Physicochemical characterization of iron oxide MNPs with different coatings
In this study we examined in mice the biodistribution and the biotransformation of MNPs with iron oxide cores of the same size but coated with APS, DEX or DMSA. The iron oxide MNPs were synthesized by co-precipitation method following the protocols described previously [46] and after core synthesis, a standard protocol was used to oxidize magnetite to maghemite activating the MNPs surface for coating [47, 48]. Although these types of MNPs have been used in previous studies of our group [30, 40, 49] we present here the physicochemical characteristics of the batches prepared for this new study, which may differ slightly from data obtained from other batches used in earlier works. In brief, TEM images revealed them to be monodisperse iron oxide MNPs ~ 12.0 ± 1.2 nm in diameter. The different coatings of these iron oxide cores produced MNPs with different surface charges, positive (APS), neutral (DEX) and negative (DMSA: Additional file 1, Fig. S1 a, b). In brief, TEM images revealed them to be monodisperse iron oxide MNPs ~ 12.0 ± 1.2 nm in diameter. The different coatings of these iron oxide cores produced MNPs with different surface charges, positive (APS), neutral (DEX) and negative (DMSA: Additional file 1: Fig. S1 a, b).
The hydrodynamic radius of the APS, DEX and DMSA coated-MNPs were 122 nm, 109 nm and 83 nm, respectively (Additional file 1: Fig. S1c) [40]. Thus, in addition to producing different surface charges, these coatings also affected the final size of the MNPs in suspension. The MNPs appeared to form small aggregates, although in all cases a single monomodal size distribution was recorded with polydispersity index (PDI) < 0.3 [48]. The Z-potentials confirmed the surface charge of the APS (+ 23 mV), DEX (-1.8 mV) and DMSA (-34 mV) coated MNPs (Additional file 1: Fig. S1d) [40]. Moreover, the M(H) hysteresis loop confirmed their superparamagnetic behavior at room temperature (RT). Finally, the saturation magnetization values of all the samples were ~ 80 Am2/kgFe (Additional file 1: Fig. S1e), consistent with previously reported values for γ-Fe2O3 nanoparticles [40] and in agreement with the chemical structure determined by Mössbauer spectroscopy of particles prepared by this methodology [50, 51].
The AC magnetic susceptibility measurements were temperature dependent (50–350 K range) and the typical relaxation phenomenon of MNPs was observed, an in-phase magnetic susceptibility maxima [χ’ (T)max] together with an out-of-phase magnetic susceptibility [χ″(T)] maximum at slightly lower temperatures (Additional file 1: Fig. S1f). The maximum of the MNPs with each of the coatings was detected at slightly different temperatures depending on the interparticle dipolar interactions and in agreement with the dynamic light scattering (DLS) results, being the DMSA coated particles the ones presenting the lower degree of aggregation. The temperature of these maxima (190–220 K range) was used as a fingerprint of the presence of the MNPs in animal tissues. Furthermore, the susceptibility per mass of iron in the form of particles was used to quantify the MNPs in tissues [43].
Experimental study of iron oxide MNP degradation in vivo
MNP degradation was studied in vivo in 5-week-old female C57BL/6 mice (Envigo Laboratories) maintained under controlled conditions at the National Center for Biotechnology (CNB) animal facility. The mice were randomly divided into four groups of 7 animals, that each received five doses (at a twice weekly frequency) of PBS (control), APS, DEX or DMSA coated MNPs (100 μl of MNPs, 2.5 mg Fe/mice) by retro-orbital i.v. injection under isoflurane anesthesia (0.5–5% inhaled). This dose schedule is the same schedule that we routinely use for MNP injection to treat tumors in mouse models of cancer [17, 52]. After administration of the MNPs, different health parameters were evaluated in all the mice, including weight, physical appearance, blood cell populations and hepatic toxicity profile. In all cases, these results were compared to the control group that received PBS alone (Fig. 1a). To evaluate the possible toxic effects of the APS, DEX or DMSA coated MNPs, the appearance of several signs of systemic toxicity was evaluated over the 15-month study period, assessing bradykinesia or lethargy, piloerection, gastrointestinal symptoms and irregular breathing. No signs of acute toxicity were observed at any of the time points analyzed after the administration of the different coated MNPs.
