# Urchin-like magnetic microspheres for cancer therapy through synergistic effect of mechanical force, photothermal and photodynamic effects – Journal of Nanobiotechnology

May 12, 2022

### Preparation and characterisation of UHMMs

UHMMs were fabricated through a two-step process, in which carbonylated polystyrene (PS) microspheres and FeSO47H2O were used to synthesise the UHMM precursor using the hydrothermal method. The UHMMs precursor has a core–shell nanostructure that comprises Fe2O3 nanorods as the shell and PS as the core. The PS microspheres served as templates and were removed when Fe2O3@PS was transferred to Fe3O4 during heat treatment in a tube furnace under an Ar/H2 (5%) atmosphere. Scheme 1 shows a schematic illustration of the UHMMs synthesis.

Figure 1a shows the morphology of the precursor-Fe2O3@PS microspheres under a transmission electron microscope (TEM). The TEM images revealed that the precursor possessed a core–shell structure; Fe2O3 nanorods were observed on the surface of carboxy PS due to the chelation reaction between Fe2+ and -COOH. The diameter of the PS core was 500 nm, and the length of the Fe2O3 nanorods was approximately 400 nm. Additionally, the temperature or heating time during the hydrothermal reaction manipulated the density and length of the Fe2O3 nanorods through material concentration (Additional file 1: Fig. S1a–i). After hyperthermia and reduction, the PS template was removed to obtain ~ 800 nm UHMMs (Fig. 1b, c); UHMMs have a uniform morphology with a hollow urchin-like structure. One hundred eighty-three particles from the TEM images were taken for the calculation. Figure 1d shows the distribution of the UHMMs particle size; most of the particles (83.98%) were smaller than 1000 nm, and 32.6% particles measured 800–900 nm, which corresponded with the result of DLS measurement in different solutions (Fig. 1e, Table 1 and Additional file 1: Fig. S4a, b).

The UHMMs structure corresponds to the set temperature and heating time. For instance, the hollow structure was disrupted by increasing the temperature above 400 °C or by overtime heating (Additional file 1: Fig. S2e and l). However, acanth sphere clusters formed at temperatures below 290 °C or elapsed times of less than 1 h due to PS melting (because the PS was not eradicated) (Additional file 1: Fig. S2b and i). Therefore, optimal annealing conditions to prepare UHMMs with good morphology and structure were 320 °C and 80 min in a tube furnace (Fig. 1c).

X-ray diffraction (XRD) patterns of the UHMMs show that the diffraction peaks were assigned to the (111), (220), (311), (222), (400), (422), (511), (440), and (533) planes of the Fe3O4 criterion card (JCPDS19-0629) (Fig. 1h). Therefore, UHMMs are highly pure Fe3O4 with a spinel structure and possess good superparamagnetic properties with a magnetisation saturation rate of 62.72 emu·g −1 (Fig. 1g). The O and Fe peaks are present in the energy dispersive spectroscopy (EDS) spectrum (Fig. 1i and k), confirming that UHMMs contain only two elements. The atomic proportions of Fe and O are 72.18% and 27.83%, respectively, (Fig. 1k and Table 2) confirming the XRD analysis results. The measured zeta potential of the UHMMs was − 1.67 mV; the UHMMs@Ce6 zeta potential of − 11.12 mV is attributed to the Ce6 from the abundant carboxyl groups (Fig. 1f).

To confirm the good structural stability and magnetism of the UHMMs, we maintained the UHMMs solution (0.5 mg/mL in deionised water) for 1 d, 30 d, and 3 months at room temperature (25 °C). The solution was also placed in a low-frequency vibrating magnetic field for 2 h and 10 h. The UHMMs maintained their original structure after 1 d and 30 d of storage (Fig. 1l, m); additionally, the magnetisation saturation was maintained at 62.70 emu g−1 (Fig. 1j) after 3 months. Therefore, the synthesised UHMMs possessed good structural and magnetic stability. The UHMMs later treated by a low-frequency vibrating magnetic field for 2 h and 10 h showed similar structural and magnetic results; thus, low-frequency vibrating magnetic fields cannot destroy the UHMMs structure (Fig. 1n, o).

The above-presented results indicate that the UHMMs synthesised in this study have ideal and steady urchin-like structures and magnetisms and can be potentially applied to cancer cell destruction with mechanical force induced by a low-frequency vibrating magnetic field.

