Based on the satisfactory results already achieved by anchoring newly synthesized mannosylated surfactants to SLNas surface [
5], the purpose of this study was to investigate if the pulmonary surfactant lining the alveolar surface could possibly form, due to its high content of phospholipids, a lipid corona around the nanoparticles that might nullify alveolar macrophage (AM) active targeting and make SLNas functionalization unnecessary. The investigation involved both mannosylated SLNas (SLNas/MS) and SLNas functionalized by MS in a blend with sodium taurocholate (SLNas/MS-ST) in order to modulate the mannose concentration on the nanoparticle surface. Indeed, ST itself can play a key role, since an activity as opener of tight junctions between alveolar epithelial cells has been described [
40], although the definitive mechanism has not been fully elucidated yet. In addition, it is also noteworthy that mannose receptors (MR) bear a cystein-rich domain that recognizes sulfonic group. SLNas obtained by using ST (SLNas/ST) or F127 (SLNas/F127), a non-ionic surfactant devoid of groups able to bind MR, were evaluated as the controls.
3.1. SLNas Physico-Chemical Properties
The study of inhaled particulate drug carriers for AM active targeting, intended to increase particle specificity for macrophages and internalization potential over the passive targeting, must not disregard the investigation of the physico-chemical features required to both reach alveolar region and be phagocytized by AM. Therefore, the obtained SLNas samples were characterized for morphology, size, surface charge, shape factor, physical state, wettability, surface properties, density, flowability, drug loading, drug release, and respirability performance.
It is well known that the efficiency of an inhaled powder formulation to settle into the deep lung depends upon particle geometrical parameters (shape, morphology, and size) that affect the powder respirability as well as macrophage uptake [
6]. The most commonly used particles for DPI device exhibit irregular shapes. In vitro inhalation studies indicated that elongated [
41,
42], needle-like [
43], porous and wrinkle [
44], as well as thin flaky shapes [
45] can improve lung deposition properties owing to reduced particle interactions [
46]. Concerning macrophage phagocytosis, particle shape has shown to affect cell uptake mechanism or rate, with roundness being a preferred characteristic compared to elongated or filamentous shapes [
6,
47,
48]. All the obtained SLNas samples exhibited rounded micro-aggregates of nanoparticles having an irregular shape with tendency to roundness, as indicated by circularity values from 0.6 to 0.83, with 1 indicating perfect roundness (
Figure 1,
Table 1).
Particle size of the samples, measured by PCS and expressed as the mean diameter of the main class (>85%), was within the range of about 300 to 750 nm, where the larger size was shown for SLNas/MS and SLNas/ST compared with that for SLNas/MS-ST and SLNas/F127 (
p < 0.05). The size measured on the samples in their solid state by means of the BET technique, d(BET), aimed to simulate the therapeutic use of powdered drug by means of a DPI device. The d(BET) values were within the range from about 700 nm, for SLNas/MS, to 1100 nm for all the other samples (
p < 0.05). This size is considered proper for the deposition onto alveolar region and AM endocytosis [
6]. The related PDI values ranged from 0.3 to 0.6, owing to the presence of minor populations in the range of 2–5 μm (<15%), owing to a multimodal size distribution of the particle aggregates. The surface charge of all the samples was negative (from −15 to −55 mV). Considering that SLNas include the same lipid matrix, the largest negative values were reasonably imparted by the negatively charged MS and ST unlike what occurred with the non-ionic surfactant F127. Extent and rate of macrophage uptake are known to be directly related to particle net charge magnitude, with negatively charged surface being more physiologically compatible than the positive ones [
6]. Moreover, unlike particles bearing positive charges, negative surface charges showed to favor particle localization within lysosomes, in which
Mycobacterium tuberculosis survives [
49,
50]. As regards SLNas density features, true density values ranged from 1.147 to 1.247 g/cm
3, bulk density from 0.031 to 0.147 g/cm
3, and the respective tapped density values from 0.038 to 0.29 g/cm
3; SLNas/F127 exhibited the highest values of both bulk and tapped density among the samples. The increased tapped density compared to the bulk density values indicates the presence of inter-particle void spaces. Tapped density represents a pivotal parameter for both flowability and aerodynamic diameter required for a respirable powder. Indeed, particles having low tapped density can efficiently aerosolize from a DPI, resulting in a high respirable fraction of inhaled therapeutics [
51]. Flowability values, expressed as Carr’s Index, were affected by SLNas surface decoration; in particular, MS provided excellent flow properties also in the presence of ST (Carr’s Index <10) [
52]. Conversely, SLNas/ST and SLNas/F127 flowability resulted fair and very poor, respectively. The flow properties of a powder as well as its de-aggregation and aerosol performance can also be influenced by inter-particulate interactions related to the specific surface area, i.e., the surface area per unit mass. Such a parameter depends on both particle size and surface roughness of the powder [
41,
53]. The average values for BET specific surface area of the SLNas samples were found to be 5–7 m
2/g.
