Electromagnetic Field‐Programmed Magnetic Vortex Nanodelivery System for Efficacious Cancer Therapy

Abstract Effective delivery of anticancer drugs into the nucleus for pharmacological action is impeded by a series of intratumoral transport barriers. Despite the significant potential of magnetic nanovehicles in electromagnetic field (EF)‐activated drug delivery, modularizing a tandem magnetoresponsive activity in a one‐nanoparticle system to meet different requirements at both tissue and cellular levels remain highly challenging. Herein, a strategy is described by employing sequential EF frequencies in inducing a succession of magnetoresponses in the magnetic nanovehicles that aims to realize cascaded tissue penetration and nuclear accumulation. This nanovehicle features ferrimagnetic vortex‐domain iron oxide nanorings coated with a thermo‐responsive polyethylenimine copolymer (PI/FVIOs). It is shown that the programmed cascading of low frequency (Lf)‐EF‐induced magnetophoresis and medium frequency (Mf)‐EF‐stimulated magneto‐thermia can steer the Doxorubicin (DOX)‐PI/FVIOs to the deep tissue and subsequently trigger intracellular burst release of DOX for successful nuclear entry. By programming the order of different EF frequencies, it is demonstrated that first‐stage Lf‐EF and subsequent Mf‐EF operation enables DOX‐PI/FVIOs to effectively deliver 86.2% drug into the nucleus in vivo. This nanodelivery system empowers potent antitumoral activity in various models of intractable tumors, including DOX‐resistant MCF‐7 breast cancer cells, triple‐negative MDA‐MB‐231 breast cancer cells, and BxPC‐3 pancreatic cancer cells with poor permeability.


Section S1. Calculating the diffusion coefficient of PI/FVIOs FITC
Labeling of PI/FVIOs with FITC. PI/FVIOs were dispersed in 1 mL DI water, and the pH was adjusted to 8.0 using sodium bicarbonate. FITC-NHS dissolved in PBS (3 mg/mL) was added to the above prepared solution. After shaking for 3 h, the solution was centrifuged at 6000 rpm for 10 min, and it was then washed with DI water several times until no fluorescence was detected in the supernatant. The as-produced PI/FVIOs FITC were then re-dispersed in PBS buffer.
Calculating the diffusion coefficient of PI/FVIOs FITC . PI/FVIOs FITC diffused from the outside of the tumor spheroid to the inside, which results in a lower grey value near the geometric centre of the tumor spheroid in the grayscale image. The efficient diffusion of PI/FVIOs FITC throughout the tumor spheroid can be evaluated using the following factors: (1) The grey value of pixels increases near the centre of the tumor spheroid, and (2) there is a reduced distance between the pixel with the minimum grey value and the centre of the tumor spheroid.
The parameter P is defined to analyze the distribution of PI/FVIOs FITC in a tumor spheroid: where, is the grey value of the point pixel in an image, is the Euclidean distance from the pixel to the geometric centre of the tumor spheroid, and N is the number of pixels in the tumor spheroid.
After normalizing both and , the modified formula can be presented as follows: where, the value is the grey value in the range [0,1], and the value is the distance in the The parameter r is proposed to evaluate the diffusion capacity of PI/ FVIOs FITC in the tumor   spheroid under Lf-EF, and it can be determined using the equation presented as follows: where, and represent the distribution of PI/FVIOs FITC in the tumor spheroid before and after Lf-EF treatment, respectively. Theoretically, the r value is expected to be less than 1, and the smaller the r value, the higher the diffusion capacity of PI/FVIOs FITC .