The photothermal conversion efficiency is closely related to the absorption coefficient at a specific wavelength. Therefore, Fig. 2b shows the absorption spectrum of 0.3-mg/mL UHMMs aqueous dispersions in the ultraviolet, visible, and near-infrared regions. The absorption coefficient (E) of the UHMMs aqueous dispersions (Fig. 2b) is 6.12 at 808 nm and calculated using Eq. (1):

$${varvec{E}}=frac{{varvec{A}}{varvec{lambda}}}{{varvec{c}}cdot {varvec{l}}}$$

(1)

where E is the absorption coefficient, is the absorption value of the UHMMs aqueous dispersions at 808 nm, c is the concentration of the UHMMs aqueous dispersions, and l is the thickness of the cuvette.

Because UHMMs are structured on Fe3O4 nanorods with good absorption at 808 nm, we used an 808 nm laser (0.38 W, 0.35 cm2) to irradiate the UHMMs aqueous dispersions (0.3 mg/mL) for 20 min and used an infrared thermal imager to collect the temperature data and images (Fig. 2a and c). After the initial 7 min of irradiation, the temperature of the UHMMs aqueous dispersions increased significantly from 25.6 °C ± 0.4 °C to 39.6 °C ± 2.4 °C. Subsequently, the temperature profile plateaued after 10 min and reached 44.0 °C ± 0.2 °C at 20 min. We obtained the photothermal conversion efficiency of UHMMs with the temperature-drop curve after removal of the 808 nm laser using Eq. (2):

$${varvec{eta}}=frac{{varvec{h}}{varvec{A}}left({{varvec{T}}}_{{varvec{m}}{varvec{a}}{varvec{x}}}-{{varvec{T}}}_{{varvec{s}}{varvec{u}}{varvec{r}}{varvec{r}}}right)-{{varvec{Q}}}_{{varvec{i}}{varvec{n}},{varvec{s}}{varvec{u}}{varvec{r}}{varvec{r}}}}{{varvec{I}}(1-{10}^{-{varvec{A}}{varvec{lambda}}})}$$

(2)

where η is the conversion efficiency calculated using the temperature-drop curve, and hA is the coefficient of the photothermal conversion efficiency. (Tmax − Tsurr) was 18.94 °C (Fig. 2c); Qin,surr = Cpmwater (Tmax − Tsurr); I is the power density of the 808 nm laser, and Aλ is the absorption value of UHMMs aqueous dispersions (0.3 mg/mL) at 808 nm.

Therefore, η was calculated as 16.50%. The control group, which used deionised water to increase the temperature from 25.8 ℃ ± 0.6 under the same conditions (Fig. 2a and c), simultaneously indicated the safety of the power of the 808 nm laser applied in our work. Furthermore, an 808 nm laser was used to re-irradiate the UHMMs solution after cooling to room temperature for three cycles; the UHMMs retained excellent photothermal efficiency from an initial ~ 25.0 °C to ~ 44 °C. Thus, the UHMMs have good photothermal conversion efficiency and stability for application in tumour photothermal therapy (PTT).

To further enhance the killing effect of UHMMs in tumour therapies, Chlorin e6 (Ce6), a photosensitiser that kills cancer cells by generating cytotoxic reactive oxygen species (ROS) under light activation was selected. A little amount of Ce6 was loaded into the UHMMs using the sedimentation method at a concentration of 1 mg/mL in ethanol solution. Different feed ratios between the UHMMs and Ce6 were selected for loading analysis (Fig. 2d). After analysis using an ultraviolet spectrophotometer and calculation from Eq. (3), the average loading rate was found to be approximately 11.51% (w/w), with a feed ratio of 1:0.5 (UHMMs: Ce6 (Fig. 2e).

$$frac{{{varvec{A}}{varvec{lambda}}}_{{varvec{b}}}-{varvec{A}}{{varvec{lambda}}}_{{varvec{l}}}}{{varvec{A}}{{varvec{lambda}}}_{{varvec{b}}}}times frac{{{varvec{c}}}_{{varvec{C}}{varvec{e}}6}}{{{varvec{c}}}_{{varvec{U}}{varvec{H}}{varvec{M}}{varvec{M}}{varvec{s}}}}times 100boldsymbol{%}$$

(3)

where Aλb is the optical density of the Ce6 blank at 400 nm; Aλl is the optical density of Ce6 in the supernatant after the loading steps; and CCe6 and CUHMMs are the concentrations of Ce6 and UHMMs in the sedimentation system, respectively.

