Synthesis and characterization of FMSPs
Herein, homemade Fe3O4·Rhodamine [email protected] SPs (FMSPs) used in this work are fabricated by coating preassembled Fe3O4·Rhodamine6G SPs with PDA. HRTEM images show that FMSPs possess an uniform size with the core size of 50 nm and the PDA shell of about 10 nm (Fig. 2a, b). The PDA shell can improve the biocompatibility, physiological stability, and colloidal stability of FMSPs. The hydrated diameter measured through DLS is 93.6 nm (Fig. 2c), which is larger than that measured in TEM images, but still within the optimal range for uptake by non-phagocytic cells [32–34]. The polydispersity index (PDI) measured by DLS is 0.186, implying the excellent dispersibility of our homemade FMSPs. Zeta potential measurement of FMSPs suspended in cell culture media reveals a weak negative charge of -12.7 ± 2.3 mV, which is more favorable for their interactions with negatively charged cell membrane and subsequent uptake . The magnetic curves of FMSPs are characterized by SQUID, and the saturation magnetization is 62 Am2/kg without any evident remanence or coercivity at 300 K, suggesting the superparamagnetic property of FMSPs (Fig. 2d). Figure 2e exhibits the high magnetic responsibility of FMSPs toward external magnetic field, which is the key parameter for maximizing the force transfer efficiency from the external magnetic field to the FMSPs-loaded cells. The optical properties of FMSPs are investigated as well. As shown in Fig. 2g, FMSPs possess a green color emission at around 556 nm under 350 nm excitation. The photograph shown in Fig. 2f further indicates their bright green fluorescence excited by the hand-hold UV lamp. PC12 cells after incubation with FMSPs exhibit a strong green fluorescence, which demonstrates that the FMSPs have an ideal capacity to label mammalian neural cells (Fig. 2h).
Minimal cytotoxicity is essential for any biomedical application. The cytotoxicity of FMSPs is firstly evaluated by standard CCK8 assay. After incubation with FMSPs for 24 h, the cell viability of PC12 cells remain above 80% at concentration range from 20 to 200 µg/ml (Fig. 3a). Measurements of cell viability after 48 and 72 h in culture demonstrate the similar trend, implying that the uptake of FMSPs do not affect cell viability and replication rate. The Additional file 2: Movie S1 show that PC12 cells are able to maintain active mitotic proliferation and neurites outgrowth after endocytosis of FMSPs. The amount of FMNPs uptake by PC12 cells is calculated by ICP-AES. After incubation with FMSPs (10 µg/ml), the cells are able to incorporate quantities of Fe up to 6.90 ± 0.07 pg/cell, corresponding to ~ 3.73 ± 0.036 × 104 FMSPs/cell. These data demonstrate that using 10 µg/ml as the working concentration not only guarantees extremely low cytotoxicity, but also ensures that sufficient FMSPs can be taken up by cells. The low cytotoxicity of Fe3O4 is one of its great advantages comparing with other MNPs [8, 22, 35, 36]. Our results are consistent with these literatures and manifest that PDA coated superparamagnetic Fe3O4·Rhodamine 6G SPs are basically noncytotoxic even under the high concentration of 200 µg/ml. Additionally, effects of FMSPs treatment on the normal differentiation capacity of DRG neuronal cells are examined as well. As a result, there is not any significant change in viability and differentiation capacity of DRG neurons can be observed even after incubation with FMSPs (10 µg/ml) for up to 72 h. DRG neurons treated by FMSPs retain their differentiation property, leading to the neurite outgrowth and complex neuronal networks formation (Fig. 3b and Additional file 3: Movie S2).
The idea of using mechanical tension mediated by MNPs to stretch the growth and elongation of axons provides a new approach for repairing neurological diseases and injuries. To this end, design and preparation of ideal MNPs appears to be a challenge to overcome. Our results suggest that FMSPs are promising candidates as biocompatible magnetic agents for magnetically-driven cell actuation.
Cellular behaviors of FMSPs internalization
The uptake behavior of FMSPs by cells is further studied. CLSM images of cells at a single focal plane exhibit the high fluorescence within the cells, verifying the internalization of FMSPs into the cells (Fig. S1 in Additional file 1: Supplementary Information). For assessing the extent of FMSPs internalization, the intracellular fluorescence intensity is measured by fluorescence microscopy and flow cytometry. As shown in Fig. S2 in Additional file 1: Supplementary Information, the cellular uptake of FMSPs depends on the FMSPs concentration in the medium. With increasing the concentration of FMSPs in the culture environment, the fluorescence intensity in cells and the intracellular amount of FMSPs increase. PC12 cells have the great uptake amount and high fluorescence levels even at low concentrations of FMSPs (10 µg/ml). Subsequently, the influence of incubation time on the uptake of FMSPs by PC12 cells is measured by fixing the FMSP concentration (10 µg/ml) but prolong the incubation duration from 1 to 6 h. As a result, the fluorescence intensity increases rapidly within 1 h and reach a plateau after 2 h incubation (Fig. S3 in Additional file 1: Supplementary Information). According to the FACS analysis in Fig. S3f and S3g, it can be seen that the uptake of FMSPs within 1 h is extremely high, followed by a significant deceleration between 1 and 2 h, then reaches saturation after 2 h.
