Introduction

The development of drug and gene delivery systems has led to the hope that they will improve the treatment of many diseases, and some of these systems have already been approved for clinical use [1]. Solid lipid nanoparticles (SLNs) are colloidal carriers developed in 1991 as an alternative delivery system to liposomes, emulsions, and polymeric nanoparticles. To produce SLNs, the liquid lipid (oil) has been replaced by a solid lipid dispersed in water or in an aqueous surfactant solution. Hence, SLNs exhibit many advantages such as long-term physical stability, high drug payload, and a long-lasting release profile. Such characteristics result in a drug delivery system capable of carrying both lipophilic and hydrophilic compounds, depending on the method of preparation and appropriate selection of lipids and surfactants [2, 3].

SLN formulations are granted as low- or non-toxic because they are produced using biodegradable compounds, which are commonly used in pharmaceuticals and cosmetics, also called GRAS (Generally Regarded As Safe) [3, 4]. More recently, cationic solid lipid nanoparticles (cSLNs) have been introduced as delivery systems for poorly water-soluble drugs, especially for gene delivery purposes aimed for treating a variety of diseases [5, 6]. Gene delivery requires that oligonucleotides (e.g., DNA, siRNA, microRNA) be delivered intracellularly; thereby, upon internalization, cSLNs must interact with the endosomal membrane to release their cargo [7]. The molecular nature of these interactions could compromise cell membrane integrity, triggering mitochondrial and lysosomes damage, and increasing the number of autophagosomes [8, 9]. Although much attention has been paid to the development of cSLNs for drug and gene delivery, little is known about its toxicological interaction to living organisms.

This study aims to evaluate the cSLN biocompatibility by delineating the toxicological profile of the cSLN formulation previously developed in our laboratory [7, 10, 11]. First, we evaluated the interaction of the cSLNs with reticuloendothelial system (RES) organs such as the spleen, liver, lungs, and also on clearance organs such as the kidneys. Second, we investigated the blood-brain barrier (BBB) as a seminal biological barrier prone to maintain brain homeostasis and neuronal function. Finally, this study assessed the potential risks of cSLNs in vivo to establish criteria for its safer application for drug and gene therapy.

Materials and methods

Materials

Reagents were purchased from Sigma-Aldrich (St. Louis, Missouri, USA), unless specified otherwise.

Production and characterization of cSLNs

Cationic SLNs were produced in water solution using microemulsion extrusion technique as previously described [7, 10, 11]. Briefly, 100 μL of a 10-mM Pluronic F68 solution and 250 μL of a 10-mM DOTAP suspension (Avanti Polar Lipids, Alabama, USA) were mixed with 650 μL of Milli-Q™ water and added to a heated tube (75 °C) containing 7 mM stearic acid. Next, cSLNs were homogenized in a heated (75 °C) mini-extruder (Avanti Polar Lipids, AL, USA) through 15 cycles containing Isopore™ 100-nm pore polycarbonate membrane (Millipore, MA, USA). Finally, the heated suspension was transferred to ice bath until stabilization of temperature and stored at 4 °C until use. To produce fluorescently labeled SLNs, we added 0.4 mol% Rhodamine-PE (Avanti Polar Lipids) to the mixture and proceeded to the mini-extruder homogenization step described above.

The Zetasizer Nanoseries (Malvern Instruments Ltd., Worcestershire, UK) was used to perform the physicochemical characterization of cSLNs. The hydrodynamic diameter and polydispersity index (PDI) were measured by dynamic light scattering and zeta potential by electrophoretic mobility.

Animals and treatment

All experimental procedures were approved by the Institutional Committee for Ethics in Animal Use (CEUA/UNICAMP, protocol no. 3985-1) and are in accordance with the ethical standards of the Brazilian Society of Laboratory Animal Science guidelines. Male Wistar rats (Rattus norvegicus) (6–7 weeks old, weighing 200–250 g) were used in this study. Animals were obtained from the Multidisciplinary Center for Biological Research (CEMIB/UNICAMP) at 4 weeks of age and kept on a standard 12/12-h light/dark cycle, in a temperature-controlled environment (22 ± 2 °C), and with access to water and chow ad libitum until reaching 6–7 weeks old. The rats were then randomly divided into three groups and injected via the tail vein (i.v.) with physiological saline (Control), sterile Milli-Q water (Vehicle), or cSLNs/cSLNs-Rhodamine PE at a volume of 5 mL/kg body weight (total lipid content 3.725 mg/mL).

