Inhalation of Dimethyl Fumarate-encapsulated solid lipid nanoparticles attenuate clinical 1 signs of Experimental Autoimmune Encephalomyelitis and pulmonary inflammatory 2 dysfunction in mice 3

32 Rationale: The FDA approved Dimethyl Fumarate (DMF) as an oral drug for Multiple Sclerosis 33 treatment based on its immunomodulatory activities. However, it also caused severe adverse effects 34 mainly related to the gastrointestinal system. Objective: Investigated the potential effects of solid 35 lipid nanoparticles (SLN) containing DMF, administered by inhalation on the clinical signs, central 36 nervous system (CNS) inflammatory response, and lung function changes in mice with 37 experimental autoimmune encephalomyelitis (EAE). Materials and Methods: EAE was induced 38 using MOG35-55 peptide in female C57BL/6J mice and were treated via inhalation with DMF39 encapsulated SLN (CTRL/SLN/DMF and EAE/SLN/DMF), empty SLN (CTRL/SLN and 40 EAE/SLN), or saline solution (CTRL/saline and EAE/saline), every 72 hours during 21 days. 41 Results: After 21 days post-induction, EAE mice treated with DMF-loaded SLN, when compared 42 to EAE/saline and EAE/SLN, showed decreased clinical score and weight loss, reduction in brain 43 and spinal cord injury and inflammation, also related to the increased influx of Foxp3+ cells into the 44 spinal cord and lung tissues. Moreover, our data revealed that EAE mice showed signs of 45 respiratory disease, marked by increased vascular permeability, leukocyte influx, production of 46 TNF-α and IL-17, perivascular and peribronchial inflammation, with pulmonary mechanical 47 dysfunction associated with loss of respiratory volumes and elasticity, which DMF-encapsulated 48 reverted in SLN nebulization. Conclusion: Our study suggests that inhalation of DMF-encapsulated 49 SLN is an effective therapeutic protocol that reduces not only the CNS inflammatory process and 50 disability progression, characteristic of EAE disease, but also protects mice from lung inflammation 51 and pulmonary dysfunction. 52 D ow naded rom http://pndpress.com /clinsci/ardf/doi/10.1042/C S202107922/cs-2021-0792.pdf by gest on 17 D ecem er 2021 C liical Scnce. This is an Acepted M ancript. ou re encuraged to se he Vrsion of R eord tat, w en puished, w ill relace his vesion. he m st up-tote-version is avilable at https://rg/10.1042/C S210792

immune changes in lungs of control mice nebulized with SLN-encapsulated DMF, showing no 105 apparent side effect through this administration route. In the context of CNS disease, inhalation of 106 SLN-encapsulated DMF was able to reduce the clinical manifestations of motor dysfunction and 107 central nervous system inflammation, also attenuating the pulmonary inflammation and dysfunction 108 after 21 days in EAE treated mice compared with the EAE control group. Finally, low grade of 109 inflammation in SLN-encapsulated DMF treated mice may be due to increased influx of After reaching 20 g and 12 weeks of age, the mice were randomly divided into 6 groups: 128 control + saline (CTRL/saline); control group + empty SLN (CTRL/SLN); control + SLN + DMF 129 (CTRL/SLN/DMF); EAE+ saline group (EAE/saline); EAE + empty SLN (EAE/SLN) and EAE + 130 7 purchased from Synth (Brazil). Poloxamer 188, CAS number 9003-11-6 and dimethyl fumarate 157 (DMF), CAS number 624-49-7) were purchased from Sigma-Aldrich (USA). Water was purified 158 from Millipore Milli Q Integral 3 Water Purification System, Millipore Corporation (USA), and its 159 resistivity was 18.2 MΩ-cm. 160 The SLN were prepared using aqueous phase, lipid phase, and surfactants. The aqueous 161 phase containing the poloxamer 188 and the lipid phase containing glyceryl monostearate, 162 hydrogenated soy phosphatidylcholine, and DMF were separately heated around 60 °C and mixed 163 homogeneously. Then, they were emulsified using a 600W ultrasonic tip stirrer for 15 minutes in a 164 discontinuous mode (1-minute of sonication with 1-minute interval) and cooled in an ice bath to 165 harden the lipid cores. The production was carried out batch-by-batch and stored at refrigerated 166 temperature (7-8° C) until administration. Previous experiments demonstrated the physical and 167 chemical stability for longer than the storage period before administration to animals (unpublished 168 results). 