Previous studies reported that the endocytosis process participated in NPs cellular uptake is an energy-dependent active transport [37–42]. To better evaluate the uptake kinetics, the cellular uptake of FMSPs under low temperature (4 °C) and ATP-depleted environments (in the presence of 2-DG) are analyzed. Fluorescent images indicate that the low temperature and ATP-depletion can inhibit the cellular uptake of FMSPs (Fig. S4a-d in Additional file 1: Supplementary Information). As shown in Fig. S4e and S4f, comparing with positive control group at 37 °C, FMSPs uptake by PC12 cells under low temperature or ATP-depletion environment reduced by 91.8 ± 0.3% (P < 0.05) and 65.6 ± 2.8% (P < 0.05) respectively. This result reveals the energy-dependent characteristic of FMSPs uptake.
Since the endocytosis of NPs in non-phagocytic cells generally include caveolae-mediated, clathrin-mediated, and micropinocytosis [43, 44]. To elucidate the specific endocytic mechanism of FMSPs uptake by cells exactly, different endocytosis inhibitors are employed to identify the endocytic pathways involved in the uptake of FMSPs (Fig. 4a and 4b). MβCD and genistein are firstly utilized to inhibit the caveolae-mediated endocytosis . It reveals a significant decrease in FMSPs uptake with the inhibition rates of approximately 37.9 ± 4.3% (P < 0.05) and 46.1 ± 3.6% (P < 0.05) in PC12 cells. This result suggests that the caveolae-mediated endocytosis is one pathway for FMSPs uptake. Then, chlorpromazine is employed to inhibit the clathrin-mediated endocytosis . As result, no obvious inhibition of FMSPs uptake is found, indicating clathrin-mediated endocytosis do not play an important role in the uptake of FMSPs (data is not shown). At last, amiloride and Cyto D are used to suppress the micropinocytosis , and the inhibition rates of approximately 45.5 ± 3.6% (P < 0.05) and 49.6 ± 0.6% (P < 0.05) are observed, indicating that micropinocytosis is a primary pathway for FMSPs uptake. To look in more detail, the trafficking of FMSPs into the cells is monitored by TEM analysis. Figure 4c shows the dispersed FMSPs in the cell membrane appearing to enter the cells by caveolae-mediated endocytosis. Figure 4d shows the cluster of FMSPs appearing to enter the cells by micropinocytosis. Neither FMSPs in nuclei nor damages at the cytoplasmatic organelles are found.
CLSM is further performed to identify the biodistribution of FMSPs within nerve cells. The FMSPs intracellular distribution is evaluated via co-incubation followed by co-staining with various organelle-specific fluorescent probes (LysoTracker, ER-Tracker, Golgi-RFP, and MitoTracker) (Fig. 5a-d). Figure 5a and 5d show that the internalized FMSPs are primarily distributed in Golgi apparatus and mitochondria of PC12 cells. Accordingly, the Pearson correlation coefficient (Rr) are 0.656 ± 0.067 and 0.624 ± 0.026 respectively, suggesting the good colocalization between FMSPs and these subcellular structures. Most importantly, our FMSPs can be observed not only in cytoplasm but also in the cone growth of developing neurites. With the help of CLSM, tomographic scanning and time-lapsed imaging are performed to clearly distinguish the intracellular distribution of FMSPs. Additional file 4: Movie S3 shows that FMSPs are transported bidirectionally within neurites, which can be detected as puncta in the growth cones, neurites, and cell bodies.
To study the exocytosis of internalized FMSPs, cells are firstly pre-incubated with FMSPs for 4 h, then washed with PBS and incubated with fresh culture media for 2, 4, 8, 12 and 24 h at 37 °C. Figure S5 in Additional file 1: Supplementary Information shows the significant decrease in intracellular FMSPs fluorescence intensity with increasing incubation time. But cells still maintain a high survival rate in the whole test period. The effective exocytosis of FMSPs further demonstrates the low cytotoxicity and high biocompatibility of FMSPs.
A comprehensive study of the neural cell behaviors on FMSPs uptake is highly indispensable for assessing the biological features and subsequent biomedical applications of FMSPs. The findings derived from this study not only provide some details on neural cell behaviors for FMSPs uptake, but also set the foundation for using the mechanical force mediated by FMSPs to guide the regeneration of axons.
