Functionalized Poly(N-isopropylacrylamide)-Based Microgels in Tumor Targeting and Drug Delivery

Over the past several decades, the development of engineered small particles as targeted and drug delivery systems (TDDS) has received great attention thanks to the possibility to overcome the limitations of classical cancer chemotherapy, including targeting incapability, nonspecific action and, consequently, systemic toxicity. Thus, this research aims at using a novel design of Poly(N-isopropylacrylamide) p(NIPAM)-based microgels to specifically target cancer cells and avoid the healthy ones, which is expected to decrease or eliminate the side effects of chemotherapeutic drugs. Smart NIPAM-based microgels were functionalized with acrylic acid and coupled to folic acid (FA), targeting the folate receptors overexpressed by cancer cells and to the chemotherapeutic drug doxorubicin (Dox). The successful conjugation of FA and Dox was demonstrated by dynamic light scattering (DLS), Fourier-transform infrared (FTIR) spectroscopy, thermogravimetric analysis (TGA), UV-VIS analysis, and differential scanning calorimetry (DSC). Furthermore, viability assay performed on cancer and healthy breast cells, suggested the microgels’ biocompatibility and the cytotoxic effect of the conjugated drug. On the other hand, the specific tumor targeting of synthetized microgels was demonstrated by a co-cultured (healthy and cancer cells) assay monitored using confocal microscopy and flow cytometry. Results suggest successful targeting of cancer cells and drug release. These data support the use of pNIPAM-based microgels as good candidates as TDDS.


Introduction
Cancer is one of the leading causes of death in the world. In 2020, the world health organization stated that the number of deaths caused by cancer reached ten million deaths worldwide [data from WHO] [1]. One of the most commonly used therapies is chemotherapy, which is delivered systematically in a non-targeted manner [2]. Over the past several decades, the development of engineered nano-and micro-systems for targeted drug delivery have received great attention thanks to their possibility to overcome the limitations of classical cancer chemotherapy, including poor solubility, targeting incapability, nonspecific action and, consequently, systemic toxicity [3,4]. For instance, the anticancer drug Gels 2021, 7, 203 3 of 17 different concentrations of p (NIPAM)-co-5%AA, as well as p (NIPAM)-co-5%AA-co-FA and the cytotoxic effect of p (NIPAM)-co-5%AA-co-FA-co-Dox were tested. Finally, the specific tumor targeting experiments that test the suggested targeting behavior of the particles qualitatively and quantitatively were carried out. These are confocal microscopy and flow cytometry.

Synthesis of p(NIPAM)-co-5%AA Microgels and Conjugation with Folic Acid and Doxorubicin
A sequential synthesis and conjugation processes were performed to generate microgel particles decorated with the targeting molecule folic acid and the anticancer drug doxorubicin. p(NIPAM)-co-5%AA were synthesized by Surfactant Free Emulsion Polymerisation (SFEP) technique as described in materials and methods to avoid toxic surfactant contamination [28,29]. Successively, EDC-NHS protocol was adopted to first bind folic acid to some of the acrylic acids of p(NIPAM)-co-5%AA microgels and then doxorubicin to the remaining acrylic acid residues. The success of the protocol was demonstrated by the UV-VIS analysis in which it was evident the characteristic peak of folic acid (340 mm) on p(NIPAM)-co-5%AA-co-FA and both folic acid and doxorubicin (485 nm) peaks on p(NIPAM)-co-5%AA-co-FA-co-Dox ( Figure 1). The amount of folic acid and doxorubicin conjugated was calculated by spectrophotometric analysis using the standard calibration curves (Figures S1 and S2).
ning calorimetry (DSC), and dynamic light scattering (DLS) to calculate the size distribution against temperature change. This is in addition to thermogravimetric analysis (TGA) and Fourier-transform infrared spectroscopy (FTIR) as confirmation of successful coupling reaction of EDC/NHS with each stage of folic acid conjugation and Dox conjugation. The cell biocompatibility of different concentrations of p (NIPAM)-co-5%AA, as well as p (NIPAM)-co-5%AA-co-FA and the cytotoxic effect of p (NIPAM)-co-5%AA-co-FA-co-Dox were tested. Finally, the specific tumor targeting experiments that test the suggested targeting behavior of the particles qualitatively and quantitatively were carried out. These are confocal microscopy and flow cytometry.