Aerodynamic diameters were calculated for all the SLNas samples on the basis of their shape, BET size, and density characteristics and found to be in the range of 210–676 nm. Despite the values should be higher than 0.5 μm to promote particle deposition onto the alveolar epithelium [
46,
54], the analysis of the respirability performance, which reflects also the degree of particle cohesiveness and de-aggregation, provided more reliable results. The percentage values of Emitted Dose (ED) and Fine Particle Fraction (FPF) were measured by GTI (
Table 2) to define the suitability for the powder to be discharged during the air actuation by DPI and the fraction deposited into alveolar region. Mannosylated samples (SLNas/MS and SLNas/MS-ST) as well as SLNas/ST showed ED values >80%, that is the threshold required by the European Pharmacopoeia [
55] and FPF data around or >40%, indicating adequate respirability performance for antibiotics administered by DPI [
56]. Only SLNas/F127 did not comply with powder respirability requirements (about 70% ED and 12% FPF); this was probably due to its very poor flowability and/or lower surface charge, which led to greater cohesiveness and lower de-aggregation ability compared with the other samples.
Particulate carriers must exhibit adequate respirability characteristics, but they also must ensure an adequate drug payload within them to achieve a therapeutic effect with a feasible dose inside the DPI. In addition, the carriers have to possess the ability to retain the drug before their phagocytosis by AM [
57]. The obtained RIF loading levels and EE values were about 9% and 35%, respectively, without significant differences among the samples (
p > 0.05) (
Table 3).
By considering that RIF was stable under the adopted preparation conditions, as previously demonstrated [
34], EE values indicate that RIF leached partly from the lipid matrix into the continuous phase of the emulsion during the preparation process or dialysis phase, as monitored in preliminary studies. The loaded RIF was found in the amorphous state within all SLNas matrices, as demonstrated by DSC analyses (
Figure 2).
Indeed, RIF characteristic endotherm at about 260 °C (
Figure 2f) was present only in the physical mixture (
Figure 2e). The thermal events of PA and CM, ranging from 50 to 70 °C, were observed in both SLNas and physical mixture thermograms and no modification occurred during the second heating run (data not shown). Thermogravimetric profiles revealed mass losses related to the degradation of SLNas components in the range of 200–600 °C. No mass losses around 100 °C, attributable to dehydration processes, were observed in the samples except for SLNas/ST, which showed a mass loss of about 4% (
Figure 2c). RIF within SLNas lipid matrices was confirmed to be in an amorphous state by XRPD patterns. Unlike the physical mixture showing RIF characteristic peaks at 9.9°, 11.1°, and 19.9° 2θ, even if with decreased peak intensity, all diffractograms exhibited the typical peaks of the matrix components except for RIF peaks, regardless of the sample (
Figure 3).
XPS analysis and wettability measurements were carried out to determine the chemical nature provided by the functionalizing moieties on the SLNas surface. Concerning XPS analysis, a technique for analyzing 5 nm-thick surface layer of the sample, the obtained data indicate the quantitative levels of the elements present on the sample surface (
Table 4).
Carbon (C1s), oxygen (O1s), and nitrogen (N1s), this latter belonging to MS, ST, and RIF, were found on the surface of all the samples, whereas sulphur (S2p) and sodium (Na1s) were detected only in SLNas samples prepared using ST as surfactant or co-surfactant agent. The lowest levels of nitrogen atoms were detected in SLNas/F127 owing to the only contribution of RIF close to the sample surface. Consequently, the highest nitrogen levels exhibited by the other samples are reasonably attributable to the functionalizing/surfactant agents MS and ST located on the particle surface.
As regards wettability, contact angles less than 90° were measured for all SLNas samples revealing that wetting of the surface was favorable (
Table 4). The surface wettability is conceivably attributable to the hydrophilicity conferred by the surfactants to the hydrophobic nanoparticle lipid matrix. The weakest water-wet properties were given by SLNas/F127 in accordance with the values assigned to the brush-like conformation adopted on hydrophobic surfaces by the PEO chains of the surfactant [
58,
59].