The dose was chosen based on the lipid content observed in the FDA-approved drug Doxil® [12] and previously explained in details [13]. All data were evaluated 24 h after one single i.v. injection. Hematological and biochemical analyses were also performed 72 h after cSLN single i.v injection.

Blood collection for hematological and biochemical analyses

For hematological and biochemical analyses, blood samples from all animals used for histopathological and immunohistochemical analysis were collected. Additionally, analysis of blood samples collected at 72 h post-cSLN injection was also included. The animals were anesthetized by intraperitoneal injection of ketamine (Dopalen®, 270 mg/kg body weight) and xylazine (Anasedan®, 30 mg/kg body weight) solution (both from Fortvale Valinhos, SP, Brazil). Then, four mL of blood was sampled by cardiac puncture and stored in either K3 EDTA tubes for hematology analysis or gel separator tubes for biochemical analysis.

The blood samples from K3 EDTA tubes were immediately analyzed using a T540 hematology system (Beckman Coulter, CA, USA). For serum extraction, the blood from gel separator tubes was centrifuged at 1500×g for 15 min at room temperature. The serum was analyzed for albumin, total bilirubin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatinine (CRE), and blood urea nitrogen (BUN), using a Cobas® 6000 C501 Clinical Chemistry Analyzer (Roche Diagnostics, Baden-Wüttenberg, Germany). At least three blood samples were evaluated, and if necessary, an additional set of blood samples was analyzed (n = 3–5/group).

Histopathology

After anesthesia, the animals (n = 3/group) were transcardially perfused with physiological saline followed by 4% paraformaldehyde (PFA) in 0.1 M phosphate buffered saline (PBS), pH 7.4. Following perfusion, the target organs (spleen, lungs, liver, and kidneys) were immediately removed, placed in 4% PFA overnight, dehydrated in ascending ethanol series, and embedded in paraffin. Sections of 5-μm thickness were processed for hematoxylin-eosin staining. All histopathological analysis was performed through light microscopy (Olympus BX51 photomicroscope, Japan).

For each tissue sample, five images were randomly taken per slide and scored for inflammation by two blinded investigators. The degree of inflammatory cell infiltration on tissue samples was scored as follows: zero for absence of inflammatory cells infiltration, one for mild infiltration, two for moderate infiltration, and three for severe infiltration of inflammatory cells.

Immunohistochemistry

In parallel to histopathological analysis, the expression of tissue neutrophils and macrophages was assessed through paraffin-embedded sections and cryosections (n = 3/group). Deparaffinized and rehydrated 5-μm-thick paraffin sections were stained with anti-CD68 (1:100; Abcam, Cambridgeshire, UK), a glycoprotein marker highly expressed by both circulating and tissue macrophages, as previously described [14].

For immunofluorescence analysis, the animals were perfused as described above and the brain, spleen, liver, lungs, and kidneys were excised and fixed in 4% PFA for 2 h (n = 3/group). Following fixation, the organs were cryoprotected sequentially in both 15% and 30% sucrose solutions for 24 h each at 4 °C, embedded in Tissue Tek O.C.T. freezing medium (Sakura Finetek, CA, USA) and frozen in N-hexane cooled at liquid nitrogen (− 60 °C). Then, 7-μm-thick sections were washed in 0.05 M Tris buffered saline (TBS, pH 7.4), permeabilized with 0.3% Triton X-100 in TBS for 10 min, and blocked in 3% bovine serum albumin in TBS for 1 h at room temperature. Prior to incubation with the primary antibody, the sections were pretreated with 0.3 M Glycine in TBS for 1 h at 4 °C to reduced fixative-induced autofluorescence. To confirm that cSLNs were in brain parenchyma, we stained the cryosections (5 μm thickness) with anti-rat endothelial cell antibody-1 (RECA-1) (1∶100, Santa Cruz Biotechnology, CA, USA), followed by Alexa Fluor 647-conjugated goat anti-mouse IgG (1∶200, Thermo Fisher Scientific, IL, USA). Neutrophils were visualized by anti-Ly6G (1:100, BioLegend, CA, USA), followed by Alexa Fluor 488-conjugated goat anti-mouse IgG (1∶200, Thermo Fisher, IL, USA). The sections were counterstained with DAPI (1 μg/mL, Thermo Fisher Scientific, IL, USA) and examined using Zeiss LSM 780-NLO confocal on an Axio Observer Z.1 microscope (Carl Zeiss AG, Germany) at the National Institute of Science and Technology on Photonics Applied to Cell Biology (INFABIC- UNICAMP). All images were collected under the same experimental conditions with pinholes set to one airy unit for each channel and using 1024 × 1024 image format.