The evolution of the body weight of the mice over the entire experiment showed no significant differences among the groups (Fig. 1b) and the leukocyte profile was very similar between the mice treated with MNPs and the untreated mice, indicating there were no infections or symptoms of toxicity in mice inoculated with the different coated MNPs (Fig. 1c). The serum from mice injected with APS-, DEX- or DMSA-MNPs showed transient increases in aspartate aminotransferase (AST) and alanine aminotransferase (ALT) relative to the controls, enzymes associated mainly with hepatic damage. The increase in AST peaked from 7 days to 1-month post-treatment (Fig. 1d), whereas the increase in ALT was only produced by DMSA-MNPs at 7 days (Fig. 1e). However, these increases did not compromise the survival of the mice and the levels detected were within the range of normal values, similar to those detected in the control female C57BL/6 mice considering their age and the blood extraction method (~ 213.57 ± 38.42 U/L).
From the results obtained, we concluded that none of the MNPs were toxic to the mice over the observation period at the doses used. After sacrificing the animals, the size, appearance and color of the internal organs was apparently normal in all cases. Indeed, similar results were observed previously where DMSA-coated NPs synthesized by decomposition in an organic medium or DMSA-MNPs obtained by co-precipitation were administered to mice, although some mild toxicity was observed at 7 days in these earlier experiments that normalized over time, supporting our conclusion that this type of MNP did not cause toxicity over a period of 90 days [29, 53].
Biodistribution of the iron oxide MNPs with different coatings in mouse tissues
To determine whether the NP coating influenced the biodistribution of the MNPs, we assessed whether the MNPs were still circulating in the blood or if they were located in the organs, and where they tended to accumulate 7 days after the last dose administered. Thus, the amount of iron in the blood was determined by ICP-OES, comparing each of the coated MNPs with the controls. The iron detected in the blood of the mice treated with the MNPs was lower than that detected in the untreated control mice (Additional file 1: Fig. S2), which would appear to reflect the regulation of iron metabolism after MNP internalization [54, 55].We previously found that in magnetic susceptibility analyses DMSA-coated NPs were not detected in the blood between 30 min of administration and up to 90 days [29]. Elsewhere the half-life of NPs with similar coatings in the blood was reported to be between minutes and 62 h depending on the coating and animal model used [56]. Hence, the blood residence time of the MNPs appears to be less than 7 days irrespective of the coating they carry. In fact, 7 days after administration the MNPs had accumulated in the different organs in which they were distributed [36, 57, 58], which led us to analyze the biodistribution of all the MNPs at this time.
The biodistribution of APS, DEX or DMSA coated MNPs was studied in C57BL/6 mice using AC magnetic susceptibility measurements after i.v. administration of five doses. This technique was especially relevant as the susceptibility maxima serves as a fingerprint of the presence of particles in a given tissue [42] and it has not only been used previously for biodistribution studies but also, to follow the biotransformation of MNPs over time [17, 29, 35, 39]. In our experiments, spleen, liver, kidney, lung, heart and thymus tissues were characterized magnetically to track MNP accumulation. The AC susceptibility signal obtained from the MNPs in the spleen and liver was detected at similar temperatures as when these MNPs were assessed in agar (compare Additional file 1: Fig. S1f with Fig. 2a, b), confirming the presence of the material administered in these tissues. We found minimal or no signal from the MNPs in lung, kidneys, brain, heart and thymus tissues, at least not within the limits of detection of the technique (~ 0.7 μgFe: Fig. 2c). As expected, no signal from the particles was found in tissues from the control (PBS-treated) mice (Additional file 1: Fig. S3).
The initial biodistribution of MNPs depends on several factors, including their half-life in blood, the mouse strain, the injection dose, repetitive administration or the induction of anesthesia [29, 59, 60]. However, the physicochemical properties of the MNPs (surface charge, coating and the size of the core) exerts the greatest influence on the time of circulation in the blood [61]. Generally, iron oxide MNPs that exhibit long blood half-lives have limited distribution into the liver cells with significant uptake into the macrophage cells of other organs like the spleen, lymph nodes and bone marrow [62]. We found that 7 days after administration all the MNPs were localized in the spleen and liver, although a correlation was observed between the MNP coating and their final organ localization. Cationic surface APS-MNPs accumulated more in the spleen than in the liver, whereas anionic surface DMSA-MNPs and neutral DEX-MNPs accumulated similarly in the liver and spleen. Previously, it was proposed that stronger uptake of MNPs by liver macrophage and endothelial cells is related to a shorter circulation time of the particles in the blood [56, 63].
