Impact of Degree of Ionization and PEGylation on the Stability of Nanoparticles of Chitosan Derivatives at Physiological Conditions

Nowadays, the therapeutic efficiency of small interfering RNAs (siRNA) is still limited by the efficiency of gene therapy vectors capable of carrying them inside the target cells. In this study, siRNA nanocarriers based on low molecular weight chitosan grafted with increasing proportions (5 to 55%) of diisopropylethylamine (DIPEA) groups were developed, which allowed precise control of the degree of ionization of the polycations at pH 7.4. This approach made obtaining siRNA nanocarriers with small sizes (100–200 nm), positive surface charge and enhanced colloidal stability (up to 24 h) at physiological conditions of pH (7.4) and ionic strength (150 mmol L−1) possible. Moreover, the PEGylation improved the stability of the nanoparticles, which maintained their colloidal stability and nanometric sizes even in an albumin-containing medium. The chitosan-derivatives displayed non-cytotoxic effects in both fibroblasts (NIH/3T3) and macrophages (RAW 264.7) at high N/P ratios and polymer concentrations (up to 0.5 g L−1). Confocal microscopy showed a successful uptake of nanocarriers by RAW 264.7 macrophages and a promising ability to silence green fluorescent protein (GFP) in HeLa cells. These results were confirmed by a high level of tumor necrosis factor-α (TNFα) knockdown (higher than 60%) in LPS-stimulated macrophages treated with the siRNA-loaded nanoparticles even in the FBS-containing medium, findings that reveal a good correlation between the degree of ionization of the polycations and the physicochemical properties of nanocarriers. Overall, this study provides an approach to enhance siRNA condensation by chitosan-based carriers and highlights the potential of these nanocarriers for in vivo studies.


Introduction
Over the last decade, advances in RNA engineering have enabled the obtaining of small interfering RNAs (siRNA) with enhanced pharmacologic activities, which has driven an increased interest in establishing therapeutic platforms for the treatment of several diseases [1]. Despite improvements, the therapeutic efficiency of siRNA is still limited by the use of efficient vectors capable of directing them to their site of action, inside the target cells [2]. In this scenario, the use of non-viral vectors has gained prominence, especially after the use of lipid nanoparticles in mRNA vaccines applied in the fight against the Coronavirus Disease 2019 (COVID-19) pandemic [3] and for the transport of siRNA in the first RNA interference (RNAi) drug approved by the US Food and Drug Administration (FDA) in 2018 [2].

Polymer Composition
The main objective of this study was to combine the diisopropylaminoethyl (DIPEA) groups with chitosan to obtain derivatives with increasing degrees of ionization and stable siRNA-nanoparticles at physiological conditions. The advantage of this approach relies on the sensitivity of DIPEA units to pH [21]. Moreover, the grafting onto the primary amino groups of chitosan gives origin to secondary amino groups and this may allow control of the physicochemical properties of the nanoparticles at neutral pH ( Figure 1). Representative 1 H and 13 C NMR spectra of derivatives are shown in Figure 2 and more detailed spectra are presented in the Supplementary Information (Figures S1-S12).
The degree of deacetylation (DDA) was determined using the 1 H NMR spectra and the Equation (1), = (1 − 3 ) × 100 (1) where IHAc is the area of the signal of the three methyl hydrogens (H7) of acetylglucosamine units (δ 2.4 ppm) and IH1 is the area of the H1 hydrogen signals (δ 4.9-5.5 ppm). After deacetylation, the DDA increased from 76% (Chc) to 96% (Chd), which confirms the success of the deacetylation reaction (Table 1). Additionally, the 13 C NMR spectrum of Chc has a very low-intensity peak for methyl carbon (C7) of the acetyl group (δ 22 ppm), which indicates a high DDA [15,22]. The main objective of this study was to combine the diisopropylaminoethyl (DIPEA) groups with chitosan to obtain derivatives with increasing degrees of ionization and stable siRNA-nanoparticles at physiological conditions. The advantage of this approach relies on the sensitivity of DIPEA units to pH [21]. Moreover, the grafting onto the primary amino groups of chitosan gives origin to secondary amino groups and this may allow control of the physicochemical properties of the nanoparticles at neutral pH ( Figure 1). Representative 1 H and 13 C NMR spectra of derivatives are shown in Figure 2 and more detailed spectra are presented in the Supplementary Information (Figures S1-S12).
The degree of deacetylation (DDA) was determined using the 1 H NMR spectra and the Equation (1), where I HAc is the area of the signal of the three methyl hydrogens (H7) of acetylglucosamine units (δ 2.4 ppm) and I H1 is the area of the H1 hydrogen signals (δ 4.9-5.5 ppm). After deacetylation, the DDA increased from 76% (Ch c ) to 96% (Ch d ), which confirms the success of the deacetylation reaction (Table 1). Additionally, the 13 C NMR spectrum of Ch c has a very low-intensity peak for methyl carbon (C7) of the acetyl group (δ 22 ppm), which indicates a high DDA [15,22]. The spectra for derivatives are characterized by the appearance of a strong signal centered at δ 1.7 ppm corresponding to methyl hydrogens of the DIPEA groups ( Figure  2a). The degrees of substitution by DIPEA (DSDIPEA) were determined using Equation (2): = 12 × 100 (2) In this equation, IHDIPEA is the integral of the signal attributed to the 12 methyl hydrogens of DIPEA (H12-H15) at δ 1.6-1.9 ppm. As seen in Table 1, by adjusting the ratio DI-PEA-Cl/Chd in the feed, DIPEA-chitosans with DSDIPEA in the range of 5 to 55% (mol/mol) could be obtained, which allows a more in-depth study of the effect of DIPEA content on the properties of chitosan as a siRNA nanocarrier. The grafting with DIPEA was also confirmed by the low-field displacement of the H1 signal from δ 5.2 to 5.3 ppm and by the emergence of new peaks at δ 3.5-4.5 ppm attributed to methine and methylene DIPEA hydrogens (Figure 2a).
The derivatives were also characterized by 13 C NMR and the appearance of new signals on spectra of the derivatives are compared to the Chd spectrum in Figure 2b. The new peaks appeared at δ 16-18 ppm, δ 42-44 ppm, and δ 55-57 ppm and are attributed to the methylic (C12-C15), methylene (C8, C9), and methine (C10, C11) carbons from DIPEA, respectively. Additionally, it is worth mentioning that no additional signal was observed at δ 65 ppm [23], even for the more substituted derivative (DIPEA55Chd), confirming the grafting at the amino groups and a good control of the polymer composition.
Selected DIPEA-derivatives were also grafted with PEG to provide stealth properties to nanoparticles, aiming to avoid the biological interactions responsible for nanoparticles  The spectra for derivatives are characterized by the appearance of a strong signal centered at δ 1.7 ppm corresponding to methyl hydrogens of the DIPEA groups ( Figure 2a). The degrees of substitution by DIPEA (DS DIPEA ) were determined using Equation (2): Mar. Drugs 2022, 20, 476