Section S2. Numerical modeling of the penetration by PI/FVIOs
A three-dimensional channel filled with ellipsoidal cells was numerically modeled (as shown in Figure S5a).
in which is the translational velocity vector of the nanomaterial i with mass . When the nanomaterials hit the cells, the contact force between them can be calculated as: The distance between a FVIO and a cell, , was determined to calculate the repulsive contact force, , for , and the cohesive contact force, , for . The symbol represents the contact stiffness between a FVIO and a cell, and is the unit vector of contact normal direction. The contact force vanishes when the distance is greater than a critical value , which indicates the detachment of the nanomaterial from the cell. In the present simulations, the magnitudes of and were set to twice that of the potential force .
The penetrations by the FVIOs through the cell-filled channel at various EF frequencies were simulated. The time evolution of the average displacements of nanomaterials is shown in Figure S5b, in which the average displacement was scaled by the nanomaterial diameter , and the time was scaled by the characteristic time , i.e., the time required for a FVIO to travel by a channel distance subjected to a potential force (with no contact force and no random force ). It is evident that FVIOs could barely move in a static magnetic field and they moved by a small displacement at low EF frequencies, i.e., = 0.03 and 0.06 kHz. The transportation of FVIOs was significantly improved by increasing the EF frequencies to 0.1 and 0.3 kHz. However, as the EF frequency was raised to 0.5 kHz, the movement of FVIOs was significantly reduced to the level of that in the static magnetic field. The difference in the transportation of nanomaterials at various EF frequencies could also be observed in the snapshots at time t/ = 1400 (as shown in Figure S5c). The present simulation results are in good agreement with the experimental observation. Thus, the random forces induced by the EF can disturb the motion of FVIOs, which increases the thermal dynamics and reduces the chances for FVIOs to be trapped on the rough and sticky surfaces of the cells. A very lowfrequency EF is close to a static magnetic field and too high a frequency can lead to the leveling-out of the disturbance by averaging over a short period of time. As a result, the best performance is achieved at an optimal EF frequency.

Section S3. Analyzing the distribution of PI/FVIOs FITC fluorescence signal in solid tumor
Lf-EF displays a gradient field distribution. To investigate the tumor penetration ability of PI/FVIOs under the influence of Lf-EF, we sliced the tumors along three mutually perpendicular axes (x, y, and z axes), and then we observed the diffusion and distribution of PI/FVIOs FITC in the tumor using a fluorescence microscope. The direction parallel to the central axis of the coil was denoted as the z-axis direction of the tumor (x-y plane); the other two directions were the x-axis (y-z plane) and the y-axis (x-z plane). The fluorescence intensity of PI/FVIOs FITC after Lf-EF actuation at the tumor x-y plane remained to be over 48%, and it was also kept at about 30% at the y-z plane and x-z plane. In contrast, the fluorescence intensity at any plane of the control group was only around 12-17%. The results show that after Lf-EF exposure, the diffusion of PI/FVIOs FITC into the tumor was significantly enhanced along the z-axis of the tumor. In the other two directions, PI/FVIOs FITC showed a limited diffusion enhancement as compared to that in the z-axis direction. However, it is worth noting that their diffusion abilities (along x and y axes) were still higher than those of the group without Lf-EF exposure. Such a result is mainly due to the larger gradient of Lf-EF along the z-axis direction (x-y plane).

Section S4. Measurement of SAR values
Quantitation of heat generation by the samples was performed using an induction heating system (M5, Xi'an SuperMag Nano-biotechnology Co. Ltd). 1 mL sample was exposed to Mf-EF (300 Oe) at various magnetic field frequencies. The temperature rise profiles were then recorded using an optical fiber thermocouple. The SAR was determined using the following equation: where, C is the specific heat of the medium (C water = 4.18 J g -1°C -1 ), ∆T/∆t is the initial slope of the time-dependent temperature curve, and m Fe is the weight fraction of Fe in the medium. The concentrations of Fe in the samples were determined using ICP-MS (Agilent 7900).
The instrument used to produce Lf-EF (30-500 Hz) (Section S5-1) is described as follows: A Cu coil with 300 turns of wires was used to generate the Lf-EF, and it was driven by a programmable alternative AC/DC power (NF, EC750SA/EC1000SA; Japan). The power supply was operated in AC mode as the voltage source and output was a sinusoidal alternating voltage (30-500 Hz, 10 V).
Section S5-1. Digital photograph of the Lf-EF instrument.
The instrument used to produce Mf-EF (220-470 kHz) (Section S5-2) is described as follows: Mf-EF was generated by an induction heating system (M5, Xi'an SuperMag Nanobiotechnology Co. Ltd). The frequency was modulated by changing the coils with the different diameter and turns.
Section S5-2. Digital photographs of the Mf-EF instrument.