To verify the successful Ce6 drug loading on UHMMs, we employed reactive oxygen fluorescent probe (DCFH-DA) to detect the ROS level of the UHMMs@Ce6, UHMMs, and Ce6 in cells (maintaining the concentration of UHMMs@Ce6, UHMMs at 0.75 mg/mL, and Ce6 at 86.25 μg/mL under light exposure for 10 min). Figure 2f shows the difference in reactive oxygen among the Ce6, UHMMs@Ce6, and UHMMs after exposure to the 808 nm laser. As expected, the amount of ROS in UHMMs@Ce6 was 1.60 ± 0.53 times higher than that in the empty UHMMs but less than that in the Ce6 group due to the cancer cells killed by photothermal irradiation with the 808 nm NIR laser. Thus, UHMMs load Ce6 and produce a weak ROS under an 808 nm NIR laser for application in PDT tumour therapy.

### In vitro toxicity of UHMMs and UHMMs@Ce6

Estimating the toxicity of UHMMs before performing subsequent experiments in vitro is imperative to guarantee safety. Firstly, laryngocarcinoma cancer cells (TU212) were incubated with RPMI1640 culture solution containing UHMMs/UHMMs@Ce6 with low to high concentrations (0.3, 0.5, and 1.0 mg/mL, PBS-dispersed) for 2, 6, 12, and 24 h. Second, we evaluated cell viability through a Cell Titer-Glo® assay, which detects the number of viable cells by quantitative determination of ATP with luminescence. Figure 3a, b show the toxicities of UHMMs and UHMMs@Ce6, respectively. At low concentration (0.3 mg/mL), TU212 cells maintained good viability for 2–24 h; 91.01% ± 6.86 and 88.08% ± 2.80 of cancer cells survived UHMMs and UHMMs@Ce6 incubation for 24 h, respectively.

However, for an increased PBS-dispersed UHMMs/UHMMs@Ce6 concentration of 0.5 mg/mL and prolonged incubation time of 24 h, the viability of the cells significantly decreased compared to the untreated group; 76.84% ± 8.77 and 71.57% ± 7.55 of cancer cells survived in the UHMMs and UHMMs@Ce6, groups, respectively. The cancer cell viability in the control group (no materials) was ~ 100% after 24 h of incubation.

The cancer cell viability decreased remarkably at UHMMs/UHMMs@Ce6 concentrations of 1 mg/mL; 74.16% ± 4.10 (UHMMs) and 55.99% ± 2.94 (UHMMs@Ce6) of cancer cells survived after 24 h of incubation. These results indicate that the toxicity of UHMMs@Ce6 was greater than that of UHMMs at the same concentration for a longer incubation time.

In addition, the L929 cell line, one of the important constituent cells of the dermis, was also tested in toxicity experiments and the obtained results were same as expected. The cells kept a good viability after co-cultured with low concentration of UHMMs and UHMMs@Ce6 dispersion in 24 h. Similarly, a high concentration of UHMMs and UHMMs@Ce6 dispersion would not be friendly to cells with long time co-culturing. Additionally, compared with UHMMs, the results showed greater toxicity with UHMMs@Ce6 (Additional file 1: Fig. S3a, b).

Thus, UHMMs/UHMMs@Ce6 possess weak toxicity to cancer cells at lower concentrations during a short incubation time (< 12 h) but show non-negligible toxicity at higher concentrations and longer incubation times, especially for UHMMs@Ce6. Part of the toxicity is attributed to the Fenton reaction of Fe3O4, which releases Fe2+ and Fe3+ into the cell culture medium. Ce6 also showed little phototoxicity to cells during incubation.