of 24
In this equation, I HDIPEA is the integral of the signal attributed to the 12 methyl hydrogens of DIPEA (H12-H15) at δ 1.6-1.9 ppm. As seen in Table 1, by adjusting the ratio DIPEA-Cl/Ch d in the feed, DIPEA-chitosans with DS DIPEA in the range of 5 to 55% (mol/mol) could be obtained, which allows a more in-depth study of the effect of DIPEA content on the properties of chitosan as a siRNA nanocarrier. The grafting with DIPEA was also confirmed by the low-field displacement of the H1 signal from δ 5.2 to 5.3 ppm and by the emergence of new peaks at δ 3.5-4.5 ppm attributed to methine and methylene DIPEA hydrogens (Figure 2a).
The derivatives were also characterized by 13 C NMR and the appearance of new signals on spectra of the derivatives are compared to the Ch d spectrum in Figure 2b. The new peaks appeared at δ 16-18 ppm, δ 42-44 ppm, and δ 55-57 ppm and are attributed to the methylic (C12-C15), methylene (C8, C9), and methine (C10, C11) carbons from DIPEA, respectively. Additionally, it is worth mentioning that no additional signal was observed at δ 65 ppm [23], even for the more substituted derivative (DIPEA 55 Ch d ), confirming the grafting at the amino groups and a good control of the polymer composition.
Selected DIPEA-derivatives were also grafted with PEG to provide stealth properties to nanoparticles, aiming to avoid the biological interactions responsible for nanoparticles disassembling outside cells. The degrees of substitution by PEG (DS PEG ) were determined using Equation (3), where the I HPEG refers to the area under the peak attributed to the 172 methylene protons of PEG repeating units (-OCH 2 CH 2 -), at δ 3.9-4.0 ppm [15]. The PEG grafting was also confirmed by the appearance of a small peak at δ 3.65 ppm (Figures S7 and S8) attributed to terminal methyl hydrogens (-OCH 3 ) of the PEG chain [19]. As shown in Table 1, DS PEG of both grafted derivatives was around 1.5% (mol/mol), and no DIPEA release was observed, which confirms the success of the synthetic route adopted in this study.
FTIR spectroscopy was also employed to characterize the derivatives and the spectra for chitosans (Ch d and Ch c ) and some selected derivatives are shown in Figure 2d. When compared to commercial chitosan (Ch c ), the deacetylated sample (Ch d ) displays a weak band assigned to the -C=O stretching (1650 cm −1 ) and the -CH symmetrical deformation (1370 cm −1 ). The decreased intensities of both bands are expected, taking into account the deacetylation reaction and this corroborates the NMR results. The substitution by DIPEA was also confirmed by FTIR, i.e., when compared to chitosan (Ch d ), the FTIR spectra of derivatives displayed a new band at 2970 cm −1 assigned to the -CH stretching originating from the DIPEA group ( Figure S17). Moreover, the DIPEA grafting led to the decreased intensity for the -NH bending vibration band (1590 cm −1 ) due to the substitution of primary amine groups [24,25].
The average molecular weights (Mw) from chitosan and its derivatives were estimated by GPC ( Figure S14) and the results are shown in Table 1. As previously reported [26], the deacetylation reaction led to the depolymerization of chitosan chains due to reaction conditions (high temperature and strong basic solution), which have reduced the Mw from 208 (Ch c ) to 141 kDa (Ch d ) and increased the molar-mass dispersity (Ð M ). Similarly, the molecular weights for derivatives tended to decrease upon DIPEA grafting (especially for the more substituted ones), a sign that main chain degradation took place during the chemical modification of Ch d , as earlier reported for DEAE-chitosan [27]. Another relevant effect of MW may be observed in the synthesis of derivatives when the data in Table 1 are compared to previous results reported by our research group [16] for DIPEA-chitosans of higher MW (200 kDa), i.e., under the same experimental conditions, chitosans of high MW need to be fed with more DIPEA-Cl (about 20-30%) to originate DIPEA-chitosans with the same DS DIPEA of derivatives listed in Table 1, which have about 100 kDa.

Compositional Effects on the Degree of Ionization (α) and Buffering Capacity (BC)
The degree of ionization (α) of polycations is closely related to the surface charge of their nanoparticles, impacting directly their physicochemical and biological properties. For instance, while a high positive charge density on polycations favors the complexation of nucleic acid, it may increase the toxicity, while nanoparticles with low charge density may have their cell uptake and colloidal stability reduced [28]. In this study, the α of derivatives is highly dependent on DIPEA content, especially for higher pH values. The degrees of ionization were determined based on the conductivity titrations of the derivatives. The conductivity traces (Figure 3a-c) allow the detection of the starting and the ending points for the deprotonation of the amino groups. The last break in the conductivity traces corresponds to the complete deprotonation of the tertiary amino groups, which occurs near a pH of 10 and this is consistent with a pKa of 8.5 reported for diisopropylamino groups [29]. The degrees of ionization calculated using Equation (4) (Section 3.2.8) are shown in Figure 3d and reveal that, at physiological pH (7.4), the low α of chitosan (6.1%) increased to 13%, 29%, 41%, and 64% for DIPEA 5 Ch d , DIPEA 15 Ch d , DIPEA 34 Ch d , and DIPEA 55 Ch d , respectively. In addition, it is worth mentioning that at pH 6.4 deacetylated chitosan is about 50% ionized, in accordance with the reported pKa value for its primary amine groups [5]. Due to the grafting by DIPEA groups, derivatives had a clear increase in their solubility at neutral pH when compared to unsubstituted chitosan that formed visible aggregates (Figure S16 Supplementary Information). It has been reported that the interaction of nanoparticles with the endosomal membrane plays an important role in the endosomal escape rate [30,31]. Thus, decreases in pH during the endocytosis [32] may increase the surface charge of nanoparticles, favoring their interaction with the endosomal membrane and the release process.
Mar. Drugs 2022, 20, x FOR PEER REVIEW 6 of 25 tosans of high MW need to be fed with more DIPEA-Cl (about 20-30%) to originate DI-PEA-chitosans with the same DSDIPEA of derivatives listed in Table 1, which have about 100 kDa.