Positively charged particles like APS-MNPs may have longer circulation times in the blood than negatively charged particles [29, 56], which could explain the differences in the accumulation of APS and DMSA coated MNPs in the liver. In addition to the influence of the surface charge of the particle, its size is also an important factor to consider as smaller particles tend to remain in the bloodstream for longer [27]. By contrast, larger MNPs (> 50 nm in diameter) were more easily sequestered by macrophages in the liver and spleen [26, 64]. The MNPs used in this study had a hydrodynamic size between 165 and 1554 nm when incubated for 24 h with mouse serum (Additional file 1: Fig. S4), such that it was expected they would mainly be sequestered by macrophages in the liver and spleen, accumulating more strongly in these organs. Another issue to be considered is the composition of the PC as there are proteins like albumin and apolipoproteins that have a stabilizing effect, while others like fibrinogen trigger particle aggregation [40, 65]. Consequently, macrophages modified the internalization rate, the endocytic pathways used and the MNP uptake times [38, 66] based on the factors to which they have been previously exposed.
To complement the MNP biodistribution studies performed by AC magnetic susceptibility, mouse liver and spleen samples from treated C57BL/6 mice were stained using the Prussian blue technique 7 days to 15 months after i.v. MNP administration, showing iron accumulation in areas of both tissues by light microscopy (the iron that corresponds to the presence of MNPs was observed in blue, while cells were observed by counterstaining with neutral red). Prussian blue staining of spleen tissue sections showed iron accumulated mainly in the red pulp of the spleen following administration of any of the MNPs studied. Stained areas in the red pulp were evident in spleen sections, even in control tissues (Fig. 3), possibly due to the storage of iron degradation products as a result of erythrocyte phagocytosis and the presence of splenic macrophages [29, 67]. This accumulation of MNPs in the red pulp was expected as it is the spleen area that specializes in filtering the blood, eliminating old erythrocytes, pathogens or foreign elements. Many of the elements that circulate in the blood, such as aged erythrocytes, pathogens and MNPs, arrive transported by the arterial blood into the reticular fiber network of the spleen red pulp, where they are first retained and later phagocytosed by the many macrophages that are located at this reticular fiber network [67,68,69]. Since the spleen red pulp is a physiological storage site for iron, erythrocytes and platelets, iron stain was even observed in the red pulp of control mice [24]. The Prussian blue staining observed at 7 days in the spleen sections reflects this conclusion, along with the magnetic susceptibility measurements (Fig. 2). MNPs accumulate in the spleen in greater proportions when they were coated with APS as opposed to DEX and lastly, DMSA (Fig. 2).
From 7 days to 6 months after APS-MNP administration, iron was observed in both the red and white spleen pulp, and subsequently, a large amount of iron was internalized into the spleen until 9 months when it began to decrease. However, the iron signal did not disappear in the spleen of mice that received APS-MNPs. There was a gradual decrease of iron staining in the white pulp of spleen sections from mice that received DEX-MNPs after 6 months, and iron was observed for up to 3 months in both the red and white pulp of the spleen in mice that received the DMSA-MNPs, after which the amount of iron began to decrease. This appearance of an iron signal in the white pulp could be related to an excessively strong increase in iron in the spleen, which was more accentuated after the administration of APS-MNPs followed by DEX- and DMSA-MNPs. Hence, the iron signal persists longer in the white pulp of spleens in mice treated with APS-MNPs. Data from mice that received polyacrylic acid-coated NPs (PAA-NPs) was consistent with our results in which iron accumulated in both the white and red splenic pulp. No iron accumulated in the white pulp of the control mice, only in the red pulp, which might reflect the storage of iron degradation products as a result of erythrocyte phagocytosis [70].