Magnetic mechanical forces induce axonal outgrowth
To prove the concept that FMSPs can be exploited to manipulate the growth and development of neurites/axons under external magnetic fields. Experiments are carried out inside a constant magnetic flux density gradient generated by one perpetual cuboid neodymium magnet, which provide about 6.0 T/m of magnetic field to the cells at the center of the dish along the direction of the magnetic field gradient (Fig. S6 in Additional file 1: Supplementary Information). The homemade FMSPs used in this work have an average Fe3O4 core diameter of 50 nm (the volume is about 6.54 × 104 nm3) and a saturation magnetization of 62 Am2/kg. Since the density (ρ) of Fe3O4 is 5.17 g/cm3, a single FMSP subjected to a force FFMSP is calculated to be ~ 1.15 × 10− 4 pN (Eq. (2)). The number of FMNPs up-taken by PC12 cells is estimated to be ~ 3.73 × 104 FMSPs per cell. Thus, the average magnetic force on single cell Fcell is calculated to be ~ 4.29 ± 0.042 pN (Eq. (3)).
PC12 cells loaded with FMSPs are used to examine the effect of magnetic forces on the growth of neurite under the external magnetic field. The inclination angles θ between the long axis of the neurites and the line drawn parallel to the magnetic field are measured (Fig. 6a). Neurite orientation is quantified by introducing the concept of orientation index (Oi). Figure 6a and Additional file 5: Movie S4 show that the neurites of PC12 cells treated with FMSPs (FMSPs+, M+) tend to be arranged in parallel with one another and grow preferentially along the direction of the magnetic force when the magnetic field is applied. In contrast, the neurite growth directions for the control neurons appear to be random with no preferable direction in the absence of magnetic stimulation. Furthermore, experimental evidences demonstrate that neither the FMSPs nor the magnetic field alone can influence the neurite growth direction. The value of Oi in the blank control group (FMSPs−, M−) is -0.032 (-0.571∼0.604), which is not statistically significant different from that obtained when the magnetic field is applied (FMSPs−, M+; Oi =-0.027, P = 0.493) or cells are treated with FMSPs (FMSPs+, M−; Oi =-0.076, P = 0.580) alone (Fig. 6b). In contrast, Oi in the treatment group (FMSPs+, M+) is 0.726 (0.171∼0.936), which is much higher than that in other groups (P < 0.05) (Fig. 6b). Concerning the length distribution of neurites along the magnetic force direction, a trend toward statistical significance (P < 0.05) with the higher value of the neurites length for cells treated with both the FMSPs and the magnetic field (43.115 µm, 26.370∼65.817 µm) compared to the control group (36.110 µm in FMSPs control group, 21.885 µm in magnetic field control group and 28.289 µm in blank control group, P < 0.05) is observed (Fig. 6c).
To determine effects of nanomagnetic force stimulation on the growth cone motility and axon elongation rate, primary DRG neurons are cultured inside microfluidic chambers. In the treatment group (FMSPs+, M+), neurons are dissociated and incubated in medium enriched by FMSPs under an external magnetic field with a magnetic gradient. Figure 7a and Additional file 6: Movie S5 show the magnetic force acting on the FMSPs-bound axons of DRG neurons allows growth cone to grow rapidly and directionally toward the distal-axon (DA) compartment (magnetic source) within a short time by crossing the microchannels. The growth of DRG neurons in a microfluidic chamber are monitored for more than 24 h by time-lapsed imaging, and the average elongation rate of the axons are 33.4 ± 8.6 µm/h throughout the observation period. In the control groups (FMSPs−, M−; FMSPs +, M− and FMSPs−, M+) without nanomagnetic force stimulation, axons emerge out of the neuron cell bodies spontaneously and develop towards all directions without any preferred orientation (Fig. 7b-d and Additional file 7–9: Movie S6-S8). At the same time, the rates of axon growth decrease significantly in these control groups (11.1 ± 3.8 µm/h in blank control group, P < 0.05; 10.9 ± 4.2 µm/h in FMSPs control group, P < 0.05 and 9.7 ± 4.2 µm/h in magnetic field control group, P < 0.05) (Fig. 7e). Thus, it is reasonable to believe that synergic combination of FMSPs and external magnetic field can improve the motility of growth cone aligning the direction of the magnetic force and accelerate the outgrowth of axons.
The axon elongation is the result of axonal framework stretches caused by the tension from growth cone under external chemical or physical stimulations . More recently, mechanical forces in the range of several pN have been reported to have the ability to influence the outgrowth process with the help of a MNPs-based technique [6, 7, 10, 11, 48]. In our system, we demonstrate that internalized FMSPs in neural cells may generate a ~ 4.29 ± 0.042 pN tension force to physically direct the neurite orientated outgrowth and the axonal elongation. Based on previous works and our findings, it is believed that upon precise control the magnitude of mechanical forces mediated by FMSPs and the duration under forces, manipulating the development of axonal is possible.