Compositional Effects on the Degree of Ionization (α) and Buffering Capacity (BC)
The degree of ionization (α) of polycations is closely related to the surface charge of their nanoparticles, impacting directly their physicochemical and biological properties. For instance, while a high positive charge density on polycations favors the complexation of nucleic acid, it may increase the toxicity, while nanoparticles with low charge density may have their cell uptake and colloidal stability reduced [28]. In this study, the α of derivatives is highly dependent on DIPEA content, especially for higher pH values. The degrees of ionization were determined based on the conductivity titrations of the derivatives. The conductivity traces (Figure 3a-c) allow the detection of the starting and the ending points for the deprotonation of the amino groups. The last break in the conductivity traces corresponds to the complete deprotonation of the tertiary amino groups, which occurs near a pH of 10 and this is consistent with a pKa of 8.5 reported for diisopropylamino groups [29]. The degrees of ionization calculated using Equation (4) (Section 3.2.8) are shown in Figure 3d and reveal that, at physiological pH (7.4), the low α of chitosan (6.1%) increased to 13%, 29%, 41%, and 64% for DIPEA5Chd, DIPEA15Chd, DIPEA34Chd, and DI-PEA55Chd, respectively. In addition, it is worth mentioning that at pH 6.4 deacetylated chitosan is about 50% ionized, in accordance with the reported pKa value for its primary amine groups [5]. Due to the grafting by DIPEA groups, derivatives had a clear increase in their solubility at neutral pH when compared to unsubstituted chitosan that formed visible aggregates (Figure S16 Supplementary Information). It has been reported that the interaction of nanoparticles with the endosomal membrane plays an important role in the endosomal escape rate [30,31]. Thus, decreases in pH during the endocytosis [32] may increase the surface charge of nanoparticles, favoring their interaction with the endosomal membrane and the release process. Considering that chitosan nanoparticles are internalized mainly via acid endocytic pathways [19,33] and that polycations that having great buffering capacities (BC) may provide an enhanced endosomal escape of the therapeutic genes by the proton sponge hypothesis [34], the BC of DIPEA-chitosan derivatives was also qualitatively evaluated ( Figure 3e). Except for DIPEA 55 Ch d , all derivatives displayed a BC higher or similar to unmodified chitosan in the pH range from 7.4 to 6.8 (highlighted in Figure 3e), which corresponds to an initial pH decay during the acidic endocytic pathway [35]. The improvement of BC in this pH range may be promising in providing siRNA with an earlier endosomal escape.

Electrophoresis
The polycation-siRNA interaction and the coacervation process to form nanoparticles is mainly driven by electrostatic attraction between the negatively charged siRNA molecules and the polycations [36]. The siRNA loading is very dependent of charge ratios (N/P ratio) in the nanoparticle core, i.e., if there are not enough positively charged amines to neutralize the phosphate groups of the nucleic acid, not all of the siRNA will be efficiently loaded by the vector and the resulting complex will have a negative charge [5,37]. Hence, an electrophoretic mobility study can be used to detect compositions (charge ratios) at which the siRNA is not efficiently complexed, which would result in formulations with low therapeutic potential [38].
At pH 7.4, the deacetylated chitosan (Ch d ) was not able to retain the siRNA in any of the N/P ratios studied and siRNA was released from all wells ( Figure 4a). However, for nanoparticles formulated with DIPEA 5 Ch d , DIPEA 15 Ch d , DIPEA 34 Ch d , and DIPEA 55 Ch d , siRNA was completely retained at decreasing N/P ratios (7.0; 5.0; 3.0, and 3.0, respectively) which is consistent with the increasing degrees of ionization for these polycations ( Figure 4b). The similar siRNA release for nanoparticles formulated with DIPEA 34 Ch d and DIPEA 55 Ch d , even having different ionization at pH 7.4, suggests that besides electrostatic interaction, conformational changes on the polymer chain may affect the condensation process and the strength of interaction.
The PEGylated derivatives (PEG-DIPEA 15 Ch d and PEG-DIPEA 34 Ch d ) had the same siRNA release behavior as non-PEGylated ones (Figure 4a), indicating that DIPEA substitution enabled chitosan to dampen the negative effect associated with PEGylation, i.e., in general there is a decrease in the strength of the interaction between polycations and siRNA due to the steric hindrance exerted by the PEG chains [5,20].
Overall, these results make clear the positive effects of DIPEA grafting showing the gradual strengthening of the siRNA/polycation interactions, which may be an advantage compared to chitosan that, under physiological pH, interacts weakly with siRNA molecules providing unstable particles at low N/P ratios.

