3D printed systems for colon-specific delivery of camptothecin-loaded chitosan micelles

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ “This is an Accepted Manuscript of an article published by Elsevier in European Journal of Pharmaceutics and Biopharmaceutics on October 2021, available at: https://doi.org/10.1016/j.ejpb.2021.07.005 .” 1 3D printed systems for colon-specific delivery of 1 camptothecin-loaded chitosan micelles 2 3 Andreia Almeidaa, b, c, †, Vicente Linaresd, †, Gloria Mora-Castañod, Marta Casasd, 4 Isidoro Caraballod, Bruno Sarmentoa, b, e* 5 6 a INEB - Instituto Nacional de Engenharia Biomédica, Universidade do Porto, Rua 7 Alfredo Allen, 208, 4200-135, Porto, Portugal 8 b i3S – Instituto de Investigação e Inovação em Saúde, Universidade do Porto, 9 Rua Alfredo Allen, 208, 4200-135, Porto, Portugal 10 c Instituto de Ciências Biomédicas Abel Salazar, Universidade do Porto, Rua 11 Jorge Viterbo Ferreira, 228, 4050-313, Porto, Portugal 12 d Department of Pharmacy and Pharmaceutical Technology, Faculty of 13 Pharmacy, Universidad de Sevilla, C/Profesor García González 2, 41012 Seville, 14 Spain. 15 e CESPU, Instituto de Investigação e Formação Avançada em Ciências e 16 Tecnologias da Saúde, Rua Central da Gandra, 137, 4585-116, Gandra, Portugal 17 † Equal contribution 18 *Corresponding author: Tel +351 226074949, [email protected] 19 20 Abstract: The use of 3D printing technology in the manufacturing of drug delivery 21 systems has expanded and benefit of a customized care. The ability to create 22 tailor-made structures filled with drugs/delivery systems with suitable drug 23 dosage is especially appealing in the field of nanomedicine. In this work, chitosan- 24 based polymeric micelles loaded with camptothecin (CPT) were incorporated into 25 3D printing systems (printfills) sealed with an enteric layer, aiming to protect the 26 nanosystems from the harsh environment of the gastrointestinal tract (GIT). 27 Polymeric micelles and printfills were fully characterized and, a simulated 28 digestion of the 3D systems upon an oral administration was performed. The 29 printfills maintained intact at the simulated gastric pH of the stomach and, only 30 released the micelles at the colonic pH. From there, the dissolution media was 31 used to recreate the intestinal absorption and, chitosan micelles showed a 32 significant increase of the CPT permeability compared to the free drug, reaching 33 2 an apparent permeability coefficient (Papp) of around 9×10-6 cm/s in a 3D 34 intestinal cell-based model. The combination of 3D printing with nanotechnology 35 appears to have great potential for the colon-specific release of polymeric 36 micelles, thereby increasing intestinal absorption while protecting the 37 system/drug from degradation throughout the GIT. 38 39 Keywords: 3D printing; polymeric micelles; oral administration; colon-specific 40 delivery; intestinal permeability. 41 42 1. Introduction 43 The oral route is the most preferred via of administration of drugs due to the 44 greatest patient compliance and ease of administration. In the market, drugs for 45 oral administration are the most used worldwide, however, some drugs present 46 poor oral availability and cannot be administered orally. This is the case of several 47 anticancer drugs, that despite being effective against cancer, cannot be 48 administered orally due to the poor aqueous solubility and low intestinal 49 permeability [1, 2]. 50 Camptothecin (CPT) was found in the bark of the Chinese tree, Camptotheca 51 acuminate [3] and, is a potent anticancer drug against several types of cancer, 52 including colorectal cancer. Its mechanism of action is based on the inhibition of 53 the nuclear DNA topoisomerase I (TOP1) by binding the complex DNA-TOP1, 54 leading to DNA damage and, subsequently, to cell apoptosis [4]. CPT belongs to 55 the Class IV of the Biopharmaceutical Classification System (BCS), which means 56 to have low solubility and low permeability. Indeed, CPT present some toxicity, 57 its physical-chemical stability is pH dependent and can suffer hydrolysis if 58 administered orally due to the chemical and enzymatical barriers of the 59 gastrointestinal tract (GIT). Additionally, the poor oral bioavailability is a result of 60 the high level of the first pass metabolism and the action of drug efflux 61 transporters, as the P-glycoproteins (P-gp) [5]. Thus, there is a great need to 62 overcome these drawbacks in order to be able to deliver CPT efficiently after an 63 oral administration. 64 Nanotechnology has changed the field of drug delivery and, particular attention 65 has been given to oral administration of drugs [6-8]. Polymeric micelles are 66 nanosystems constituted by a hydrophobic core and a hydrophilic shell widely 67 3 used in the encapsulation of hydrophobic drugs. These systems are described 68 as potential oral vehicle of delivery since they can protect the drug from the harsh 69 environment of the GIT, increasing its stability. Polymeric micelles are attractive 70 also due to the high drug encapsulation capacity, where a therapeutic dose can 71 be reached easily, avoiding side-effects and avoid efflux pumps of the epithelium 72 [9-11]. 