Assessment of BBB integrity

Evans blue dye extravasation method (n = 6/group) was used to test if the cSLNs were capable to disrupt the integrity of the BBB. The Evans blue dye extraction and quantification were performed accordingly to Wang and Lai (2014) with minor modifications. Briefly, a 2% solution of Evans blue dye diluted in physiological saline (5 mL/kg of body weight) was injected into the lateral tail vein 24 h after cSLN injection. Two hours later, the animals were anesthetized and transcardially perfused with physiological saline. The brain hemispheres were homogenized in a 1:2 ratio (wt:vol) with 50% trichloroacetic acid, and the debris were removed by centrifugation at 10,000 rpm for 20 min at 4 °C. The supernatant was added to a 96-well plate (30 μL per well supplemented with 90 μL of 95% ethanol) and measured at 620 nm using a microplate reader TP-reader (ThermoPlate, Guangdong, China).

Statistics

The significance levels of the numerical data between control and treatment groups were determined using one-way ANOVA followed by Bonferroni’s post-hoc test (for p < 0.05, p < 0.01 and p < 0.001) GraphPad Prism 5.0 (GraphPad Software, CA, USA). Data are presented as the means ± standard error of the means (SEM).

Data availability

All data generated or analyzed during this study are included in this published article and Additional file 1.

Results

Produced cSLNs presented diameter of 108 ± 6.1 nm and PDI values ≤ 0.1, indicating a monodisperse and homogeneous formulation. The positive zeta potential (+ 46.8 ± 1.8 mV) was due to the presence of the cationic lipid DOTAP. These results agree with those obtained by our group in previous studies [7, 10, 11, 15].

After cSLN physicochemical characterization, we performed the hematological analysis of rats after cSLN administration (Table 1). Blood compatibility is an essential property for intravenously administered nanomaterials and should always be assessed before their clinical use to delineate putative nanoparticle toxicological profile [16, 17]. No effect of the cSLNs was observed (at 24 h or 72 h) on erythrocyte parameters or on platelet numbers (Table 1). We observed slight increases in the total white blood cell (WBC) counts in relation to control 24 h after i.v. injection of cSLNs. The cSLN-induced increase of WBCs was associated with an elevation of circulating neutrophils compared to control (p < 0.01) and vehicle groups (p < 0.01). The 24-h neutrophilia was though transitory since the number of neutrophils decreased after 72 h; the values of neutrophils counts remained within the references ranges for rats [18] although still statistically higher than in both the control (p < 0.01) and vehicle (p < 0.001) groups. Therefore, the number of neutrophils is considered physiologically normal 72 h after systemic cSLN exposure.

Table 1 Cationic SLN effects on hematological parameters of male Wistar rats
Full size table

Typically, an increased circulating number of neutrophils are associated with inflammation and parallel migration to tissues. To investigate this, we performed histopathological analysis after 24 h of the lungs, liver, spleen, and kidneys (Fig. 1), as well as, immunohistochemistry for detecting tissue macrophages (Fig. 2) and neutrophils (Fig. 3).

Fig. 1
figure 1

Histopathological analysis of tissue sections stained with hematoxylin-eosin following cationic solid lipid nanoparticle (cSLN) injection. Representative photomicrographs of the lungs, liver, spleen, and kidneys from a–d Control (saline), e–h Vehicle (sterile Milli-Q water), and i–l cSLN-treated animals 24 h after intravenous injection of 5 mL/kg b.w.. Scale bar = 100 μm. Arrows indicate inflammatory infiltrate