3.2. Lipid Corona Formation upon SLNas Contact with Pulmonary Surfactant
Once nanoparticles administered in a dry powder form reach pulmonary alveoli, pulmonary surfactant is the first lung component they encounter as a monolayer or oligolamellar layer at the air-liquid interface of the fluid lining alveoli [
60]. Pulmonary surfactant is also present in the case of human respiratory infections even if with reduced availability of lipids and proteins [
12]. In order to evaluate the possible occurrence of interactions between SLNas samples and pulmonary surfactant, SLNas were treated with Curosurf, a commercial substitute of the natural pulmonary surfactant, and analyzed for size, PDI, zeta potential, and elemental composition. Moreover, the effect of the lipid corona around the nanoparticles on in vitro drug release was assessed.
The interactions between pulmonary surfactant and nanoparticles can be assessed by means of changes in size and zeta potential of the nanoparticles. After the treatment with Curosurf, all the samples increased significantly (
p ≤ 0.05) in size by 1.5 to 5-fold (
Table 5), except SLNas/F127 (
p > 0.05). This finding could be presumably related to the insufficient affinity of Curosurf for F127 functional groups as reported by other authors [
23].
As regards zeta potential, the negative value exhibited by Curosurf (−32 ± 3 mV) was not expected to modify significantly the negative surface charge of SLNas samples and this impaired the contribution of such a measurement to corona formation assessment.
The key role played by the phospholipid fraction in the pulmonary surfactant may justify the use of the EDX analysis focused on phosphorous element detection. The elemental analysis performed by EDX by means of the single point method provided the semi-quantitative composition of each SLNas sample before and after treatment with Curosurf. Moreover, each corresponding ESEM image showed, for all the samples treated with Curosurf, the presence of distinct particles surrounded by a poorly dense layer ascribable to the pulmonary surfactant. Only few aggregates were observed (
Figure 4). The presence, in the samples treated with Curosurf, of phosphorous belonging exclusively to phospholipids as well as sodium chloride contained in Curosurf saline solution was reasonably due to the formation of a phospholipid corona around the particles. Phosphorous was detectable in all the SLNas samples, without significant differences among them (
p < 0.05), with the exception of SLNas/F127 (level below the detection limit of <0.05%) (
Table 6). Unlike sodium and chlorine X-ray emissions due to both material impurities and ST, phosphorous levels were below the detection limit in all the untreated samples (data not shown). Hence, EDX analysis further confirms the lipid corona formation due to interactions between the pulmonary surfactant substitute and the SLNas surface decorated with both MS and ST.
3.3. Effect of Pulmonary Surfactant on Drug Release and Macrophage Activity
Lung surfactant, especially its phospholipid fraction, may impact drug release before AM internalization, as well as particle agglomeration, and cellular uptake due to lipid corona formation on the SLNas surface [
61]. SLNas designed to be taken up by AM should be able to retain most of the active principle upon contact with lung fluids until its entry into AM. Based on that, in vitro drug release from SLNas samples was evaluated in SLF containing or not Curosurf (
Figure 5).
Drug release in pure SLF was gradual from the lipid matrix of mannosylated SLNas (SLNas/MS and SLNas/MS-ST;
Figure 5a,b) and reached RIF percentages of about 40% in 3 h. In contrast, about 60% RIF was released from the control samples SLNas/ST and SLNas/F127 (
Figure 5c,d). Drug retaining capability exhibited by mannosylated samples could be attributed to a closer settlement of the functionalizing agent lipophilic tale within the particle lipid phase which may generate a firmer SLNas matrix as observed in a previously published research [
5]. Therefore drug spreading over the lung fluid before AM uptake should not occur for mannosylated SLNas, since 50–75% of inhaled particles should be taken up by AM in 2–3 h [
47]. RIF release in SLF containing Curosurf decreased by 20–25% (
p < 0.05) for all the samples, except for SLNas/F127, that is the only sample that proved to fail in interacting with the pulmonary surfactant substitute. The observed improvement in drug retention, attributable to an additional resistance to drug diffusion provided by lipid corona layer, could be considered as an advantage as it leads to a greater amount of drug available within the lipid matrix to be transported into AM. After AM phagocytosis, lipid materials are known to be degraded inside macrophages by lipolytic enzymes, allowing drug intracellular leakage and antibacterial activity [
62].
Besides the conflicting results concerning the effects of protein/lipid corona on biological interactions between nanoparticles and cells, it is reasonable to expect that the activity generated by the surface chemistry of particles coated by the pulmonary surfactant layer might undergo some changes upon contact with AM membrane and its MR. Consequently, the effect of the pulmonary surfactant layer on SLNas ability to be phagocytosed with or without impact on MR interactions was investigated on MH-S cell line, referred as an in vitro model similar to human primary cells [
63]. All the samples and pure RIF were preliminarily subjected to cytotoxicity test according to the incubation time.
