Methods

An organic-solvent free method based on ultrasounds, using arachidic or stearic acid as solid lipid and Tween^® 60 as surfactant, was used to produce SLN¹⁷. In short, lipid and surfactant were melted at 85 °C in a water bath. A solution of Lys was prepared using ultra-pure water, heated at the same temperature and 4.4 mL of this solution added to the melted lipids. The resulting emulsion was sonicated using a probe-type sonicator for 5 min at 80% amplitude (model VCX-130 with a VC 18 probe, Sonic & Materials Inc., Newtown, CT, USA) to form a nanoemulsion and afterwards allowed to cool down at room temperature (RT), rendering an aqueous suspension of SLN, which was stored at RT until further use.

An organic-solvent free method based on HPH⁴⁵, using stearic acid as solid lipid and Tween^® 60 as surfactant, was used to produce SLN. In short, 1600 mg of lipid and 260 mg of surfactant were melted at 80 °C in a water bath. A solution of Lys at 6.6 mg/mL was prepared using ultra-pure water and heated at the same temperature. The melted lipids were mixed with 60 mL of Lys solution, at the same temperature, to form an emulsion. This emulsion was pre-homogenized, for 5 min at 12,000 rpm (T25 digital, ULTRA TURRAX^®-IKA^®-Werke GmbH & Co. KG, Staufen, Germany), and then kept at 80 °C in a water bath until further processing. Using a 10 mL syringe, 9 mL of the previous emulsion were placed in the HPH chamber (SPCH-10-EP Ultra High-Pressure Homogenizer, Stansted Fluid Power Products Ltd, Stansted, UK) and homogenized at 5.5 Bar and 80 °C. The resulting nanoemulsions were collected and either placed in the homogenizer for re-processing (additional cycle) or allowed to cool at RT to enable SLN formation. After the adequate HPH cycles were performed, the SLN dispersion were placed in glass vials and stored at RT until further usage. Three different number of cycles were tested: 1 cycle, where the resulting emulsion was stored after the first cycle; 2 cycles, where the resulting emulsion was subjected to another HPH cycle; and 3 cycles, where the resulting emulsion was subjected to 2 consecutives additional HPH cycles.

An organic-solvent free method based on HPH⁴⁵, using stearic acid as solid lipid and Tween^® 60 as surfactant, was used to produce SLN similarly to the previous method. In short, 275 g of lipid and 65 g of surfactant were melted at 80 °C in a water bath. A 3 L solution of Lys at 6.6 g/L was prepared using deionized water and heated at the same temperature. The latter solution was added to the melted lipids and the resulting emulsion homogenized (homogenizer Rannie™ model Bluetop, Copenhagen, Denmark) at 200 Bar. The resulting nanoemulsion was stored at RT until further usage.

The SLN formulations were characterized in terms of mean diameter and size distribution profile (polydispersity index—PdI) by dynamic light scattering (DLS) using a 90Plus Particle Size Analyzer (Brookhaven Instruments Corporation, Holtsville, NY, USA) and in terms of zeta potential by electrophoretic light scattering using a ZetaPALS Zeta Potential Analyzer (Brookhaven Instruments Corporation, Holtsville, NY, USA). A wavelength of 660 nm, a temperature of 20 °C, and a detection angle of 90° were used in the measurements and the refractive indexes were 1.59 and 1.33, for the SLN and the solvent (water), respectively¹⁷. All samples were diluted using ultra-pure water.

Lysine encapsulation efficiency (%Lys) was determined as the difference between the total amount of Lys added in the preparation of SLN and the amount of Lys outside the SLN (Eq. 1). The percentage of Lys in the SLN composition, Lys loading capacity (%LC), was also calculated (Eq. 2).

Firstly, the SLN suspensions were diluted 100× and filtrated with Amicon^® Ultra Centrifugal Filters Ultracell-50 kDa (EMD Millipore, Darmstadt, Germany) at 2200×g for 33 min using an Allegra^® X-15R centrifuge (Beckman Coulter, Pasadena, CA, USA). The supernatant was then collected for quantification using a previously developed, high-performance liquid chromatography-based method for the purpose of quantifying Lys, described elsewhere⁴⁶.

The abomasal medium was mimicked using fasted state simulated gastric fluid (FaSSGF)⁴⁷, prepared according to manufacturer specifications, with a pH of 1.6, supplemented with pepsin at 1 mg/mL. Both stearic and arachidic acid SLN were incubated, in triplicate, in the previously mentioned medium, at 39 °C, under light stirring⁴⁸. The same medium without any SLN was also incubated and used as control, also in triplicate. Aliquots were collected from each replica at 0 h (t₀) and at 1 h (t_f) of incubation^49,50 for further analysis. Both samples and controls were assessed in terms of size using DLS and compared with the values obtained for SLN without any contact with the abomasal mimicking medium (t_i).