Effect of the Degree of Ionization and PEGylation on Nanoparticle Properties: Size, Morphology, Surface Charge, and Colloidal Stability of Nanocarriers
The control of the surface charge, size, and stability of nanocarriers is pivotal for obtaining promising vectors for gene therapy. In general, siRNA nanocarriers with sizes smaller than 500 nm, slightly positive zeta potentials, and the ability to avoid unwanted interactions during the in vitro application are more prone to efficiently transport the siRNA, to be internalized by cells and to provide high knockdown levels [39,40]. For chitosan nanoparticles, these characteristics depend greatly on the N/P ratio, pH, and ionic strength in which they are prepared [18,41].
To adjust the properties of the nanoparticles for in vivo applications, their formulations were prepared and evaluated at pH 7.4 and ionic strength of 150 mmol L −1 . For the siRNA carriers based on DIPEA-chitosan, the N/P ratios needed to obtain vectors on a nanoscale decreased with increasing DS DIPEA . Concurrently, the zeta potential (ζ) of nanocarriers increases with the DIPEA content, in close agreement with the degree of ionization of the polycations ( Figure 5). Nanoparticles smaller than 250 nm were obtained only when the Mar. Drugs 2022, 20, 476 8 of 24 N/P ratio was positive, i.e., the condensation process is efficient only after negative charges of siRNA molecules had been completely neutralized. Moreover, the PEG graft induced changes to ζ and mainly to the hydrodynamic diameter (Dh) of nanoparticles and, also, nanoparticles formulated with PEGylated derivatives were smaller than 200 nm, using 50% less polymer than the non-PEGylated ones (Figure 5a vs. Figure 5b). These nanoparticles had low polydispersities (Figure 5f), and values smaller than 0.2 were often obtained. Overall, while nanoparticles formulated with PEGylated-only chitosan (PEG-Ch d ) formed complexes composed of large aggregates with a negative charge surface ( Figure S18), the derivatives have formed nanoparticles with promising sizes (100-200 nm) and zeta potentials (6-12 mV) to be applied as vectors for gene therapy. Notably, the importance of the degree of ionization (α) of these derivatives for obtaining nanovectors of adequate Dh and ζ was also observed in a previous study [16], in which it was shown that high MW DIPEA-chitosans with low DS DIPEA (15%) are more prone to formulate efficient siRNA nanocarriers (with small size, positive ζ, and improved stability) at pH 6.3 (α ≥ 80%) than at pH 7.4 (α < 35%), due to an increasing degree of ionization with decreasing pH.  The control of the surface charge, size, and stability of nanocarriers is pivotal for obtaining promising vectors for gene therapy. In general, siRNA nanocarriers with sizes smaller than 500 nm, slightly positive zeta potentials, and the ability to avoid unwanted interactions during the in vitro application are more prone to efficiently transport the siRNA, to be internalized by cells and to provide high knockdown levels [39,40]. For chi the importance of the degree of ionization (α) of these derivatives for obtaining nanovectors of adequate Dh and ζ was also observed in a previous study [16], in which it was shown that high MW DIPEA-chitosans with low DSDIPEA (15%) are more prone to formulate efficient siRNA nanocarriers (with small size, positive ζ, and improved stability) at pH 6.3 (α ≥ 80%) than at pH 7.4 (α < 35%), due to an increasing degree of ionization with decreasing pH. SEM-FEG microscopy images (Figure 6a,b) pointed to nanoparticles with sphericallike morphologies and sizes in accordance with the light scattering analysis (100-200 nm), results that are very close to those reported for siRNA nanoparticles formulated with DEAE-chitosans at lower pH values [15,17]. Spherical nanoparticles may have favored cellular uptake in comparison to cubical and rod-shaped nanoparticles [42]. Additionally, SEM-FEG microscopy images (Figure 6a,b) pointed to nanoparticles with sphericallike morphologies and sizes in accordance with the light scattering analysis (100-200 nm), results that are very close to those reported for siRNA nanoparticles formulated with DEAEchitosans at lower pH values [15,17]. Spherical nanoparticles may have favored cellular uptake in comparison to cubical and rod-shaped nanoparticles [42]. Additionally, the cellular uptake may be enhanced by using positively charged nanoparticles that interact with the endosomal membrane during their internalization [43]. When the DIPEA 34 Ch d nanoparticles formulated at N/P ratio 10 were evaluated at pH 6.3 (early endosome pH), their zeta potential (17.6 mV) was almost twice as high as that at pH 7.4 (9.6 mV), which suggests a favored interaction with the internal endosomal membrane during the acid internalization pathway, in agreement with the ionization results discussed above ( Figure S19). in the first 30 min after their preparation. Although the PEGylation version (PEG-DI-PEA15Chd) had prolonged the stability of nanoparticle sizes for up to 2 h of analysis, after 24 h, the sizes increased to micrometers (Figure 6c). On the other hand, the most substituted chitosans (PEG-DIPEA34Chd and DIPEA55Chd) were able to keep their small sizes (100-200 nm) even after 24 h of analysis, with polydispersities lower than 0.4, suggesting that degrees of ionization higher than 40% are more appropriate to confer stability at neutral pH at physiological ionic strength ( Figure S20). To evaluate the stability of nanoparticles in solution, their hydrodynamic diameter was monitored over time in the presence and absence of bovine serum albumin (BSA). At N/P ratio 10, nanoparticles displayed slightly positive zeta potentials (Figure 5c,d), Dh smaller than 250 nm (Figure 5a,b) and a complete siRNA entrapment (Figure 4a). So, this N/P ratio was chosen for the stability studies carried out under physiological conditions (pH 7.4 and ionic strength of 150 mmol L −1 ), as shown in Figure 6. Contrary to deacetylated chitosan, which forms large aggregates at pH 7.4 [15], the DIPEA-derivatives formed siRNA nanocarriers with good colloidal stability proportional to the DS DIPEA and further improved by the PEGylation. In the absence of BSA, DIPEA 15 Ch d nanoparticles displayed poor stability characterized by the occurrence of large aggregates (greater than 1000 nm) in the first 30 min after their preparation. Although the PEGylation version (PEG-DIPEA 15 Ch d ) had prolonged the stability of nanoparticle sizes for up to 2 h of analysis, after 24 h, the sizes increased to micrometers (Figure 6c). On the other hand, the most substituted chitosans (PEG-DIPEA 34 Ch d and DIPEA 55 Ch d ) were able to keep their small sizes (100-200 nm) even after 24 h of analysis, with polydispersities lower than 0.4, suggesting that degrees of ionization higher than 40% are more appropriate to confer stability at neutral pH at physiological ionic strength ( Figure S20).
The adsorption of proteins is one of the major obstacles faced by nanoparticles during their biodistribution, since it may alter their physicochemical properties and favor their destruction via an opsonization process that lowers the vector transfection efficacy [44].
Hence, the interaction between nanoparticles and albumin (40 g L −1 ) was probed by monitoring the size distribution of nanoparticles/BSA solution over time under physiological conditions [45]. The experiment was performed using the most substituted DIPEA-chitosan derivatives and their PEGylated versions. The nanoparticle size traces show a BSA peak centered at 8 nm and the non-PEGylated version of DIPEA 34 Ch d nanoparticles showed a peak centered at 600 nm corresponding to BSA/nanocarrier aggregates (Figure 6d-f), while, for the PEG-DIPEA 34 Ch d , no aggregates were observed and the sizes remained between 100 and 200 nm, even after four hours of study, indicating an inhibition of the BSA/nanoparticle interactions (Figure 6e). The BSA adsorption is expected considering that the number of negatively charged groups on the BSA molecule increases with pH, therefore interaction with positive surfaces cannot be neglected, since the BSA charge at pH 7.4 is slightly negative [46]. Interestingly, the size distribution of nanoparticles formulated by the most substituted derivative (DIPEA 55 Ch d ) was minimally affected by BSA presence, i.e., the peak attributed to nanoparticles was kept centered at 100-200 nm and no larger structures were observed, even after four hours of incubation (Figure 6f). These results indicate that DIPEA helps to maintain the structure of the nanoparticles and the vectors with the highest DIPEA content are the most promising to remain stable in biological systems.

