How safe are magnetic nanomotors: From cells to animals

Abstract

Helical magnetic nanomotors can be actuated using an external magnetic field and have potential applications in drug delivery, colloidal manipulation, and bio-microrheology. Recently, they have been maneuvered in biological environments such as vitreous humour, dentinal tubules, peritoneal fluid, stromal matrix, and blood, which are promising developments for clinical applications. However, their biocompatibility and biodistribution are vital parameters that must be assessed before further use. An extensive quantitative evaluation has been performed for these parameters for the first time through in vitro and in vivo experiments. Investigations of cell death, proliferation, and DNA damage ascertain that the motors are non-toxic. Also, an unbiased transcriptomic analysis affirms that the motors are not genotoxic till 20 motors/ cell. Toxicity studies in mice reveal that the motors show no signs of toxicity up to a dose of 55 mg/ kg body weight. Further, the biodistribution studies show that they remain in the blood circulation after injection and at later stages possibly adhere to the walls of the blood vessel because of adsorption. However, perfusion with physiological saline decreases this adsorption/adhesion. Overall, we demonstrate the biocompatibility of nanomotors in live cellular and organismal systems, and a systemic biodistribution analysis reveals organ-specific retention of motors.

1. Introduction

Recent advances in the controlled motion of nanoparticles, manipulated through external energy sources (chemical [1,2], magnetic [3,4], acoustic [5] or biological [6,7]) for controlled navigation, referred to as “micro-nanomotors”, have opened up new avenues for biological applications [8–10]. From a biological perspective, targeted navigation will allow local bio-rheological measurements [11,12], payload delivery (such as drugs [13–15], genetic material [8,16,17]), and mechanical force application [18–20] in biological environments. Compared to conventional methodologies, these motors potentially allow access to hard-to-reach locations and cavities in the body with significantly better specificity and speed than traditional passively diffusing molecules or drug carriers.

Of these micro-nanomotors, helical magnetic nanomotors (NMs) [4,21,22] can be propelled by an externally applied rotating magnetic field, a scalable, non-invasive form of actuation with minimal effects on biological systems. Spatiotemporal manipulation and multifunctionality of NMs can be used for movement in blood [23], peritoneal cavity [3], bile duct [24], mucus [25], and vitreous humour [26], generating magnetic hyperthermia [27], active colloidal manipulation [19], manoeuvering inside living cells [28] and tissue-like microenvironments [29], and measurement of local viscosity [11,12], thereby making them ideally suitable for a wide variety of in vivo theragnostic applications.

However, before any new entity or formulation is used for biological or clinical applications, it is imperative that their toxicity and biodistribution profile is established, an essential requirement for nanoparticles. While effective in in vitro studies and PK/PD prediction based on molecular composition, many nanoparticles have shown no apparent bioactivity and even turned out to be lethal in in vivo studies [30]. This is due to the complexity of the physiological responses in a whole animal system compared to studies in isolated parts or cellular systems. Thus, an exhaustive evaluation of toxicity and biodistribution of any new formulation in both in vitro and in vivo animal models is crucial since these models are broadly representative of the corresponding systems in humans. In the case of any nanoparticle, these parameters are influenced by shape, size, material composition, and the route of administration [31–38]. In addition to these, the mode of actuation also affects the biocompatibility of micro-nanomotors. Chemically-powered motors can be toxic depending upon the fuel source. On the other hand, biologically-powered motors are affected by immunotoxicity due to foreign proteins/organisms present on them. The NMs used in this study are inorganic particles actuated by an external magnetic field, an energy source that is not detrimental to the biological environment. Here, we explore the in vitro and in vivo biocompatibility of these NMs, with due consideration to the factors mentioned above, to estimate a safe dosage for biological applications.