169 170 Dynamic light scattering and polydispersity 171 The droplet size and polydispersity analysis were performed with a dynamic light-scattering 172 (DLS) Zetasizer Nano system ZS (Malvern Instruments, Worcestershire, UK), with He-Ne 4 mW 173 laser source at 633 nm using a recording angle of 173 o . For analyzes, the samples were diluted 174 tenfold with Milli-Q purified water. The results represent the mean ± the standard deviation of at 175 least ten determinations. The analyzes were carried out at 25°C ± 0.2°C. 176 177 Zeta potential 178 The zeta potential was determined with a Zetasizer Nano system ZS (Malvern Instruments, 179 Worcestershire, UK). The samples were diluted with Milli-Q purified water. The results represent 180 the mean ± the standard deviation of at least ten determinations. The analyzes were carried out at 181 25°C ± 0.2°C. 182

Evaluation of EAE clinical evolution 209
Mice were monitored daily for body weight and clinical disease, as previously described 30 . 210 The clinical signs were scored as follows: 0) No clinical signs; 1) Tail paresis; 2) Tail paralysis; 3) 211 Paresis of hind limbs, uncoordinated movement; 4) Paralysis of a hind leg; 5) Paralysis of both hind 212 legs; 6) Paralysis of the hind legs and paresis of the anterior ones; 7) Paralysis of the hind legs and a 213 paralyzed anterior paw; 8) Paralysis of the hind and anterior legs; 9) Dying and 10) Death 31 . 214 215

Vascular permeability 216
To evaluate the blood plasma leakage through gaps produced in the endothelium of post-217 capillary venules in the organs, a modified protocol based on the Evans Blue (Dinâmica, SP, Brazil) 218 dye was used as a marker of vascular albumin extravasation, as previously described 32 . First, mice 219 were anesthetized intraperitoneally with a mixture of ketamine (60 mg/kg, Laboratório Cristália, 220 SP, Brazil) and xylazine (10 mg/kg, Rompun®, Bayer Briefly, mice from each group were anesthetized using an i.p. injection of a mixture of ketamine (60 235 mg/kg) and xylazine (10 mg/kg). To observe leukocyte-endothelial interactions, circulating 236 leukocytes were fluorescently labeled by retro-orbital plexus administration of Rhodamine 6G dye 237 (200 μL, 0.3 mg/kg) (Sigma, SP, Brazil). Rhodamine 6G is a fluorochrome that provides selective 238 labeling of leukocyte and platelet mitochondria. A midline skin incision measuring 2-3 cm was 239 performed, and detachment of the paravertebral musculature and laminectomy were performed at 240 the level of the thoracic vertebrae to create a spinal imaging window measuring 0.5-1.0 cm in 241 length. After the surgery, the mice were transferred to the microscope stage and were maintained at 242 37 °C using a heating pad (Fine Science Tools Inc., Canada). 243 To assess the leukocyte-endothelium interactions in the spinal cord microcirculation, the 244 fluorescent leukocytes were visualized under a Zeiss Imager M.2 (20x long-distance objective lens; 245 Göttingen, Germany) equipped with a fluorescent light source (epi-illumination at 510-560 nm, 246 using a 590 nm emission filter) 26 . Rolling cells were defined as cells that rolled at speed slower 247 than that of blood flow. Cells were considered to be adherent when they remained stationary for at 248 least 20s.After analysis, mice were sacrificed by decapitation following intraperitoneal injection of 249 ketamine (60 mg/kg) and xylazine (10 mg/kg) according to approval by the Institutional Animal 250 Care and Use Committees at the Federal University of Minas Gerais, as previously reported [35][36][37] . 251 252

Spinal cord and lung histological analysis 253
Mice were perfused intracardially using 10% formalin in PBS, and the tissues (spinal cord 254 and lung) were maintained in the fixative until they were prepared for paraffin embedding. Spinal 255 cords were sectioned transversely, and the thoracic-lumbar level was analyzed (the same region 256 used for intravital microscopy). Serial spinal cord and lung sections (4 μm) were cut and were 257 stained using hematoxylin and eosin (HE). The slides were photographed using the Olympus BX51 258 microscope (20X objective) equipped with Image-Pro Express 4.0 software (Media Cybernetics, 259 MD, USA). 260 On the 21st day after induction, all experimental groups were anesthetized intraperitoneally with a 303 mixture of ketamine (60 mg/kg) and xylazine (10 mg/kg), and the trachea of each animal was 304 exposed and cannulated with a 1.7 mm polypropylene catheter. Washing was done by injecting 1 305 mL aliquots of sterile PBS, injected and collected 3 times each, obtaining 1.7 -2.0 mL of final 306 volume recovered from the wash. At the end of the experiment, mice were decapitated after 307 intraperitoneal injection of ketamine (60 mg/kg) and xylazine (10 mg/kg). 