In Vitro Studies of Polymers and Nanoparticles Cytotoxicity, Cellular Uptake and Gene Knockdown
The cytotoxicity is one of the essential biological properties that limits the application of polycations as nucleic acid carriers. While non-viral vectors based on synthetic polycations such as PEI, poly(amidoamine) and poly(B-amino ester) may exhibit a more prominent cytotoxicity [47][48][49], chitosan is known for its good cytocompatibility [5]. The cytotoxicity of chitosan derivatives and their nanoparticles were evaluated in RAW 264.7 and NIH/3T3 cell lines using the MTS assay (Figure 7). Both cell lines displayed viabilities close to 100% even after their treatment with polycation concentrations up to 0.5 g L −1 . For nanoparticles, the viabilities remained above 80% after their incubation with nanoparticles formulated at an N/P ratio of 10. Moreover, compared to previous studies, while 0.3 g L −1 of chitosan substituted by 57% of DEAE led to cell viabilities at about 60%, in the presence of a DIPEA 55 Ch d (0.5 g L −1 ) derivative, they remained near to 90%, which suggests that DIPEA derivatives are less cytotoxic and their vectors more suitable for gene therapy [50].
Aiming at evaluating the internalization of the nanocarriers by RAW 264.7 macrophages, the vectors were loaded with siRNAs labeled with fluorophores groups (FAM or Cy5) and the cells were evaluated by confocal microscopy (Figure 8). Due to its size and negatively charged backbone, the free siRNA was not able to enter the cells (Figure 8a upper panel). However, when carried by lipofectamine and by chitosan derivatives, red dots (siRNA-Cy5) could be clearly observed around the nuclei of cells. Additionally, unlike lipofectamine, the siRNA carried by DIPEA-chitosans tended to be distributed more throughout the cytoplasm of the cell. The same behavior was observed for siRNA-FAM transported by PEG-DIPEA 15 Ch d labeled with RITC (Figure 8b), i.e., the siRNA (green dots) extended throughout the whole cytoplasm area, as well as the polycation labeled with rhodamine (red dots). Moreover, the merged image revealed some yellow dots in the cytoplasm, provided by the green (siRNA) and red (polymer) dots overlap, which indicates the presence of intact nanoparticles in addition to the presence of free siRNA and polymer. This result suggests that the nanoparticles can be disassembled at different times inside the cell, so the siRNA is being continuously released by the nanoplatforms developed in this study.
The potential of DIPEA-chitosan derivatives as siRNA nanocarriers was evaluated by the GFP and TNFα knockdown in HeLa and RAW 264.7 cells, respectively (Figures 9 and 10). After the treatment of HeLa-GFP with PEG-DIPEA 15 Ch d nanoparticles loaded with siRNA anti-GFP, GFP was not visible in several cells, as indicated by the red arrows in the confocal microscopy images (Figure 9). Moreover, when the GFP area in the images were measured and normalized by the DAPI area, a reduction of about 30% in GFP expression was seen for HeLa treated with PEG-DIPEA 15 Ch d nanoparticles in comparison with non-treated cells, a value very close to the lipofectamine result ( Figure S23). This result strongly suggests the efficient internalization of carriers displayed in Figure 8b, and reinforces the potential of PEG-DIPEA 15  Aiming at evaluating the internalization of the nanocarriers by RAW 264.7 macrophages, the vectors were loaded with siRNAs labeled with fluorophores groups (FAM or Cy5) and the cells were evaluated by confocal microscopy (Figure 8). Due to its size and negatively charged backbone, the free siRNA was not able to enter the cells (Figure 8a upper panel). However, when carried by lipofectamine and by chitosan derivatives, red dots (siRNA-Cy5) could be clearly observed around the nuclei of cells. Additionally, unlike lipofectamine, the siRNA carried by DIPEA-chitosans tended to be distributed more throughout the cytoplasm of the cell. The same behavior was observed for siRNA-FAM transported by PEG-DIPEA15Chd labeled with RITC (Figure 8b), i.e., the siRNA (green dots) extended throughout the whole cytoplasm area, as well as the polycation labeled with rhodamine (red dots). Moreover, the merged image revealed some yellow dots in the cytoplasm, provided by the green (siRNA) and red (polymer) dots overlap, which indicates the presence of intact nanoparticles in addition to the presence of free siRNA and polymer. This result suggests that the nanoparticles can be disassembled at different times inside the cell, so the siRNA is being continuously released by the nanoplatforms developed in this study. The nanoparticles formulated with the most substituted derivatives (DIPEA 34 Ch d and DIPEA 55 Ch d ) and that have exhibited the most promising physicochemical properties, i.e., improved siRNA condensation, small sizes, positive surface, and enhanced colloidal stability, were tested for their ability to knockdown the TNFα cytokine (Figure 10a). The expression of TNFα cytokine by RAW 264.7 cells was evaluated using the ELISA assay and compared with cells treated with LPS only (100% of TNFα expression). In general, the results showed that gene knockdown is dependent on the N/P ratio, DS DIPEA , and the PEG grafting. The transfection efficiency mediated by non-PEGylated derivatives (DIPEA 34 Ch d and DIPEA 55 Ch d ) was greatly dependent on the N/P ratio and a significant knockdown of TNFα was observed only for N/P ≥ 5.0. These results correlate well with the physicochemical properties observed for nanoparticles, which exhibited positive zeta potentials of about +10 mV and Dh values in the range of 100-250 nm only for N/P ratios higher than 5 ( Figure 5). In addition, it is noteworthy that, although the physicochemical properties of the nanoparticles were mainly influenced by the degree of ionization of the polycations (as shown in the sections above), all DIPEA-chitosans showed buffering capacity at endosomal pH range (Figure 3), thus the increase of polymer concentration to formulate the nanocarriers at high N/P ratios may favor the endosomal escape of the nanovectors by the proton sponge hypothesis, since the buffering capacity is proportional to the non-protonated amino groups at neutral pH. At N/P ratio 10, the increase in DS DIPEA content, from 34 to 55%, promoted a decrease of about 30% in TNFα expression. Besides, PEGylation provided a significant improvement in transfection efficiency of DIPEA 34 Ch d and the knockdown was increased from 25% to 65% (N/P ratio 10), suggesting that both the decrease in size and the increase in colloidal stability were responsible for the higher efficiency of the PEG-DIPEA 34 Ch d nanoparticles. Moreover, in the presence of the FBS proteins, PEG-DIPEA 34 Ch d and DIPEA 55 Ch d derivatives built siRNA nanoparticles that were able to reduce TNFα expression by 50-65% when compared to LPS-treated cells only, which is consistent with their resistance to protein adsorption displayed in Figure 6, and with their significant cell uptake observed by confocal microscopy (Figure 8). On balance, these results have confirmed that chitosan can be associated with DIPEA and PEG groups for formulating promising siRNA nanocarriers at physiological conditions (pH 7.4, I = 150 mmol L −1 ) that are worthy of evaluation in in vivo studies. The potential of DIPEA-chitosan derivatives as siRNA nanocarriers was evaluated by the GFP and TNFα knockdown in HeLa and RAW 264.7 cells, respectively (Figures 9  and 10). After the treatment of HeLa-GFP with PEG-DIPEA15Chd nanoparticles loaded with siRNA anti-GFP, GFP was not visible in several cells, as indicated by the red arrows in the confocal microscopy images (Figure 9). Moreover, when the GFP area in the images result strongly suggests the efficient internalization of carriers displayed in Figure 8b, and reinforces the potential of PEG-DIPEA15Chd as a siRNA nanocarrier.   The nanoparticles formulated with the most substituted derivatives (DIPEA34Chd and DIPEA55Chd) and that have exhibited the most promising physicochemical properties, i.e., improved siRNA condensation, small sizes, positive surface, and enhanced colloidal stability, were tested for their ability to knockdown the TNFα cytokine (Figure 10a). The expression of TNFα cytokine by RAW 264.7 cells was evaluated using the ELISA assay and compared with cells treated with LPS only (100% of TNFα expression). In general, the results showed that gene knockdown is dependent on the N/P ratio, DSDIPEA, and the PEG grafting. The transfection efficiency mediated by non-PEGylated derivatives (DIPEA34Chd . Statistical analysis (a) was performed using the unpaired Student test (at a significance level of 0.05) and comparisons between the only LPS-treated cells and the cells treated by siRNA/nanoparticles (black symbols) are shown and, also, between the same samples in the presence and absence of FBS (red symbols). ** p < 0.01, *** p < 0.001, and NS = no significant differences between the means.

