Customized hybrid and NIR-light triggered thermoresponsive drug delivery microparticles synthetized by photopolymerization in a one-step flow focusing continuous microreactor

Customized hybrid and NIR-light triggered thermoresponsive drug delivery microparticles synthetized by photopolymerization in a one-step flow focusing continuous microreactor Isabel Ortiz de Solorzano†‡§*, Gracia Mendoza†‡,, Manuel Arruebo†‡§*, Victor Sebastian†‡§ †Department of Chemical Engineering. Aragon Institute of Nanoscience (INA) and Instituto de Ciencia de Materiales de Aragón (ICMA), Universidad de Zaragoza-CSIC, University of Zaragoza, Campus Río Ebro-Edificio I+D, C/ Poeta Mariano Esquillor S/N, 50018-Zaragoza, Spain ‡Aragon Health Research Institute (IIS Aragón), 50009 Zaragoza, Spain §Networking Research Center on Bioengineering, Biomaterials and Nanomedicine, CIBER-BBN, 28029-Madrid, Spain *Corresponding author: [email protected]; [email protected]. KEYWORDS: Thermoresponsive; photopolymerization; bupivacaine; microparticles; LED; microfluidics; coaxial; PNIPAm, Hollow gold nanoparticles, NIR-light triggered, drug delivery. ABSTRACT Photopolymerization is a selective technique that takes advantage of light-sensitive molecules to initiate and propagate monomeric structures to render covalently bonded macromolecular materials known as polymers. Herein, we present a novel one-step microfluidic synthesis of customized hybrid-thermoresponsive Poly(Nisopropylacrylamide) (PNIPAm) based microparticles (MPs) containing plasmonic hollow gold nanoparticles (HGNPs) and bupivacaine (BVP) used as a model drug. Those hybrid microparticles were prepared using a flow-focusing microreactor coupled to a UV LED device built with a simple outer PTFE tubing and an inner flexible capillary. Different tubing characteristics and flow rate ratios were altered in order to control the size of the resulting microparticles. In addition, components such as monomer, crosslinker and photoinitiator concentrations, as well as LED intensity and irradiation time were tuned to obtain different MPs and their characteristics and polymerization rates were compared by Gel permeation Chromatography (GPC). Thermoresponsive properties were analyzed and the presence of HGNPs was confirmed in light-activated triggered drug release applications. Bupivacaine loading and release studies were demonstrated with the resulting hollow and solid microparticles (which were obtained depending on the polymerization rate used) and their temperature responsiveness was assessed using a NIR laser when HGNPs were present in the constructs. Finally, cytotoxicity studies, cell-cycle arrest and apoptotic induction were carried out to certify their suitability for further biomedical applications to be used as triggerable drug depots. 1 Introduction Engineered polymeric microparticles (MPs) have become very interesting multifunctional platforms in biomedicine.1 Biosensing2, drug delivery3 or tissue engineering4 are some of the areas in which MPs standout as advanced materials at the cutting edge of functional polymers. However, their biological suitability depends on their properties, which have a direct relationship with their size, composition, structure and configuration. Thus, controlling them has turned into the key challenge for researchers of the field.5,6 Conventional discontinuous polymerization methods, such as emulsion6,7, dispersion8 or spraying9, do not allow a good control over the final MPs properties. This usually leads to polydispersity and low reproducibility10,11,12 even at reduced production scales due to the heterogeneous distribution of reactants and temperatures and an insufficient mixing. In the last years, the production of monodisperse particles for biomedical use has stimulated great interest in a myriad of scientific and engineering fields to take advantage of their benefits in controlling drug release rates and obtaining reliable results. Continuous microfluidic devices have introduced different approaches to overcome these previous limitations.13,14 Micro-sized channels permit to handle fluid flows at the microscale very precisely leading to highly reproducible synthesis and narrow size distributions in the resulting MPs due to an efficient mixing driven by molecular diffusion.15,16 Microfluidic systems offer also some advantages against classical systems17: homogenous reaction conditions, portability, high sensitivity, low energy consumption, highly integrated multifunction and easy scalability. Droplet microfluidics has been demonstrated to be one of the most efficient microfluidic approaches thanks to the facile control of the reaction conditions and fast MP formation18. Two types of flow focusing devices have been considered in the droplet based microfluidics approach to produce spherical MP19: 1) a chip-based flow focusing device with a T or Ψ shaped microchannel integrated on a chip, and 2) a capillary-based flow focusing device (coaxially aligned microcapillaries). Glass-made microchips and capillaries are very expensive, as well as fragile materials.20,21 Glass manufacturing has the limitations of a complicated welding and gas tight sealing, a difficult integration in a macro-system (connecting to the macro scale) and also any posterior modification is not economically viable. On the other hand, the design and fabrication of microfluidic silicon-based chips could be expensive, labour-intensive and requires clean room facilities. Furthermore, polymerization reactions can easily promote chip blockage, precluding a stable flow. Consequently, while some problems can be overcome by careful choice of the reaction conditions, there is a need for low-cost, modular components that can be easily assembled into flow-focusing devices by users lacking proficiency or access to microfabrication facilities22. In this sense, polymeric PTFE coaxially aligned capillaries provide a real