73 Chitosan is a natural polymer extensively used in the production of polymeric 74 micelles due to the ease of modification into an amphiphilic polymer with self- 75 assembly properties. Moreover, chitosan has mucoadhesive properties and has 76 the ability to temporarily open the tight junctions of the epithelium, properties that 77 can enhance the intestinal permeability of drugs [12]. 78 Three-dimensional printing (3DP) is gaining interest in pharmaceuticals by virtue 79 of its ability to manufacture medicines with freeform geometries and flexibility in 80 dose, drug release, etc. Thanks to its versatility, 3DP for medicines has become 81 an interesting tool for drug personalization specially for paediatric and geriatric 82 population due to their particular physiological needs [13-16]. 83 Several 3DP techniques have been successfully employed for medicine 84 manufacturing. Between them, fused deposition modelling (FDM) is one of the 85 most common for its easy use, acceptable level of accuracy and good 86 reproducibility. However, one of the main drawbacks is the impossibility to print 87 thermolabile drugs due to the high temperatures achieved both during the drug 88 loading process of the thermoplastic filament, and the 3D printing process [17]. 89 Consequently, for thermo-sensitive pharmaceutical formulations, FDM does not 90 appear to be the best option. Nevertheless, we have already shown the potential 91 of combining two 3DP techniques (FDM and injection volume filling (IVF)) in one 92 bioprinter to overcome this limitation [18]. IVF is based on a syringes system 93 which works with low pressure injections, being useful for dispersions, 94 suspensions, and low-viscosity pastes. As a result, 3D printed systems filled with 95 a liquid or semisolid pharmaceutical formulation, called printfills, were developed. 96 Thanks to the combination of FDM and IVF, the drug and a pH-sensitive polymer 97 were incorporated in the scaffold produced by FDM at once, in a continuous and 98 automated process. 99 Targeted drug delivery together with dose personalization can improve the 100 patient´s standard of living [19, 20]. Therefore, the ease to design modified drug 101 4 release systems is a valuable advantage of 3DP for medicines. In this sense, 102 colonic drug release becomes a potential area for 3DP, especially for the 103 treatment of colonic diseases like Crohn disease, cancer, or ulcers [21, 22]. 104 The aim of this study was to protect chitosan-based micelles loading CPT from 105 the harsh environment of the GIT, allowing the release of micelles in the colon, 106 improving the intestinal permeability of CPT. Free CPT and loaded into micelles 107 [23] were incorporated in printfills manufactured through an automated 108 manufacturing 3D printing process. 109 110 2. Materials and methods 111 2.1. Materials 112 Polylactic acid (PLA) was chosen as biodegradable thermoplastic filament for 113 FDM (grade Ingeo 3D850, peak melt temperature 165-180ºC, melting point 114 220ºC, density 1.25 g/cm3), supplied by Leon3D (Spain). Eudragit FS30D 115 (Evonik, Germany) was employed as delayed release polymer. CPT was 116 purchased from Hangzhou ROYAL Import & Export Co., Ltd; 3.5 kDa cutoff 117 dialysis membrane (regenerated cellulose, Spectra/Por® 3 Dialysis Tubing, 118 Fisher Scientific) was used to produce the micelles. 119 120 2.2. Methods 121 2.2.1. Production of camptothecin-loaded polymeric micelles 122 Polymeric micelles based on chitosan were obtained by a dialysis method, as 123 previously described [23, 24], however, the CPT association efficiency (AE) was 124 optimized and improved from the last published work [23]. Briefly, 10 mg of O- 125 methyl-O′-succinylpolyethylene glycol-chitosan-oleic acid (mPEG-CS-OA 126 polymer) were dissolved in 4 mL of dimethyl sulfoxide (DMSO):HCl 0.1 M (9:1) in 127 a glass tube and left overnight at RT until its complete dissolution. CPT, 128 previously dissolved in DMSO at 1 mg/mL, was added to the prior solution at 129 different drug loadings (w/w). The mixture continued under stirring and, after 6 h, 130 dialyzed with three-cycle renewal against water to remove the DMSO and the 131 non-associated CPT. Afterwards, CPT-loaded micelles were centrifuged during 132 20 min at 3110 g to remove the drug precipitate during dialysis, frozen with 0.5% 133 (w/v) trehalose as a cryoprotectant, and freeze-dried to achieve the micellar 134 powder. 