Deacetylated Chitosan
The commercial chitosan (Ch c ) was deacetylated as previously described [54] to obtain the starting chitosan (Ch d ), which was then used for the synthesis of derivatives. Briefly, 5.0 g of Ch c was solubilized in 250 mL of acetic acid (2%, v/v) under magnetic stirring overnight. Next, 300 mL of NaOH (50%, w/v) was added to the solution to obtain a finely divided aqueous suspension of chitosan which was heated (100 • C) and kept in an N 2 atmosphere. After 90 min under intense magnetic stirring in a reflux system, the reaction mixture was poured into 4 L of preheated water (50 • C) and allowed to cool at room temperature. The resulting deacetylated sample was washed several times with deionized water by the decantation process until neutral pH. After that, the precipitated chitosan was recovered by filtration, solubilized in acetic acid (250 mL; 2%, v/v), and the deacetylation process was repeated once more aiming to obtain a high DDA for the starting chitosan. After sedimentation and filtration, the deacetylated chitosan (Ch d ) was dried by lyophilization.

DIPEA-Chitosan Derivatives
The different DIPEA derivatives were obtained by adjusting the ratios of 2-Chloro-N,N-diisopropylethylamine hydrochloride (DIPEA-Cl) to glucosamine of deacetylated chitosan (Ch d ) according to Table 1. The procedure adopted for DIPEA 34 Ch d synthesis is described below. First, 584 mg of Ch d was solubilized under magnetic stirring in 36 mL of 0.1 mol L −1 HCl overnight. Then, the solution was heated to 70 • C and the pH was adjusted to pH 12 with 5 mol L −1 NaOH. Next, 526 mg of DIPEA-Cl was added into the reaction mixture and kept under intense magnetic stirring for 90 min (at 70 • C). During the reaction time, the pH was monitored and 5 mol L −1 NaOH was added whenever necessary to maintain the pH at 12. The polymer was purified by dialysis (MWCO 3.5 kDa) on the first day against NaOH (0.05 mol L −1 ) and on the following days against deionized water, with a daily change of solvent until neutral pH. Finally, the polycation was recovered by lyophilization. All other DIPEA-chitosan derivatives were obtained by the same procedure.

PEGylated Chitosan Derivatives by Disulfide Bonds
The chitosan derivatives were grafted with 2 kDa O-(2-mercaptoethyl)-O-methylpolyethylene glycol (PEG-SH) by activation with N-succinimidyl 3-(2-pyridyldithio) propionate (SPDP) [15]. The procedure for PEG-DIPEA 34 Ch d synthesis is described below. Initially, 200 mg of DIPEA 34 Ch d were solubilized with 10 mL of acetic acid (0.03 mol L −1 ) and diluted with PBS buffer (pH 7.4) to obtain a polymer concentration of 10 g L −1 . Next, the pH of the mixture was adjusted to 7.4 with NaOH (5 mol L −1 ) and 3.8 mg of SPDP (solubilized in 1 mL of DMSO) was added dropwise into the polymeric solution and kept under magnetic stirring at room temperature. After 3 h, 24.5 mg of PEG-SH (solubilized in 1 mL of PBS) was added to the reaction mixture and stirred for another 16 h at 40 • C. The derivative was purified by dialysis (MWCO 14 kDa) against PBS and an NaOH solution (pH 8-9) for three days with a daily change of solvent. Finally, the PEG-DIPEA 34 Ch d was recovered by lyophilization. Other PEG-grafted derivatives were obtained by the same procedure. In this reaction, the molar ratio between the PEG-SH and the polymer primary amine groups (NH 2 ) was fixed at 0.02 and the PEG-SH/SPDP ratio was fixed at 1.0 (Table 1).

Rhodamine-Labeled Chitosan Derivative
Aiming to evaluate the internalization of the nanoparticles by the cells, a chitosan derivative was labeled with rhodamine B isothiocyanate (RITC) following a previously reported method [55,56]. Then, 120 mg of PEG-DIPEA15Chd was solubilized in 15 mL of acetic acid (0.1%, v/v) and the pH of the solution was set at 6.2 by the addition of 0.1 mol L −1 NaOH. Next, 3.5 mL of a methanolic RITC solution (0.2%, m/v) was added dropwise to the mixture and kept under magnetic stirring in an N 2 atmosphere for 18 h. After that, the polymer was purified by dialysis (MWCO 3.5 kDa) against NaOH solution (0.05 mol L −1 ) for one day and against deionized water for 10 days, with a daily change of solvent. Finally, the RITC/PEG-DIPEA 15 Ch d was recovered by lyophilization, washed with acetone by a Soxhlet system, and dried at room temperature. The degree of substitution by RITC (DS RITC ) on the derivative was determined by UV-visible spectroscopy using the RITC as the standard to build the analytical curve ( Figure S13, Supplementary Information), as previously described [56].