alternative thanks to their ease of fabrication and modification, flexibility, re-usability and excellent chemical and mechanical properties and optical transparency.23,24 Microparticle precipitation can be provoked by ionic crosslinking25, temperature induced gelation26 or by polymerization27. Polymerization can be carried out upon heat assisted radical generation or by UV irradiated photopolymerization among other techniques.28 In this sense, photopolymerization represents a fast process having a precise control over the characteristics of the final polymer after the induction by UVlight lamp exposure20 avoiding the potential problems that high temperature may provoke on biomolecules used during the synthesis. The use of light emitting diodes (LEDs) holds great promise in the field of photopolymerization due the following advantages29: 1) low energy consumption, 2) no ozone release, 3) low heat generation, 4) low operating costs and maintenance, 5) high lifetimes, 6) easy and safe handling, 7) 100% output immediately obtained after turn-on, and 8) easily controlled intensity. Then, the assembly of LED devices and microfluidics to control the polymerization process is an interesting alternative to deal with the controlled production of polymer MPs. Polymerization of droplets can be accomplished upon incorporating a curing agent directly into the droplet, or in the continuous phase. The photoinitiator solubility in the disperse or continuous phase will be the key value to select the polymerization approach. Droplet polymerization can be accomplished during droplet formation, inside the microreactor chip30,31, or in a totally separated step off-chip32,33,34. Time, intensity of light, and reaction components determine the composition, size and properties of the synthetized MPs.35,36 MPs prevail as drug delivery vehicles, considering that a large number of products based on polymer MPs have already been commercialized. To name a few: 1) Trelstar® injectable microspheres loaded with triptorelin pamoate and Enantone LP® loaded with Leuprorelin and Lupron® depot loaded with leuprolide acetate for prostate cancer treatment. 2) Sandostatin LAR® depot loaded with octreotide acetate and Somatulin L loaded with lanreotide for acromegaly treatment. 3) Risperdal Consta® depot loaded with risperidone for the treatment of schizophrenia as well as for the longer-term treatment of Bipolar I disorder. 4) Nutropin® depot loaded with recombinant human growth hormone as growth hormone regulator; and 5) Vivitrol® depot loaded with naltrexone for treating alcohol dependence and preventing relapse to opioid dependence. One of the most promising MPs so far for biomedical applications are gel structures.37 Their capability to respond to several external stimuli such as temperature38 or pH39 make them extremely interesting in biomedicine. Among them, microgels based on thermoresponsive polymers are widely studied. One of them, poly(N- isopropylacrylamide) (PNIPAm) presents a broad variety of opportunities in order to customize the final product according to the desired application40,41,42. Crosslinking, monomer and initiator concentration as well as flow rate ratio, irradiation time and intensity of the UV-light source, lead to MPs with completely different properties43. Mechanical and thermoresponsive properties of resulted MPs are determined by the fine-tuning of the synthesis variables. Synthesis of hybrid PNIPAm MPs has been conducted in conventional batch type reactors, showing the possibility to combine the function of both thermoresponsive polymers and stimuli sensitive inorganic NPs. For example, Ekici et al.44 incorporated magnetic iron NPs in hyaluronic acid-PNIPAm nanogels obtaining a double magnetism-temperature responsive system. Moreover, microfluidics allows the precise and controlled encapsulation in MPs of different components from fluorescent dyes45 to live cells46,47 that can be useful in drug delivery studies or biosensing48. In general, the loading of cargos capable to respond to external stimuli such as light, ultrasound or magnetic fields are very interesting to explore the versatility of hybrid MPs. Superparamagnetic Fe3O4 NPs (SPIONs) have also been used as triggers for achieving a burst release of lipophilic substances encapsulated in oil cores from core-shell PNIPAm microparticles after induction by alternating magnetic fields49. Also, gold nanostructures have been widely used for light-triggered drug release from PNIPAm microcapsules42. Wang et al.50 recently showed the possibility to activate shape transition of Au nanorods-PNIPAm hybrid MPs using high power NIR laser irradiation. Those hybrid MPs were produced in several stages, resulting in a semicontinuous process: 1) Drop formation in continuous flow, 2) Drop sedimentation in stagnant conditions for 5 minutes after collection in order to fine tune the shape and 3) Drop photopolymerization in stagnant conditions under 365 nm UV light during 5 min. Kim et al.51 also used the same microfluidic approach to generate hollow PNIPAm microcapsules containing gold nanorods. But in this last case the double-emulsion droplets were incubated in a glass vial for 12 hours at room temperature to complete the polymerization, turning it also as a semi-continuous process, which represents a drawback to increase the productivity. Although a high control over the size and thermoresponsive properties of the MPs were achieved by the reported methods50,51, the production of PNIPAm hybrid MPs co-loading in the same MP drugs and metal nanoparticles (HGNPs) using a single flow device in continuous fashion by photopolymerization is still lacking. The development of a new flow synthesis approach that enables to address previous drawbacks would accelerate the translation of hybrid MPs applications for their future biomedical use. Herein, we present a facile, low-cost, and efficient method of producing drug