135 5 136 2.2.2. Characterization of camptothecin-loaded micelles 137 Mean particle size, polydispersity index (PdI), and surface charge were assessed 138 by dynamic light scattering (DLS) and Laser Doppler Micro-electrophoresis 139 (ZetaSizer Nano ZS, Malvern, UK), respectively, with an angle of 173° at 25 °C. 140 Polymeric micelles were analyzed freeze-drying resuspended in MilliQ water at 141 1% (w/v) concentration. Morphological analysis CPT-loaded mPEG-CS-OA 142 micelles was assessed by transmission electron microscopy (TEM). Micelles 143 were prepared on a grid treated with uranyl acetate and then examined in a JEM- 144 1210 Transmission Electron Microscope (JEOL Ltd., Tokyo, Japan) operating at 145 120 kV. The AE was determined by the indirect method by high-performance 146 liquid chromatography (HPLC) and calculated as described in Eq. (1). Drug 147 loading (DL) was obtained according to Eq. (2). Briefly, the HPLC analysis was 148 performed in a Shimadzu UFLC Prominence System (USA) equipped with two 149 Pumps LC-20AD, a column oven CTO-20AC, an autosampler SIL-20AC, a 150 System Controller CBM-20A, a degasser DGU- 20A5, a RF-10Axl fluorescence 151 detector coupled to the LC System and a LC Solution, Version 1.24 SP1 152 (Shimadzu), and the method used for CPT quantification was described by 153 Martins et al., [25]. 154 155 𝐴𝐸% = 𝑤𝑒𝑖𝑔ℎ𝑡 𝑜𝑓𝐶𝑃𝑇 𝑎𝑑𝑑𝑒𝑑−𝑤𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝑓𝑟𝑒𝑒 𝐶𝑃𝑇 𝑤𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝐶𝑃𝑇 𝑎𝑑𝑑𝑒𝑑 ×100 Eq. (1) 156 𝐷𝐿% = 𝑤𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝐶𝑃𝑇 𝑎𝑑𝑑𝑒𝑑−𝑤𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝑓𝑟𝑒𝑒 𝐶𝑃𝑇 𝑡𝑜𝑡𝑎𝑙 𝑚𝑖𝑐𝑒𝑙𝑙𝑒𝑠 𝑤𝑒𝑖𝑔ℎ𝑡 ×100 Eq. (2) 157 158 2.2.3. 3D printed systems production 159 Printfills (3D printed systems) were manufactured employing a bioprinter 160 (Regemat 3D S.L., Spain) which combines FDM and IVF techniques. Two 161 different batches of printfills were designed by the software Regemat 3D 162 DESIGNER. First batch contained free CPT, whereas the second one CPT 163 encapsulated within polymeric micelles. 164 165 2.2.3.1 Preparation of drug printable ink 166 Suitable drug printable inks were prepared for each batch: Firstly, free CPT 167 dissolved in DMSO with a concentration of 5 mg/mL was diluted in a dissolution 168 6 of 10 mL DMSO:dH2O at a ratio of 7:3. Secondly, to obtain printable ink with 169 lyophilized CPT loaded-micelles, 4.5 mL of dH2O were employed to rehydrate 170 and resuspend the micelles. Both printable inks were prepared with the minimum 171 light exposure to reduce the interaction with the drug. 172 173 2.2.3.2. 3D printing process 174 Printing settings for each batch are shown in Tables S1 and S2 in the 175 supplementary materials. Due to the different composition of the two drug inks, 176 systems prepared with micelles required more injection points and lower deposit 177 speed in order to ensure the same volume incorporated in the structure. Also, the 178 amount of Eudragit FS30D was higher to ensure the top face sealing of the printfill 179 (Table S1 and Table S2). 180 According to the digital design of the 3D printed system, printfills were built layer 181 by layer (Figure S1). FDM starts the printing process building the scaffold in the 182 platform by extruding 2 bottom solid layers. The extrusion continues with parallels 183 lines separated by 1.8 mm and 2 mm for free CPT batches and CPT 184 encapsulated, respectively. The next layer is built with perpendicular lines respect 185 to the lower one, resulting in a quadrilateral mesh of 1.8×1.8 mm and 2×2 mm 186 for each batch. This process continues until the18th layer of the scaffold when the 187 two automated syringes of IVF inject inside the scaffold. The first syringe 188 contained printable ink with the free CPT and injected 200 µL in the 18th layer in 189 4 different points of the mesh, whereas 12 injection points were necessary for 190 micelles. Again, FDM continues extruding a couple of layers to finish the 191 scaffolds. Finally, IVF with the second syringe injected the set volume of the 192 delaying polymer in the last layer of the system sealing the top face of the printfill. 193 All 3DP process from the FDM extrusion until the last injection of IVF was a 194 continuous and automatic manufacturing. Once the printfills were finished, were 195 dried in a drying oven (632 plus Nahita, Auxilab S.L., Navarra, Spain) at 29 ºC for 196 24 h to evaporate the solvent (Figure S2 in the supplementary materials). 197 198 2.2.4. Characterization of the 3D printed systems 199 2.2.4.1. Physical characteristics of 3D printed systems 200 7 Printfills were weighed in a precision balance (Scaltec, type SBC31) and 201 thickness and diameter were measured using a digital micrometer (Comecta, 202 SA). 203 204 2.2.4.2. Scanning electron microscopy (SEM) 205 The interior mesh and transversal section of the delaying polymer were evaluated 206 in the Microscopy Service of the CITIUS in the University of Seville by using 207 scanning electron microscopy (SEM) with a FEI TENEO electronic microscope 208 (FEI Company, USA), operating at 5 kV. Printfills were coated with a 12 nm thin 209 Pt/Pd layer with Leica EM SCD500 high vacuum sputter coater. 