Hydrogen ( 1 H) and Carbon ( 13 C) Nuclear Magnetic Resonance (NMR) Spectroscopy
The polymer compositions were determined by 1 H NMR spectroscopy using a 400 MHz 400/54 Premium Shielded NMR spectrometer (Agilent Technologies, Santa Clara, CA, USA) at 70 • C, with a 4.0 s acquisition time and relaxation delay of 1.0 s. For each spectrum, about 64 or 128 scans were accumulated. The 13 C NMR analysis of selected polymers was performed at 20 • C on a 400 MHz Avance II NMR spectrometer (Bruker, Billerica, MA, USA) at a 13 C-frequency of 150.9 MHz with a 12 • pulse width. The acquisition time was 0.4 s with a relaxation delay of 2.0 s. For each 13 C NMR spectrum, about 20,000 scans were accumulated. In both analyses, the samples were solubilized in DCl/D 2 O (1:20) to obtain a polymer concentration of 10 g L −1 ( 1 H NMR) or 70 g L −1 ( 13 C NMR). Before dissolution, the polymers were dried at 60 • C under reduced pressure.

Fourier Transform Infrared (FTIR)
The starting chitosan and selected DIPEA derivatives were analyzed with a Vertex 70 Attenuated total reflectance-Fourier transform spectrometer, (Bruker, Ettlingen, Germany). The polymeric samples were dried at 60 • C under reduced pressure and kept inside a desiccator for four hours before analysis. For each sample, 32 scans were collected at 4 cm −1 resolution from 4000 to 400 cm −1 and the spectra were analyzed with the OriginPro 8.5 software.

Gel Permeation Chromatography (GPC)
The GPC was used to estimate the molecular weights and molar-mass dispersity (Ð M ) of the polymers, using monodisperse pullulan samples with molecular weights ranging from 6.2 to 805.0 kDa as standard to build the analytical curve ( Figure S14, Supplementary Information). The measurements were carried out on an LC-20A chromatograph with RID-10A refractive index detection (Shimadzu, Kyoto, Japan). Two linked columns (SB-803 HQ and SB-805-HQ, Shodex, New York, NY, USA), with dimensions of 0.8 cm × 30 cm, were applied for the separation of the polymeric chains. The polymers were dissolved in acetic buffer (0.3 mol L −1 acetic acid/0.2 mol L −1 sodium acetate) pH 4.5 at a final concentration equal to 5 g L −1 . The analysis was performed using the same acetic buffer as eluent with a flow rate of 0.8 mL min −1 and at a temperature of 40 • C.

Ionization Degree and Buffering Capacity
The degree of ionization (α) and the buffering capacity (BC) were evaluated by conductometric and potentiometric titrations. First, the proper amount of polymer (previously dried under reduced pressure at 60 • C) needed for the 2.4 × 10 −4 mol of amine groups were solubilized in 40 mL of HCl (0.1 mol L −1 ) with the ionic strength adjusted to 150 mmol L −1 with NaCl. Next, the polymeric solutions were titrated with a standardized NaOH (0.1 mol L −1 ) solution and the pH and conductivity were monitored. The BC was estimated by the NaOH consumed in a specific pH range, based on the acid-base titration curve. The α as a function of pH was estimated using Equation (4), where V X is the volume of NaOH added to achieve a specific pH value and V 1 and V 2 are the volumes at the beginning (α = 100%) and the end (α = 0) of the deprotonation process, respectively. Both titrations were performed at the same time to obtain V 1 and V 2 (conductometric) and pH (potentiometric).

siRNA Nanoparticle Formulation
Nanoparticles were self-assembled following the complexation method [19,57]. First, a stock solution (approx. 1.0 g L −1 ) was prepared by the solubilization of the polymer sample in 0.1 mol L −1 HCl followed by its dilution in a 50 mmol L −1 phosphate buffer pH 7.4, whose ionic strength (I) was set to 150 mmol L −1 with NaCl. Next, the appropriate volume of this polymeric stock solution was vortexed with siRNA in phosphate buffer to build the nanocarriers, which were kept at room temperature for 30 min before measurement. The final concentration of siRNA in the nanocarrier solution varied from 0.37 to 3.70 µmol L −1 , depending on the analysis. The volume of the polymeric stock solution was adjusted to build nanoparticles at varied N/P ratios (N = amino groups of polymer, P = phosphate groups of siRNA), calculated using Equation (5). Where n siRNA is the number of siRNA moles in the solution and the term 42 refers to the number of phosphate groups per siRNA molecule. The terms m pol and mm pol refer respectively to the polymer mass (grams) in the solution and the average molecular mass (g mol −1 ) of repeating units present in the polymer chain. β is the mean number of amino groups present in the polymer (for example, each 100 units of DIPEA 34 Ch d have 130 amines, then its β is 1.30). The mm pol and the β are found based on polymer composition, which is determined by 1 H NMR spectroscopy. The n siRNA was calculated based on siRNA concentration and its molecular mass.

Electrophoresis Assay: Evaluation of siRNA-Polycation Interaction
The free siRNA and the nanocarriers formulated at an increasing N/P ratio (1 to 10) were applied to 0.8% (m/v) agarose gel containing ethidium bromide (0.4 µg mL −1 ) for siRNA staining. The volume of nanocarrier solution was fixed at 10 µL and its siRNA concentration was 3.70 µmol L −1 . Before their application in the agarose gel, the nanocarriers were mixed with 1.6 µL of a loading dye solution (0.25% bromophenol blue and 25% Ficoll ® 400, m/v). The electrophoresis was performed at 80 V for 75 min using TAE as running buffer. Finally, the resulting gel was documented by a UV transilluminator system.

Zeta Potential (ζ) and Hydrodynamic Diameter (Dh) of Nanoparticles
The zeta potential and the hydrodynamic diameter of nanoparticles were measured using a Zetasizer Nano ZS with a red laser (λ = 633 nm) and detection optics at 173 • (Malvern Panalytical; Malvern, UK) at 25 • C. The nanocarriers were formulated with varied N/P ratios and the siRNA concentration was fixed at 0.37 µmol L −1 . The sample volume was set at 1 mL and the results were expressed as the mean ± SD based on three independent replicates. The Dh measurements were based on Z-average size. Aiming to evaluate the colloidal stability of the nanocarriers, they were formulated as described above and their Dh was measured over time at room temperature. Additionally, to evaluate the nanoparticle behavior in the presence of proteins, their size distribution was monitored over time after the addition of bovine serum albumin (BSA) at a final concentration equal to 40 g L −1 . Between measurements, the samples were kept at 37 • C.