loaded HGNPs-PNIPAm hybrid MPs, capable to tune the structure of the drug carrier MPs in order to achieve different pharmacokinetics in drug delivery applications. A capillarybased flow focusing device with coaxially aligned micro capillaries was coupled to a UV-LED to facilitate the controlled droplet formation and fast polymerization at different synthesis conditions obtaining customized MPs for purpose-specific drug delivery applications. Flow dynamics, reagents ratio and UV exposure time were studied to achieve two types of MPs that behave in a different way, accordingly to the type of drug delivery targeted. The efficient simultaneous loading of a drug and HGNPs is also described. HGNPs were selected as trigger NPs for on-demand light-responsive drug delivery. On its part, bupivacaine was used as a model drug in order to test the pharmacokinetic response of MPs with different shell thicknesses together with the reversible temperature-dependence of hybrid MPs sizes. HGNPs immobilized in the hydrogel network enable the localized heating of the MPs after NIR light illumination, providing the remote control in the permeability of the shell and the triggered release of bupivacaine. Cytotoxicity analyses were carried out with the materials developed in two different cell lines studying their subcytotoxic concentration and their effect in cell membrane and cell cycle. 2 Materials and methods 2.1 Materials N-isopropylacrylamide ≥99% (NIPAM), N,N’-methylenebis(acrylamide) 99% (BIS), Span® 80, hexadecane, ReagentPlus®, 99%, cobalt chloride hexahydrate (ACS reagent grade), sodium citrate tribasic dihydrate (98%), poly(vinylpyrrolidone) (PVP, Mw = 55000 Da), sodium borohydride (99%), gold(III) chloride hydrate (50% Au basis) and bupivacaine hydrochloride monohydrate (99%) were purchased from Sigma-Aldrich. 2,2 Diethoxyacetophenone (DEAP) was purchased from Acros Organics. 2.2 Synthesis of hollow gold nanoparticles (HGNPs) Hollow gold nanoparticles (HGNPs) were synthetized following previous works developed in our group.52,53 In brief, 400 mL of deionized (DI) water (0.1 % of 0.35 M cobalt chloride hexahydrate) and 1.6 mL of 0.1 M sodium citrate trihydrate were deoxygenated in a two-necked round-bottom flask by bubbling the solution with argon gas for 45 min. Both, 2 mL of 1 wt.% of PVP and 400 µL of 1.0 M sodium borohydride, were added to the previous solution under magnetic stirring forming cobalt NPs. Argon flux and stirring were kept for 15 min. After that, 380 mL of the resulting NPs dispersion was transferred to a beaker containing 120 mL of DI water and 180 µL of 0.1 M gold (III) chloride hydrate under stirring. Magnetic stirring under room temperature conditions was sustained for 30 min until a complete oxidation of the residual cobalt was produced. A green colored dispersion indicates the presence of HGNPs, which were subsequently washed by several centrifugation steps. 2.3 Microreactor characteristics To produce HGNPs-PNIPAm hybrid microparticles, a coaxial capillary microfluidic device was assembled (Figure 1). The dimensions of the inner and outer capillaries were modified to tune the resulting microparticle sizes. The inner and outer capillaries were made of PEEK (hydrophilic) and PTFE (hydrophobic), respectively. The inner capillary diameter was varied from 150 µm or 25 µm, whereas outer capillary diameter was altered from 560 µm to 790 µm. The dispersed phase (aqueous) was injected at low rates of flow (2.5-30 µL/min) in order to avoid co-laminar flows and unsteady microparticle formation. The inner capillary was coaxially localized under an optical microscope in order get an axisymmetric flow-focusing device where the dispersed phase is surrounded symmetrically by the continuous phase. The coaxial capillary reactor was supported in a polymer housing fabricated by 3D-printing to avoid capillaries misalignment and to favor fluid dynamic reproducibility. In the downstream flow after droplet formation, a UV irradiation provided by a 4.6 W LED (365 nm wavelength) was used to activate the photoinitiator and promote the formation of radicals to polymerize the monomers inside the resulted droplets. LED intensity was modulated in order to obtain the optimum polymerization rate. Two different syringe pumps (Harvard Apparatus PHD ULTRA™) at selected flow rates were used to control the drop residence time of the injected reagent streams. Finally, microparticles were collected in a water recipient in order to avoid the blockage of the outer tubing. promote a fast polymerization during the drop flow in the microfluidic system. The residence time of formed droplets is usually larger in dripping mode because low flow rates (inner and outer phases) are required. In addition, droplets produced under dripping mode are pinched off near the inner capillary tip, whereas in jetting regime, droplets are pinched off from an extended thread generated downstream of the inner capillary tip. Droplet formation in co-flowing microfluidic systems is sensitive to the viscous shear stress of the external flow and the capillary pressure resisting the deformation of the internal phase.20 The dimensionless capillary number (Ca) compares the relative importance between the friction force and the surface tension. Then, we considered previous studies of droplet formation in dripping mode using similar fluids54, to select the fluid dynamic conditions to get stable droplet formation. Several factors such as stream composition or capillary dimensions were tuned in order to control the morphology and the polymerization efficiency of the system. The effect of those parameters will be discussed in the following sections. 