Cell-Membrane-Coated and Cell-Penetrating Peptide-Conjugated Trimagnetic Nanoparticles for Targeted Magnetic Hyperthermia of Prostate Cancer Cells

Prostate malignancy represents the second leading cause of cancer-specific death among the male population worldwide. Herein, enhanced intracellular magnetic fluid hyperthermia is applied in vitro to treat prostate cancer (PCa) cells with minimum invasiveness and toxicity and highly specific targeting. We designed and optimized novel shape-anisotropic magnetic core–shell–shell nanoparticles (i.e., trimagnetic nanoparticles - TMNPs) with significant magnetothermal conversion following an exchange coupling effect to an external alternating magnetic field (AMF). The functional properties of the best candidate in terms of heating efficiency (i.e., Fe3O4@Mn0.5Zn0.5Fe2O4@CoFe2O4) were exploited following surface decoration with PCa cell membranes (CM) and/or LN1 cell-penetrating peptide (CPP). We demonstrated that the combination of biomimetic dual CM-CPP targeting and AMF responsiveness significantly induces caspase 9-mediated apoptosis of PCa cells. Furthermore, a downregulation of the cell cycle progression markers and a decrease of the migration rate in surviving cells were observed in response to the TMNP-assisted magnetic hyperthermia, suggesting a reduction in cancer cell aggressiveness.


XPS Analysis
Figures S5 and S6 show, respectively, the wide scan and the high-resolution XPS spectra collected on SH, SS, SHS, and SSH MNPs. The high-resolution spectra are shown together with the results of the best fitting procedure.  and corresponding elemental peak deconvolution of Zn 2p, Co 2p, Fe 2p, Mn 2p (from left to right). The background was subtracted to the Zn 2p region of samples SHS and SSH for a better comparison with the same regions of the other two samples. The small feature present in the Mn 2p region of sample SH is due to indium substrate (namely In3p 3/2 peak), as discussed in the text.

S5
As reported in the main text, fitting of Co 2p, Fe 2p, and Mn 2p data is not straightforward, as multiplet splitting effects must be taken into account, together with usual spin-orbit coupling, shake-up and plasmon loss features. In the present case, moreover, fitting is complicated also by the overlap of different transitions in the same spectral region.
In this work, three overlaps were evaluated. The first one between the Mn2p 1/2 peak and a transition owing to the indium substrate, namely In3p 3/2 . This feature is clearly visible in the spectrum of sample SH, which does not contain Mn ( Figure S5). The problem was solved focusing only on Mn2p 3/2 peak in the analysis of Mn2p region: this is possible because the Mn2p 1/2 transition is well-separated from the Mn2p 3/2 one.
The second overlap is between the Co2p peak and a LMM Auger transition of Fe. This spectral feature was deconvoluted to avoid overestimating the relative amount of Co in the samples, according to the following procedure. First, this region was fitted for the SS sample, which does not contain Co, so that the only contribution in the 770 -810 eV range is related to the Fe Auger peak. The best fit was achieved using three components, which have no specific chemical or physical meaning, but help to reproduce the experimental trend due to the Fe LMM transition. Then, these components were included in the fitting of the other three samples, which contain both Co and Fe. All the parameters determining the shape and the relative intensities of these components were fixed to the values obtained from the previous fit, letting only the overall intensity to change together with the peaks of Co2p, to reach the best fit of the whole region. In this way, the contribution of Fe Auger transition can be separated from the Co2p peak.
Finally, the last overlap that should be considered is between the Fe2p region and an LMM Auger transition of Co. However, the contribution of Co Auger peak was neglected for two reasons: 1) there are no samples containing Co but not Fe, so the previous procedure cannot be applied in this case, and 2) the amount of Co is much lower than the one of Fe, thus the overestimation of the Fe relative amount is expected to be limited.

Fitting of Mn 2p peaks
The broad and asymmetric peak around 641 eV corresponds to the Mn2p 3/2 photoemission in the case of SHS, SS, and SSH samples, while for the SH sample no Mn features are present and only a signal due to the indium substrate is detected, as discussed above. According to the literature 1 , the Mn2p 3/2 photoemission peak is decomposed in six contributions centered at

Fitting of Fe 2p peaks
For each sample, the XPS spectrum of Fe2p displays two main asymmetric peaks, Fe2p 3/2 and Fe2p 1/2 , centered at 711.0 eV and 724.5 eV, respectively ( Figure S5). These contributions are mainly due to Fe 3+ cations. Moreover, a small shoulder centered at about 708.5 eV can be associated with the presence of Fe 2+ species. Considering that both Fe(III) and Fe(II) signals undergo multiplet splitting, the Fe2p 3/2 peak was fitted with eight components: 710.2, 711.

Fitting of Co 2p peaks
In the case of SH, SHS, and SSH samples, the Co2p spectrum has been fitted considering both Co2p 3/2 and Co2p 1/2 contribution ( Figure S5), while the corresponding spectrum of SS sample, which does not contain Co, was used to evaluate the contribution of Fe Auger peak, as described above. In the Co2p spectrum of samples SH and SSH, the features with binding energy positioned at about 780.6 and 786.5 eV were attributed to Co2p 3/2 and its satellite. Accordingly, the components centered at 796.3 and 802.8 eV correspond to the Co2p 1/2 core level and its satellite, respectively. In the Co2p spectrum of the SHS sample these peaks are not clearly distinguishable because the Fe Auger component is predominant. Due to multiplet splitting, Co fitting was achieved via deconvolution of each Co2p 3/2 and Co2p 1/2 core-level photoemission peak into four components: 780.4, 782.5, 785.5, 787.1 eV, and 796.4, 798.3, 801.2, 803.2 eV, respectively. These components correspond to the presence of Co(II) in the samples 1 .