Morphology
The nanoparticles were formulated as described above for the ζ and Dh measurements at an N/P ratio of 10. Next, 1 µL of nanocarrier solution was dripped over a silicon plate that was kept inside a desiccator for solvent evaporation. Then, the nanoparticles were coated with a carbon layer and evaluated by an FSM-6701F Field Emission Gun-Scanning Electron (FEG-SEM) microscope (JEOL, Akishima, Japan). The samples were examined under an accelerating voltage of 2.0 kV.

Cytotoxicity of Polymers and Nanoparticles
RAW 264.7 or 3T3/NIH cells were seeded in a 96-well plate at a density of 20,000 cells per well one day before their treatment with the nanoparticles or polymer solutions. Then, for evaluation of polymer cytotoxicity, the culture medium was replaced by 200 µL of polymeric solution diluted in complete medium at increasing concentrations (0.02, 0.1, and 0.5 g L −1 ). For evaluation of nanocarrier cytotoxicity, the cells received 50 µL of nanoparticle solution (formulated in phosphate buffer at varied N/P ratios) plus 150 µL of complete medium. The final concentration of siRNA applied to cells was 200 nmol L −1 . After 24 h of incubation, the cell viability was evaluated using a CellTiter96 ® Aqueous One Solution kit (Promega Corporation, Fitchburg, MA, USA) following the manufacturer instructions. For this purpose, the medium was replaced by 100 µL of incomplete medium plus 20 µL of MTS, the cells were incubated for 3 h and then the absorbances were measured at 490 nm on an Elx 808 microplate reader (Agilent Technologies, Winooski, VT, USA). The following controls were used: cells treated with 3% (m/v) sodium dodecyl sulfate (SDS), DMEM/HCl solution, the phosphate buffer used for nanoparticle formulation, and the non-treated cells. The cell viability was determined by comparison with non-treated cells (100% of cell viability), and the results were expressed as the mean ± SD for three independents assays. Statistical analysis was performed using the unpaired Student t-test following the Holm-Sidak method at a significance level of 0.05.

Confocal Microscopy: Cellular Uptake and GFP Knockdown
In the cellular uptake study, RAW 264.7 macrophages were seeded at a density of 6.0 × 10 4 cells per well in a 24-well plate containing a square glass coverslip at the bottom and kept overnight in incubation. The next day, the cells were washed with PBS buffer prior to the addition of 400 µL of siRNA-nanocarrier solution plus 600 µL of incomplete medium.
The nanocarriers were formulated at N/P ratio 10 and the final concentration of siRNA was 50 nmol L −1 . Here, siRNAs labeled with cyanine 5 (siRNA-Cy5) or 5 -carboxyfluorescein (siRNA-FAM) were used. After 4 h of incubation, the cells were fixed by 1 mL 4% (m/v) paraformaldehyde for 15 min and nuclei stained by 1 mL of 4 ,6-diamidine-2 -phenylindole (DAPI, 1 mg L −1 ) for 10 min. After washing the cells with PBS, the coverslips were mounted on glass slides containing glycerol and the images were captured by an LSM 710 microscope (Zeiss, Oberkochen, Germany) using ZEN 2010 software.
In the GFP knockdown study, HeLa-GFP cells (6.0 × 10 4 cells/well) were grown in a square glass coverslip as described above for RAW 264.7 macrophages. After washing with PBS buffer, 400 µL of nanoparticles (N/P ratio 10, pH 7.4, and I = 150 mmol L −1 ) plus 600 µL of DMEM were added. In this experiment, the nanoparticles were formulated with siRNA anti-GFP, whose final concentration was 200 nmol L −1 . After 4 h of incubation, the medium was replaced by complete medium and the HeLa-GFP cells were incubated for another 44 h. Next, the cells were fixed, nuclei stained and analyzed as described for the cellular uptake study of RAW 264.7 macrophages.

Transfection Efficiency Study
The transfection efficiency was evaluated by the knockdown of TNFα protein on RAW 264.7 macrophages treated with nanocarriers loaded with siRNA anti-TNFα, as previously reported [16]. Briefly, RAW 264.7 cells were seeded in a 24-well plate at a density of 1.8 × 10 5 cells per well and incubated overnight. Then, the cells were washed with PBS buffer and received 450 µL of nanoparticles/siRNA solution plus 1.35 mL of medium (with or without 10% FBS). The nanocarriers were formulated at various N/P ratios (1 to 10) and the final concentration of siRNA was 200 nmol L −1 . After 5 h of incubation, the medium was replaced by 1 mL of complete medium and the cells were incubated for a further 19 h. After that, the medium was removed and the macrophages received 300 µL of complete medium with lipopolysaccharide (LPS; 100 ng mL −1 ) from Escherichia coli. The plate was incubated for 4 h, then the supernatant was collected, centrifuged (12,000 RCF), and stored (−20 • C) for TNFα quantification using a Murine TNFα Standard TMB ELISA Development kit (Peprotech, Rocky Hill, CT, USA), following the manufacturer instructions. The relative TNFα expression was determined by comparison with cells treated with LPS only (100% of TNFα expression). As controls, cells non-stimulated by LPS and cells treated with siRNA only were used. Experiments were performed in triplicate and expressed as the mean ± SD. Statistical analysis was performed using the unpaired Student t-test following the Holm-Sidak method at a significance level of 0.05.

Conclusions
Diisopropylethylamine-chitosan derivatives with varied compositions were synthesized to develop a potential carrier for siRNA delivery. To overcome the low affinity of chitosan with siRNA at neutral pH, the degree of ionization of the polycations was tuned by the content of diisopropylethylamine groups. Vectors with degrees of ionization varying from 30 to 65% were shown to form positively charged siRNA nanoparticles with good colloidal stability at pH 7.4. The insertion of DIPEA increased the siRNA-polycation interactions and allowed positively charged nanoparticles with small sizes at N/P ratios 5 and 10 to be obtained. Derivatives were also grafted with PEG and stability studies showed that nanoparticles having spherical-like shapes retained their sizes in the range of 100-200 nm with low polydispersities 24 h after their formulation and revealed good stability even in a protein medium. Chitosan-derivatives and their siRNA nanocarriers showed low cytotoxicity (cell viability above 85%) in different cell lines. Confocal microscopy showed a successful uptake of nanocarriers by RAW 264.7 macrophages and a promising ability to silence GFP in HeLa cells. Additionally, the siRNA nanocarriers were able to promote TNFα knockdown in LPS-stimulated macrophages, even in FBS medium, and the results showed close agreement with nanocarrier physicochemical properties. Overall, this study provides new systems based on chitosan with an enhanced siRNA condensation process and highlights the potential of these derivatives for in vivo studies.