Gene expression mediated by magnetic mechanical forces
To further determine the effects of mechanical signals mediated by FMSPs on gene expression profile in neural cells, mRNA transcriptome sequencing and bioinformatics analysis on cell samples from different experimental groups are performed. The mechanism of magnetic mechanical force on promoting axonal regeneration from the perspective of genetics is further expected to elucidate. Since only the cooperation of FMSPs and magnetic fields can affect the regeneration of axons, cells from the treatment group (FMSPs+, M+) and the blank control group (FMSPs−, M−) are selected for gene sequencing. As presented in Fig. 8a, volcano plot is used to assess gene expression variation between treatment and blank control group. After the pairwise comparison, a total of 89 mRNAs display differential expression including 43 up-regulated mRNAs. The level of expression changes for differentially expressed mRNAs is exhibited in heatmap (Fig. 8b). The GO enrichment analysis is utilized to annotate genes and analyze the biological processes of genes. From Fig. 8c it can be seen that a total of 43 up-regulated mRNAs are mainly involved in regulation of chemokine secretion process (GO: 0090197, 0090196, 0090195), central nervous system neuron differentiation and axonogenesis process (GO: 0021953, 0021952, 0021955), as well as regulation of mitotic cell cycle phase transition process (GO: 1900087, 1901990, 1902808, 1901987). Meanwhile, three target genes including Csf1r, Cdh11 and Ppp1r1c are identified and screened among the biological processes highly-correlated axon growth from up-regulated GO terms. The three identified differentially expressed mRNAs are further validated with reverse transcription-quantitative real-time PCR (RT-qPCR) analysis in PC12 cells. The expression level of Cdh11, Csf1r and Ppp1r1c are significantly up-regulated in treatment group compared with the blank control group (P < 0.01) (Fig. 9a-c). Furthermore, the lysates from samples are conducted the western blot. As shown in Fig. 9d-f, the protein level of Cdh11, Csf1r, and Ppp1r1c are remarkably increased in treated samples compared with the blank control sample (*P < 0.05, **P < 0.01). The findings further verify the accuracy of sequencing data and in line with the bioinformatics results.
Cdh11 gene encodes a type II classical cadherin (Cadherin-11) from the cadherin superfamily, integral membrane proteins that mediate calcium-dependent cell-cell adhesion. The expression of Cdh11 is associated in growth cones movement in the period of axonal migration . Cadherin-11 is an axon growth-promoting factor, involved in growing motor and sensory axons in the mouse embryo in a previous study [50, 51]. In the regulation of axon growth, Cadherin-11 play an important role in adhesive interactions occurring between growing axons. Cadherin-11 promote the fasciculation course of motor axon bundles  and the construction of neural networks  by regulating interaction of growth cones for neighboring axons. The up-regulated expression of Cdh11 gene and the increased translation of cadherin-11 protein in the treatment group are consistent with the observations of rapid and targeted growth of axons, which suggests that mechanical forces mediated by FMSPs can promote Cdh11 gene expression.
Csf1r gene encodes the receptor for colony stimulating factor 1. The colony stimulating factor 1 (Csf1) is a cytokine which controls the production, differentiation, and function of macrophages. This receptor mediates most of the biological effects of this cytokine. After a peripheral nerve lesion, the neuron undergoes a number of degenerative processes, the so-called Wallerian degeneration, followed by attempts at regeneration. Breakdown of the axon distal to the site of injury is initiated 48 to 96 h after transection. Deterioration of myelin begins, and the axon becomes disorganized. Macrophages, as immune cells, gathered around the damaged areas can phagocytose myelin and axonal debris  which is conducive to decrease inflammatory immune response[54–56], as well as promoted axonal regeneration by releasing a large number of regeneration related factors, including extracellular matrix proteins, growth factors, cytokines and chemokines[31, 57–59]. Our results indicate that the combination of FMSPs and magnetic field can improve the biological function of Csf1 cytokines by up-regulating the expression of Csf1r gene, and in turn contribute to axonal regeneration through Csf1 dependent macrophages activation .
Ppp1r1c gene encodes protein phosphatase 1 (PP1) regulatory inhibitor subunit 1C, belongs to a group of PP1 inhibitory subunits that are themselves regulated by phosphorylation. PP1 is a major serine/threonine phosphatase that regulates a variety of cellular functions and play a critical role in the regulation of neuronal morphology. PP1 is proved to promote axonal growth by dephosphorylation of the microtubule-associated proteins . However, Ppp1r1c as an up-regulated gene, contribute to the development of axon growth in the current study. The possible reason of this seemingly contradictory effect may be heterogeneity of samples between different experiments. There are few Ppp1r1c studies associated with axon growth. Thus, the potential molecular mechanism of Ppp1r1c for axon growth needs more evidences.