308 The collected wash liquid was centrifuged in 5 mL tubes at 4°C for 5 min at a speed of 309 1500 rpm, forming a cell pellet used for total and differential cell count. The cell pellet was 310 resuspended in 100 μl of 3% BSA diluted in PBS, and the total number of leukocytes was 311 determined by counting the cells stained with Turk's solution in Neubauer's chambers. Total counts 312 were made on optical microscopes with a 40X objective. For the differential count, the slides were 313 prepared by cytospin (Cytospin 3, Shandon) and stained with May-Grunwald-Giemsa, as previously 314 described 38, 39 . The differential counts were performed in optical microscopes with a 100X 315 objective using immersion oil. The SLN used in the present work was characterized physiochemically, as shown in Table 1. 389 The unloaded (empty) SLN had 157 nm of average diameter, while the DMF-loaded SLN showed 390 315 nm of average diameter (Table 1). Thus, DMF increased the oil phase volume in the preformed 391 nanoemulsions, leading to structural changes in the entire system and, therefore, increasing droplets 392 diameter. The diameter values obtained by dynamic light scattering using a Zetasizer Nano system 393 ZS (Malvern Instruments, Worcestershire, UK) indicate that our delivery system of DMF is within 394 the acceptable nanometric range for a nanoparticle. The polydispersity analysis using the dynamic 395 light-scattering using a  ( Figure 2B). There were no statistical differences observed between the control and vehicle groups 439 ( Figure 2). 440 To further investigate the effects of DMF-encapsulated SLN in CNS of EAE induced mice, 441 we evaluate the integrity of the blood-brain barrier (BBB) in the brain and spinal cord tissues using 442 the Evans Blue protocol ( Figure 3A and B). In the brain, there were an increase in vascular leakage 443 in the EAE/saline and EAE/SLN groups compared to CTRL/saline and CTRL/SLN/DMF groups 444 ( Figure 3A). At the same time, the EAE-induced mice that received nebulized DMF-encapsulated 445 SLN presented a reduction of cerebral plasma leakage compared to EAE/saline and EAE/SLN 446 groups ( Figure 3A  the EAE-induced mice showed a significant increase in the signal of grey intensity from the 460 13 th d.p.i. when compared to CTRL/saline group. It is important to highlight that on the 13 th d.p.i. 461 the signal intensity for the saline EAE group was even more significant compared to CTRL/saline 462 group ( Figure 4C and D). However, once the inhalation of DMF-encapsulated SLN was able to 463 reduce the white matter deterioration at 20 th d.p.i. (Figure 4C and D), no statistical differences were 464 observed when comparing EAE/SLN/DMF and EAE/saline groups, suggesting that it may 465 contribute to attenuated mice clinical signs during DMF-encapsulated SLN treatment. 466 Despite the protection observed by DMF-encapsulated SLN inhalation in mice, like reduced 493 clinical signs and brain inflammation of EAE, but increased leukocyte rolling and adhesion in 494 spinal cord vasculature, we further investigated the T regulatory lymphocyte (Tregs) population. 495

Inhalation of DMF-encapsulated SLN increases leukocyte influx into the spinal cord 468 microvasculature but attenuates brain inflammation during EAE induced by MOG35-55 in
We aimed to use the transgenic mice for the Foxp3, a transcription factor marker of Tregs, 496 expressing GFP fluorescence (Foxp3 GFP + ) to quantify the Treg cells in the spinal cord and lung 497 tissues by immunofluorescence (Figure 6). Our results showed that in the spinal cord tissues from 498 CTRL/saline and EAE/saline mice groups, the Foxp3+ expression was very similar and low grade 499 ( Figure 6A and B). However, the EAE mice treated with DMF-encapsulated SLN presented an 500 increased influx of the Foxp3 + cells when compared with the CTRL/saline and EAE/saline mice 501 groups ( Figure 6A and B), also confirmed by the higher Foxp3 mRNA expression of 502 EAE/SLN/DMF treated mice in relation to EAE/saline group ( Figure 6C). We also evaluated the 503 impact of DMF-encapsulated SLN inhalation in lungs from mice submitted to EAE disease ( Figure  504 6D-F). In the same way, as found in brain tissues, the CTRL/saline and EAE/saline mice groups 505 exhibited similar counts of Foxp3+ cells in the lungs ( Figure 6D and E). However, the number was 506 higher in EAE/SLN/DMF treated mice group ( Figure 6D and E), also confirmed by higher Foxp3 507 mRNA expression when compared to EAE/saline mice group ( Figure 6F). 508 509