Traditionally, biodistribution has been estimated through inductively-coupled plasma mass spectrometry (ICP-MS) [39], fluorescence imaging [39,40] or positron emission tomography (PET) [41] for particles to be used in biological systems. At the same time, toxicity is assessed by analysing changes in various physiological parameters such as liver and kidney function and blood cell profiles. Among various classes of micro-nanomotors, the toxicity of chemically-powered ones moving in the GI tract has been performed in mice by histopathology and blood-biochemistry analysis [39,42]. In comparison, studies with magnetic nanomotors have been limited to in cellulo studies using MTT assay [23,43]. In a few studies, adhesion and migration of cells have been used to evaluate the cytotoxicity of magnetic helical micro-machines [44,45]. We believe this is a critical gap concerning practical applications of magnetic nanomotors, considering the recent break-throughs achieved with them as potential in vivo therapeutic agents [3,11,12,19,23–29].

Here for the first time, we perform the toxicological assessment of micro-nanomotors beyond cell-level viability measurement assays and study how magnetic NMs influence cell viability, cellular physiology, and affect gene expression. Our studies reveal the motors to be non-toxic to both the cellular phenotype and gene expression till doses of up to 20 motors/ cell. In this work, for the first time, we also establish that the NMs are non-toxic in mice up to concentrations of 55 mg or 8 × 10¹⁰ motors/ kg body weight using the Up-and-Down method according to Organisation for Economic Co-operation and Development (OECD) guidelines for testing chemicals. The biodistribution profile of the NMs is also encouraging as they are found to be in the circulating system immediately after injection and get reversibly adsorbed on the blood vessels after an extended duration. This gives us an estimate of the operational window for working with these motors under in vivo settings. The results of these tests using NMs reiterate the importance of exhaustive studies on toxicity and biodistribution of micro-nanomotors and establish the quantitative limits for their applicability as safe theragnostic vehicles.

2. Results and discussion

2.1. Synthesis and characterization of gold-coated helical magnetic nanomotors

Previous studies from our group have reported that NMs can be manoeuvered inside living cells, blood [23] dentinal tubules [46], and discriminate between cancer and non-cancer cells [29] in a co-culture system. These demonstrations make NMs an ideal candidate for further applications in biological systems. We used the NMs with similar compositions for further biological testing so that they could be developed further. Towards this, we first scaled up the synthesis of NMs, as large volumes are needed for biological applications.

As reported previously, helical structures made of silica (SiO₂) were synthesised using Glancing Angle Deposition (GLAD) technique (Fig. 1A) [4]. The structures had a total length of 2 μm with two helical turns, and iron was incorporated into the structure during fabrication to enable magnetic properties. The NMs were then conformally coated with silica using Plasma Enhanced Chemical Vapour Deposition (PECVD) to reduce aggregation and further sputter-coated with gold for multifunctionality. Gold is a biocompatible material that provides a signature in ICP-MS with minimal interference from any other material, thus helping to estimate biodistribution studies accurately. The coating also provides an additional platform for conjugating biomolecules using thiol chemistry to develop multifunctional nanomotors. Lastly, the gold coating also increases optical scattering, allowing improved visualization for bright-field microscopy. The different materials used in the fabrication of NMs and their roles are detailed in table S1. The motors were then magnetised perpendicular to their longitudinal axis using a Neodymium magnet.

Fig. 1.