The intestinal medium was mimicked using fed state simulated intestinal fluid (FeSSIF)⁴⁷, prepared according to manufacturer specifications, with a pH of 5, supplemented with pancreatin at 3 mg/mL. Both stearic and arachidic acid SLN were incubated, in triplicate, in the previously mentioned medium, at 39 °C, under light stirring⁴⁸. The same medium without any SLN was also incubated and used as control, also in triplicate. Aliquots were collected from each replica at 0 h (t₀) and at 2 h (t_f) of incubation⁵⁰, for further analysis. Both samples and controls were assessed in terms of size using DLS and compared with the values obtained for SLN without any contact with the intestinal mimicking medium (t_i).

The SLN stability in the bloodstream was assessed using fresh bovine blood. Bovine blood was collected from the jugular vein of Holstein Friesian cattle during their slaughter at a commercial abattoir (PEC Nordeste—Indústria de Produtos Pecuários do Norte, Penafiel, Portugal). Blood was harvested and stored in 0.5 L Schott flasks, previously coated with EDTA. The collection and use of this animal byproduct was authorized by the Portuguese Directorate-General of Food and Veterinary Medicine of the Ministry for Agriculture and Sea (authorization number: N.12.006.UDER). Both stearic and arachidic acid SLN were incubated in this medium, in triplicate, at 39 °C, under light stirring, protected from the light and sealed to avoid oxidation (only in contact with the atmosphere during aliquot collection). The same medium without any SLN was also incubated, in triplicate, and used as a negative control. Aliquots were collected at 0, 3, 6, 9, 12 and 24 h of incubation. Both samples and controls were centrifuged at 2000×g for 15 min to separate blood components, and the supernatants collected for further analysis. All samples and controls were assessed in terms of size using DLS and compared with the values obtained for SLN without any contact with any of the medium (t_i).

Ruminal contents were obtained from three multiparous (number of parity = 2) adult Holstein cows, dry and not pregnant that were fitted with a ruminal cannula (10 cm diameter; Bar Diamond Inc., Parma, ID, previously approved and licensed by the Portuguese Directorate-General of Food and Veterinary Medicine of the Ministry for Agriculture and Sea, permit 0421/000/000/2015). The cows were housed at the Vairão Agricultural Campus of School of Medicine and Biomedical Sciences, University of Porto (Vila do Conde, Portugal). Cows were handled in strict accordance with good animal practices as defined by the national authority and European Union Directive 2010/63/EU. Experimental animal procedures were approved by the Local Animal Ethics Committee of School of Medicine and Biomedical Sciences, University of Porto, licensed by the Portuguese Directorate-General of Food and Veterinary Medicine of the Ministry for Agriculture and Sea, and conducted by trained scientists (FELASA category C). This study was also performed in accordance with ARRIVE guidelines. Each cow was fed a TMR, based on corn silage or haylage, and always had fresh drinking water available. Ruminal contents were collected before the morning meal from the 4 quadrants of the rumen and placed in a 4 L pre-warmed (39 °C) thermal jug. In the laboratory, the ruminal digesta of each cow was homogenized and strained through 4 layers of linen cloth at 39 °C in O₂-free CO₂ atmosphere. The interval between the collection of the ruminal contents and incubation never exceeded 60 min. One part of the strained ruminal fluid, pH 6.3, was diluted anaerobically into 4 parts of Kansas State Buffer⁵¹ and mixed at 39 °C in an O₂-free CO₂ atmosphere. Thirty milliliters (25 mL of buffered ruminal fluid and 5 mL of each SLN formulation) were dispensed anaerobically into 125 mL serum bottles (Sigma-Aldrich Inc., St. Louis, MO, USA) containing 250 mg (dry matter) of wheat straw, sealed with butyl rubber stoppers and aluminum crimp caps (Sigma-Aldrich Inc., St. Louis, MO, USA), and incubated in a water bath at 39 °C. Each SLN formulation was added to previously prepared serum bottles and incubated in quadruplicate. Blanks, containing no SLN, were also incubated in quadruplicate, to be used as controls. Two of these replicates of each sample and blanks were placed in an ice bath immediately after the addition of the SLN formulation to serve as negative controls (t₀ samples, where t represents the time and ₀ indicates that no incubation occurred). Fermentations were stopped after 24 h by cooling the bottles in an ice-slurry bath (t_f samples, where t represents the time and _f indicates that the incubation occurred until the end of the designated time period). The samples were subsequently compared with SLN that did not contact the rumen inoculum (t_i samples, where t represents time and _i represents the SLN characterized after production)¹⁷.