Abnormal pulmonary inflammation and dysfunction occur during EAE manifestation and 510
are controlled by inhalation of DMF-encapsulated SLN in mice. 511 Lungs have been implicated in the generation of an immune response, including in the 512 context of EAE challenged mice 50 . Because the airways were used as the route of administration of 513 inhaled DMF-encapsulated SLN, we investigated the possible effects on lung tissue such as 514 inflammation (protein leakage, leukocyte influx, and cytokine production), lung injury, and possible 515 respiratory mechanical dysfunction in mice. First, we evaluated the pulmonary vascular 516 permeability using two distinct protocols: the Evans Blue protocol was used in lung tissues, and 517 total protein quantification by Bradford protocol from samples of bronchoalveolar lavage fluid 518 (BALF) recovered after 21 st day of EAE induction in mice. Our results showed that the DMF-519 encapsulated SLN inhalation in CTRL/SLN/DMF mice group altered neither lung permeability by 520 Evans's Blue ( Figure 7A) nor protein leakage into the airways by Bradford quantification (Figure  521 7B). Mice belonging to the EAE/saline and EAE/SLN groups presented increased pulmonary 522 vascular permeability ( Figure 7A and B) characteristics of lung diseases when compared to the 523 CTRL/saline and CTRL/SLN/DMF mice groups. However, inhalation of DMF-encapsulated SLN 524 in EAE mice was able to reduce lung permeability, as confirmed by Evan's Blue and Bradford 525 protocols, when compared to EAE/saline and EAE/SLN groups ( Figure 7A and B). 526 Bronchoalveolar lavage was performed to evaluate the leukocyte populations in the airways. 527 The total leukocyte counts in BALF recovered from CTRL/SLN and CTRL/SLN/DMF mice groups 528 were very similar compared to the unchallenged mice CTRL/saline group ( Figure 7C), presenting 529 low numbers of macrophages, neutrophils, and lymphocytes ( Figure 7D-F). However, EAE/saline 530 and EAE/SLN mice groups showed a marked increase in leukocyte numbers in BALF ( Figure 7C) 531 compared to their respective controls. This leukocyte population was composed predominantly of 532 macrophages with a smaller mixed-population consisting of neutrophils and lymphocytes ( Figure  533 7D-F). The EAE mice that received inhalation with DMF-encapsulated SLN showed a reduction in 534 the number of total leukocytes from BALF ( Figure 7C), reducing the macrophage, neutrophil, and 535 lymphocyte numbers to the basal levels ( Figure 7D-F) compared to the EAE/saline and EAE/SLN 536 mice groups. 537 The ELISA protocol was performed to assess the lung cytokines concentration. Our results 538 showed that the inhalation of SLN empty or SLN/DMF in control mice did not interfere with TNF-539 α, IL-17, and IL-10 levels quantified in lung samples ( Figure 7G-I). However, EAE/saline mice 540 group presented increased TNF-α and IL-17 levels in lungs when compared to CTRL/saline and 541 CTRL/SLN groups ( Figure 7G and H), as well as a lower in IL-10 concentration when compared to 542 CTRL/SLN/DMF mice group ( Figure 7I). On the other hand, inhalation of DMF-encapsulated SLN 543  Figure 7G and H), 544 with increased IL-10 when compared to EAE/saline mice group ( Figure 7I). 545 To further evaluate the lung injury, tissue pathology based on a score that characterizes the 546 inflammatory manifestations of the lung sites was performed in three regions: perivascular 547 inflammation, parenchymal inflammation, and peribronchial inflammation. Control mice that 548 received inhalation of DMF-encapsulated SLN, CTRL/SLN/DMF, showed no apparent pathological 549 changes ( Figure 8A-D) in the perivascular (arrowheads), parenchymal, and peribronchial (arrows) 550 regions depicted in (Figure 8E), as well as observed in CTRL/saline mice group. Our results also 551 demonstrated that EAE/saline and EAE/SNL mice groups revealed an increase in lung 552 inflammation, marked by intense perivascular (Figure 8A and E, arrowhead) and peribronchial 553 ( Figure 8C and E, arrowhead) inflammation, and less intense in the parenchyma areas ( Figure 8B  554 and E) when compared to the CTRL/saline and CTRL/SLN/DMF mice groups. On the other hand, 555 the EAE/SLN/DMF mice group presented a reduction in the perivascular ( Figure 8A and E, 556 arrowhead) and peribronchial ( Figure 8C and E, arrowhead) inflammation, also observed in the 557 parenchyma areas ( Figure 8B and E), when compared to the EAE/saline mice group. In general, the 558 total score indicated that inhalation DMF-encapsulated SLN was able to revert the pulmonary 559 inflammatory process evidenced in EAE/saline and EAE/SLN mice groups ( Figure 8D). 