A) Fabrication process of helical magnetic nanomotors (NMs). B) (i) SEM image of NMs. (Scale bar = 2 μm). (ii) TEM image of an NM with distributions of various elements ?silicon (Si), oxygen (O), silver (Ag), iron (Fe), titanium (Ti), and gold (Au). (Scale bar = 500 nm). C) Schematic image of (i) a triaxial Helmholtz coil setup and (ii) the mechanism of propulsion under a rotating magnetic field. D) Time-lapse images showing propulsion of an NM in deionised water at an actuating field strength of 20 G and frequency of 40 Hz. (Scale bar = 3 μm). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Scanning electron microscopy (SEM) images of the motors confirmed the large-scale uniformity of the fabrication process, a necessary factor for biological applications (Fig. 1B–i). The transmission electron microscopy (TEM) images reveal the structure of NMs, showing distinct distributions of silicon, oxygen, iron, silver, titanium and gold (Fig. 1B–ii). A triaxial Helmholtz coil was used to employ a rotating magnetic field (Fig. 1C–i) which applies a torque on the motors. This allows them to propel due to the rotational-translational coupling by their corkscrew structure (Fig. 1C–ii). The motion was recorded at a field strength of 20 G where the propulsion velocity depends upon the frequency at which the magnetic field rotates; at a frequency of 40 Hz, the velocity was 11 μm/s (Fig. 1D, Supplementary video 1). Further, we did not observe any precession motion, which implies that the magnetisation angle was close to the normal of the helices [47,48].

2.2. Nanomotors are cytocompatible in MDA-MB-231 and A549 cell lines

First, to evaluate long term cytocompatibility, the in vitro toxicity of NMs was probed, for which two different cell lines were used- the human epithelial breast cancer cell line MDA-MB-231 and the human epithelial lung cancer cell line A549. The effect of motors (used in two different amounts, 1:2 and 1:20 to cell numbers) on these cells was evaluated using live-cell imaging and quantified by monitoring changes in proliferation index, apoptosis, and gene expression analysis.

It is well established that toxicity may trigger apoptosis, a form of programmed cell death with characteristic changes in cell morphology. We used flow cytometry to quantitatively estimate the percentage of apoptotic cells five days after incubating the cells with NMs (Fig. 2A, fig. S1A). The cells were stained with FITC-conjugated Annexin V and propidium iodide (PI) to record the same. Annexin V binds to phosphatidylserine, a phospholipid present in the cell membrane. In live cells, phosphatidylserine is oriented towards the cytosol, which prevents the binding of the FITC-Annexin V. However, in apoptotic cells, phosphatidylserine flips and is displayed on the surface of the cell. This allows binding of FITC-Annexin V, resulting in green fluorescence; a larger number of FITC labelled cells correspond to higher cell death.

Fig. 2.

A) Quantitative analysis of MDA-MB-231 cell state by flow cytometry after five days of incubation with helical magnetic nanomotors (NMs). The graphs show the percentage of (i) normal, (ii) early apoptotic, (iii) late apoptotic, and (iv) necrotic cells based on staining with FITC-Annexin and PI. Data represented as mean ± SEM. (n = 3). B) Normalised proliferation index of the cells after five days of incubation with NMs. Data represented as mean ± SEM. (n = 4). The multiplicity-adjusted p-values are shown on the graphs. C) Fluorescence microscopy images of the cells after 48 h of incubation with NMs (number of cells: number. Of motors = 1: 20). The cell nuclei are stained cyan by Hoechst 33342, the live cells are stained green by Calcein-AM, and the dead cells are stained red by propidium iodide. (Scale bar = 200 μm). Puromycin was used as a control in these studies as it induces significant cell death, evident by high PI and low calcein staining. D) Fold change in the expression of genes (i) TP53, (ii) P21, (iii) PCNA, and (iv) CXCR4, 48 h after incubating the cells with NMs (no. of cells: no. of motors = 1: 20). The different colours represent the independent biological replicates. The bar diagram denotes the mean value of the replicates. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Unlike this, PI, a positively charged dye, does not permeate live cells. It can, however, enter dead cells as the membrane integrity is compromised, and it binds to DNA by intercalating between the bases, giving a strong red fluorescence. Overall, dual labelling with FITC-Annexin V and PI implies that normal cells will not show fluorescence in FITC and PI channels. Cells in the stage of early apoptosis show fluorescence in the FITC channel alone, while cells in the stage of late apoptosis show fluorescence in both FITC and PI channels. On the other hand, necrotic cells show fluorescence in the PI channel alone [49]. For A549 cells, there was no significant difference between the percentage of cells showing apoptosis between control and treatment groups (fig. S1A). In the case of the MDA-MB-231 cell line, a slight difference between the control cells and cells incubated with a high concentration of motors was seen (Fig. 2A). However, the magnitude of this difference is just around 1–2 %. Thus, we can infer that the NMs do not induce significant cell death even after long-term exposure.