Synthesis of p(NIPAM)-co-5%AA Microgels and Conjugation with Folic Acid and Doxorubicin
A sequential synthesis and conjugation processes were performed to generate microgel particles decorated with the targeting molecule folic acid and the anticancer drug doxorubicin. p(NIPAM)-co-5%AA were synthesized by Surfactant Free Emulsion Polymerisation (SFEP) technique as described in materials and methods to avoid toxic surfactant contamination [28,29]. Successively, EDC-NHS protocol was adopted to first bind folic acid to some of the acrylic acids of p(NIPAM)-co-5%AA microgels and then doxorubicin to the remaining acrylic acid residues. The success of the protocol was demonstrated by the UV-VIS analysis in which it was evident the characteristic peak of folic acid (340 mm) on p(NIPAM)-co-5%AA-co-FA and both folic acid and doxorubicin (485 nm) peaks on p(NIPAM)-co-5%AA-co-FA-co-Dox ( Figure 1). The amount of folic acid and doxorubicin conjugated was calculated by spectrophotometric analysis using the standard calibration curves ( Figure S1 and S2).

Size of Microgels
The effect of temperature change on the size of p(NIPAM)-co-5%AA, p(NIPAM)co-5%AA-co-FA, and p(NIPAM)-co-5%AA-co-FA-co-DOX was studied by dynamic light scattering analysis (DLS) (Figure 2 and Figure S3). The size of the three microgel particles showed typical microgel behavior [30]. Below the VPTT (volume phase transition temperature) (34 ºC), the particles were swollen and configure a large size. At 34 • C (VPTT), the three microgels underwent a sharp decrease in size as the hydrogen bonds between the polymer particles and water molecules break due to energy gained under higher temperature [30,31], causing the polymer-polymer interactions to dominate. Hence, water molecules were expelled from microgel particles, causing the microgel to collapse and deswell [30,32].
At 15 • C, p(NIPAM)-co-5%AA had an average diameter of 701 nm while that of p(NIPAM)-co-5%AA-co-FA had an average diameter of 451 nm particle size. This was because FA, being a large molecule with several hydrophobic aromatic moieties, tended to decrease the hydrophilicity of the particle and decrease the hydrogen bonding with water molecules, hence it contained less water than that of the AA one. Further conjugating the Gels 2021, 7, 203 4 of 17 particles with Dox molecules had increased the length of the hairy layers, hence causing the particle to increase in size at an average diameter of 1500 nm. Doxorubicin, being another bulky molecule with several hydrophilic groups, had helped the microgels to swell and reach the micro-scale. ature [30,31], causing the polymer-polymer interactions to dominate. Hence, water molecules were expelled from microgel particles, causing the microgel to collapse and deswell [30,32]. At 15 °C, p(NIPAM)-co-5%AA had an average diameter of 701 nm while that of p(NIPAM)-co-5%AA-co-FA had an average diameter of 451 nm particle size. This was because FA, being a large molecule with several hydrophobic aromatic moieties, tended to decrease the hydrophilicity of the particle and decrease the hydrogen bonding with water molecules, hence it contained less water than that of the AA one. Further conjugating the particles with Dox molecules had increased the length of the hairy layers, hence causing the particle to increase in size at an average diameter of 1500 nm. Doxorubicin, being another bulky molecule with several hydrophilic groups, had helped the microgels to swell and reach the micro-scale.
Attaching FA to the microgel particles, the microgel's VPTT was unaffected but the size of the microgel was reduced even further. In the case of p(NIPAM)-co-5%AA-co-FAco-DOX, a rapid and sharp decrease in size was observed. At 50 ºC, particles of the three microgels p(NIPAM)-co-5%AA, p(NIPAM)-co-5%AA-co-FA, and p(NIPAM)-co-5%AAco-FA-co-Dox were deswollen to an average size of 247, 177, and 433 nm, respectively. Moreover, calculating the deswelling degrees between the minimum and maximum temperature is rather challenging. This is because the size of p(NIPAM)-co-5%AA-co-FA, tends to fluctuate greatly from 300 to 504 nm, then dropping back again to 400. The decrease in size of p(NIPAM)-co-5%AA-co-FA in comparison to p(NIPAM)-co-5% AA is due to the decreased hydrophobicity of the particles because of the hydrophobic rings in the molecular structure of folic acid. The hydrophobic structure of the molecule decreases the hydrogen bonding between the particle and water and hence decreases the amount of water entrapped within the particles. After adding Dox with a complex structure and large molecules, the particle size tends to increase due to elongated hairy structures [30].
It is worth mentioning that the overall PDI (polydispersity index) of p(NIPAM)-co-5%AA was 0.057, which indicated the highly satisfactory consistency between particle size and distribution. Attaching FA molecules to the above-mentioned microgels decreased this consistency and increased the overall PDI to reach 0.503, which was fairly satisfactory. However, the conjugation of the bulky Dox molecules had increased the overall PDI to 0.833. The reason for this increase in PDI was the fact that Dox is a bulky molecule. When Dox is chemically conjugated to p(NIPAM)-co-5%AA-co-FA, it can either attach to FA Attaching FA to the microgel particles, the microgel's VPTT was unaffected but the size of the microgel was reduced even further. In the case of p(NIPAM)-co-5%AA-co-FAco-DOX, a rapid and sharp decrease in size was observed. At 50 ºC, particles of the three microgels p(NIPAM)-co-5%AA, p(NIPAM)-co-5%AA-co-FA, and p(NIPAM)-co-5%AAco-FA-co-Dox were deswollen to an average size of 247, 177, and 433 nm, respectively. Moreover, calculating the deswelling degrees between the minimum and maximum temperature is rather challenging. This is because the size of p(NIPAM)-co-5%AA-co-FA, tends to fluctuate greatly from 300 to 504 nm, then dropping back again to 400. The decrease in size of p(NIPAM)-co-5%AA-co-FA in comparison to p(NIPAM)-co-5% AA is due to the decreased hydrophobicity of the particles because of the hydrophobic rings in the molecular structure of folic acid. The hydrophobic structure of the molecule decreases the hydrogen bonding between the particle and water and hence decreases the amount of water entrapped within the particles. After adding Dox with a complex structure and large molecules, the particle size tends to increase due to elongated hairy structures [30].
It is worth mentioning that the overall PDI (polydispersity index) of p(NIPAM)-co-5%AA was 0.057, which indicated the highly satisfactory consistency between particle size and distribution. Attaching FA molecules to the above-mentioned microgels decreased this consistency and increased the overall PDI to reach 0.503, which was fairly satisfactory. However, the conjugation of the bulky Dox molecules had increased the overall PDI to 0.833. The reason for this increase in PDI was the fact that Dox is a bulky molecule. When Dox is chemically conjugated to p(NIPAM)-co-5%AA-co-FA, it can either attach to FA moiety or to the unreacted AA, which gives the microgel versatility to have free end FA moiety on the surface of the microgel to target the folate receptor.