In liver sections there was a reduction in the number of iron clumps (Fig. 4), as well as an increase in the size of the iron deposits over time, which might be explained by the formation of phagocytic cell clusters in the liver parenchyma [24]. In liver sections stained with Prussian blue iron complexes, a specific homogeneous distribution of iron throughout the liver sections was evident after short times (7 days or 1 month of MNP administration), which shifted to an accumulation close to the blood or bile ducts in mice treated with MNPs from 1 to 3 months. At longer times, between 6 and 15 months, the amount of iron observed in the liver decreased in all the sections analyzed (Fig. 4, bottom right insets). Apart from the accumulation of iron in the spleen and liver, we found no structural or histopathological changes in any of these tissues.
In summary, these results suggested that the type of coating plays a crucial role in the biodistribution of the MNPs, probably due to the changes induced in their physicochemical properties: surface charge, state of aggregation, hydrodynamic size and interaction with biological media [71,72,73]. In most cases, these differences dictate the amount of MNPs that accumulate in the different organs.
Short term intracellular degradation of iron oxide MNPs with different coatings in endolysosomal vesicles
Although MNPs accumulated in the liver and spleen, regardless of their coating, their proportions in these organs and their degradation over time did seem to be influenced by the coating. Hence, the intracellular degradation of MNPs was studied within 24 h after internalization in two different macrophage lines: RAW 264.7 cells [30], a murine circulating macrophage-like cell line; and NCTC1469 cells, a mouse liver-derived macrophage-like cell line [29, 35]. Biodegradation studies identified a loss of the magnetic properties of MNPs after administration, which is correlated with an increase in iron metabolism suggesting their active degradation [74, 75]. The availability of iron derived from MNPs depends on the mechanisms by which nanoparticles are internalized by cells and how this internalization influences their degradation [35]. The highly proteolytic properties of endolysosomes, such as low pH, high ionic strength and the presence of various catabolic enzymes are primarily responsible for the degradation of nanomaterials [76, 77]. MNP degradation first occurs at the level of the PC associated with the MNP surface after coming into contact with blood or other biological fluids [40]. The coating or functionalization of the MNPs then degrades and finally, the metallic core disintegrates [78]. Each of these processes is influenced by the nature of the MNPs, the type of cell in which degradation occurs and the cell’s metabolic state [79].
To assess whether the MNPs undergo different rates of intracellular degradation, we monitored the size of the cells magnetic MNP core within the endolysosomes 24 h after RAW 264.7 and NCTC1469 were exposed to these particles. Fractions were isolated that were enriched in the expression of bona fide endolysosomal markers and hence, of these organelles (Additional file 1: Fig. S5). Nevertheless, cell viability was not affected by MNP treatment even after a 24 h incubation with MNP iron concentrations up to 125 μgFe/ ml (Additional file 1: Fig. S6). To follow the core size reduction as an indication of MNP degradation by TEM, the mean particle size of over 100 MNPs was measured at 24 h (Additional file 1: Fig. S7). Considering the core size of the MNPs in water as the baseline (12.0 ± 1.2 nm: Fig. 5), there were differences in the intralysosomal degradation of the APS-MNPs between cell types, with a greater reduction of the core size in RAW 264.7 cells (8.8 ± 1.3 nm) than in NCTC1469 cells (10.3 ± 1.5 nm). DMSA-MNPs were rapidly degraded in both cell types, with a higher degradation rate than APS-MNPs in the liver-derived macrophage NCTC1469 line (9.7 ± 1.9 nm: Fig. 5 a, b). Finally, poor in vitro degradation of DEX-MNPs was evident in both cell types, perhaps related to their weaker internalization in these cells (Additional file 1: Fig. S8).
The influence on the reduction in the core size of each MNP was corroborated by field-dependent magnetization M(H) measurements. At RT, the M(H) curves of the MNP loaded endolysosomes displayed slight differences between cell types and MNP coatings (see Fig. 5). Common features of the endolysosome loaded MNP M(H) hysteresis curves relative to the MNP suspensions included a paramagnetic contribution over time (i.e.: a lineal increase in magnetization in the high magnetic field region), which was probably related to the presence of Fe ions, a decrease in saturation magnetization of the endolysosomes from the cells exposed to MNPs for 24 h and a reduction in magnetic susceptibility (dM/dH slope at low magnetic field) as a consequence of the reduction in MNP size. In general, there was more degradation of the DMSA-MNPs in both cell types, reflected in the reduced size of the core detected in TEM images and by the magnetization analyses (Fig. 5c). APS-MNPs were degraded more severely in the RAW 264.7 cells than in the hepatic NCTC1469 cells, indicative of greater degradation in spleen macrophages (Fig. 5d). Finally, no significant degradation of DEX-MNPs was observed in either of the two cell types analyzed (Fig. 5e), which could be related to the poor internalization of these particles given their almost neutral surface charge (Additional file 1: Fig. S1d).