3.2 Microparticles Morphology and dimensions It is well accepted that the size of droplet formation in a co-flowing microfluidic device depends on the capillaries dimensions, and inner and outer flow rates. These parameters are highly important because they can modulate the competition between viscous shear stress of the external stream and capillary pressure resisting deformation of the internal stream. In this work, we have selected two different inner and outer capillaries to tune the size of MPs under a stable droplet formation in dripping regime. Figure 2 shows the averaged MPs diameter obtained under different capillary dimensions and flow rates. In general the most important effect is observed when the external PTFE capillary was modified. Figure 2: Average MPs size depending on flow ratio and inner diameter of the microreator channels. a) and b) show the results for the smallest external PTFE tubing (560 µm) and large (150 µm) and small (25 µm) internal capillary, respectively. c) and d) present the results for the larger PTFE tubing (790 µm) and large (150 µm) and small (25 µm) internal capillary, respectively. Same colors represent the same inner flow. For each of them, from left to right continuous flow increases. Inverted microscope images of MPs synthesized with flow ratio of 5100 L/min are shown in: e) ID: 150 µm and OD 560 µm; f) ID: 25 µm and OD 560 µm; g) ID: 150 µm and OD 790 µm; h) ID: 25 µm and OD 790 µm. i) Summary and comparison of MPs diameter synthesized with two different flow ratios and all combinations of internal and external tubing diameters. Smaller outer capillary yielded to microparticles between 400 and 500 µm (Figure 2 b) and a)); while a larger outer capillary diameter generated MPs over 700 µm in diameter (Figure 2 c) and d)). These results can be rationalized by the droplet formation mechanism. The junction close to the inner capillary tip is obstructed by the inner fluid protrusion. The continuous flow close to the protrusion is restricted as the protrusion is enlarged, establishing a pressure gradient across the protrusion. The droplet is pinched off once the pressure gradient in the continuous flow is sufficiently high to overcome the capillary pressure inside the dispersed drop. Viscous shear forces are usually sufficiently large to pinch off the droplet before it grows to block the outer capillary55. Then, the droplet diameter is usually smaller than the outer channel diameter. In addition, the droplet diameter is inversely proportional to the average velocity of the carrier flow because the drag force increases as the continuous phase velocity does56. This fact confirms that the MPs diameter increases as the flow rate of the outer stream is decreased when the inner flow rate is kept constant (the drug force is reduced). On the other hand, the inner capillary size mainly affects MPs polydispersity. A more confined droplet formation benefited from a smaller capillary, generating narrower diameter deviation. This is reflected in the different coefficient of variance (CV) of MPs sizes obtained with the four systems studied. While MPs CVs obtained with larger inner capillary (790 µm) have values between 15 nm (when inner flow rate is 10 µL/min) to 100 nm (when continuous phase gets to 30 µL/min), a maximum variance of 30 nm is achieved in experiments when using 25 µm inner capillary. Similar results were previously observed with analogous platforms.57 Dripping frequency production (f) for all samples ranges between 7.5 Hz to 0.2 Hz depending on the inner flow rate and final size of MPs. Slow flow rate (2.5 µL/min) and large final MPs (~700 µm) lead to slow droplet formation and frequencies around 0.2 Hz. However, increasing the inner flow rate up to 30 µL/min and confining the size of the outer capillary, and thus the final MPs size (~400 µm), provoked a faster droplet formation and frequencies over 7.5 Hz. It is of paramount importance to highlight that this low dripping frequency was required in order to get a complete polymerization reaction in continuous fashion, a fact that is novel in this work. According to the aforementioned polymerization requirements, it is essential to achieve certain level of polymerization rate in order to obtain stable MPs with a competent functionality for a potential biomedical use. This fact, together with the fluid dynamic restrictions required to achieve a stable droplet generation limit the available phase flow rates. In this work a single LED source has been considered, however using several LEDs in series would potentially increase the irradiation time allowing the use of larger flow rates. It has been demonstrated that droplet size in coaxial flow platforms with flow ratios within our working limits do not change significantly21. However, it can still have small effects on the final MPs sizes. Statistical analysis of our results (Figure S1) showed some significant differences in almost all flow ratio combinations suggesting that as mentioned before, keeping the same inner flow, the increase in the continuous flow rate for all samples led to a slight decrease in the final MPs size. In all cases, monodisperse MPs were obtained with variations less than 15 % of their size in the worst-case scenario (150 µm inner capillary) but with average CV of 2.5 % in the most stable cases (25 µm inner capillary). 