### Cancer cell killing efficiency of UHMMs mediated by single mode therapy

UHMMs possess good photothermal conversion efficiency. Therefore, to evaluate their photothermal damage effect, TU212 cells were incubated with UHMMs-RPMI1640 culture medium solution and irradiated with an 808 nm laser (1.08 W/ cm2). The viability of the TU212 cells treated by “UHMMs + laser” was associated with the irradiation time and the concentration of PBS-dispersed UHMMs. Firstly, the concentration of PBS-dispersed UHMMs was maintained at 0.75 mg/mL. Figure 3c shows the viability of TU212 cells after treatment with the 808 nm laser for 5, 10, and 15 min; 10.44% ± 9.80 cells survived after 15 min of irradiation, and the cell viabilities of the remaining two groups was 94.49% ± 8.79 (5 min) and 58.49% ± 6.85 (10 min). Secondly, keeping the irradiation time of 808 nm laser for 10 min, 89.61% ± 9.93, 58.49% ± 6.85, and 13.12% ± 12.42 of cancer cells retained activity at 0.5, 0.75, and 1.0 mg/mL UHMMs dispersions, respectively (Fig. 3d).

These results indicate the concentration dependency of UHMMs when used for photothermal effects to damage cancer cells. The group of “RPMI1640 + laser” and “UHMMs (no laser)” had no apparent adverse effect on cell viability, as both showed high and regular cell activity during the experiments.

Hoechst/Propidium Iodide (PI) fluorescence double staining qualitative detection also indicates the damage to cancer cells corresponding to the UHMMs concentration of the irradiation time (Additional file 1: Fig. S6 and S7). The dead cells are shown as red because the cell membrane lost biological functions and PI stained cell nucleus; contrarily, the survived cells are shown as blue. Thus, the dead cell (red colour) count was increased with the increased UHMMs concentration or irradiation time. These results confirm that UHMMs have good photothermal efficiency for killing cancer cells using 808 nm laser and promising tumour photothermal therapy applications.

A simple device (Additional file 1: Fig. S5) with two permanent magnets was developed to generate a low-frequency vibration, which is a magnetic field that can produce a sine vibration. The details of the working principle of the low-frequency vibration magnetic field generator can be found in Experimental Section and work reported in previous study [44]. Herein, we designed three groups to explore the cancer cell-killing effects of UHMMs in VMF with different incubation times, vibration times, and dispersion concentrations. TU212 cells were incubated with different concentrations of UHMMs (0.3, 0.5, and 1.0 mg/mL, RPMI1640 medium) for 1, 3, and 6 h and exposed to VMF for 0.5, 2, and 4 h. As expected, the cancer cell killing efficiency of UHMMs with VMF was significantly greater than that of the untreated and control groups. Figure 3e demonstrates the cancer cell killing efficiency when co-cultured with UHMMs in VMF in different situations. Remarkably, the killing efficiency was correlated with the UHMMs concentration or the vibration time; 17.18% ± 4.37%, 22.89% ± 2.55%, and 34.00% ± 6.32% of cancer cells were killed when the UHMMs concentration increased from 0.3–1.0 mg/mL during 2 h of vibration. Further, 11.84% ± 6.57%, 22.00% ± 2.55%, and 29.00% ± 5.53% of cancer cells were killed (efficiency was positively correlated with treatment time) for vibration time increased from 0.5–4 h using 0.5 mg/mL UHMMs (Fig. 3f).

Figure 3g shows viable cell activities of 84.00% ± 7.63%, 77.11% ± 2.55%, and 75.60% ± 6.90% when incubated with UHMMs for 1, 3, and 6 h, respectively when the concentration of UHMMs was kept at 0.5 mg/mL and VMF vibration time for 2 h. Therefore, cancer cell killing efficiency can be improved in VMF when co-cultured with UHMMs for a short time. However, cancer cell activity remained above 95% in the control group, which only added UHMMs. There was no significant difference compared with the no-treatment groups. Qualitative detection with Hoechst/propidium iodide fluorescence double staining also yielded results consistent with prior quantitative analysis (Additional file 1: Fig. S8–S10).