It should be noted that 24 h is a short period to draw meaningful conclusions about the degradation of each MNP. However, it was impossible to monitor degradation beyond 24 h due to the multiplication of the cells in vitro and thus, we were only able to analyze this 24 h period of intralysosomal degradation. To study the disappearance of MNPs after their administration and the differences in degradation kinetics, degradation experiments in vivo should be performed over longer time periods, establishing the differences between coatings at the organ level.
Degradation of iron oxide MNPs in the liver
To get a deeper understanding of the intracellular degradation of MNPs with different coatings in hepatic macrophage cells they were studied at longer times after administration to C57BL/6 mice. This analysis focused on liver macrophages as they showed greater differences in MNP degradation over short periods and approximately 30–99% of the MNPs administered accumulate in the liver after administration [80, 81]. To perform this analysis, specific F4/80 staining was studied in liver tissue sections between 7 days and 15 months from the administration of the last dose of the MNPs. After counterstaining the sections with Prussian blue and studying the co-localization of F4/80 with the iron signal, any decrease in this co-localization in hepatic macrophage cells was considered a sign of degradation.
The liver is a complex network of interrelated cells, with specialized epithelial cells, hepatocytes, representing approximately 60–80% of its parenchymal cells. Other liver cells include: Kupffer cells (KCs) and mobile macrophages, hepatic sinusoidal endothelial cells, hepatic stellate cells, biliary epithelial cells (cholangiocytes), resident immune cells (dendritic cells, natural killer cells, and lymphocytes) and circulating blood cells in transit through the liver. KCs represent 80–90% of the total body macrophage population and they are responsible for most of the phagocytic activity in the liver [82,83,84].
Using immunohistochemistry with F4/80 markers on paraffin embedded sections, we assessed whether the MNPs were located fundamentally in the liver macrophages or KCs, and where they were degraded [36, 61, 85]. Mainly DMSA and APS coated MNPs were partitioned early into KCs in the liver (see Fig. 6). When Prussian Blue staining/iron co-localization with the KC F4/80 macrophage immunolabeling was studied, a decrease in the iron staining in KCs was observed over time. Greater iron staining in KCs was observed at short degradation times for DMSA-MNPs, between 7 days (42.22%) and 1 month (77.78%). However, the increase in the presence of iron in KCs was not observed until 14 days with APS coated MNPs (55.55%) and 6 months for DEX coated MNPs (62.50%). When DMSA-MNP intracellular degradation was observed, 1 month after administration, there were fewer KCs in which iron was detected, whereas this decrease in KCs containing iron was not observed until 6 and 12 months for APS- and DEX-MNPs, respectively. These data suggested that DMSA-MNPs were degraded more rapidly in the liver than APS or DEX coated MNPs. Finally, when degradation was analyzed at the final time point (15 months), iron detected in KCs administered DMSA-MNPs (10.52% of the KCs containing iron) was similar to the iron detected KCs administered PBS (9.47% of the KCs containing iron). In the case of APS and DEX coated MNPs, the blue iron signal still co-localized with hepatic macrophage labeling after 15 months (21.82% for APS-MNPs and 38.46% for DEX-MNPs), suggesting a slower degradation of these MNPs.
The importance of KCs in innate immunity and the degradation of intracellular iron has been highlighted [61, 86]. KCs are liver resident macrophages that engulf and destroy pathogens, as well as other foreign bodies and materials in the blood. These macrophages are also involved in erythrocyte recycling and apoptotic cell digestion [80], and they are thought to recognize MNPs as foreign material and internalize them through multiple receptors [37]. The uptake and retention of MNPs by KCs is strongly correlated with their surface charge, the nature of their chemical coating and their size [87]. Larger particles are usually phagocytosed more easily by this cell type, while MNPs with strongly cationic and anionic surface charges adsorb a quantity of serum proteins to form their PC and can aggregate, interacting more readily with macrophages in vitro. Most surface-neutral ligands adsorb less serum proteins to their surface and they are therefore less efficiently absorbed by phagocytic cells than more charged nanoparticles [88] (Additional file 1: Fig. S4).