The contents of the flasks were transferred to Falcon tubes and centrifuged at 500×g for 5 min to separate any remaining feed and the heaviest microorganisms, such as protozoa, from the SLN. The supernatant was collected and centrifuged at 30,000×g for 20 min to separate the SLN from the bacteria and other microorganisms that did not deposit in the first centrifugation. Both supernatants and deposits were collected for characterization in terms of size, PdI and zeta potential. This procedure was also performed on rumen inoculum and buffer mixture without the addition of any SLN (blanks), and these results were used as controls to verify the separation of the SLN and rumen inoculum¹⁷.

The L929 mouse fibroblastic cell line was selected for the in vitro biosafety characterization of the SLN, as they are recommended for biomaterial safety evaluation for biomedical devices and food applications³⁷.

Cells were cultivated in DMEM, supplemented with 10% (v/v) FBS, 1% (v/v) penicillin/streptomycin mixture and 1% (v/v) amphotericin B, at 37 °C in 5% CO₂ atmosphere. The medium was replaced every 3 days and when cells were confluent, they were detached from the culture flasks with a scrapper (Nunc™ Cell Scrappers, Thermofisher Scientific, Waltham, MA USA). After this physical detachment, cells were centrifuged at 300×g for 5 min in a Heraeus™ Multifuge™ X1R centrifuge (Thermo Fisher Scientific; Waltham, USA) and then resuspended in fresh DMEM. Counting of viable cells was performed using a Neubauer chamber (Improved Neubauer Bright-Line, Boeco; Germany) in a Motic^® AE2000 Binocular Inverted Microscope (Motic Electric Group Co., Ltd; Xiamen, Fujian, China)⁵².

To assess the impacts of SLN in cellular viability, a MTT assay was performed⁵³. Cells were firstly collected and seeded in 96-well plates (5 × 10⁴ cells per well) with 100 µL of fresh culture medium. After cellular adhesion, the medium was discarded and replaced by fresh DMEM medium, supplemented with SLN or free Lys. The SLN were added to achieve the final concentrations of 0.125, 0.25, 0.5, 1 and 2 mg of SLN per mL of medium, and free Lys was added to achieve the same concentrations that would be found in the respective mass of SLN. Cells with no SLN nor free Lys were included to be used as positive controls and cells incubated with 1% triton X-100 were used as negative controls. The cells were incubated, in quadruplicate, for all the aforementioned conditions for 24 h, at 37 °C in 5% CO_2. After incubation, the medium was completely discarded and replaced by 100 µL of a 0.5 mg/mL MTT solution. The plate was incubated for 2 h, at 37 °C in 5% CO_2, after which the MTT solution was discarded and 200 µL of DMSO were added to solubilize the formazan crystals formed inside the cells. Finally, the absorbance was read at 570 and 630 nm using a Synergy™ HT Multi-Mode Microplate Reader (Biotek Instruments, Winooski, VT, USA). The absorbance at 630 nm was subtracted from the absorbance at 570 nm to remove interferences from the medium, and the cellular viability was calculated by comparing the treated cells with the control cells. The results were normalized and compared with the positive controls (in theory, 100% of viability).

The crystallinity degree and lipid polymorphism of the SLN were evaluated by differential scanning calorimetry (DSC) using a PerkinElmer Pyris 1 DSC 200 F3 Maia^® (Netzschm, MA, USA). Samples were weighed (2–5 mg) directly on aluminum pans and an empty pan was used as a reference. All samples were scanned between 20 °C and 90 °C, with a heat rate of 10 °C per min and cooled down at 40 °C per min. This analysis was performed in the SLN, both loaded and not loaded with Lys, the bulk materials that constitute them and mixtures of these materials with Lys. The values of onset (°C), melting point (°C) and enthalpy (ΔH) were calculated from the data using OriginPro (OriginLab), whereas the recrystallization index (RI) was calculated using the Eq. (3)⁴¹.

Firstly, 2 mL of SLN dispersion, supplemented with 1% AEROSIL^®, were frozen overnight at − 80 °C (Deep Freezer GFL^®, Burgwedel, Germany) and were afterwards lyophilized using a laboratory freeze drier (LyoQuest-85 plus v.407, Telstar^® Life Science Solutions, Terrassa, Spain) for 72 h at − 80 °C under a pressure of 0.4 mbar. The lyophilized SLN were resuspended in the same volume of water (2 mL) and characterized in terms of size, PdI and zeta potential as previously described.

To evaluate the shelf-life of the developed SLN formulations, they were stored at RT and assessed, after 1, 2 and 4 wks, in terms of their size, size distribution profile, zeta potential and %Lys, using the previously described methods.

All statistical analyses were performed with IBM SPSS Statistics (SPSS 27.0, Armonk, NY, USA). Univariate one-way analysis of variance (ANOVA) was performed to compare multiple groups of independent samples. When the effect was statistically significant (P ≤ 0.05), the differences between the respective groups were compared with a post-hoc test (Pairwise, P ≤ 0.05).