A scanning electron microscope (SEM) was used to observe damage to laryngocarcinoma cells after 0.5–4 h of vibration to explore the definite mechanism of cancer cell death caused by UHMMs by generating mechanical force under VMF. The SEM images in Fig. 3i show the destruction of TU212 cells caused by mechanical force with UHMMs (0.5 mg/mL) in VMF (maintaining the maximum value of magnetic intensity at 400 mT and the frequency at 2 Hz). The particles attached to fusiformis cells are UHMMs, and the black holes on the surface of cancer cells show the destruction of the cell membrane by UHMMs. The diameter of the black holes enlarged to ~ 5 μm, and their number increased after 4-h exposure to VMF. Furthermore, lactate dehydrogenase (LDH) leakage corresponds to the extent of cell membrane destruction; 11.03% ± 1.06, 18.37% ± 2.81%, and 27.96% ± 6.04% LDH was released to the cell medium after VMF treatment for 0.5, 2, and 4 h, respectively (Fig. 3h).

Therefore, the mechanical force generated by UHMMs under VMF is an effective in vitro inducer of tumour death. The cancer cell killing efficiency relies on the mechanical force generated by UHMMs in VMF, which destroys cell membrane integrity and releases the contents of the cell.

### Cancer cell killing efficiency of UHMMs mediated by multimode therapy (photothermal, mechanical force, and photodynamic effects)

Next, we used UHMMs@Ce6 (1 mg UHMMs containing 0.115 mg Ce6) to investigate the cancer cell killing efficiency of the combination therapy of the magneto-mechanic force, photothermal, and photodynamic effects under VMF and the 808 nm laser. Furthermore, to prevent the reduction of the photothermal effect caused by UHMM accumulation under a low-frequency vibrating magnetic field, an 808 nm laser was applied before using VMF. We created seven in vitro group experiments: RPMI1640 (no treatment), Ce6 + laser, UHMMs + VMF, UHMMs + laser, UHMMs@Ce6 + laser, UHMMs + laser + VMF, and UHMMs@Ce6 + laser + VMF.

First, we maintained the concentration of UHMMs/UHMMs@Ce6 dispersion (dispersed in RPMI1640) at 0.75 mg/mL. Second, TU212 cells were co-cultured with UHMMs and UHMMs@Ce6 solution for 3 h. TU212 cells containing UHMMs or UHMMs@Ce6 solution were irradiated for 10 min with an 808 nm laser (1.08 W/cm2) or vibrated with VMF for 2 h. Cell Titer-Glo® assay was used to test the viability of the cancer cells.

The TU212 cancer cells exhibited more serious destruction with group than the other groups after the VMF and laser combination treatments with UHMMs@Ce6. Approximately 98.83% ± 1.30% of cells were killed after multimode treatment (Fig. 4b), indicating the elimination of almost all cancer cells by magneto-mechanic force, photothermal, and photodynamic effects. However, the UHMMs@Ce6 + laser and UHMMs + laser + VMF groups using two modes to kill TU212 cells also showed moderate injury; 79.87% ± 6.48% of cancer cells were killed under the combination of photothermal and photodynamic effects, and 59.00% ± 5.20% were killed by the photothermal effect and mechanical force. In addition, the variable killing effects (Fig. 4b) for single treatments in groups , , and were 73.00% ± 7.36%, 72.00% ± 9.40%, and 58.00% ± 6.85%, respectively. The killing efficiency of combined photodynamic and photothermal effects is greater than the sum of a single photodynamic (Ce6) or photothermal (UHMMs) treatment. Thus, photothermal damage could strengthen the photodynamic killing effect. The main reasons that caused efficient destruction are probably attribute to: firstly, the permeability of cancer cells’ membrane was increased by photothermal effect, more ROS generated by Ce6 could enter in weak cancer cells and aggravate the destruction. Secondly, the dark green color of Ce6 may play a little photothermal effect to kill cancer cells. However, the strengthen effect induced by photothermal effect is limited, which will be discussed in more detail in the discussion section. Additionally, compared with the viability of the no-treatment group, cancer cell activity showed no noticeable difference among UHMMs, UHMMs@Ce6, and Ce6 groups without laser or VMF, revealing good UHMMs biocompatibility.

The images of Hoechst/propidium iodide fluorescence double staining qualitative detection indicated the same result among the seven groups (Fig. 4a). The blue and red colours represent live and dead cells, respectively. The number of red cells gradually increases with single, double, and multimode treatments. Hence, the results illustrate that the multimode therapy with UHMMs@Ce6 is a promising therapeutic strategy for tumour damage.