3.3 Polymerization efficiency and thermoresponsive properties Photopolymerization reactions depend on a set of different variables (monomer, photoinitiator and crosslinker concentrations and LED irradiance and time of irradiation) and some of them were studied using this co-flowing microfluidic device. Considering that the main goal of this research was to produce MPs with on-demand light-responsive drug delivery ability and allow a remote control of the polymeric shell permeability, we attempted the analysis of volume ratio shrinkage in the resulted MPs. This parameter was also selected as a key factor to analyze the polymerization efficiency and thermo-responsive properties. It must be highlighted that either an unsuccessful or limited polymerization yield can induce MP collapse, aggregation or a scarce volume shrinkage upon a temperature variation in the lower critical solution temperature (LCST) range. On the other hand, crosslinking agent entails a key role in the shrinkage behavior of any hydrogel. Higher concentration of crosslinking molecules among formed polymer chains leads to steric difficulties to achieve a complete shrinkage of the hydrogel from its swollen hydrated state to its shrunken dehydrated form. Figure 3: Volume change with temperature for a) MPs synthetized with 250 mg/mL monomer concentration and 40 µL DEAP and different crosslinker concentrations; b) to g) Time-lapse microscopy images of MPs synthetized with the same conditions at different temperatures. All samples were synthetized under flow ratio of 5/100 µL/min, inner capillary diameter: 25 µm and outer capillary: 560 µm. Figure 3 a) represents the volume ratio with temperature obtained for three different samples containing 25, 50 and 75 monomer-crosslinker NIPAM-BIS ratio, respectively. The results confirm that all MPs produced show a fast response to environmental temperature. It is relevant that dramatic changes occur between 32-39ºC (LCST of PNIPAm is 32ºC). This may be attributed to polydispersity in the polymer chain lengths and crosslinking degree58,59. It was also confirmed that the highest the NIPAM-BIS ratio, the largest the rate of volume change of the photo-polymerized MPs is. The volume change in PNIPAm MPs is rationalized because MPs are in a swollen and hydrophilic state below the LCST that is switched to a shrunken and hydrophobic state above the LCST. The state transition was studied in a time-lapse inverted microscope at different temperatures. Figures 3 b) to g) depict representative optical images at different environmental temperatures, where the dramatic reduction of MPs size above the LCST is clearly observed as well as the reversibility of this process once the temperature decreases. Similar behavior was also observed when monomer concentration and monomer/crosslinker ratios (NIPAM/BIS) were modified (Figure S2) confirming the thermoresponsive behavior of the final polymer conforming MPs. Polymerization efficiency is related to the polymerization rate and it has direct influence in the length of the polymeric chains formed and the possibility to collapse and reduce their volume and size under high temperatures. Figure 4 shows the effect of monomer (NIPAM) and photoinitiator (DEAP) concentration on final molecular weight of resulted MPs and their ability to reduce their volume under temperature variations. Figure 4: a) Volume ratio observed in samples synthesized with different monomer and photoinitiator concentrations; b) GPC results for three different samples with different DEAP/monomer ratio; c) Volume ratio shown for two samples polymerized with two different LED light irradiation intensities. Figure 4 a) demonstrates that the photoinitiator (DEAP) concentration is crucial to achieve a desired grade of polymerization that is high enough to endow MPs formation with the appropriate rate of volume change and thermosensitive behavior. A high DEAP concentration leads to a high concentration of radical species that initiate polymerization. However, the fastest the polymerization is, the shorter the polymer chains formed and then, the less the volume change when the environmental temperature increases. This observation was also confirmed by Gel Permeation Chromatography (GPC) (Figure 4 b). The GPC chromatogram provides the molecular weight distribution of the resulting species, labeling the main molecular weight peaks (Figure 4-b). According to GPC analysis, the molecular weight of the resulted polymer is lower as the photoinitiator concentration is increased because a high density of monomer building blocks is activated by the radical species. This fact was also evidenced by inverted microscopy, where optical images taken at different environmental temperatures depicted a significant shrinkage as the photoinitiator content was decreased (Figure S3). On the other hand, the activity of the photoinitiator molecules is determined by their stimulation source, the UV-LED light in our system. The irradiance of LED source determines the velocity of the polymerization process leading to different polymer chain lengths and thus to different photothermal properties. There is an intensity threshold that determines the minimum intensity necessary in order to obtain rigid and mechanically stable MPs below the one where no enough polymer chains were formed, and MPs did not keep their shape after exiting the microfluidic platform. Intensities from 0.2 to 0.9 A were used in our study finding the threshold described before set at 0.7 A. At intensities of 0.2 and 0.4 A no MPs formation took place and easily breakable MPs were obtained, respectively. However, above 0.7 A no significant changes were observed when UV-LED light intensity was increased to 0.9 A (Figure 4 c) in our system. As the polymerization is initiated at the interface of the aqueous/organic phase of detached droplets, the diffusion of radical species that conduct the polymerization reaction might have a relevant role in the MPs structure. In these terms, hollow or solid MPs could be produced depending on the radical species droplet internalization. Kim et al.31 corroborated the relationship of a high polymerization ratio with the improved mechanical strength of the resulted MPs. Long irradiation times, high DEAP concentration and high irradiation intensity favored to obtain a solid structure and mechanically resistant MPs. However, the highly solid MPs obtained under these conditions resulted in low shrinkage ratios (~24 v/v%). The polyethylene glycol diacrylate (PEGDA) photopolymerization reaction from a droplet microfluidic device emulsification process was also studied by Filatov et al.36 showing the importance of the photoinitiator concentration and the reaction time in the final