210 211 2.2.5. Camptothecin release from the printfills 212 The dissolution studies were carried out using a USP Apparatus I Sotax AT7 213 smart (Allschwil, Switzerland) at 37 ± 0.5 °C with a rotation speed of 50 rpm. 214 Printfills were exposed to different switch solutions (SS) to mimic the intestinal 215 transit according to Schellekens et al., [26]. Their methodology was modified to 216 prolong the colon residence time, as shown in Table S3 in the supplementary 217 materials. The different SS (Table S4 in the supplementary materials) were added 218 from one phase to another to obtain the required pH. Samples were withdrawn at 219 specific interval times: 1, 2, 2.5, 3.5, 4.0, 4.2, 4.3, 4.5, 4.8, 5.0, 5.3, 5.5, 6.0, 6.5, 220 7.0, 7.5, 8 hours. The amount of CPT released from the printfills at each time- 221 point was quantified by HPLC, as previously described [25]. The results are 222 present as a cumulative percentage of CPT released at each analyzed time 223 (n=6). Drug release data (Mt/M∞≤0.8) was also analysed using Microsoft Excel 224 2010 (Microsoft, Albuquerque, NM, USA) according to zero order, Higuchi [27], 225 Korsmeyer et al., [28] and Peppas and Sahlin [29], following the Equations (3) – 226 (5): 227 228 𝑀𝑡 𝑀∞= 𝑏𝑡12 ⁄ Eq. (3) 229 𝑀𝑡 𝑀∞= 𝑘𝑘𝑡𝑛 Eq. (4) 230 𝑀𝑡 𝑀∞= 𝑘𝑑𝑡𝑚+ 𝑘𝑟𝑡2𝑚 Eq. (5) 231 232 8 where Mt/M∞ is the drug released fraction at time t (the drug loading was 233 considered as M∞), b and kk are kinetic constants characteristic of the 234 drug/polymer system, t is the release time, n is the release exponent that depends 235 on the release mechanism and the shape of the matrix tested, kd and kr are the 236 diffusion and relaxation rate constants, respectively, m is the purely Fickian 237 diffusion exponent for a device of any geometrical shape which exhibits controlled 238 release. Determination coefficient (r2) was used to test the applicability of the 239 release models, calculated in linear regression for zero order, Higuchi and 240 Korsmeyer equations or quadratic regression for Peppas and Sahlin model. 241 242 2.2.6. Cell culture 243 C2BBe1 [Caco-2 clone] (human colorectal adenocarcinoma) cells were acquired 244 from ATCC and HT29-MTX mucus-producing cells were kindly provided by Dr. T. 245 Lesuffleur (INSERM U178, Villejuif, France). Both cell lines were maintained in 246 complete medium comprising Dulbecco’s Modified Eagle’s Medium (DMEM, 247 ATCC) supplemented with 10% (v/v) Fetal Bovine Serum (FBS, Biochrom) and 248 1% (v/v) antibiotic-antimitotic mixture (final concentration of 100 U/mL penicillin 249 and 100 U/mL streptomycin, Biowest). Human intestinal fibroblast (HIF) primary 250 cells were obtained from ScienCell and maintained in culture with Fibroblast 251 Medium (FM) supplemented with 2% FBS, 1% of Fibroblast Growth Supplement 252 (FGS), and 1% penicillin/streptomycin solution (all supplements from ScienCell). 253 HPMEC-St1.6R cells were kindly provided by Professor C. James Kirkpatrick 254 (Institute of Pathology, Johannes Gutenberg University of Mainz, Germany) and 255 were maintained with M199 medium (Sigma), supplemented with 20% FBS 256 (Biochrom), 1% (v/v) antibiotic-antimitotic mixture (final concentration of 100 257 U/mL penicillin and 100 U/mL streptomycin, Biowest), 25 µg/mL of Endothelial 258 Cell Growth Supplement (ECGS) and 25 µg/mL of heparin (both from Sigma), 259 and 100 µg/mL L-glutamine (LabClinics). All the cells were cultures in 75 cm2 T- 260 flasks, inside an incubator at 37 ºC and 5% CO2 and 95% relative humidity. 261 262 2.2.7. Metabolic assay using drug and micelle released medium 263 To assess the safety of the micelles and the free drug to be placed in the 264 permeability assay, the released media from the printfills of the CPT and CPT- 265 loaded micelles was used to perform a resazurin assay. Briefly, 2×104 Caco-2 266 15 Drug release data (Mt/M∞< 0.8) from printfills with CPT-loaded micelles were 447 analyzed according to zero order, Higuchi, Korsmeyer–Peppas and, Peppas and 448 Sahlin equations (Table 3). In the case of printfills with free CPT, the fast drug 449 release did not allow to perform this study. Determination coefficient values 450 indicates a poor fit for the different kinetic models, that can be explained because 451 drug release profile reaches an asymptote at around 35% w/w. Thus, in order to 452 obtain information about drug release mechanism from this profile, data has been 453 analyzed until 260 min, i.e., before reaching asymptote. In this case, the best fit 454 corresponds to Peppas and Sahlin equation, with a clear prevalence of Kr over 455 Kd. This indicates that the drug release mechanism was controlled mainly by 456 erosion. The n value from Korsmeyer equation indicates a super case II transport, 457 confirming that that the mechanism of drug release was erosion/relaxation 458 controlled. 459 460 461 Figure 2 – Cumulative CPT release from the printfills in simulated gastrointestinal fluids. 462 Squares represent the release of CPT-loaded micelles and circles represent the release 463 of free CPT. Error bars represent mean ± SD (n = 6). 