Cellular toxicity in response to foreign particles may also manifest as a reduction in the proliferation of cells. Hence, we calculated the proliferation index or rate as an indicator of change in cell growth.

Proliferation index is the ratio of the total number of live cells (at a particular time point) to the number of cells at the initial time point and thus quantifies the rate of cell division in a population. For both the cell lines, the proliferation index of cells incubated with the NMs (for five days) was not significantly different from that of the control cells (Fig. 2B, fig. S1B) at both the high and low concentrations of motors. This shows that even after long term exposure, the NMs (at the given concentrations) do not affect cell proliferation.

Further, we employed a dual staining approach based on staining with fluorescent dyes Calcein-AM and PI using confocal microscopy to determine cell death qualitatively. Calcein-AM is converted by active esterases, present in the cytoplasm of live cells into Calcein, which gives strong green fluorescence. As mentioned before, PI permeates the cell membrane only in dead cells, giving a red fluorescence. The fluorescence images show strong cytocompatibility (high green and low red signal) for both the cell lines (Fig. 2C, fig. S1C) at both the concentrations of NMs after 48 h of incubation.

Based on these observations, we surmised that the motors have no long-term effects on cell proliferation. Given that any foreign material might show some signs of stress on cells, which cannot be captured by examining the differences in cell proliferation, we decided to evaluate changes in the expression of some genes, that occurs during cellular stress. The cells were incubated with a high concentration (1:20 ratio) of NMs for 72 h, after which quantitative gene expression analysis was performed by real-time PCR for recording changes in the expression of TP53, P21, PCNA, and a receptor, CXCR4. An increase in expression of TP53 and P21 and a reduction in expression of PCNA is indicative of the presence of cellular stress and damage to the DNA. TP53 is a tumor suppressor protein and facilitates the repair of the DNA damage [50–53]. P21 is a cyclin-dependent kinase inhibitor that regulates the cell cycle, DNA damage repair and the apoptosis pathway [54–57]. Proliferating cell nuclear antigen (PCNA) regulates DNA replication and DNA damage repair [58–60]. C-X-C chemokine receptor type 4 (CXCR4) is a protein expressed in the cell membrane that plays a critical role in cell adhesion and migration. Expression of CXCR4 is induced under conditions of high oxidative stress and DNA damage [61,62]. In Fig. 2D, a decrease in the expression of P21 is seen in MDA-MB-231 cells. In fig. S1D, a decrease in the expression of P21 and an increase in the expression of CXCR4 is seen in A549 cells. However, these effects, observed after 72 h of incubation, are not significantly reflected in the proliferation index and apoptosis assay performed after five days. The increased concentration of nanomotors in the vicinity of cells may elicit such a transient response due to the foreign body effect. This response did not translate to long-term cytotoxicity as the cells resolve the aforementioned effect after some time if the foreign body is inert or non-toxic.