core-shell structure and MPs resistance. However, a plateau of time reaction exists at very high photoinitiator concentration where increasing photoinitiator concentration does not lead to faster polymerization reaction. On the other hand, when the monomer concentration increases, more monomer molecules are available at the droplet surface leading to thicker MP shells and thus more mechanically resistant MPs. This is explained due to a faster polymerization process with the same reaction time. Considering our previous remarks, it was devised the formation of two types of MPs with a different internal structure and good mechanical properties, where the density of polymeric chains was tuned in order to modulate the void fraction and the rate of volume shrinkage. Figure 5 shows some of the most representative MPs produced by tuning the polymerization rate, either by modifying the flow rate of the continuous outer phase or the photoinitiator concentration. MPs produced at the opposite conditions were discarded because they unfeasible applicability. That is, MPs produced with the smallest flow rate and photoinitiator concentration (Figure 5-a), because they were not mechanically stable. On the other hand, MPs produced at the highest flow rate and photoinitiator concentration (Figure 5-d), because of the high density of polymeric chains and rigidity would seriously affect the rate of volume shrinkage. As a result, intermediate conditions were selected as the proper ones in order to load the cargo and study the temperature assisted drug release (Figure 5 b-c). Figure 5: From left to right MPs synthesized with monomer concentration of 150 mg/mL (a) and b)) and 250 mg/mL (c) and d)). Images a) and c) were synthetized with 40 µL of photoinitiator (DEAP); while 500 µL of photoinitiator (DEAP) were used in images b) and d). Transition temperature in drug delivery applications may lead to different drug release profiles which could be appropriate for a broad variety of disease treatments. This tuning of LCST can be achieved by combining different monomers in the polymeric structure. It has been previously demonstrated that the presence of diverse functional groups at the side chains of the polymer chain results in variations of hydrophobicity and thus polymer-water interactions60. Acrylamide monomer has been widely used in order to increase the transition temperature in PNIPAm-based polymers. The lack of isopropyl groups in its side chain increases water-polymer interactions making more energy necessary to fold the polymer chains. In our proposed microfluidic platform, this varied monomer combinations driven to obtain customized thermoresponsive MPs with desired LCST is easy to carry out. We demonstrated that adding Acrylamide (AAm) as co-monomer the LCST of the resulting MPs can be easily modified (See Figure S4). LCSTs for MPs with different acrylamide percentages are summarized in Table 1. corresponding bupivacaine concentration. High cell viability (> 70%) was observed for all MPs concentration tested for both solid and hollow MPs exudates showing more compatibility than free BVP in human fibroblast. Following studies were developed at a MPs concentration of 0.5 mg/mL as it was considered as subcytotoxic dose following the ISO 10993-5 standard, which states a viability of 70% as the threshold for considering non-cytotoxic concentrations.62 Cell apoptosis studies by flow cytometry were selected to show the potential cell membrane effect caused by the MPs exudes (Table 2). The incubation of the two cell lines (TPH1 macrophages and human dermal fibroblasts) at the subcytotoxic concentration (0.5 mg/mL) with MPs exudates loaded with HGNPs and BVP did not show remarkable changes compared to non-treated samples. Only macrophages showed a slight increase in necrosis and late apoptosis for hollow and solid loaded MPs, respectively (<3%) and a consequent decrease in cellular viability (<4%). Figure 8: Cytotoxicity results for two cell lines: a) Human fibroblasts and b) macrophages for two different shell thickness MPs denominated solid and hollow MPs loaded with HGNPs and Bupivacaine; and free bupivacaine. Mean values and SD obtained from fives samples. Table 2: Apoptosis results obtained by flow cytometry Control Exude solid MPs Exude hollow MPs Macrophages (%) Necrosis 4.43 2.09 7.41 Late apoptosis 1.72 5.74 2.58 Early apoptosis 0.59 1.44 1.14 Viability 93.26 90.74 88.87 Fibroblasts (%) Necrosis 0.51 0.33 0.38 Late apoptosis 1.99 2.47 2.01 Early apoptosis 3.05 3.18 2.76 Viability 94.44 94.02 94.85 Cell cycle studies are depicted in Table 3. Cell treatment for 24 h with HGNPs-BVP loaded PNIPAm MPs exudates at subcytotoxic doses (0.5 mg/mL) did not display accentuated effects on cell cycle. A slight increase in G2 phase in macrophages (<2%) and in S phase in fibroblasts (<4%) are the only changes observed after treatment compared to control samples. Therefore, at the doses tested the exudates released by the MPs here prepared did not show cytotoxicity, cell-cycle arrest or apoptotic induction. Table 3: Cell cycle results obtained by flow cytometry Control Exude solid MPs Exudate hollow MPs Macrophages (%) G1 55.72 53.44 54.11 S 13.32 13.49 13.41 G2 30.96 33.08 32.48 Fibroblasts (%) G1 45.72 43.83 44.19 S 31.34 35.99 35.17 G2 22.94 20.18 20.64 4 Conclusions Summing up, in this work we obtained thermosensitive light-responsive hybrid MPs loaded with an anesthetic drug using an innovative one-step continuous synthesis method. A simple, versatile and highly productive microfluidic synthesis was developed for this purpose. The influence of the reaction conditions in the final MPs characteristics was analyzed. The presence of plasmonic HGNPs in the MPs was demonstrated as potential trigger for drug delivery applications in biomedicine. In order to test drug loading and release behavior, bupivacaine was selected as a drug of interest and the interaction, loading capacity and release studies confirmed the suitability of the MPs obtained for reaching tunable drug delivery profiles. Finally, cytotoxicity assays showed a subcytotoxic dose of 0.5 mg/mL further used to study the cell metabolism and cell cycle showing no remarkable influence in any of the cell lines studied. Altogether, make this microfluidic device suitable for MPs synthesis for further local treatment applications such as the treatment of chronic pain where an anesthetic depot could be externally activated to release its cargo on demand. Further improvements such as different capillary sizes, LED irradiation arrangements or introduction of other drug molecules or inorganic nanoparticles would be the next steps to expand the possibilities of this novel microfluidic device. 