464 465 Table 3 – Mathematical modelling and drug release kinetics from the enteric printfills 466 with CPT-loaded micelles up to 260 min. 467 Zero order Higuchi Korsmeyer Peppas and Sahlin r2 b (min-0.5) r2 Kk (min-n) n r2 Kd(min-m) Kr(min-2m) r2 0.8586 0.0362 0.8995 1.9113x10-6 2.0311 0.818 0.1245 -0.0025 0.9417 b, Higuchi kinetic constant; n, release exponent; kk, Korsmeyer kinetic constant; kd, Peppas 468 diffusion kinetic constant; kr, Peppas relaxation kinetic constant; r2, determination coefficient. 469 16 470 3.4. In vitro cytocompatibility of polymeric micelles 471 In order to assess the safety and biocompatibility of the micelles and free CPT 472 loaded in the printfills and released in the simulated gastrointestinal fluids (GIF), 473 a resazurin assay was performed in Caco-2 cells during 4, 6 and 24 h of 474 incubation. The micelles and free drug present in the dissolution medium after 6h 475 were used to evaluate its cytotoxic potential. As we described before, the release 476 of the micelles and drug only occurs at intestinal pH, thus, it only makes sense to 477 evaluate safety by this way. 478 As can be seen in Figure 3, both polymeric micelles and free drug present in the 479 dissolution media showed no cellular cytotoxicity against Caco-2 cells, a 480 colorectal cancer cell line. Moreover, the metabolic activity of these cells 481 increased up to 100%, which may be due to the presence of simulated GIF. 482 Actually, according to the ISO 10993-5 guideline [38], it is reported that above 483 70% of cell viability is considered as safe, when testing in vitro formulations. 484 Moreover, since we are working with a relatively low CPT concentration (< 10 μg) 485 due to the slow CPT release and, based on previous published works [35, 39- 486 41], it was expected no cytotoxic effect of the free CPT and CPT-loaded micelles 487 present in the dissolution medium of the in vitro CPT release. 488 489 Figure 3 – Cellular viability of the dissolution medium from the printfills containing CPT- 490 loaded micelles (blue columns) and free CPT (pink columns) after 4, 6 and 24 h of 491 incubation with Caco-2 cells. Error bars represent mean ± SD (n = 6). 492 493 17 3.5. Permeability of camptothecin after simulated gastrointestinal 494 passage 495 Printfills were prepared and filled with CPT-loaded micelles and free CPT. Then, 496 they were placed in a simulated GIF to release both micelles and drug at the 497 intestine to mimic oral intake. The dissolution medium at 6 h since the dissolution 498 assay started was used to perform the permeability study. 499 To evaluate the CPT permeability, two different intestinal models were used, the 500 2D Caco-2 standard model and a 3D model, as previously described. As can be 501 seen in Figure 4, the CPT permeability from the micelles was higher than the free 502 drug in both models with significative differences at the last incubation times. The 503 CPT permeability in the 3D model was around 27% which, compared to the 504 standard model (20%), an enhanced intestinal permeability was verified. Since 505 the 3D model represents the human intestine more closely, we are facing a drug 506 permeability closer to the in vivo situation. Moreover, the Papp presented 507 significative differences between micelles and free drug, where in the 3D model 508 the Papp of CPT from the micelles reached values around 9×10-6 cm/s, which 509 represents a great increase of the CPT permeability and bioavailability (Figure 510 4C). The free drug maintained the Papp around 5×10-6 cm/s in both models, which 511 also describes the permeability of the free drug as being almost half of the 512 permeability of the drug by the micelles, in the case of the 3D model. Silva et al., 513 described CPT permeability of 4×10-6 cm/s and 8×10-6 cm/s for a similar system 514 in a Caco-2 and a co-culture 2D model, respectively [35], which is very close to 515 the results found in this work. 516 As it is widely known, chitosan has the ability to open the tight junctions of the 517 epithelium [42]. In addition, chitosan has mucoadhesive properties and, in a more 518 realistic model, as the 3D model, chitosan micelles can stick to the mucus layer 519 and release the drug, also contributing to the CPT permeability enhancement and 520 bioavailability. Indeed, other authors already described the interaction of chitosan 521 with mucus layer, leading to a permeability increase [43, 44]. 522 The TEER maintained stable during the experiment for both micelles and free 523 drug, which is an indicative of the safety of the tested formulation [45]. The slight 524 decrease observed after 1 h of the experiment starting may be related to the 525 permeability increase, possible due to the opening of the tight junctions of the 526 epithelium [46], which then increasing again in the end in the experiment. 527 18 Overall, chitosan micelles and free drug were maintained stable after simulated 528 oral administration and were protected within the printfills and only released at 529 the intestinal pH, where the permeability was tested. Micelles significantly 530 increased the CPT permeability in both models, especially in the 3D model, that 531 better mimics what occurs in vivo. We are facing an increase in the bioavailability 532 of CPT after an oral administration, which can be very useful for the treatment of 533 various diseases, as the cancer therapy. 