2.3. Transcriptomics analysis show signs of phagocytosis but not cytotoxicity in the MDA-MB-231 cell line

We have examined cytocompatibility through different phenotypes such as the absence of cell death, normal cell proliferation, and lack of apoptosis. However, as mentioned above, the absence of effect on cellular proliferation does not mean the motors are inert to living cells. To record the unbiased effect of motors on cells, MDA-MB-231 cells were incubated with a high concentration of the NMs for 72 h. The cellular RNA was then isolated, and transcriptome analysis was performed on them by mRNA sequencing. A total of 1713 differentially expressed genes (DEGs) were significantly altered in cells treated with nanomotors compared to control cells (based on absolute log2 fold change ≥1 and p-value ≤0.05) (Fig. 3A, B). Among these, 884 genes were up-regulated, and 829 were down-regulated. The up-regulated gene with the lowest p-value was PTGS2 (prostaglandin-endoperoxide synthase 2), which encodes the COX-2 enzyme. This enzyme synthesises prostaglandins, a key inflammatory molecule that also stimulates cell proliferation and angiogenesis [63], showing the pro-inflammatory response of cells to the motors. The down-regulated gene with the lowest p-value was CGB8 (chorionic gonadotropin subunit beta 8). This gene encodes a glycoprotein hormone, human chorionic gonadotropin, shown to be involved in tumor initiation, growth, and metastatic outgrowth [64].

Fig. 3.

Differential expression profile of genes between control cells and cells treated with helical magnetic nanomotors (NMs) shown as A) Volcano plot, and B) Heat map. C) The top five up-regulated (red) and downregulated (green) genes from the KEGG pathway analysis- (i) phagosomal pathways; (ii) platelet activation; (iii) PI3K-Akt signalling; (iv) NF-κB signalling; (v) cytokine-cytokine receptor interaction; (vi) steroid biosynthesis, and (vii) Rap1 signalling. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis (Fig. 3C) on the transcriptome data revealed the phagosome pathway, which may indicate changes in the expression of proteins that mediate the uptake of NMs by the cells. We also observed significant changes in expression for pathways contributing to cellular injury or damage, such as PI3K-Akt signalling, NF-κB signalling, cytokine-cytokine receptor interaction, steroid biosynthesis, and Rap1 signalling pathways. However, the fold change in the genes involved in these pathways is minimal in the cells treated with NMs compared to the control cells. Table S2 lists the significant DEGs that influence the biological processes of cell proliferation, apoptosis and inflammation. Interestingly, we observe significant changes in expression for both pro and anti-inflammatory genes, implying that, as a whole, NMs do not elicit a considerable immune response. The fold changes may be due to the crowding of the cells by the high concentration of motors. Nanoparticles could cause meta-bolic or structural disturbances in the cells, which may lead to transient changes in pro and anti-inflammatory genes, which are known to be the primary responses to any external agent. However, the amount of inflammation and whether it sustains depends upon the particles’ dosage, size, shape, and surface chemistry [65]. Though we observed changes in inflammatory genes after 72 h in this work, these effects were either resolved internally or were not significant enough to cause cytotoxicity, as evidenced by the proliferation studies after five days. We, therefore, conclude that up to 20 nanomotors per cell show good biocompatibility.

2.4. Biodistribution analysis in Balb/c indicate adsorption of motors on walls of blood vessels

Once we established that NMs are quite inert and do not induce significant intracellular damage or stress, we examined the in vivo biocompatibility and biodistribution of intravenously injected NMs in Balb/c mice. Biodistribution analysis was performed by quantifying the amount of gold in various organs through ICP-MS, which, through back-calculation, would then yield the number of motors present in the sample/tissue. Initially, we injected the motors through the tail vein, the conventional strategy for intravenous administration of any drug/particle. However, we observed that due to the comparatively large dimensions of nanomotors and the increased chances of clump formations (due to the presence of magnetic material), the motors were getting lodged in the narrow tail vein itself (fig. S2), as 90 % of the injected NMs were detected at this location. Similar results have been reported for nanoparticles [66]. Thus, it was concluded that tail vein injection is not an ideal strategy to administer NMs intravenously in mice.