Full size image
Fig. 2
figure 2

Immunodetection of the glycoprotein CD68, a marker of macrophages, in tissues. Representative photomicrographs of macrophages (brown color) 24 h after intravenous injection of (a–c) control and (d–f) cationic solid lipid nanoparticles (cSLNs). Scale bar 100 μm. rb respiratory bronchioles, cv centrilobular vein, pv portal hepatic vein, wp white pulp, rp red pulp, g glomeruli

Full size image
Fig. 3
figure 3

Immunofluorescent detection of Ly6G, a marker of neutrophils, in rat tissue sections. Representative images for neutrophils (Ly6G, green) 24 h after intravenous injection of a–d control and e–h cSLNs in rats (5 mL/kg b.w.); nuclei were stained with DAPI (blue). Scale bar = 200 μm

Full size image

Histopathological analysis showed normal cytoarchitecture in all organs of control (Fig. 1a–d) and vehicle (Fig. 1e–h) groups. However, in the lungs, cSLN treatment led to inflammatory cells agglomeration at three sites: at the last segment of the non-respiratory air conducting airway periphery, at the terminal bronchiole, and dispersed among the alveoli (Fig. 1i). In the liver, the inflammatory infiltrate was particularly next to the hepatic portal area, and some lobules showed the branch of the hepatic vein, and bile ductules were widened (Fig. 1j). In the spleen, we found a higher density of lymphoid cells forming the white pulp, which was visible by the darker blue nodules that resulted from the closed together hematoxylin-stained lymphocyte nuclei; besides, the diameter of lymphoid nodules in cSLN-treated samples was larger than in control and vehicle samples. A higher density of lymphoid cells was also observed in the red pulp (Fig. 1k) as compared to both controls. No tissue alteration was observed in the kidneys (Fig.1l) unless for some higher cellularity exhibited on glomeruli. These alterations reflected in the increased inflammation score from the spleen > liver > lungs > kidneys regarding the most affected organs by inflammatory infiltration (see Additional Table 1).

Once triggered, the inflammatory response can induce activation and recruitment of neutrophils and monocytes to the site of injury. Compared to control groups (Fig. 2a–c), Fig. 2 shows a profuse presence of macrophages (brown color) in the lungs (Fig. 2d), liver (Fig. 2e), spleen (Fig. 2f), and absence in the kidneys (Fig. S1).

The presence of tissue neutrophils is depicted in Fig. 3. Compared to the control (Fig. 3a–d), Ly6G-positive cells were visually more abundant in the lungs, liver, and spleen of the cSLN group. In the lungs, the majority of immunolabeling (green color) was found scattered throughout the alveolar interstitium and at points of the respiratory bronchiole and alveolar sac lining (Fig. 3e). In the liver, Ly6G-positive cells were found distributed at midway between the hepatic portal region and the central vein area (Fig. 3f); just a few labeled neutrophils could be found in both controls. In the spleen, Ly6G immunolabeling was also prominent in the splenic red pulp and inside the white pulp (Fig. 3g). In the white pulp, Ly6G staining was diffused at regions populated by T lymphocytes into lymphoid follicles and close to the central arteriole; similarly, a diffuse staining was observed at the region of B lymphocyte dominium close to the marginal zone. Nevertheless, a mild Ly6G labeling was observed at the marginal zone proper, i.e., at the interface between the white and red pulp. In the kidneys, Ly6G-labeled cells were found inside the glomeruli of cSLN-treated rats, whereas there was none in the glomeruli of control ones (Fig. 3h).

Next, we examined serum biochemical markers of hepatic and renal functions to evaluate the toxicity of cSLNs after systemic exposure (Table 2). In this study model and experimental design, the cSLNs did not influence liver biomarkers (i.e., albumin, total bilirubin, ALT, and AST levels). Regarding the kidney biomarkers, we did not observe alterations in creatinine levels but found a transitory ~ 38% decrease in BUN levels at 24 h in the cSLN-treated animals, compared with the control group (p < 0.01). The levels of the BUN returned to the control levels after 72 h, which indicate that the function alteration was short-lived. Moreover, the histopathological examination of kidneys did not exhibit abnormality in the cSLN-treated group.