### Photothermal conversion of UHMMs with in vivo 808 nm NIR laser

The good photothermal conversion efficiency of UHMMs under an in vitro 808 nm NIR laser was previously demonstrated. To further identify the photothermal conversion ability of UHMMs for solid tumour treatment applicability, 5 mg/mL of PBS-dispersed UHMMs were directly injected into a tumour by intratumoural injection. Nude mice bearing the tumours were exposed to an 808 nm NIR laser (1.08 W/cm2) for 10 min (the tumour site was right under the 808 nm laser apparatus, Fig. 5a) after 12 h of feeding.

Figures 5b, c show the temperature curve and infrared thermal images, respectively. The temperature of the red line, which was injected with UHMMs, showed a significant and rapid increase for the first 2 min under 808 nm laser irradiation, from 37.27 ± 0.25 ℃ to 48.53 ± 1.0 ℃. After 5 min the temperature reached 49.43 ± 0.91 °C and remained steady until 10 min (49.3 ± 1.2 ℃). Notably, the temperature exceeded 42 °C for at least 9.5 min during 10 min of irradiation; thus, the photothermal effect of UHMMs induced by the 808 nm laser effectively damaged the solid tumour. Correspondingly, the temperature of the control group injected by 1 × PBS and treated with the 808 nm laser for 10 min showed a security range (Fig. 5b) and reached 40.43 ℃ ± 0.5 ℃ after 10 min. Thus, the power density and irradiation time of the 808 nm laser in this work are safe.

### In vivo efficiency of laryngocarcinoma solid tumour damage with multimode therapy (photothermal, mechanical force, and photodynamic treatments)

Multimode therapy with UHMMs/UHMMs@Ce6 showed good cancer cell-killing effects using a laser, VMF, or in vitro combination. Therefore, in the current study, eight groups of mice were used for in vivo cancer therapy (Additional file 1: Fig. S11).

Thirty-two tumour-bearing nude mice (tumour volume ≈ 4 × 4 mm) were randomly divided into 8 groups and each group has 4 mice: PBS (no treatment), Ce6 + laser, UHMMs + VMF, UHMMs + laser, UHMMs@Ce6 + VMF, UHMMs@Ce6 + laser, UHMMs + VMF + laser, and UHMMs@Ce6 + VMF + laser. To ensure optimal efficiency in tumour damage, VMF was applied first before treatment with the 808 nm laser.

Figures 6a and Additional file 1: S12–S19 show the changes in tumour size with different treatments over 15 days. For group , 0.575 mg/mL Ce6 solution (1 × PBS) was injected into the tumour by intratumoural injection; the group was then exposed to 808 nm NIR laser (1.08 W/cm2) for 10 min after 12 h of feeding. The tumour growth curve of “Ce6 + Laser” showed was gradually inhibited compared with the PBS group after 7 d. On day 7, the sizes of the tumours in groups and were 205.22 ± 26.38 and 96.90 ± 24.68 mm3, respectively. Fifteen days later, the difference between the tumour sizes of groups and increased (641.43 ± 89.55 and 324.80 ± 29.45 mm3, respectively). On day 15, the weight of the tumour indicated that photodynamics had a negative influence on tumour growth (Fig. 6c). The tumour weights of groups and were 0.502 ± 0.030 and 0.293 ± 0.066 g, respectively. Notably, the curve of group was slightly steeper after 7 d, as Ce6 was easily metabolised in vivo; additionally, photodynamic effects were invalidated by the 808 nm laser during the second treatment.

For groups and , 5 mg/mL UHMMs or UHMMs@Ce6 dispersion (1 × PBS) was directly injected into the tumour by intratumoural injection; the groups were exposed to VMF (400 mT, 2 Hz) for 2 h after 12 h of feeding. The tumour sizes of these groups showed a growth curve similar to that of the VMF treatment and significantly inhibited tumour growth by magneto-mechanical force five days later. On day 7, the size of the tumours in groups and were 117.63 ± 18.51 and 83.14 ± 42.52 mm3, respectively. Figure 6a and Additional file 1: Figs. S12, S14, S16 show the significant difference in tumour size between the PBS (group , 205.22 ± 26.38 mm3) and groups and . After 15 days of VMF treatment, the tumour volumes of groups , , and were 641.43 ± 89.55, 351.68 ± 44.02 and 300.92 ± 32.02 mm3, respectively. On day 15, the weight of the tumour indicated that the magneto-mechanical force induced by VMF significantly inhibited tumour growth (Fig. 6c). The tumour weights of groups , and were 0.502 ± 0.030, 0.327 ± 0.031 and 0.232 ± 0.058 g, respectively. The tumour growth curve initially rose gently but rapidly increased after 7 d for several reasons. First, as the tumour volume increased, the pre-set amount of UHMMs located at a fixed place inside or beside the solid tumour could not transmit the mechanical force to the new-born tumour under VMF; therefore, the inhibition caused by the magneto-mechanical force was weakened, and the tumour growth rapidly increased. These results indicate that the magneto-mechanical force induced by UHMMs and VMF inhibited tumour growth, especially during the early period. However, the magneto-mechanical force should be combined with other therapies to obtain a curative effect in solid tumour therapy because of the weakness of the magneto-mechanical force in the late period.