The differences in the MNP coating and the associated differences in the composition of the PC may explain why DEX-MNPs reside longer in KCs, as these MNPs were phagocytosed more slowly by macrophages due to their neutral charge [89, 90]. Regarding the influence of PC on the internalization of MNPs with different coatings in liver macrophages, we previously observed that the PC associated with APS and DMSA coated MNPs was more diverse in terms of size and composition, with the presence of complement proteins and immunoglobulins that favor the opsonization of MNPs by macrophages [91]. These results suggested that not only do MNPs accumulate in a higher proportion in the liver or spleen, depending principally on their coating and influenced by the hydrodynamic size of the MNPs, but also, that the MNPs seem to have different degradation rates in the liver at least, with anionic DMSA-MNPs being degraded faster than the cationic APS-MNPs.
Long-term degradation of iron oxide MNPs with different coatings
In light of the above, we evaluated MNP degradation over 15 months by monitoring the magnetic susceptibility measurements in both the spleen and liver, the tissues in which the MNPs mainly accumulated. The long-term changes to the MNPs accumulated in these tissues was assessed by AC susceptibility from 7 days to 15 months post-administration. In general, MNPs were detected in the spleen tissues throughout the post-administration period analyzed, whereas MNPs were only detected in liver samples until 3- or 6-months post-administration depending on the NP coating (Figs. 7, 8). These results were consistent with the Prussian Blue staining data obtained for both these organs.
In addition, a new contribution from a different iron-containing species was detected 14 days after MNP administration in the spleen tissue, with a maximum in the out-of-phase susceptibility located at 8–10 K, accompanied by a maximum at slightly higher temperatures in the in-phase susceptibility. This signal was also observed in some of the hepatic tissue samples at longer time points and it corresponded to the typical signal of ferritin, the iron storage protein that enables iron to accumulate in a biomineral form inside the protein cage [43, 45, 75]. Moreover, a paramagnetic signal was observed in the in-phase magnetic susceptibility at low temperature in some samples of liver tissue. This may be attributed to other iron-containing species in which this element is not part of any mineral or biomineral form.
In this type of measurement, the height of the magnetic susceptibility maxima when plotted per tissue mass is directly related to the iron concentration in the tissue (mFe/msample). A decrease in the height of the susceptibility maximum implies a decrease in the number of particles in the tissue and/or their degradation. Here, the general trend detected was a decrease in the height of the susceptibility maximum over time for all MNPs and in both organs, indicating the disappearance of the particles over time. In addition, an increase in the signal corresponding to the presence of ferritin was observed in the spleen samples. As well as the decrease in the maximum height associated with the particles over time, it was interesting to evaluate possible changes in the location and temperature of such maxima. A change in the shape or temperature location of the MNP’s signal in the spleen and liver was observed for all the particles (Additional file 1: Fig. S9 and S10), which was also consistent with the continued degradation of the particles over time. In general, a variation of the temperature corresponding to the maximum of susceptibility was seen with time for the three types of coatings. This variation was greater in the liver with respect to that observed in the spleen, indicating stronger or faster degradation of the MNPs. The data suggested that the liver broke down the particles to a smaller average size over the same time.
Regarding the temperature location of the different susceptibility maxima, the highest temperatures were observed in the spleen for DEX and APS coated MNPs, while for DMSA-MNPs these temperatures were similar in the liver and spleen. Hence, APS and DEX coated MNPs appear to agglomerate more in the spleen [65] than those coated with DMSA. Furthermore, although the amount of DMSA coated MNPs decreased over time in the spleen, the width of the susceptibility signal did not vary as much as with the other particles, suggesting that their particle size distribution was better maintained over time. This could be due to the particles remaining in the spleen being excreted, reducing the number of intact particles in this organ or alternatively, the MNPs were not degraded concomitantly but rather, particle by particle. We consider that their translocation elsewhere from spleen is less probable.