Electrophoretic Mobility
Electrophoretic mobility (Em) of microgel particles is mainly affected by three factors: the size of microgels, solvent viscosity, and dielectric constant [33]. The latter two factors are needed to be kept at a minimum to measure the Em of microgel particles accurately across the temperature range, hence the usage of DI water as the dispersant [34]. The three microgels, p(NIPAM)-co-5%AA, p(NIPAM)-co-5%AA-co-FA, and p(NIPAM)-co-5%AA- Electrophoretic mobility (Em) of microgel particles is mainly affected by three factors: the size of microgels, solvent viscosity, and dielectric constant [33]. The latter two factors are needed to be kept at a minimum to measure the Em of microgel particles accurately across the temperature range, hence the usage of DI water as the dispersant [34]. The three microgels, p(NIPAM)-co-5%AA, p(NIPAM)-co-5%AA-co-FA, and p(NIPAM)-co-5%AAco-FA-co-Dox, showed an increase in their magnitude of Em (|Em|) as the temperature increased from 15-60 ºC (Figures 3 and S4). At 15 °C, p(NIPAM)-co-5%AA had a negative electrophoretic mobility of average -0.946 µmcm/Vs. While that of p(NIPAM)-co-5%AA-co-FA average Em is −0.401 µmcm/Vs, which showed that conjugating p(NIPAM)-co-5%AA to FA resulted in a decrease in its Em. p(NIPAM)-co-5%AA-co-FA-co-Dox had an average Em of -0.0364 µmcm/Vs, this was due to the positive charge density of Dox, as well as, the bulky structure of the particle that causes the negative charges from the sulphate ions to be masked [33,35].
At 37 °C, the particle size dramatically decreased, which causes an increase in the surface charge density, hence an increase in electrophoretic mobility. In the case of p(NIPAM)-co-5%AA, the increase in electrophoretic mobility around VPTT was sharp. This was because the negative charges were exposed, while in case of p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox, it was suggested that the complex structure of the particle had masked some of the charges causing the increase in Em to be steep.

Thermogravimetric Analysis (TGA)
TGA (Thermogravimetric Analysis) analysis in Figure 4 shows the thermostability of microgel particles, in terms of mass percentage retained against temperature under ambient atmosphere. p(NIPAM) was thermally stable up till 250 °C where afterwards it started to decrease in mass. This was because the microgel gets burnt in the presence of oxygen At 15 • C, p(NIPAM)-co-5%AA had a negative electrophoretic mobility of average -0.946 µmcm/Vs. While that of p(NIPAM)-co-5%AA-co-FA average E m is −0.401 µmcm/Vs, which showed that conjugating p(NIPAM)-co-5%AA to FA resulted in a decrease in its E m. p(NIPAM)-co-5%AA-co-FA-co-Dox had an average E m of -0.0364 µmcm/Vs, this was due to the positive charge density of Dox, as well as, the bulky structure of the particle that causes the negative charges from the sulphate ions to be masked [33,35].
At 37 • C, the particle size dramatically decreased, which causes an increase in the surface charge density, hence an increase in electrophoretic mobility. In the case of p(NIPAM)co-5%AA, the increase in electrophoretic mobility around VPTT was sharp. This was because the negative charges were exposed, while in case of p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox, it was suggested that the complex structure of the particle had masked some of the charges causing the increase in E m to be steep.