Table 2 Changes of serum biochemical parameters in cSLN-treated rats
Full size table

For delivery to the central nervous system (CNS), nanoparticles should innocuously cross the BBB, which controls the paracellular trafficking of substances and the transcellular transport. To evaluate if cSLNs reached the brain, we i.v. injected cSLNs labeled with Rhodamine-PE through the lateral tail vein of rats. The brain tissue analysis of cSLN-treated rats indicated the presence of small particulate signals (green dots) (Fig. 4b), which were not found on the control brain section (Fig. 4a). Also, the endothelial cell marker RECA-1 confirmed that the cSLNs did not remain located into the brain vessels and was found in the brain parenchymal tissue.

Fig. 4
figure 4

Uptake of fluorescently labeled cationic solid lipid nanoparticles (cSLNs) in brain parenchyma. Representative confocal microscopy images from brain sections 24 h after a sterile Milli-Q water (vehicle) and b cSLN injection (green dots, arrowheads). Vessels were visualized using endothelial cell marker RECA-1 (red). Nuclei were stained with DAPI (blue). Scale bar 100 μm

Full size image

After demonstrating that the cSLNs are capable of crossing the BBB, we investigated whether cSLNs do not disrupt the BBB integrity. The integrity of BBB was evaluated by determining extravasation of the Evans blue dye in brain tissue 24 h after cSLN injection. The amount of dye extracted from brain hemispheres of the cSLN-treated animals did not differ significantly from that in the brain of control and vehicle groups (Fig. 5a). The images of coronal brain sections harvested 2 h following infusion of Evans blue, 26 h after cSLN injection, demonstrated the absence of dye accumulation in brain parenchyma. The sectioned brain of cSLN-treated rats showed similar color as control and vehicle groups, confirming the integrity of BBB (Fig. 5b).

Fig. 5
figure 5

Evaluation of Evans blue dye, a marker of vascular permeability, in rat brain samples. Quantitative analysis of Evans blue dye extravasation in a brain tissue homogenate 24 h after physiological saline (control), sterile Milli-Q water (vehicle), and cSLN injection (5 mL/kg b.w.). b Representative coronal brain sections 2 h after intravenous Evans blue injection in control, vehicle and cSLNs. Values expressed as mean ± SEM (n = 6 animals/group). Student’s t test

Full size image

Discussion

Solid lipid nanoparticles (SLNs) are usually considered to be a safe drug and gene delivery system because they are composed of well-known biocompatible and GRAS excipients [19]. However, just a few studies have directly addressed the in vivo toxicity of SLNs. Therefore, this study aimed to determine the toxicological profile of cSLNs in rats assessing hematological parameters to gain a systemic insight into cSLN toxicity, particularly into inflammation in specific organs. In general, inflammation is believed to provide protection for the body, but an exaggerated and persistent inflammatory response can lead to more serious and often irreversible lesions to organs [20]. In addition, we evaluated the BBB integrity to investigate the consequences of the cSLN delivery into the CNS.

Blood biocompatibility studies identified some alterations, even though no death or abnormal clinical signs were observed in the animals through the time-window of 72 h. Four alterations were identified: (i) neutrophilia; (ii) increases in the population of peripheral tissue pools of neutrophils and macrophages in the lungs, liver, and spleen; (iii) increase in the cellularity of both the splenic white and red pulps, represented by lymphocytes and neutrophils; and (iv) decrease in BUN levels. These alterations were transient, and most of the parameters were reestablished at 72 h. One possibility is that, at earlier intervals, the high concentration of cSLNs in the blood activated the recruitment of neutrophils and macrophages to promote the phagocytosis of foreign particles to minimize organs damage. Such event configures the cornerstone of the innate immune response that does not require specific antigenic detection but rapid combat to extraneous invaders into the organism [21]. These results are in accordance with other studies that described an increased number of neutrophils [22] and neutrophil activation [23] produced by cSLNs. In addition, cSLNs induced activation of polymorphonuclear neutrophils by Ca2+ influx, oxidative stress, degranulation, and formation of neutrophil extracellular traps [23]. Interestingly, we did not observe alterations in erythrocyte parameters (RBCs and MCV) as reported by Silva et al. [22].