We then decided to inject motors in the abdominal aorta, one of the largest blood vessels in the body (Fig. 4A). The ease of performing the surgery required to access the abdominal aorta was the primary deciding factor in choosing this specific blood vessel (Fig. 4B). 10⁸ NMs in 200 μL of PBS was injected, and the biodistribution of motors in the mice, thus obtained after various time points, is shown in Fig. 4C. The sham group was injected with 200 μL of the nanomotor carrying medium, which in this case is PBS. We observed that 10 mins after injection of NMs, they get distributed across the body, as shown by the measured value (≈70 %) in the carcass of the mice. Among the organs, the motors are primarily present in the heart (≈14 %), abdominal aorta (≈10 %) and the rest of the motors (≈6 %) were detected in the liver, kidneys, intestine, and brain. The presence of NMs in highly vascularised tissues is anticipated, which our findings have captured. After 4 h of injection of NMs, their distribution alters, and only 40 % of them are localised in the carcass, while their presence in the abdominal aorta increases (≈35 %). The number of motors detected in the heart decreases (≈1 %), whereas the amount seen in the liver increases (≈7 %). The rest of the motors (≈15 %) are distributed among other organs, mainly kidneys, intestine, and brain. These observations indicate that 4 h after injection of nanomotors, their amount in circulation decreases as they are possibly getting adsorbed in the blood vessels or the narrow capillaries.

Fig. 4.

A) Experimental protocol for biodistribution studies. B) Location and schematic of the abdominal aorta. C) Biodistribution of helical magnetic nanomotors (NMs) 4 h after injection into the abdominal aorta. The analysis was performed through ICP-MS of various organs (n = 5). D) Biodistribution of NMs after perfusion of the mice with physiological saline, which was performed 4 h after injection of the motors (n = 2). The organs referred to as others on the left in panels C and D are expanded in the graph on the right.

We next investigated whether the motors could be dislodged from the location where they were adsorbed. For this, 4 h after injection of motors into the abdominal aorta, we perfused the mice with physiological saline, and ICP-MS was performed on the perfused fluid along with other organs (Fig. 4D). The motors did dislodge from their location in various organs, as evidenced by the change in the distribution of motors after perfusion. However, only a fraction (≈5 %) is removed from the body through the perfused fluid; a significant share of the injected motors (≈75 %) remained in the carcass. In the case of other highly vascularised organs like the liver and kidneys, the presence of NMs considerably decreases after perfusion. This implies that perfusion of mice with physiological saline removes the motors from the locations at which they are adsorbed, possibly by altering their environment.

Nanomotors do not show abnormalities in serum parameters in Balb/c mice but show few signs of tissue injury at 55 mg/kg.

Finally, we test the biocompatibility of NMs in Balb/c mice through injection into the abdominal aorta. We decided to conduct the testing in female mice as they are more sensitive to toxicity than male mice [67] Here, we use the Up-and-Down method according to OECD guidelines for testing chemicals [68,69], a methodology involving fewer animals. The material composition of various concentrations of NMs used in this study (corresponding to the up-and-down sequence) is listed in table S3. Limitations in fabrication capped the maximum possible dose at 55 mg/kg, corresponding to 1.6 × 10⁹ NMs/animal. Though up-and-down methods define monitoring of mice for 48 h, we observed them up to seven days after injection of motors for any sign of delayed toxicity. The mice were sacrificed after seven days, and histopathology of various organs was performed (Fig. 5). After seven days, the complete blood profile (table S4) and serum parameters (Table 1) of the animals were also examined for signs of toxicity. For serology assessment, the levels of the enzymes alanine transaminase (ALT), aspartate transaminase (AST), alkaline phosphatase (ALP), and creatinine were examined as indicators of damage to the liver and kidney.

Fig. 5. Hematoxylin and Eosin stained sections of various organs of Balb/c mice, seven days after injection of indicated concentrations of helical magnetic nanomotors into abdominal aorta.

(i) Brain (black arrow indicates the pyramidal neurons in the grey matter, blue arrow indicates the glial cells in the white matter); (ii) Heart (black arrow indicates myocardial fibres, blue arrow indicates loss of cellular and nuclear detail); (iii) Kidney (black arrow indicates glomerular tuft, blue arrow indicates hemolysis); (iv) Liver (black arrow indicates central vein, blue arrow indicates hepatocytes); (v) Lung (black arrow indicates alveoli, blue arrow indicates blood vessel), and (vi) Spleen (black arrow indicates red pulp, white arrow indicates white pulp). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Table 1. Serum parameters seven days after injection of nanomotors into the abdominal aorta of Balb/c mice.