Thermogravimetric Analysis (TGA)
TGA (Thermogravimetric Analysis) analysis in Figure 4 shows the thermostability of microgel particles, in terms of mass percentage retained against temperature under ambient atmosphere. p(NIPAM) was thermally stable up till 250 • C where afterwards it started to decrease in mass. This was because the microgel gets burnt in the presence of oxygen until it reached a plateau at 400 • C and p(NIPAM) was turned to ashes (which is the remaining mass). p(NIPAM)-co-5%AA experienced a similar sigmoid curve as plain p(NIPAM), but showed higher thermal stability as it decreased in mass at 290 • C and reached a plateau at 440 • C. p(NIPAM)-co-5%AA-co-FA, and p(NIPAM)-co-5%AA-co-FA-co-Dox exhibited similar behavior in thermal stability to one another. The steady decrease in mass over a wide range of temperatures indicates that FA led to an increase in thermal stability and slow decomposition for the p(NIPAM) microgels, this was due to the chemical conjugation of FA to p(NIPAM)-co-5%AA. FA is a thermal stable moiety and degrades slowly at high temperatures, and as such, FA sustained p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox microgels up to 40% of their masses at 600 • C [36]. It can then be concluded that FA had been chemically conjugated to p(NIPAM)-co-5%AA microgels due to the high thermal stability. FA to p(NIPAM)-co-5%AA. FA is a thermal stable moiety and degrades slowly at high temperatures, and as such, FA sustained p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox microgels up to 40% of their masses at 600 °C [36]. It can then be concluded that FA had been chemically conjugated to p(NIPAM)-co-5%AA microgels due to the high thermal stability.

Differential Scanning Calorimetry (DSC)
The thermal behavior of p(NIPAM) and p(NIPAM)-co-5%AA undergoes two stages, these are melting of crystallization (micro-melting) and the melting point of the sample. Differential Scanning Calorimetry (DSC) showed that the first stage melting of crystallization occurs at 116 °C for p(NIPAM), while it occurred further in p(NIPAM)-co-5%AA at 153 °C. Furthermore, a series of endothermic peaks at 411 °C for p(NIPAM) and 404 °C for p(NIPAM)-co-5%AA indicating their melting points was registered. Further heating exhibited two-step exothermic peaks for p(NIPAM), but one for p(NIPAM)-co-5%AA. p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox exhibited a lower crystallization melting point at an endothermic peak of 116 °C for p(NIPAM)-co-5%AA-co-FA and 131 °C for p(NIPAM)-co-5%AA-co-FA-co-Dox. Melting points of p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox were 154 and 145 °C, respectively and did not exhibit any exothermic peaks like the other two microgels ( Figure 5). This indicates that the change in thermal behavior in p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox was due to the moieties that were chemically conjugated to p(NIPAM)-co-5%AA. Moreover, the existence of only one melting point in each p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox indicated the purity of the sample and that nothing else was co-existing with these microgels.

Differential Scanning Calorimetry (DSC)
The thermal behavior of p(NIPAM) and p(NIPAM)-co-5%AA undergoes two stages, these are melting of crystallization (micro-melting) and the melting point of the sample. Differential Scanning Calorimetry (DSC) showed that the first stage melting of crystallization occurs at 116 • C for p(NIPAM), while it occurred further in p(NIPAM)-co-5%AA at 153 • C. Furthermore, a series of endothermic peaks at 411 • C for p(NIPAM) and 404 • C for p(NIPAM)-co-5%AA indicating their melting points was registered. Further heating exhibited two-step exothermic peaks for p(NIPAM), but one for p(NIPAM)-co-5%AA. p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox exhibited a lower crystallization melting point at an endothermic peak of 116 • C for p(NIPAM)-co-5%AAco-FA and 131 • C for p(NIPAM)-co-5%AA-co-FA-co-Dox. Melting points of p(NIPAM)co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox were 154 and 145 • C, respectively and did not exhibit any exothermic peaks like the other two microgels ( Figure 5). This indicates that the change in thermal behavior in p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox was due to the moieties that were chemically conjugated to p(NIPAM)-co-5%AA. Moreover, the existence of only one melting point in each p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox indicated the purity of the sample and that nothing else was co-existing with these microgels.