To investigate the extension of the cSLN-induced inflammation in biological tissues, we performed histopathological examination of the organs of the RES. We analyzed these organs because intravenously administered lipid-based nanoparticles follow the common fate of non-functionalized nanoparticles, that is, opsonization, and accumulation in RES organs, such as the liver, spleen, and lungs [24, 25]. Infiltration of inflammatory cells was observed at the peribronchiolar (terminal bronchiole) area and hepatic portal system, suggesting that a single intravenous injection of cSLNs was enough to recruit circulating WBCs to these organs. CD68 and Ly6G immunohistochemistries demonstrated that macrophages were the main cell infiltrated in the liver and spleen, while neutrophil infiltration was more prominent in the lungs and spleen 24 h after of cSLN injection. The population of inflammatory cells that migrate to organs constitutes pools able to exert different function when challenged [26]. Circulating blood flows from the periphery of the hepatic lobule where portal branches are located to the central lobule. In stressful situations, it becomes evident why inflammatory infiltrate accumulates at the periphery, as shown in the present findings. Interestingly, spleen and liver microcirculations are formed by sinusoids that have a very tortuous path and an enlarged lumen which together favors low circulation and retention of particles along their course. In addition, capillary sinusoids of the spleen and liver present multiple fenestrations and discontinuities which allow NPs to escape from capillary bed and get access into the organs’ parenchyma. It is likely that such characteristics would ground the inflammatory response greater in the spleen and liver than in the other organs studied. On the other hand, it is known that the lungs have one of the more developed capillary networks and thus an especially larger area exposed to xenobiotics circulating in the blood. Despite endothelium lining of pulmonary capillaries is continuous and not fenestrated, it is extremely thin and liable to be disrupted by particulate xenobiotics and be captured and removed by interstitial and alveolar macrophages. Regarding the kidneys, few Ly6G-positive cells were found in the glomeruli but without permanent damage to the kidneys, as demonstrated by the histopathological examination. Our results were similar to those that showed that lipid-based delivery systems are not entirely inert since they can induce an inflammatory response [27,28,29]. For example, the intravenous injection of cationic lipid-based nanoparticles in mice (10 mg/kg) triggered an inflammatory response by elevating both Th1 and Th17 cytokine expression [30]. Taken together, our results suggest that cSLNs were not toxic since the side effects observed were short-lived and did not lead to permanent damage to the organs, as demonstrated by the biochemical parameters of liver and kidney function 72 h after cSLN injection. Cationic SLNs can be used to deliver their cargo to across the BBB [31, 32]. Therefore, after identifying that the injection of the cSLNs synthetized in our laboratory led to low-intensity inflammation in different organs, we decided to evaluate the cSLN effects on brain vascular permeability. It is well known that systemic inflammatory response can alter circulatory properties, increase vessel permeability, and modulate the immune system [33] as it was shown here in organs with different types of capillaries. Even non-cytotoxic concentration of cationic nanoparticles can induce cell membrane permeability [34, 35]. Hence, we started evaluating the blood vessel integrity following the Evans blue dye extravasation into the brain tissue, which under physiological conditions cannot permeate the blood-brain barrier [36]. After 2 h from the intravenous injection of the dye, and 26 h following cSLNs i.v. infusion, cSLNs did not induce BBB injuries, demonstrated by the absence of Evans blue into brain tissue. These results indicate that cSLN transient inflammation found in some tissues was not enough to compromise the BBB integrity. In addition, it has been demonstrated that lipid carriers can cross the BBB without disrupting the structural barrier due to their lipophilicity nature, even without surface functionalization [37,38,39]. Besides the highly lipophilic characteristic, our cSLN formulation is positively charged, which probably contributed to their permeation across the BBB and accumulation in brain parenchyma. It is noteworthy that cSLNs appear to be intracellularly since they seem in close contact with nuclei. Further studies would be useful in elucidating the exact location of this formulation in the brain.

Conclusions

In summary, this study provides evidence that cSLNs induce transient toxic effects in the blood with signs of acute inflammation in the reticuloendothelial cell–containing organs. The transient inflammation should not be neglected; it could have an important implication in the therapeutic use of cSLNs, not only from the negative perspective, but also transient inflammation could be desired in some therapeutic application (e.g., as a vaccine adjuvant). However exploratory, this study provides some insight into the underlying mechanism of cSLN in vivo toxicity, providing compelling evidence that can be useful for the pharmaceutical field and in regulatory toxicology. Preclinical animal studies are a necessary step for reliable translation of findings from animals to humans. Further efforts are required to identifying the mechanisms underlying cSLNs and cell/biological system interactions.