534 535 Figure 4 – Intestinal permeability and respective TEER values of the CPT-loaded 536 micelles and free CPT after being released in the simulated GIF. (A) CPT permeability 537 across Caco-2 standard model; (B) CPT permeability across a 3D intestinal model; (C) 538 apparent permeability coefficients (Papp) of CPT-loaded micelles and free CPT across 539 Caco-2 monoculture model and across the 3D model. All experiments were conducted 540 from the apical to basolateral side in HBSS at 37 °C. Error bars represent mean ± SD (n 541 = 3). *p<0.05, ***p<0.001 and ****p<0.0001 denotes a significant difference by two-way 542 analysis of variance (ANOVA) Sidak’s multiple comparison test. 543 544 3.6. Cellular integrity of the intestinal models 545 The TEER represents the integrity of the monolayer as previously mentioned, 546 and thus, TEER values during 21 days of culture can be seen in the 547 supplementary material (Figure S4). This can guarantee that the monolayer was 548 formed and ready for the permeability studies. Also, during the permeability 549 assay, the TEER was registered and, as described above, the values were 550 maintained at the end of the experiment, meaning the permeability obtained was 551 not due to the monolayer disruption. Moreover, to ensure the integrity of the 552 membrane after the permeability experiment, hematoxylin and eosin (H&E) 553 staining was performed for both models and conditions (Figure 6). For this assay, 554 two dyes were used, hematoxylin and eosin, which are basic and acidic dyes, 555 19 respectively. Hematoxylin stains acidic structures in purple, as the nucleus and, 556 eosin dyes basic structures in pink, as the cytoplasm and extracellular matrix [47]. 557 Taking this into account, in Figure 6 is possible to observe a continuous 558 monolayer in both models, corroborating with the TEER values, i. e., the cellular 559 membrane integrity was maintained. Also, it is possible to visualize some portions 560 of the monolayer where the membrane suffered disruption, but it was due to the 561 handling procedure, which happens sometimes. In the case of the 3D model, it is 562 possible to see a collagen layer with fibroblast embedded, as described in the 563 methods section. Absence of cytotoxicity of chitosan-based micelles was also 564 demonstrated by others authors [48-50]. Overall, the TEER values and the H&E 565 staining support the hypothesis that the increased permeability is not due to the 566 membrane disruption because, as demonstrated, it remains intact and tightly 567 formed during and after the permeability assay. 568 569 570 Figure 6 – H&E staining to address the cellular integrity when cells were exposed to the 571 dissolution medium containing CPT-loaded micelles and free CPT for the permeability 572 assay. Cytoplasm stained with pink and nucleus with purple. The Transwell® membrane 573 is transparent and is right below the cellular monolayer. Scale bar = 100 µm. 574 575 Conclusion 576 Chitosan-based micelles loaded with CPT presenting an average diameter <200 577 nm were efficiently filled into 3D systems covered by an enteric layer. It was 578 possible to verify the Eudragit top layer by SEM, as well as, the presence of the 579 20 polymeric micelles and free drug inside of the printfills. The dissolution assay 580 demonstrated the printfills are able to maintain the micelles intact until they reach 581 the intestinal pH. From the dissolution media mimicking a digestion upon oral 582 administration of the printfills, it was possible to verify a significant increase of the 583 intestinal permeability in both Caco-2 and 3D intestinal models. Thus, printfills 584 filled with polymeric micelles were able to maintain them intact along the 585 simulated gastrointestinal digestion and, specifically delivered it at the colonic pH. 586 In this regard, it was possible to substantially increase the CPT intestinal 587 absorption and, consequently, its oral availability avoid the enzymatic 588 degradation and metabolization. This work showed a tremendous potential for a 589 colonic-specific delivery of CPT-loaded micelles with proven improved intestinal 590 absorption. 591 592 Acknowledgements 593 This article is based upon work from COST Action CA15216 “European Network 594 of Bioadhesion Expertise: Fundamental Knowledge to Inspire Advanced Bonding 595 Technologies” supported by COST (European Cooperation in Science and 596 Technology) and by FCT - Fundação para a Ciência e a Tecnologia/ Ministério 597 da Ciência, Tecnologia e Ensino Superior in the framework of the project “Institute 598 for Research and Innovation in Health Sciences” (UID/BIM/04293/2019). Andreia 599 Almeida (SFRH/BD/118721/2016) would like to thank to FCT, Portugal for 600 financial support. The authors also acknowledge the support of the following i3S 601 Scientific Platforms: Biointerfaces and Nanotechnology (FCT - 602 UID/BIM/04293/2019) and Histology and Electron Microscopy (PPBI-POCI-01- 603 0145-FEDER-022122). The authors would like to thank Ministerio de Ciencia, 604 Innovación y Universidades of Spain (Grant RTI2018-095041-B-C31) for financial 605 support. 