Group was treated by photothermal irradiation under an 808 nm laser and was very effective against solid tumours. A 5 mg/mL UHMM dispersion (1 × PBS) was directly injected by intratumoural injection, and the group was exposed to an 808 nm laser (1.08 W/cm2) for 10 min after 12 h of feeding. Three days after 808 nm laser irradiation, the skin of the tumour side turned dark purple then black (Additional file 1: Fig. S15). The tumour was markedly inhibited and decreased to 48.63 ± 22.49 mm3 on day 7 (Fig. 6a). Furthermore, the tumour growth curve was flat and decreased slightly from days 7 to 12. However, the tumour size increased again after 11 days and finally approached 98.50 ± 50.04 mm3. The tumour volume of PBS (group , 641.43 ± 89.55 mm3) was six times larger than that of UHMMs + laser (group ) (Fig. 6a). On day 15, the tumour weight of group (0.082 ± 0.029 g) was significantly different from that of group (Fig. 6c). These results indicate that UHMMs photothermal therapy induced by an 808 nm laser was satisfactorily treated the solid tumour but did not eliminate the tumour, which grew again 11 days later (Additional file 1: Fig. S15).

Tumour growth was further inhibited with dual-mode treatment. Figure 6b and Fig. 6c show that the curative effect was better in group , which used photothermal and photodynamic treatments under an 808 nm NIR laser than in groups or (Additional file 1: Figs. S12, S13, S15, and S17). Similarly, 5 mg/mL UHMMs@Ce6 dispersion (1 × PBS) was directly injected into the tumour by intratumoural injection; the group was exposed to an 808 nm NIR laser (1.08 W/cm2) for 10 min after 12 h of feeding. On day 7, the tumour size of group (60.60 ± 46.25 mm3) was much smaller than that of group (205.22 ± 26.38 mm3) but close to that of group (48.63 ± 22.49 mm3). However, on the last day, the tumour size of group (53.46 ± 68.54 mm3) was smaller than that of group (98.50 ± 50.04 mm3). The tumour weight of group (0.051 ± 0.033 g) was also lower than that of group (0.082 ± 0.029 g).

Additionally, the tumour showed better inhibition for combined photothermal and magneto-mechanical forces than for a single treatment. In group , 5 mg/mL UHMMs dispersion (1 × PBS) was directly injected into the tumour by intratumoural injection for 12 h. Subsequently, the tumour-bearing nude mice were exposed to VMF (400 mT, 2 Hz) for 2 h and irradiated by an 808 nm laser (1.08 W/cm2) for 10 min. On day 15, the tumour sizes of groups , and were 324.80 ± 29.45, 98.50 ± 50.04 and 18.29 ± 22.72 mm3 (Fig. 6a, Additional file 1: Figs. S12, S14, S15 and S18). The tumour weight of group (0.022 ± 0.005 g) was also significantly lower than that of groups and (0.327 ± 0.031 and 0.082 ± 0.029 g, respectively) (Fig. 6c).

These results indicate that dual-mode treatments of photothermal with photodynamic or magneto-mechanical force have an encouraging curative effect on solid tumour inhibition. Furthermore, the curative effect of tumour inhibition with dual-mode therapy was better than that of a single treatment with UHMMs or UHMMs@Ce6.