Fourier Transform Infra-Red Spectroscopy (FTIR)
The FTIR (Fourier Transform Infra-Red Spectroscopy) spectra of the three microgels The FTIR (Fourier Transform Infra-Red Spectroscopy) spectra of the three microgels are shown in Figure 6, while the peaks and their assignments are mentioned in Table 1. The FTIR of p(NIPAM)-co-5%AA showed a peak at 3417 cm −1 of the hydroxyl group of the carboxylic acid and the C=O in the carboxylic acid group. The sulphate ions were expressed at 1130 cm −1 . It was also worth noting that some peaks that were available in p(NIPAM) were shifted in p(NIPAM)-co-5%AA, these include 3283, 2972, 2933, 2876, 1632, 1538, 1457, and 1386 cm −1 . Figure 5. The DSC graph of p(NIPAM), p(NIPAM)-co-5%AA, p(NIPAM)-co-5%AA-co-FA, and p(NIPAM)-co-5%AA-co-FA-co-Dox under nitrogen atmosphere at a temperature range of rtp-600 °C.

Fourier Transform Infra-Red Spectroscopy (FTIR)
The FTIR (Fourier Transform Infra-Red Spectroscopy) spectra of the three microgels are shown in Figure 6, while the peaks and their assignments are mentioned in Table 1. The FTIR of p(NIPAM)-co-5%AA showed a peak at 3417 cm −1 of the hydroxyl group of the carboxylic acid and the C=O in the carboxylic acid group. The sulphate ions were expressed at 1130 cm −1 . It was also worth noting that some peaks that were available in p(NIPAM) were shifted in p(NIPAM)-co-5%AA, these include 3283, 2972, 2933, 2876, 1632, 1538, 1457, and 1386 cm −1 . Finally, p(NIPAM)-co-5%AA-co-FA-co-Dox had few additional functional groups that are expressed exclusively for Dox in its spectrum including the 13C-H and COH stretch of Dox occurring at 1377 and 1209 cm −1 , which are very unique to Dox [37].
The FTIR results of p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox, and the shift in wavenumbers that were observed in the spectra ( Figure 5 and Table Finally, p(NIPAM)-co-5%AA-co-FA-co-Dox had few additional functional groups that are expressed exclusively for Dox in its spectrum including the 13C-H and COH stretch of Dox occurring at 1377 and 1209 cm −1 , which are very unique to Dox [37].
The FTIR results of p(NIPAM)-co-5%AA-co-FA and p(NIPAM)-co-5%AA-co-FA-co-Dox, and the shift in wavenumbers that were observed in the spectra ( Figure 5 and Table 1) were confirmatory results that the moieties were chemically conjugated and that FA and Dox were not ionically interacting with the p(NIPAM)-co-5%AA microgels, as it would have diffused out through the dialysis step.

Biocompatibility of p(NIPAM)-co-5%AA and p(NIPAM)-co-5%AA-co-FA Microsystems
Viability assay was initially performed on cells treated with microgels without any anticancer drug conjugated, used as a control, in order to verify their biocompatibility. Therefore, CCK-8 (Cell counting kit-8) assay was performed on normal (HB2) and tumor (MDA-MB 231) cells treated for 24h with different concentrations (15,31,46,62,77, and 93 µg/mL) of p(NIPAM)-co-5%AA or p(NIPAM)-co-5%AA-co-FA microgels. Cells treated with doxorubicin (5, 10. 15, 20, 25, and 30 µM) were used as positive control. As reported in Figure 7, microgel particles alone or conjugated with folic acid do not alter the cell viability of both normal and tumor cells (viability of around 100%), also if used at high concentration (92.88 µg/mL). Furthermore, cell viability was also maintained at a higher concentration (100 µg/mL) of p(NIPA)-co-5%AA until 48 h of treatment and the cell morphology was not altered as suggested by acridine orange assay (Figures S5 and S6).

Qualitative Uptake of p(NIPAM)-co-5%AA-co-FA-co-Dox
Once established the biocompatibility of microgels, cell internalization uptake was initially investigated by fluorescence microscopy by incubating MDA-MB 231 cells with a fluorescence variant of microgel particles over time, as reported in supporting information (Figures S7 and S8). The green fluorescence relative to microgels appeared localized in specific areas, probably corresponding to the Golgi apparatus or the endoplasmic reticulum after 1 h of incubation ( Figures S7c-c" and S8). Table 1. FTIR peaks of p(NIPAM), p(NIPAM)-co-5%AA, p(NIPAM)-co-5%AA-co-FA, and p(NIPAM)co-5%AA-co-FA-co-Dox and their assignments with references. in Figure 7, microgel particles alone or conjugated with folic acid do not alter the cell viability of both normal and tumor cells (viability of around 100%), also if used at high concentration (92.88 µg/mL). Furthermore, cell viability was also maintained at a higher concentration (100µg/mL) of p(NIPA)-co-5%AA until 48 h of treatment and the cell morphology was not altered as suggested by acridine orange assay (Figures S5 and S6).