SLNas cytotoxicity was investigated at increasingly sample amounts (from 0.125 to 1 mg/mL) and a significant dose-dependent cell cytotoxicity was found for the mannosylated SLNas samples starting from the dose of 0.5 mg/mL (
p < 0.05) (
Figure 6a).
Since excipient concentrations and RIF loading levels are recognized as safe (non-cytotoxic) in the obtained SLNas (
Figure 6b), the dose-dependent cell apoptosis could be reasonably attributed to the increase of particle amounts [
64] or particle clustering phenomena [
65,
66]. With respect to SLNas/F127, such an effect may have been precluded by F127 surfactant recognized to be capable of inhibiting nanoparticle cell adhesion [
67].
Based upon the cytotoxicity results, the study of SLNas macrophage internalization was performed on 0.25 mg/mL SLNas dose, corresponding to about 22.5 μg/mL RIF for all the samples, that is higher than the minimum inhibitory concentration against
Mycobacterium tuberculosis strain (1 μg/mL) [
68]. Intracellular RIF concentrations were determined on SLNas samples up to an incubation time of 3 h that is considered consistent with both AM phagocytosis rate [
47,
69] and the occurrence of intracellular RIF [
5]. The internalization data were compared with those obtained following both MR inhibition and pulmonary surfactant co-treatment (
Figure 7).
Fast and complete drug translocation within AM was achieved by cell exposure to SLNas/MS and reached 100% intramacrophagic RIF at 0.5 h incubation (
Figure 7a). This maximum drug concentration was maintained intracellularly until drug degradation occurs, leading to a concentration decrease by one-half at 3 h incubation (
p < 0.05). Conversely, SLNas/MS-ST sample behaved as the controls SLNas/ST and SLNas/F127 and showed RIF translocation values less than 50%, regardless of the incubation time (
Figure 7b–d). The greatest internalization ability of SLNas/MS is therefore the consequence of SLNas mannose recognition by MR, which are highly expressed on MH-S cell line [
70]. By contrast, ST seemed to not contribute to AM internalization through mechanisms of active targeting. Comparison between drug translocation efficiency provided by SLNas/MS and that by SLNas/MS-ST therefore gives an indication of the key role played by mannose density on nanoparticle surface [
71]. MR-mediated uptake by macrophages was investigated by cell pre-treatment with
d-mannose to saturate the corresponding receptors [
5,
72,
73]. MR inhibition resulted in a relevant decrease (about 50%) of intracellular RIF only by cell exposure to SLNas/MS at 0.5 h (
p < 0.005) and at 1 h (
p < 0.01), before drug degradation occurrence at 3 h incubation (
Figure 7a), confirming the considerable contribution from the active targeting process. Otherwise, RIF transport across macrophage membrane was only slightly affected (10% intracellular RIF decrease) by MR inhibition upon cell contact with SLNas/MS-ST (
p < 0.05 at 1 and 3 h;
Figure 7b) and SLNas/ST (
p < 0.05 at 3 h;
Figure 7c), suggesting a marginal contribution from MR-mediated cell uptake. Insignificant changes in RIF translocation were detected for SLNas/F127 (
p > 0.05;
Figure 7d), regardless of the incubation time. It follows that a passive targeting could be assumed as the main mechanism for the macrophage uptake of these latest SLNas samples.
Concerning the effect of pulmonary surfactant on SLNas cell uptake, the interactions between mannose groups on SLNas/MS surface and macrophage MR were negligibly impaired by the presence of the pulmonary surfactant substitute, regardless of the incubation time (
p > 0.05) (
Figure 7a). This finding suggests that the adsorbed lipid corona layer does not completely mask the functional groups on nanoparticle surface as observed also by other authors [
74]. Corona formation appeared instead to promote macrophage internalization when passive targeting regulates the process, i.e., for SLNas/MS-ST and SLNas/ST samples, leading in general to increased RIF intracellular concentrations (from about 10 to 20%) (
p < 0.05;
Figure 7b,c). Different mechanisms could be hypothesized for an increased intracellular uptake generated by pulmonary surfactant coating on nanoparticles, including recognition promotion by macrophages and hydrophobic interactions or fusion between nanoparticle surface and macrophage membrane similarly to the process involving liposomes [
31]. The lack of a significant uptake modification provided by pulmonary surfactant on SLNas/F127 (
p > 0.05) proved further the absence of lipid corona.