru, Am. J. Anal. Chem., 2012, 2012, 188–194.
51. D. Drescher, G. Orts-Gil, G. Laube, K. Natte, R. W. Veh, W. Oesterle and J. Kneipp, Anal. Bioanal. Chem., 2011, 400, 1367–1373.
52. D. A. Fletcher and D. Mullins, Nature, 2010, 463, 485–492.
53. S. Dhamgaye, F. Devaux, R. Manoharlal, P. Vandeputte, A. H. Shah, A. Singh, C. Blugeon, D. Sanglard and R. Prasad, Antimicrob. Agents Chemother., 2012, 56, 495–506.
54. C. Lambert, T. Sarna and T. G. Truscott, J. Chem. Soc., Faraday Trans., 1990, 86, 3879–3882.
55. G. Savini, P. Prabhawasat, T. Kojima, M. Grueterich, E. Espana and E. Goto, Clin. Ophthalmol., 2008, 2, 31–55.
56. Z. B. Hu, T. Cai and C. L. Chi, Soft Matter, 2010, 6, 2115–2123.

---

Footnote

† Electronic supplementary information (ESI) available: Supplemental reaction schemes and additional microgel characterization data. See DOI: 10.1039/c6ra00809g

---