	Control	Sham	1.75 (mg/kg) 0.5 × 108 NM	5.5 (mg/kg) 1.6 × 108 NM	17.5 (mg/kg) 5 × 108 NM	55 (mg/kg) 16 × 108 NM	Reference range
ALT (U/L)	47.3, 60.4	1.6, 46.1	27.1, 31.2	32, 37	35.4, 37.6	32.6, 36.3	17, 77
AST (U/L)	175, 217	28, 271	126, 145	96.9, 107	186, 313	154, 232	54, 298
ALP (U/L)	11.2, 282.7	7.4, 19	110, 177	180, 220	179	97.8, 124.6	35, 96
Creatinine (mg/dL)	0.1, 0.3	0.07, 0.1	0.4	0.3, 0.4	0.3	0.4	0.2, 0.9

While no mortality was observed in any of the test groups, a reduction in the weight of mice was recorded at 48 h for all animals (possibly due to surgical procedure), which restored to normal levels after seven days (fig. S3). The mouse injected with the highest dose of 55 mg/kg showed elevations in the complete and differential leukocyte count (table S4) seven days post-injection of motors. This pointed towards an immune response triggered by the large number of NMs; however, there was no change in the serum parameters (Table 1). H&E stained sections of the brain, liver, lung, and spleen show typical architecture in most animals. Mouse injected with 55 mg/kg showed a loss of cellular and nuclear detail in the histopathological section of the heart with the loss of cross striations of the myocardial fibres. The kidney sections of the mouse also showed slight haemolysis when compared to the control group. When nanoparticles are injected into the bloodstream, plasma proteins instantly adsorb to their surface, creating a protein corona, which affects their interactions with various components of the blood and the endothelial cells. This interaction may have deleterious effects such as haemolysis which is determined by the particles’ dosage, geometry, porosity, and surface chemistry [70] and can be local where the high concentrations of the propellers are achieved; hence these effects were observed in the kidney only. Further, histological alterations of the cardiac tissue at a dosage of 55 mg/kg may result from the oxidative stress induced by the interaction of a large number of NMs with the membrane structure [71]. These effects were recorded to be non-lethal in the experimental duration, which made us conclude that the NMs are biocompatible in Balb/c mice up to 55 mg/kg body weight concentration.

We can infer from this part of the study that NMs are biocompatible in Balb/c mice up to 55 mg/kg body weight concentration which is equal to 1.6 × 10⁹ NMs/ animal. However, there are indications of tissue injury induced by the motors through activation of the immune response, obstruction in the blood vessels leading to slight damage in the heart, and minor hemolysis in the kidneys. These indications do not lead to significant damage to the tissues in seven days, as evidenced by the normal serum parameters.

3. Conclusions

Given that any change in the size, shape and composition of nanoparticles influences their biocompatibility and biodistribution, it is essential that the helical magnetic nanomotors (NMs) are assessed for these parameters before deploying them in biological systems. The standard cellular compatibility assays rely on cell survival quantification and in vivo compatibility assessment on animal survival monitoring. We have extended these standard approaches by evaluating a number of other biological parameters hitherto reported in nano-bio studies. For cellular studies, we used live-cell imaging, flow cytometry and biased/unbiased gene expression analysis, while for animal studies, we used multiple doses based on the Up and Down method alongside temporal toxicity profiling, in a detailed manner that has not been performed previously in a single study.