Evaluation of the MNP concentration over time as an indication of their partial or total degradation in the liver and spleen
To evaluate the degradation of MNPs, the variation in the iron species was quantified from the AC magnetic susceptibility measurements, as described previously [42, 43, 92]. The AC magnetic susceptibility was determined using the sample magnetic moment under an alternating magnetic field at different temperatures, presenting an in-phase or real component (χ') and an out-of-phase or imaginary component (χ"). In a biological sample, all the magnetic species present can contribute to the AC magnetic susceptibility, yet the MNPs are the only species that significantly contribute to χ" [92]. The typical signal of the iron oxide MNPs used for biomedical applications in the out-of-phase susceptibility component has the form of a maximum, and the temperature location of this maximum depends on the size distribution and aggregation of the particles. In addition, the height of this maximum is a surrogate indicator of the number of particles in the sample. This characteristic is very useful to study the presence of NPs in biological samples. Therefore, the amount of iron corresponding to the NPs and to ferritin was quantified from the out-of-phase susceptibility data, while the amount of paramagnetic iron ions in each tissue was quantified from the in-phase susceptibility when this signal was observed [43, 93].
The number of particles was estimated in the spleen and liver for up to 15 months and in general, the iron concentration in the form of particles with the same coating was always higher in the spleen than in the liver at the same time point (see Fig. 9a, b). Nevertheless, as the liver is a much larger organ than the spleen, the total iron content in the form of particles was distinct in the liver and spleen (Fig. 9c). In terms of MNPs coated with APS, the total iron mass corresponding to MNPs was similar in the spleen and liver, while for MNPs coated with DEX or DMSA the MNP iron mass in the liver was 5 or 9 times greater than in the spleen for approximately 7–14 days. The concentration of particles in both tissues decreased systematically over time, suggesting that the MNPs accumulated in these organs were degraded. In the long term, it was interesting that 9 months after their administration none of the three types of MNPs were detected in liver tissues.
To compare the rate of degradation between these organs without taking into account the effect of the initial accumulation of particles, the total iron mass was determined at different times and normalized to the iron mass quantified at 7 days for each organ. In general, degradation in the liver was faster than in the spleen and similar for the three types of MNPs in this tissue (Fig. 9d). There was no evidence that the speed of MNP degradation in the liver was affected by the type of coating. However, there were small differences in the time when MNPs were no longer detected, although this was probably due to the fact that the MNPs coated with DEX and DMSA initially accumulated more intensely than the APS-MNPs, requiring slightly longer for their complete degradation. Moreover, the differences in the initial accumulation may be related to the different MNP coatings. The time for complete clearance of the MNPs in the liver was consistent with the immunohistochemical analysis of these tissues, which suggested the faster clearance of DMSA coated MNPs in the liver than those coated with APS or DEX. Previous studies into the long-term in vivo fate of gold/iron oxide heterostructures (NHs) showed that their accumulation in the liver was greater in NHs coated with an amphiphilic polymer (PC-NHs) than those with poly(ethylene glycol), PEG-NHs [36]. This is due to the effect of some coatings like PEG in reducing opsonization by macrophages, which enhances the circulation time of the NPs [94, 95]. By contrast, the negatively charged amphiphilic polymer is not as effective in preventing macrophage uptake in the liver and spleen [27, 36, 59].
Other studies that monitored the bioassimilation of empty copper sulfide NPs (CuS-NPs) or NPs with a flower-like core of iron oxide (iron oxide@CuS-NPs) finally coated with PEG, showed their accumulation mainly in the liver and spleen following i.v. administration to 6-week-old Balb/C mice [96]. Regarding the degradation of these particles up to 6 months post-administration, they appeared to remain intact within the liver and spleen in TEM images for up to 7 days post-administration. However, 3 months after administration no intact hybrid particles were detected, although structures similar to ferritin were detected that were indicative of their degradation [96]. Ferritin also appeared here, probably reflecting the degradation of the MNPs although at times that differed from those indicated previously, which might vary for different reasons including the particle coating.