606 607 Author Contributions 608 A.A. and V.L., Conceptualization, Methodology, Writing - original draft; M.C., 609 Conceptualization, Methodology; G.M., Conceptualization; I.C., 610 Conceptualization, Supervision, Writing - review & editing; B.S., 611 Conceptualization, Supervision, Writing - review & editing. 612 613 21 Declaration of Interest: none. 614 615 References 616 [1] L. Tu, G. Wang, N. Qi, W. Wu, W. Zhang, J. Feng, Multi-functional chitosan 617 polymeric micelles as oral paclitaxel delivery systems for enhanced bioavailability 618 and anti-tumor efficacy, Int. J. Pharm., 578 (2020) 119105. 619 [2] G. Qu, S. Hou, D. Qu, C. Tian, J. Zhu, L. Xue, C. Ju, C. Zhang, Self-assembled 620 micelles based on N-octyl-N’-phthalyl-O-phosphoryl chitosan derivative as an 621 effective oral carrier of paclitaxel, Carbohydr. Polym., 207 (2019) 428-439. 622 [3] M.E. Wall, M.C. Wani, Camptothecin and taxol: discovery to clinic—thirteenth 623 Bruce F. Cain Memorial Award Lecture, Cancer Res., 55 (1995) 753-760. 624 [4] Y. Pommier, Topoisomerase I inhibitors: camptothecins and beyond, Nat. 625 Rev. Cancer, 6 (2006) 789-802. 626 [5] Y. Xue, C. Ma, I. Hanna, G. Pan, Intestinal Transporter-Associated Drug 627 Absorption and Toxicity, Adv. Exp. Med. Biol. , (2019) 361-405. 628 [6] G. Chen, D. Svirskis, W. Lu, M. Ying, Y. Huang, J. Wen, N-trimethyl chitosan 629 nanoparticles and CSKSSDYQC peptide: N-trimethyl chitosan conjugates 630 enhance the oral bioavailability of gemcitabine to treat breast cancer, J. Control. 631 Release, 277 (2018) 142-153. 632 [7] R. He, C. Yin, Trimethyl chitosan based conjugates for oral and intravenous 633 delivery of paclitaxel, Acta Biomater., 53 (2017) 355-366. 634 [8] R. Kumar, A. Sirvi, S. Kaur, S.K. Samal, S. Roy, A.T. Sangamwar, Polymeric 635 micelles based on amphiphilic oleic acid modified carboxymethyl chitosan for oral 636 drug delivery of bcs class iv compound: Intestinal permeability and 637 pharmacokinetic evaluation, Eur. J. Pharm. Sci., 153 (2020) 105466. 638 [9] W.Y. Hu, Z.M. Wu, Q.Q. Yang, Y.J. Liu, J. Li, C.Y. Zhang, Smart pH- 639 responsive polymeric micelles for programmed oral delivery of insulin, Colloids 640 Surf. B, 183 (2019) 110443. 641 [10] X. Wang, Y. Guo, L. Qiu, X. Wang, T. Li, L. Han, H. Ouyang, W. Xu, K. Chu, 642 Preparation and evaluation of carboxymethyl chitosan-rhein polymeric micelles 643 with synergistic antitumor effect for oral delivery of paclitaxel, Carbohydr. Polym., 644 206 (2019) 121-131. 645 [11] C. He, Y. Jin, Y. Deng, Y. Zou, S. Han, C. Zhou, Y. Zhou, Y. Liu, Efficient 646 oral delivery of poorly water-soluble drugs using carnitine/organic cation 647 22 transporter 2-mediated polymeric micelles, ACS Biomater. Sci. Eng. , 6 (2020) 648 2146-2158. 649 [12] R. Kumar, A. Sirvi, S. Kaur, S.K. Samal, S. Roy, A.T. Sangamwar, Polymeric 650 micelles based on amphiphilic oleic acid modified carboxymethyl chitosan for oral 651 drug delivery of bcs class iv compound: Intestinal permeability and 652 pharmacokinetic evaluation, European Journal of Pharmaceutical Sciences, 153 653 (2020) 105466. 654 [13] Á. Aguilar-de-Leyva, V. Linares, M. Casas, I. Caraballo, 3D Printed Drug 655 Delivery Systems Based on Natural Products, Pharmaceutics, 12 (2020) 620. 656 [14] C. Karavasili, A. Gkaragkounis, T. Moschakis, C. Ritzoulis, D.G. Fatouros, 657 Pediatric-friendly chocolate-based dosage forms for the oral administration of 658 both hydrophilic and lipophilic drugs fabricated with extrusion-based 3D printing, 659 Eur. J. Pharm. Sci., 147 (2020) 105291. 660 [15] X. Xu, P. Robles-Martinez, C.M. Madla, F. Joubert, A. Goyanes, A.W. Basit, 661 S. Gaisford, Stereolithography (SLA) 3D printing of an antihypertensive 662 polyprintlet: Case study of an unexpected photopolymer-drug reaction, Addit. 663 Manuf., 33 (2020) 101071. 664 [16] H. Herrada-Manchón, D. Rodríguez-González, M.A. Fernández, M. Suñé- 665 Pou, P. Pérez-Lozano, E. García-Montoya, E. Aguilar, 3D printed gummies: 666 Personalized drug dosage in a safe and appealing way, Int. J. Pharm., 587 (2020) 667 119687. 668 [17] G. Kollamaram, D.M. Croker, G.M. Walker, A. Goyanes, A.W. Basit, S. 669 Gaisford, Low temperature fused deposition modeling (FDM) 3D printing of 670 thermolabile drugs, Int. J. Pharm., 545 (2018) 144-152. 671 [18] V. Linares, M. Casas, I. Caraballo, Printfills: 3D printed systems combining 672 fused deposition modeling and injection volume filling. Application to colon- 673 specific drug delivery, Eur. J. Pharm. Biopharm., 134 (2019) 138-143. 674 [19] R. Govender, S. Abrahmsén-Alami, A. Larsson, S. Folestad, Therapy for the 675 individual: Towards patient integration into the manufacturing and provision of 676 pharmaceuticals, Eur. J. Pharm. Biopharm., 149 (2020) 58-76. 