5 Acknowledgments Financial support from the ERC Consolidator Grant program (ERC-2013-CoG-614715, NANOHEDONISM) is gratefully acknowledged. CIBER-BBN is an initiative funded by the VI National R&D&i Plan 2008-2011 financed by the Instituto de Salud Carlos III with the assistance of the European Regional Development Fund. 6 Associated content Supporting figures are available in Supporting Information File. Additional information of the reversible MPs swelling under laser irradiation is presented in Supporting Movie_S1. 7 References (1) Choi, A.; Seo, K. D.; Kim, D. W.; Kim, B. C.; Kim, D. S. Recent Advances in Engineering Microparticles and Their Nascent Utilization in Biomedical Delivery and Diagnostic Applications. Lab Chip 2017, 17 (4), 591–613. (2) Jung, S.; Choi, C. H.; Lee, C. S.; Yi, H. Integrated Fabrication-Conjugation Methods for Polymeric and Hybrid Microparticles for Programmable Drug Delivery and Biosensing Applications. Biotechnol. J. 2016, 11 (12), 1561–1571. (3) Xu, Q.; Hashimoto, M.; Dang, T. T.; Hoare, T.; Kohane, D. S.; Whitesides, G. M.; Langer, R.; Anderson, D. G. Preparation of Monodisperse Biodegradable Polymer Microparticles Using a Microfluidic Flow-Focusing Device for Controlled Drug Delivery. Natl. Inst. Heal. Public Access 2010, 5 (13), 1575– 1581. (4) Jiang, W.; Li, M.; Chen, Z.; Leong, K. W. Cell-Laden Microfluidic Microgels for Tissue Regeneration. Lab Chip 2016, 16 (23), 4482–4506. (5) Berkland, C.; King, M.; Cox, A.; Kim, K.; Pack, D. W. Precise Control of PLG Microsphere Size Provides Enhanced Control of Drug Release Rate. J. Control. Release 2002, 82 (1), 137–147. (6) Agrawal, G.; Agrawal, R. Stimuli-Responsive Microgels and Microgel-Based Systems: Advances in the Exploitation of Microgel Colloidal Properties and Their Interfacial Activity. Polymers (Basel). 2018, 10 (4). (7) Chan, K. M. C.; Li, R. H.; Chapman, J. W.; Trac, E. M.; Kobler, J. B.; Zeitels, S. M.; Langer, R.; Karajanagi, S. S. Functionalizable Hydrogel Microparticles of Tunable Size and Stiffness for Soft-Tissue Filler Applications. Acta Biomater. 2014, 10 (6), 2563–2573. (8) Cirillo, G.; Iemma, F.; Spizzirri, U. G.; Puoci, F.; Curcio, M.; Parisi, O. I.; Picci, N. Synthesis of Stimuli-Responsive Microgels for in Vitro Release of Diclofenac Diethyl Ammonium. J. Biomater. Sci. Polym. Ed. 2011, 22 (4–6), 823–844. (9) Bazzano, M.; Latorre, D.; Pisano, R.; Sangermano, M.; Woerner, M. NanoStructured Polymeric Microparticles Produced via Cationic Aerosol Photopolymerization. J. Photochem. Photobiol. A Chem. 2017, 346, 364–371. (10) Vladisavljević, G. T.; Williams, R. A. Recent Developments in Manufacturing Emulsions and Particulate Products Using Membranes. Adv. Colloid Interface Sci. 2005, 113 (1), 1–20. (11) Charcosset;, C.; Fessi, H. MEMBRANE EMULSIFICATION AND MICROCHANNEL EMULSIFICATION PROCESSES. Rev. Chem. Eng. 2005, 21, 1–32. (12) De Geest, B. G.; Urbanski, J. P.; Thorsen, T.; Demeester, J.; De Smedt, S. C. Synthesis of Monodisperse Biodegradable Microgels in Microfluidic Devices. Langmuir 2005, 21 (23), 10275–10279. (13) Choi, C. H.; Weitz, D. A.; Lee, C. S. One Step Formation of Controllable Complex Emulsions: From Functional Particles to Simultaneous Encapsulation of Hydrophilic and Hydrophobic Agents into Desired Position. Adv. Mater. 2013, 25 (18), 2536–2541. (14) Ma, S.; Thiele, J.; Liu, X.; Bai, Y.; Abell, C.; Huck, W. T. S. Fabrication of Microgel Particles with Complex Shape via Selective Polymerization of Aqueous Two-Phase Systems. Small 2012, 8 (15), 2356–2360. (15) Nosrati, Z.; Li, N.; Michaud, F.; Ranamukhaarachchi, S.; Karagiozov, S.; Soulez, G.; Martel, S.; Saatchi, K.; Häfeli, U. O. Development of a Coflowing Device for the Size-Controlled Preparation of Magnetic-Polymeric Microspheres as Embolization Agents in Magnetic Resonance Navigation Technology. ACS Biomater. Sci. Eng. 2018, 4 (3), 1092–1102. (16) Xu, J. H.; Zhao, H.; Lan, W. J.; Luo, G. S. A Novel Microfluidic Approach for Monodispersed Chitosan Microspheres with Controllable Structures. Adv. Healthc. Mater. 2012, 1 (1), 106–111. (17) Sebastian, V.; Arruebo, M. Microfluidic Production of Inorganic Nanomaterials for Biomedical Applications; Elsevier Inc., 2018. (18) Dong, H.; Tang, G.; Ma, T.; Cao, X. One-Step Fabrication of Inorganic/Organic Hybrid Microspheres with Tunable Surface Texture for Controlled Drug Release Application. J. Mater. Sci. Mater. Med. 2016, 27 (1), 7. (19) Seo, K. D.; Kim, D. S.; Sánchez, S. Fabrication and Applications of ComplexShaped Microparticles via Microfluidics. Lab Chip 2015, 15 (18), 3622–3626. (20) Choi, C.-H.; Jung, J.-H.; Kim, D.-W.; Chung, Y.-M.; Lee, C.