Qualitative Uptake of p(NIPAM)-co-5%AA-co-FA-co-Dox
Once established the biocompatibility of microgels, cell internalization uptake was initially investigated by fluorescence microscopy by incubating MDA-MB 231 cells with a fluorescence variant of microgel particles over time, as reported in supporting information ( Figure S7 and S8). The green fluorescence relative to microgels appeared localized in specific areas, probably corresponding to the Golgi apparatus or the endoplasmic reticulum after 1 h of incubation ( Figures S7 c-c'' and S8).
Fluorescence microscopy was also adopted to investigate the specific tumor targeting of microgels functionalized with folic acid. The folate receptor (FR) is overexpressed in the majority of human tumors, like breast, and, in particular, MDA-MB231 cells produce high FR concentration [40]. Therefore, a co-culture experiment was performed by seeding HB2 and green-labelled MDA-MB 231 cells together and incubating them with p(NIPAM)-co-5%AA-co-FA-co-Dox microgels or doxorubicin alone as control (identified Fluorescence microscopy was also adopted to investigate the specific tumor targeting of microgels functionalized with folic acid. The folate receptor (FR) is overexpressed in the majority of human tumors, like breast, and, in particular, MDA-MB231 cells produce high FR concentration [40]. Therefore, a co-culture experiment was performed by seeding HB2 and green-labelled MDA-MB 231 cells together and incubating them with p(NIPAM)-co-5%AA-co-FA-co-Dox microgels or doxorubicin alone as control (identified by the doxorubicin red auto-fluorescence, Figures 8c-c"' and S9c-c"'). Nuclei of both cells were stained with DAPI (blue fluorescence, Figures 8a-a"' and S9a-a"') so that HB2 healthy cells were identified by blue fluorescence alone, while MDA-MB 231 tumor cells were individuated by both blue and green fluorescence. Following the microgel particles cellular uptake over time, it was evident the presence of the red fluorescence (corresponding to doxorubicin conjugated to the particles) exclusively in tumor cells already at the shortest incubation time (30 min, Figure 8a-d) and more and more at the following incubation times (1, 2, and 4 h, Figure 8a'-d',a"-d",a"'-d"'). On the other hand, red fluorescence was totally absent in correspondence of HB2 cells (white arrows in Figure 8d-d"), suggesting a specific tumor targeting of p(NIPAM)-co-5%AA-co-FA-co-Dox microparticles. The red fluorescence, relative to doxorubicin, began to appear in HB-2 cytoplasm in 4 h, as expected by static in vitro system. On the contrary, the soluble form of the doxorubicin was inside both normal and tumor cells already after 30 min of treatment, suggesting that microparticles, conjugated with folic acid, were responsible for the selectively for cancer cells ( Figure S9). The co-localization of the blue (nuclei) and the red (doxorubicin) fluorescence in tumor cells (Figure 8) suggested that the drug was released from the microgels and entered into the nuclei, which can intercalate into the DNA causing cell death. On the other hand, microgels fluorescence signal was always localized in the cytoplasm (Figures S7 and S8).

Quantitative Uptake Study
Differential microgel particles cellular uptake between normal and tumor cells was furthermore investigated by the quantitative flow cytometric analysis, following the red autofluorescence of conjugated doxorubicin (Figures 9 and S10). Initially (30 min), there were no significant differences in p(NIPAM)-co-5%AA-co-FA-co-Dox internalization between HB2 (breast healthy cells) and MDA-MB-231 (breast cancer cells). After 1 h of incubation, the uptake gap started to increase, suggesting a specific tumor targeting due to the conjugated folic acid, reaching the maximum value after 4 h of treatment: the microgels internalization in tumor cells was 60% against the 14% of internalization into normal cells.
bicin was inside both normal and tumor cells already after 30 min of treatment, suggesting that microparticles, conjugated with folic acid, were responsible for the selectively for cancer cells ( Figure S9). The co-localization of the blue (nuclei) and the red (doxorubicin) fluorescence in tumor cells (Figure 8) suggested that the drug was released from the microgels and entered into the nuclei, which can intercalate into the DNA causing cell death. On the other hand, microgels fluorescence signal was always localized in the cytoplasm (Figures S7 and S8).