By contrast, slower degradation was seen in the spleen than in the liver and even 15 months after their administration, the three types of MNPs could be detected there. The degradation speed was similar for the three coatings in the spleen (Fig. 9d). Recent studies carried out on 17 commercial particles with different physicochemical properties (size, coating and surface charge) showed faster degradation of smaller particles and for those with a negative Z-potential, indicating that the type and structure of the coating strongly influences MNP degradation [27]. Here, we included an additional parameter that should be taken into account in future degradation studies, the amount of particles that initially accumulate within an organ, as this may be a key parameter to determine the time for complete particle clearance in an organ. The differences in the rate of degradation between APS and DMSA coated MNPs could be explained primarily by their surface charge, which dictates how quickly MNPs are trapped by macrophages. Our results suggest new directions to control NP degradation by indicating that the composition of the synthetic surface affects the residence time of NPs in the body.
Analysis of other iron-containing species in the liver and spleen
The kinetics of ferritin accumulation were followed through a quantitative analysis of the iron concentration stored in the form of this protein (see Fig. 10). Ferritin presents a superparamagnetic behavior at room temperature and a characteristic out-of-phase magnetic susceptibility maxima at around 8–10 K, in the same AC field conditions as those used here [45]. A reference ferritin from mouse liver was used for the quantitative analysis and in general, there was a trend towards higher concentrations of iron stored in the form of ferritin in all the spleens studied over time, including the controls. This may be explained by the natural iron accumulation in the spleen over the life time of the animals [55, 97].
At the shortest time points studied, 7 and 14 days after administration, no ferritin was detected in the spleens of any of the treated groups, whereas a small amount was detected in the controls (Additional file 1: Fig. S11). This may be due to the strong signal from the particles masking the smaller ferritin contribution (Fig. 8). Between 1 and 9 months, the signal corresponding to ferritin for the DMSA- and DEX-MNPs was very similar to the controls, whereas the APS-MNPs produced slightly more ferritin than the controls in that period of time. The difficulty in assessing iron accumulation in the form of ferritin is a consequence of the small amount of iron in the form of particles that accumulate in the organs (0.3–0.8 μgFe-MNPs/mgtissue) and that is potentially degraded into ferritin relative to the amount of endogenous ferritin (9 μgFe-ferritin/mgtissue). This difference (> tenfold) hinders accurate comparisons given the variability in the animals. At the last time point, 15 months, the amount of ferritin in the mice administered DMSA- and DEX-MNPs was higher than in the controls and in those administered APS-MNPs (Fig. 10a).
A completely different behavior was observed in the liver and no ferritin signal was detected in control organs at any point in the experiment. In the treated mice, it was possible to detect a small signal corresponding to ferritin in some of the liver samples long after MNP administration (6 months in the mice administered DMSA and DEX coated MNPs, and at 12 months in the APS-MNP mice). This behavior was consistent with the degradation of MNPs and the partial accumulation of ferritin. At the last time point measured, 15 months after MNP administration, no ferritin signal was detected, which probably reflects the complete clearance of MNPs as no sign of the MNPs was detected 9 months after their administration (Fig. 10b).
After quantifying the MNPs accumulated in the liver from the out-of-phase susceptibility, the paramagnetic iron that corresponds to iron ions not forming part of any mineral species was quantified from the in-phase susceptibility. In general, a similar amount of paramagnetic iron was observed (Fig. 10c) independent of the time after administration or the particle coating (Additional file 1: Fig. S11), indicating that such paramagnetic species were probably part of the basal iron species and that they were not formed as a result of MNP degradation. In the case of iron oxides, dissolution and recrystallization to form other magnetic nanoparticles has only been proved for the very specific case of stem cells [98]. In general, given the iron metabolism pathways present in the organism, the iron atoms released from the particles become part of the body iron, mineralizing in the form of ferritin, as observed here or being uptaken by other proteins related to the iron storage and transport [75].
In summary, after a period in the blood, MNPs injected intravenously accumulate in the spleen and liver, being taken up by macrophages in these organs. The amount of particles that accumulated in the different tissues is affected by their coating, a key parameter to understand the time needed for the complete clearance of the particles from the different organs. MNPs were degraded at different rates depending on the organ, degrading faster in the liver than in the spleen. In particular, the MNPs had been completely degraded in the liver after 15 months while some particles still remained in the spleen. Depending on the residence time in the organism required, a tailored coating could be designed to fulfil the needs of any future application.