-S. Novel One-Pot Route to Monodisperse Thermosensitive Hollow Microcapsules in a Microfluidic System. Lab Chip 2008, 8 (9), 1544–1551. (21) Kanai, T.; Ohtani, K.; Fukuyama, M.; Katakura, T.; Hayakawa, M. Preparation of Monodisperse PNIPAM Gel Particles in a Microfluidic Device Fabricated by Stereolithography. Polym. J. 2011, 43 (12), 987–990. (22) Xi, W.; Kong, F.; Yeo, J. C.; Yu, L.; Sonam, S.; Dao, M.; Gong, X.; Lim, C. T. Soft Tubular Microfluidics for 2D and 3D Applications. Proc. Natl. Acad. Sci. 2017, 114 (40), 10590–10595. (23) Ren, K.; Dai, W.; Zhou, J.; Su, J.; Wu, H. Whole-Teflon Microfluidic Chips. Proc. Natl. Acad. Sci. 2011, 108 (20), 8162–8166. (24) Nge, P. N.; Rogers, C. I.; Woolley, A. T. Advances in Microfluidic Materials, Functions, Integration and Applications. Chem. Rev. 2013, 113 (4), 2550–2583. (25) Hu, Y.; Wang, S.; Abbaspourrad, A.; Ardekani, A. M. Fabrication of Shape Controllable Janus Alginate/PNIPAAm Microgels via Microfluidics Technique and Off-Chip Ionic Cross-Linking. Langmuir 2015, 31 (6), 1885–1891. (26) Velasco, D.; Tumarkin, E.; Kumacheva, E. Microfluidic Encapsulation of Cells in Polymer Microgels. Small 2012, 8 (11), 1633–1642. (27) Seiffert, S.; Weitz, D. A. Controlled Fabrication of Polymer Microgels by Polymer-Analogous Gelation in Droplet Microfluidics. Soft Matter 2010, 6 (14), 3184. (28) Li, W.; Zhang, L.; Ge, X.; Xu, B.; Zhang, W.; Qu, L.; Choi, C. H.; Xu, J.; Zhang, A.; Lee, H.; Weitz, D. A. Microfluidic Fabrication of Microparticles for Biomedical Applications. Chem. Soc. Rev. 2018, 47 (15), 5646–5683. (29) Dietlin, C.; Schweizer, S.; Xiao, P.; Zhang, J.; Morlet-Savary, F.; Graff, B.; Fouassier, J. P.; Lalevée, J. Photopolymerization upon LEDs: New Photoinitiating Systems and Strategies. Polym. Chem. 2015, 6 (21), 3895–3912. (30) Sivakumaran, D.; Mueller, E.; Hoare, T. Microfluidic Production of Degradable Thermoresponsive Poly(: N -Isopropylacrylamide)-Based Microgels. Soft Matter 2017, 13 (47), 9060–9070. (31) Kim, Y. S.; Lee, H. M.; Kim, J. H.; Joo, J.; Cheong, I. W. Hydrogel Adsorbents of Poly(N-Isopropylacrylamide-Co-Methacryloyloxymethyl-12-Crown-4) for Li+recovery Prepared by Droplet Microfluidics. RSC Adv. 2015, 5 (14), 10656– 10661. (32) Cheng, C. J.; Chu, L. Y.; Ren, P. W.; Zhang, J.; Hu, L. Preparation of Monodisperse Thermo-Sensitive Poly(N-Isopropylacrylamide) Hollow Microcapsules. J. Colloid Interface Sci. 2007, 313 (2), 383–388. (33) Ge, X. H.; Huang, J. P.; Xu, J. H.; Luo, G. S. Controlled Stimulation-Burst Targeted Release by Smart Decentered Core-Shell Microcapsules in Gravity and Magnetic Field. Lab Chip 2014, 14 (23), 4451–4454. (34) Mou, C. L.; Ju, X. J.; Zhang, L.; Xie, R.; Wang, W.; Deng, N. N.; Wei, J.; Chen, Q.; Chu, L. Y. Monodisperse and Fast-Responsive Poly(N -Isopropylacrylamide) Microgels with Open-Celled Porous Structure. Langmuir 2014, 30 (5), 1455– 1464. (35) Park, S.; Kim, D.; Ko, S. Y.; Park, J. O.; Akella, S.; Xu, B.; Zhang, Y.; Fraden, S. Controlling Uniformity of Photopolymerized Microscopic Hydrogels. Lab Chip 2014, 14 (9), 1551–1563. (36) Filatov, N. A.; Nozdriukhin, D. V.; Bukatin, A. S. The Kinetic Study of Solidification PEGDA Microparticles in Flow-Focusing Microfluidic Chip. J. Phys. Conf. Ser. 2017, 917 (4), 1–5. (37) Chen, J.; Huang, K.; Chen, Q.; Deng, C.; Zhang, J.; Zhong, Z. Tailor-Making Fluorescent Hyaluronic Acid Microgels via Combining Microfluidics and Photoclick Chemistry for Sustained and Localized Delivery of Herceptin in Tumors. ACS Appl. Mater. Interfaces 2018, 10 (4), 3929–3937. (38) Jeong, W. C.; Kim, S. H.; Yang, S. M. Photothermal Control of Membrane Permeability of Microcapsules for On-Demand Release. ACS Appl. Mater. Interfaces 2014, 6 (2), 826–832. (39) Homayun, B.; Sun, C.; Kumar, A.; Montemagno, C.; Choi, H. J. Facile Fabrication of Microparticles with PH-Responsive Macropores for Small Intestine Targeted Drug Formulation. Eur. J. Pharm. Biopharm. 2018, 128 (March), 316–326. (40) Huang, S.; Lin, B.; Qin, J. Microfluidic Synthesis of Tunable Poly-(NIsopropylacrylamide) Microparticles via PEG Adjustment. Electrophoresis 2011, 32 (23), 3364–3370. (41) Zhang, K.; Wu, W.; Guo, K.; Chen, J.; Zhang, P. Synthesis of TemperatureResponsive Poly(N-Isopropyl Acrylamide)/ Poly(Methyl Methacrvlate)/Silica Hybrid Capsules from Inverse Pickering Emulsion Polymerization and Their Application in Controlled Drug Release. Langmuir 2010, 26 (11), 7971–7980. (42) Budhlall, B. M.; Marquez, M.; Velev, O. D. Microwave, PhotoAnd Thermally Responsive PNIPAm-Gold Nanoparticle Microgels. Langmuir 2008, 24 (20), 11959–11966. (43) Heida, T.; Neubauer, J. W.; Seuss, M.; Hauck, N.; Thiele, J.; Fery, A. Mechanically Defined Microgels by Droplet Microfluidics. Macromol. Chem. Phys. 2017, 218 (2), 1–19. (44) Ekici, S.; Ilgin, P.; Yilmaz, S.; Aktas, N.; Sahiner, N. Temperature and Magnetic Field Responsive Hyaluronic Acid Particles with Tunable Physical and Chemical Properties. Appl. Surf. Sci. 2011, 257 (7), 2669–2676. (45) Seiffert, S.; Thiele, J.; Abate, A. R.; Weitz, D. A. Smart Microgel Capsules from Macromolecular Precursors. 2010, 01238 (14), 1–9. (46) Akbari, S.; Pirbodaghi, T.; Kamm, R. D.; Hammond, P. T. A Versatile Microfluidic Device for High Throughput Production of Microparticles and Cell Microencapsulation. Lab Chip 2017, 17 (12), 2067–2075. (47) Steinhilber, D.; Rossow, T.; Wedepohl, S.; Paulus, F.; Seiffert, S.; Haag, R. A Microgel Construction Kit for Bioorthogonal Encapsulation and PH-Controlled Release of Living Cells. Angew. Chemie - Int. Ed. 2013, 52 (51), 13538–13543. (48) Celetti, G.; Natale, C. Di; Causa, F.; Battista, E.; Netti, P. A. Functionalized Poly(Ethylene Glycol) Diacrylate Microgels by Microfluidics: In Situ Peptide Encapsulation for in Serum Selective Protein Detection. Colloids Surfaces B Biointerfaces 2016, 145, 21–29. (49) Wang, W.; Liu, L.; Ju, X. J.; Zerrouki, D.; Xie, R.; Yang, L.; Chu, L. Y. A Novel Thermo-Induced Self-Bursting Microcapsule with Magnetictargeting Property. ChemPhysChem 2009, 10 (14), 2405–2409. (50) Wang, J.; Zhu, X.; Wei, L.; Ye, Y.; Liu, Y.; Li, J.; Mei, T.; Wang, X.; Wang, L. Controlled Shape Transformation and Loading Release of Smart Hemispherical Hybrid Microgels Triggered by ‘Inner Engines.’ ChemistrySelect 2018, 3 (15), 4067–4074. (51) Kim, B.; Soo Lee, H.; Kim, J.; Kim, S. H. Microfluidic Fabrication of PhotoResponsive Hydrogel Capsules. Chem. Commun. 2013, 49 (18), 1865–1867.