Quantitative Uptake Study
Differential microgel particles cellular uptake between normal and tumor cells was furthermore investigated by the quantitative flow cytometric analysis, following the red autofluorescence of conjugated doxorubicin (Figures 9 and S10). Initially (30 min), there were no significant differences in p(NIPAM)-co-5%AA-co-FA-co-Dox internalization be-  After 6 and 8 h, the amount of p(NIPAM)-co-5%AA-co-FA-co-Dox inside MDA-MB 231 cells increased slowly (66 and 75%, respectively), suggesting the reaching of the maximum cell internalization. By contrast, it increased inside normal cells, as expected for longer incubation time in a static in vitro system. In summary, the particle uptake ratio at 0.5, 1, 2, 4, 6, 8, and 24 h was 1.7, 2.2, 2.6, 4.3, 2.3, 1.3, and 1.8, respectively. This showed that the maximum difference in particle uptake was a ratio of 4.3 after 4 h of incubation, suggesting that p(NIPAM)-co-5%AA-co-FA-co-Dox targeted MDA-MB 321 cancer cells due to the recognition between folate and its receptor. On the contrary, in HB2 healthy cells, which present lower FR expression, the microparticles uptake was time-delayed, suggesting again a specific particles tumor targeting. The decrease registered at 24 h of incubation for both normal and tumor cells (30 and 56%, respectively) was correlated to the death of cells that initially had internalized particles.

Cytotoxicity Assay
The cytotoxic effect of doxorubicin conjugated to microparticles was evaluated on After 6 and 8 h, the amount of p(NIPAM)-co-5%AA-co-FA-co-Dox inside MDA-MB 231 cells increased slowly (66 and 75%, respectively), suggesting the reaching of the maximum cell internalization. By contrast, it increased inside normal cells, as expected for longer incubation time in a static in vitro system. In summary, the particle uptake ratio at 0.5, 1, 2, 4, 6, 8, and 24 h was 1.7, 2.2, 2.6, 4.3, 2.3, 1.3, and 1.8, respectively. This showed that the maximum difference in particle uptake was a ratio of 4.3 after 4 h of incubation, suggesting that p(NIPAM)-co-5%AA-co-FA-co-Dox targeted MDA-MB 321 cancer cells due to the recognition between folate and its receptor. On the contrary, in HB2 healthy cells, which present lower FR expression, the microparticles uptake was time-delayed, suggesting again a specific particles tumor targeting. The decrease registered at 24 h of incubation for both normal and tumor cells (30% and 56%, respectively) was correlated to the death of cells that initially had internalized particles.

Cytotoxicity Assay
The cytotoxic effect of doxorubicin conjugated to microparticles was evaluated on normal HB 2 and MDA-MB 231 tumor cells by a viability assay.
As shown in Figure 10, 5 µM of the drug conjugated to p(NIPAM)-co-5%AA-co-FAco-Dox induces cell mortality on MDA-MB 231 cells (48% of mortality) and the viability decreases in a concentration-dependent way, reaching the maximum efficiency at 15 µM, so that, at higher concentration, the plateau state was registered (around 37% at 20, 25, and 30 µM of Dox). These data suggest that conjugation protocol does not alter the structure and functionality of conjugated drug and, furthermore, that microsystems, can release the drug inside cells. On the contrary, the viability of healthy cells after incubation was around 66% for all the drug concentrations used, confirming again the specific targeting of p(NIPAM)-co-5%AA-co-FA-co-Dox to tumor cells. The small mortality of 33% registered in this case was due to the long treatment time in a static system (24 h). Doxorubicin alone was used as a positive control.
The differences in toxicity among different cell lines and microgels was probably due to the specific targeting of microgels to tumor cells, recognizing the folate receptor overexpressed by MDA-MB 231 cells. This brought a diverse cell internalization between tumor and normal cells as suggested by flow cytometry analysis, and therefore, to a distinct cytotoxic effect. It is worth mentioning that the biocompatibility of p(NIPAM) was previously tested by Mohsen et al. [41] when it showed cell viability over 90% at concentrations up to 3 mg/mL. cytotoxic effect. It is worth mentioning that the biocompatibility of p(NIPAM) was previously tested by Mohsen et al. [41] when it showed cell viability over 90% at concentrations up to 3 mg/mL.

Conclusions
Although in the last years, cancer research has seen significant progress in the understanding, diagnosis, treatment, and prevention, low selectivity of the chemotherapeutic agents and consequently high side effects often occur. In this context, a novel drug delivery system that aims to specifically target cancer cells was designed and synthesized. Based on the tumor characteristic, p(NIPAM)-co-5%AA microgel particles were covalently conjugated to folic acid that is overexpressed in the majority of tumor cells (targeting agent) and to the anti-cancer drug Doxorubicin through EDC/NHS coupling reaction. The advantage of covalently tethering DOX, rather than loading it by self-assembly, is that the amount of DOX conjugated to the microgel is taken up almost completely. While the other self-assembly systems have either low entrapment efficiency (in case of synthetic polymers), or are not feasible to scale up (such as micelles) [42,43]. Moreover, tethering