677 [20] T.C. Okwuosa, C. Soares, V. Gollwitzer, R. Habashy, P. Timmins, M.A. 678 Alhnan, On demand manufacturing of patient-specific liquid capsules via co- 679 ordinated 3D printing and liquid dispensing, Eur. J. Pharm. Sci., 118 (2018) 134- 680 143. 681 23 [21] A. Goyanes, H. Chang, D. Sedough, G.B. Hatton, J. Wang, A. Buanz, S. 682 Gaisford, A.W. Basit, Fabrication of controlled-release budesonide tablets via 683 desktop (FDM) 3D printing, Int. J. Pharm., 496 (2015) 414-420. 684 [22] C.I. Gioumouxouzis, A.-T. Chatzitaki, C. Karavasili, O.L. Katsamenis, D. 685 Tzetzis, E. Mystiridou, N. Bouropoulos, D.G. Fatouros, Controlled release of 5- 686 fluorouracil from alginate beads encapsulated in 3D printed pH-responsive solid 687 dosage forms, AAPS PharmSciTech, 19 (2018) 3362-3375. 688 [23] A. Almeida, M. Araújo, R. Novoa-Carballal, F. Andrade, H. Gonçalves, R.L. 689 Reis, M. Lúcio, S. Schwartz, B. Sarmento, Novel amphiphilic chitosan micelles 690 as carriers for hydrophobic anticancer drugs, Mater. Sci. Eng. C, 112 (2020) 691 110920. 692 [24] A. Almeida, D. Silva, V. Gonçalves, B. Sarmento, Synthesis and 693 characterization of chitosan-grafted-polycaprolactone micelles for modulate 694 intestinal paclitaxel delivery, Drug Deliv. Transl. Res., 8 (2018) 387-397. 695 [25] S.M. Martins, T. Wendling, V.M. Gonçalves, B. Sarmento, D.C. Ferreira, 696 Development and validation of a simple reversed-phase HPLC method for the 697 determination of camptothecin in animal organs following administration in solid 698 lipid nanoparticles, J. Chromatogr. B, 880 (2012) 100-107. 699 [26] R. Schellekens, F. Stuurman, F. Van der Weert, J. Kosterink, H. Frijlink, A 700 novel dissolution method relevant to intestinal release behaviour and its 701 application in the evaluation of modified release mesalazine products, Eur. J. 702 Pharm. Sci., 30 (2007) 15-20. 703 [27] T. Higuchi, Mechanism of sustained‐action medication. Theoretical analysis 704 of rate of release of solid drugs dispersed in solid matrices, J. Pharm. Sci., 52 705 (1963) 1145-1149. 706 [28] R.W. Korsmeyer, R. Gurny, E. Doelker, P. Buri, N.A. Peppas, Mechanisms 707 of solute release from porous hydrophilic polymers, Int. J. Pharm., 15 (1983) 25- 708 35. 709 [29] P.L. Ritger, N.A. Peppas, A simple equation for description of solute release 710 II. Fickian and anomalous release from swellable devices, J. Control. Release, 5 711 (1987) 37-42. 712 [30] M.H. Macedo, A.S. Barros, E. Martínez, C.C. Barrias, B. Sarmento, All 713 Layers Matter: Innovative Three-dimensional Epithelium-Stroma-Endothelium 714 Intestinal Model for Reliable Permeability Outcomes, Submitted, (2021). 715 24 [31] M.H. Macedo, E. Martinez, C.C. Barrias, B. Sarmento, Development of an 716 improved 3D in vitro intestinal model to perform permeability studies of 717 paracellular compounds, Front. Bioeng. Biotechnol. , 8 (2020) 1076. 718 [32] J.P. Rao, K.E. Geckeler, Polymer nanoparticles: preparation techniques and 719 size-control parameters, Prog. Polym. Sci., 36 (2011) 887-913. 720 [33] J.W. Hickey, J.L. Santos, J.-M. Williford, H.-Q. Mao, Control of polymeric 721 nanoparticle size to improve therapeutic delivery, J. Control. Release, 219 (2015) 722 536-547. 723 [34] H. Yuan, L.-J. Lu, Y.-Z. Du, F.-Q. Hu, Stearic acid-g-chitosan polymeric 724 micelle for oral drug delivery: in vitro transport and in vivo absorption, Mol. 725 Pharm., 8 (2011) 225-238. 726 [35] D.S. Silva, A. Almeida, F.G. Prezotti, W.M. Facchinatto, L.A. Colnago, S.P. 727 Campana-Filho, B. Sarmento, Self-aggregates of 3, 6-O, O’-dimyristoylchitosan 728 derivative are effective in enhancing the solubility and intestinal permeability of 729 camptothecin, Carbohydr. Polym., 177 (2017) 178-186. 730 [36] L. Tu, G. Wang, N. Qi, W. Wu, W. Zhang, J. Feng, Multi-functional chitosan 731 polymeric micelles as oral paclitaxel delivery systems for enhanced bioavailability 732 and anti-tumor efficacy, International journal of pharmaceutics, 578 (2020) 733 119105. 734 [37] L. Li, N. Liang, D. Wang, P. Yan, Y. Kawashima, F. Cui, S. Sun, Amphiphilic 735 polymeric micelles based on deoxycholic acid and folic acid modified chitosan for 736 the delivery of paclitaxel, Int. J. Mol. Sci., 19 (2018) 3132. 737 [38] ISO/EN10993-5. International Standard ISO 10993-5 Biological Evaluation 738 of Medical Devices—Part 5: Tests for Cytotoxicity: In Vitro Methods, in, 739 International Organization for Standardization, Geneva, Switzerland, 2009. 740 [39] P. Laskar, S. Samanta, S.K. Ghosh, J. Dey, In vitro evaluation of pH-sensitive 741 cholesterol-containing stable polymeric micelles for delivery of camptothecin, J. 742 Colloid Interface Sci., 430 (2014) 305-314. 743 [40] P. Luo, Y. Luo, J. Huang, W. Lu, D. Luo, J. Yu, S. Liu, Incorporation of 744 camptothecin into reduction-degradable supramolecular micelles for enhancing 745 its stability, Colloids Surf. B, 109 (2013) 167-175. 746 [41] Y. Du, L. Ling, M. Ismail, W. He, Q. Xia, W. Zhou, C. Yao, X. Li, Redox 747 sensitive lipid-camptothecin conjugate encapsulated solid lipid nanoparticles for 748 oral delivery, Int. J. Pharm., 549 (2018) 352-362. 749