Finally, to identify whether the multimode therapy (magneto-mechanic force, photothermal, and photodynamic effects) could better inhibit solid tumours than either dual-mode therapy or single therapy, the tumour-bearing nude mice in group were directly injected with 5 mg/mL UHMMs@Ce6 dispersion (1 × PBS) by intratumoural injection. After 12 h of feeding, this group was exposed to VMF (400 mT, 2 Hz) for 2 h vibration and later irradiated for 10 min with an 808 nm NIR laser (1.08 W/cm2) immediately. Five days after treatment with VMF and the 808 nm NIR laser, the bulging solid tumours flattened, and the colour turned black (Additional file 1: Fig. S19). These tumours were almost eliminated, and only a small scar appeared on the skin appeared after treatment for 15 d (Fig. 6b and Additional file 1: Fig. S19). Simultaneously, the average weight of the tumours in group (0.012 ± 0.001 g) differed significantly from those of the other groups and was extremely significantly different to that of group (Fig. 6c). The results indicate that multimode therapy with magneto-mechanic force, photothermal, and photodynamic effects induced by UHMMs with an 808 nm NIR laser and VMF and had an excellent curative effect on solid tumours.

The weights of mice in groups showed a normal growth curve after treatment with the 808 nm laser and VMF for 15 d (Fig. 6b). However, the weights of mice in groups , , , and treated by VMF were lower than those in group (PBS) during the later period. This difference was attributed to the 2-h daily vibration in the VMF apparatus, which negatively influenced the diet and physical health of the mice.

The weight and histological structure of the major organs of the mice showed no significant differences among the different groups, except for the spleen (Fig. 6d, f, Additional file 1: Figs. S20–S24). The spleen is the largest in vivo peripheral immune organ and plays an important role in tumour immunity. Thus, spleen weight was closely correlated with tumour size. The nude mice of group treated with the 808 nm laser and VMF had the smallest spleen (0.450% ± 0.027% of total body weight); the spleens of the mice of group accounted for 0.864% ± 0.058% of the total body weight. The weights of the tumours in group were lower than those in the other groups.

H&E staining of tumours in the groups with different treatments showed significant differences (Fig. 6h). First, the PBS group cells in the solid tumour were plump and had a clear nucleus; further, the structure of cells was completely and tightly packed. However, after a single photodynamic treatment, the cancer cells were loosely packed but maintained a normal morphology. In groups and , UHMMs/UHMMs@Ce6 were dispersed in the tumour, and the morphology was beside the nanoparticles changed significantly. The gap between the cells was enlarged, and UHMMs/UHMMs@Ce6 destroyed the integrity of the cell structure.

Furthermore, the tumour cells showed ablation and necrosis after treatment with the 808 nm NIR laser because of the high temperature. When dual-mode or multimode treatment was applied, the tumour size decreased, and the morphology of the solid tumour cells also showed similar changes. Black nanoparticles surrounded the cells; the lack of clear boundaries among the cells was due to multiple fragmented cells.

Simultaneously, the morphology and structure of the liver sections in group with H&E staining showed no significant differences (Additional file 1: Fig. S25). However, the spleen sections with H&E staining showed little dissimilarity (Additional file 1: Fig. S25). For instance, the spleen weight of the tumour-bearing nude mice in group that were injected with PBS due to almost complete tumour elimination was lower than that in group . The tightness of the spleen cells was higher in groups (UHMMs + VMF + Laser) and (UHMMs@Ce6 + VMF + laser) and correlated with tumour volume.

Routine blood tests, and hepatic and renal functions tests were carried out at 1st, 7th and 21st days after percutaneous injection with UHMMS@Ce6 dispersion (1 × PBS, 5 mg/mL) in normal ICR mice. These tests included white blood cell (WBC), red blood cell (RBC), haemoglobin (HGB), platelet (PLT), percentage of lymphocytes (LYM %), mean corpuscular haemoglobin (MCH), haematocrit (HCT), plateletcrit (PCT), mean corpuscular volume (MCV)), and total protein (TP), albumin (ALB), globulin (GLOB), alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatinine (Cr), urea (UREA), total bilirubin (TBIL), albumin/globulin (A/G), respectively. The results are showed in Additional file 1: Fig. S26. Most routine blood, hepatic and renal functions tests indexes were normal, and no statistical difference was found compared with the normal group, which further confirmed the safety of UHMMS@Ce6 in vivo.