560 Furthermore, we found that the treatment by inhalation of the control group with DMF-encapsulated 561 in SLN did not cause any toxic or adverse effects on the mice airways. 562 Regarding pulmonary function, the inflammatory process generally leads to profound 563 alterations of lung mechanics, such as volumes, elasticity and flow loss, is an excellent predictive of 564 lung disease manifestation. We used invasive spirometry to evaluate the security of our DMF-565 encapsulates SLN formulation in control mice and mice with EAE. Our results revealed that 566 inhalation in CTRL/SLN and CTRL/SLN/DMF mice groups does not cause functional respiratory 567 loss when compared to CTRL/saline group, as observed by the regular pattern of the Pressure x 568 Volume curve ( Figure 9A The BBB is formed by the junction of the endothelial cells, pericytes, and astrocytes 51, 77 , 644 being responsible for restricting the movement of external organisms and toxic substances, allowing 645 only the transport of selected molecules, such as oxygen, carbon dioxide, and glucose, utilizing 646 transport proteins 77 . It is already known that the probable mechanism associated with BBB 647 integrity dysfunction in MS is related to endothelial damage mediated in part by the increase of pro-648 inflammatory cytokines, such as IL-1β, IL-17, IL-22, and TNF-α  in treated mice. The effects of DMF on pulmonary pathologies have already been previously 688 described but not in pulmonary dysfunction associated with EAE 89-92 . DMF acts on the HCAR2, a 689 G protein-coupled receptor associated with the cell membrane, promoting its various effects on the 690 organism, such as inhibition of neutrophil infiltrates 58 . DMF treatment is effective in reversing 691 hemodynamic changes, reduction of inflammation, oxidative damage, and fibrosis in experimental 692 models of pulmonary arterial hypertension and pulmonary fibrosis 90 . As in the CNS, in the lung, 693 DMF can inhibit the pro-inflammatory signaling triggered by NF-κB activation and, consequently, 694 attenuating the inflammatory process 90, 92 . Indeed, we found reduced TNF-α and IL-17 levels in 695 the lung from the DMF-encapsulated SLN mice group, suggesting an anti-inflammatory role of 696 DMF locally administered in mice lungs. Finally, DMF-encapsulated SLN inhalation in control 697 mice demonstrated it to be safe when administered by airway route in mice. 698 Curiosity, we also observed that the EAE/SLN group showed some similarities with the 699 EAE/SLN/DMF group, exclusivity in the assessment of cerebral and pulmonary cytokine levels. 700 Some studies have shown the anti-inflammatory effects of poloxamer-188, which is one of the SLN 701 constituents 93-95 . In this aspect, the EAE/SLN group, besides having similar faith to the 702 EAE/SLN/DMF group, had no significant differences compared to the EAE mice treated with saline 703 solution. Moreover, the EAE/SLN group also showed an increase in animal clinical signs, 704 indicating that the empty SLN themselves cannot attenuate the symptoms of EAE mice. 705 In summary, the present study reveals for the first time that nebulized DMF-encapsulated 706 SLN acts in the CNS inflammation process but also may have a relevant functional role directly in 707 the lungs, ameliorating lung inflammation and function. Finally, DMF locally attenuated the 708 pulmonary inflammatory infiltration and could lead to less immune cell autoreactivity 50 and, in 709 turn, the lower autoimmune response against the mice BBB in CNS. Altogether, our results suggest 710 that inhaled SLN containing DMF may represent a potential alternative therapeutic protocol for MS 711 treatment. 712 713

Data Availability Statement 714
All data related to this study is already present in the paper.