Conclusions
Although in the last years, cancer research has seen significant progress in the understanding, diagnosis, treatment, and prevention, low selectivity of the chemotherapeutic agents and consequently high side effects often occur. In this context, a novel drug delivery system that aims to specifically target cancer cells was designed and synthesized. Based on the tumor characteristic, p(NIPAM)-co-5%AA microgel particles were covalently conjugated to folic acid that is overexpressed in the majority of tumor cells (targeting agent) and to the anti-cancer drug Doxorubicin through EDC/NHS coupling reaction. The advantage of covalently tethering DOX, rather than loading it by self-assembly, is that the amount of DOX conjugated to the microgel is taken up almost completely. While the other self-assembly systems have either low entrapment efficiency (in case of synthetic polymers), or are not feasible to scale up (such as micelles) [42,43]. Moreover, tethering the DOX and conjugating it with a targeting moiety, ensures that DOX targets only cancer cells and shall be intracellularly released upon degradation of the microgel particles by relevant enzymes. Unlike other self-assembly systems, the DOX can be released in the bloodstream. Accordingly, it is suggested that calculating the needed doses of covalently tethered Dox can be easier and more accurate than a physically entrapped one.
The new delivery system was then characterized and tested for targeting ability and capability to release the conjugated drug inside cells.
Several characterization studies were carried out, including UV-Vis analysis, DLS, TGA, DSC, and FTIR to demonstrate the successful conjugation of FA and Dox to p(NIPAM)co-5%AA microgel and that the new microgels retain microgel behavior [44].
The appearance of the typical FA and Doxo peaks in UV-VIS analysis ( Figure 1) and the variation in size (DLS analysis, Figure 2) demonstrated a variety of microgel composition due to FA and Doxo conjugation. These data were confirmed by not only the variation of TGA curves (Figure 4), but also by the alteration of the DSC profiles of the microgels ( Figure 5), shifting both the melting point and the crystallization melting point; furthermore, any exothermic peaks (that are present in p(NIPA) and p(NIPAM)-co-5%AA) were not registered. At the same time, also the FTIR profiles changed probably due to the different functional groups of the folic acid and doxorubicin. Taken together, these data confirmed the success of the conjugation, as demonstrated also by cytotoxic assay performed on normal and tumor cells ( Figure 10) and the targeting studies (Figures 8 and 9).
The uptake and localization studies of p(NIPAM)-co-5%AA-co-FA-co-Dox were performed using flow cytometry and fluorescence microscopy, while viability assay was carried out to investigate the cytotoxicity of the drug conjugated to developed microgels. Co-culture experiment demonstrated the drug release and the specific targeting of the microcomplex exclusively to the tumor cells by an active targeting that probably could be increased in vivo by a passive targeting based on the enhanced permeability and retention effect (EPR effect). Besides, viability assay results show higher cell viability for healthy cells incubated with p(NIPAM)-co-5%AA-co-FA-co-Dox than the cancer ones. Also, it is shown that at higher concentrations (25 µm and above), healthy cells were more viable when incubated with p(NIPAM)-co-5%AA-co-FA-co-Dox than when incubated with soluble form Dox. Therefore, these data suggest that p(NIPAM)-co-5%AA-co-FA-co-Dox are good candidates as delivery systems to increase the specific tumor targeting probably reducing general side effects, even if more in vivo studies need to be carried out.

Synthesis of p(NIPAM)-co-5%AA
A Surfactant Free Emulsion Polymerisation (SFEP) technique was used for the preparation of p(NIPAM)-co-5%AA as described previously and in accordance with literature [27][28][29]41]. Briefly, a three-neck lid was then fitted to the reaction vessel, which was placed onto a hot plate stirrer and heated to 70 • C with continuous stirring under N2 atmosphere. Potassium persulphate initiator (0.5 g) was dissolved in 800 mL of distilled water. The crosslinker N,N -methylenebisacrylamide 99% (0.5 g) (BS, Sigma Aldrich, Gillingham, UK), N-isopropylacrylamide (NIPAM, Sigma Aldrich, Gillingham, UK) 97% monomer (4.75 g) and acrylic acid (AA, Sigma Aldrich, Gillingham, UK) co-monomer (0.25 g) were dissolved in 200 mL of distilled water while stirring gently with a magnetic stirrer. After all the reagents were dissolved, they were transferred into the reaction vessel containing the initiator. The reaction was run for 6 h with constant stirring and under nitrogen. After 6 h, the microgel dispersion was allowed to cool down to room temperature, then dialyzed (MW cut-off was 12-14,000 kDa) in fresh distilled water for 7 days. the cells were washed with PBS and then fixed with 3.7% formaldehyde (in PBS) for 5 min at room temperature, followed by three washes with PBS. Nuclei were stained in the dark with DAPI solution (dilution of 1:10,000 in water) for 15 min at room temperature. Samples were analyzed by fluorescence microscopy (Leica, Buccinasco (MI), Italy) and confocal microscope (FLUOVIEW FV10i-LIV, Olympus, Italy).