We investigated the in vitro biocompatibility of nanomotors (both low and high concentrations) on MDA-MB-231 and A549 cell lines. The cytocompatibility of motors was confirmed using live-cell imaging, estimation of proliferation index, apoptosis assay, and unbiased gene expression analysis. Our findings revealed that while no effect is seen on the gross cellular viability, using biased and unbiased gene expression analysis, we recorded that cells exposed to NMs show signs of activation of stress pathways as well as pathways linked to phagocytosis and cytokine signalling.

For in vivo toxicity studies, we injected NMs into the abdominal aorta of Balb/c mice, and the motors were biocompatible in the mice till 55 mg/kg body weight. Minor signs of tissue injury at the maximum dose of 55 mg/kg were recorded, which resolved by day 7 of the experimentation. Thus, it is advisable to use concentrations of NMs below this limit, which translates to 8 × 10¹⁰ NMs / kg body weight, which acts as a base for clinical use of the motors. This value corresponds to 4.8 × 10¹² NMs for an adult human (having a bodyweight of 60 kg). The hafnium oxide nanoparticles (NBTXR3) that were clinically approved for intratumoral administration was used at a dosage of 7.35 g for an adult human [72]. In comparison, we predict the magnetic NMs to be safe for intravenous administration up to a dosage of 3.3 g for an adult human.

Another essential question addressed in the study was that of the pharmacokinetic profile of the motors, which we examined through biodistribution studies. While the standard approach of tail vein injection in the mice was found to be limiting in our approach, perhaps due to the shape/geometry of the particles, we were surprised to find that the observations made by us were similar to what is reported in the literature for other nanoparticles as well. In a majority of studies, while maximal recovery of injected particles was reported from the liver and spleen, similar to what we recorded, the total recoverable particles were typically around 10 % of what was injected. This could be due to possible retention of the particles in the vasculature, which can only be captured if the entire carcass was examined, but this aspect was not highlighted by any study to the best of our knowledge.

These findings forced us to explore alternate route/s of injection, especially those which can ensure better distribution, and for this, the motors were injected into the abdominal aorta of Balb/c mice. The abdominal aorta was selected as it is one of the larger arteries, thereby ensuring rapid dispersal of the motors; also, it was highly amenable to surgical procedures. We used inductively - coupled plasma mass spectrometry (ICP-MS) of the organs using the gold present on the NMs as a measure to quantify the absolute number of motors present in the specific organ. Given the sensitivity of ICP-MS, it is the technique of choice for such studies compared to other potential approaches like fluorescence imaging and positron emission tomography (PET). Typically, biodistribution is performed at a single time-point post-injection for nano-bio studies; keeping in mind the potential changes in the effect and distribution of the particles over time, we examined both the parameters at different time intervals post-injection. Our findings revealed the obvious, but sparsely reported aspect that in 10 mins after injection, the motors spread all over the body, with considerable presence in sites of a large number of capillaries. However, after 4 h, the number of motors in circulation decreases as potentially the NMs are adhering to walls of the blood vessels or getting adsorbed in the narrow capillaries. Therefore, we have a potential window of 10 mins to 4 h that allows the motors to reach the concerned organ. This aspect of biodistribution has not been examined in such detail before and highlights one of the novel findings of this study. On similar lines, temporally resolved effects of NMs were also evaluated in mice and detailed histopathological, blood cell profiling, and serological assessments revealed that in the short term, the NMs show no signs of toxicity. However, certain tissues show signs of injury over a longer duration, even though no signs in injury based on serological profiling was recorded.

Overall, we have undertaken a comprehensive study on the toxicity and biodistribution of nanomotors, not addressed before in such detail. These studies bring helical magnetic nanomotors closer to becoming a true “fantastic voyager”.

Materials and methods, list of significant differentially expressed genes, cytocompatibility study on A549 cell line, additional figures and tables on in vivo toxicity and biodistribution. Video showing propulsion of a nanomotor in deionised water. Supplementary data to this article can be found online at https://doi.org/10.1016/j.bioadv.2022.213048.

Data Availability

Data will be made available on request.