Enhancing Magnetic Hyperthermia at the Cell Membrane by Anchoring 92R‐Functionalized Magnetic Nanoparticles to Low‐Endocytic CCR9 Surface Receptors
David Egea‐Benavente, Isabel Corraliza‐Gorjón, Thomas S. van Zanten, María del Puerto Morales, Leonor Kremer, Domingo F. Barber
TL;DR
This paper introduces a new method to improve cancer treatment using magnetic nanoparticles anchored to cell membranes for better heat generation.
Contribution
A novel strategy using antibody-functionalized magnetic nanoparticles targeting CCR9 receptors to enhance magnetic hyperthermia therapy.
Findings
Antibody-functionalized magnetic nanoparticles targeting CCR9 receptors improve heat generation efficiency.
Localized hyperthermia at the subcellular level increases tumor cell death without macroscopic temperature rise.
Abstract
Magnetic hyperthermia therapy (MHT) is a promising cancer treatment that has demonstrated efficacy in phase I and II clinical trials for glioblastoma and prostate cancer. MHT relies on heat generated by magnetic nanoparticles (MNPs) when exposed to alternating magnetic fields (AMFs). The heat output depends not only on the intrinsic properties of MNPs but also on extrinsic factors such as the extracellular and intracellular environments. Aggregation of MNPs under certain conditions can significantly reduce therapeutic efficiency. To overcome this limitation, we present a strategy to enhance MHT by modulating MNP‐cell interactions. We functionalized dimercaptosuccinic acid (DMSA)‐coated MNPs with the 92R antibody (DMSA–MNPs@92R), which selectively binds to the low‐internalization chemokine receptor CCR9, overexpressed in certain tumors. Exposure of CCR9+ MOLT‐4 cells to DMSA–MNPs@92R…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5
FIGURE 6
FIGURE 7
FIGURE 8
FIGURE 9| Activation | Antibody | Blocking | HD size (nm) | PDI | ζ‐pot (mV) | |
|---|---|---|---|---|---|---|
| DMSA–MNPs | — | — | — | 41 ± 4 | 0.16 | ‒35 ± 8 |
| DMSA–MNPs@Gly | Yes | — | Gly | 77 ± 21 | 0.32 | ‒20 ± 1 |
| DMSA–MNPs@92R | Yes | 92R | Gly | 146 ± 19 | 0.42 | ‒23 ± 1 |
| DMSA–MNPs@IC | Yes | IC | Gly | 139 ± 11 | 0.40 | ‒24 ± 3 |
- —Consejo Superior de Investigaciones Científicas10.13039/501100003339
- —Spanish Ministry of Science and Innovation10.13039/501100004837
- —Severo Ochoa Program for Centres of Excellence
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsNanoparticle-Based Drug Delivery · Ultrasound and Hyperthermia Applications · Characterization and Applications of Magnetic Nanoparticles
Introduction
1
In oncology, hyperthermia (HT) is defined as the exposure of tumor tissue to supra‐physiological temperatures (>37°C) within a defined therapeutic window (39–45°C), to provoke tumor cell death and consequently, tumor regression [1]. HT is considered mild in the 39–43°C range, whereas >45°C thermal ablation occurs and causes important side effects [2]. High temperatures enhance antitumor responses of the immune system, including improved antigen presentation, increased leukocyte trafficking, and T cell, dendritic cell, and NK cell activation [3].
HT has synergic effects when combined with chemotherapy and/or radiotherapy to treat cancer [4, 5], as witnessed in clinical trials on glioblastoma, melanoma, or breast cancer [6]. Nevertheless, one drawback of HT is its lack of specificity, potentially provoking thermal damage in tumor and healthy tissues, not dissimilar to the nonspecific toxicity of systemic chemotherapy or radiotherapy [7, 8, 9]. The therapeutic use of magnetic nanoparticles (MNPs), and especially iron oxide nanoparticles (IONPs), is a promising strategy to design more personalized and efficient approaches to treat cancer, offering important advantages like better drug targeting, lower drug concentrations and consequently, fewer and milder side effects [10, 11, 12, 13]. MNPs may be catalogued as anticancer theragnostic agents, with uses in diagnosis (as MRI agents) and therapy (as targeted drug carriers). Indeed, IONPs have potential as anticancer agents by themselves, and they have enabled novel therapeutic strategies like magnetic hyperthermia therapy (MHT) to be developed. This approach is based on the heat released by MNPs in response to alternating magnetic fields (AMFs), thereby triggering cell death. This heat is produced in a precise, local manner, such that temperature only rises where MNPs are present and only when AMFs are activated [14, 15]. Moreover, the induction of MHT is minimally invasive, it has no depth restrictions, and its side‐effects are minimal. This makes it a promising strategy for deep tumors or those in organs where surgical removal is risky, like the brain [16].
The heat released by MNPs in an AMF is due mainly to two mechanisms of energy dissipation: Néel relaxation and Brownian relaxation. The former results from the inversion of magnetic moments between the two directions of the magnetization axis, the second is due to physical rotation of MNPs [17, 18]. Brownian relaxation normally contributes more significantly to the temperature rises associated with MNPs, which can be quantified according to the specific absorption rate (SAR), in terms of W/g_Fe_ (see Equation 1), whereby C* v
- is the specific heat of the medium, m_Fe_ the Fe concentration and ΔT/Δt the increase in temperature over time [19, 20] (Equation 1)
Unfortunately, MNPs heat dissipation is significantly dampened in biological media relative to water [21], primarily due to the higher viscosity of the medium and interactions with proteins or other elements. Consequently, the effective anisotropy (Κ eff) of the magnetic cores is lower in this situation and MNPs mobility is partially restricted, in turn interfering with Brownian relaxation [22]. These changes significantly affect the performance and efficacy of MHT, along with how MNPs interact with cells remain the primary bottleneck for the successful application of this therapy. Previous studies in vitro demonstrated that the magnetic properties of MNPs are modified when they interact with cells and that the heat they release decreases dramatically, as witnessed when MNPs were loaded into different cell lines (e.g., Pan02 [23], SKOV‐3 [24], or MG‐63 [25]). One possible explanation is that MNPs are taken up by lysosomes or other vesicles, where physical immobilization, dipolar interactions and subsequent clustering, or MNPs aggregation occurs [22, 24, 26, 27]. Indeed, the immobilization of MNPs due to their cellular interactions is reversible. Thus, when MNPs are internalized by cells and they accumulate in lysosomes, Brownian relaxation is blocked although Néel relaxation is not affected. However, after cell lysis and MNPs recovery, Brownian relaxation is restored to its initial values [26]. MNPs interact with hematopoietic cells (e.g. the T‐cell leukemia Jurkat cell line) very differently to other cell lines, forming a sheath around the membrane as opposed to being internalized. This capping process is due to nonspecific interactions and the aggregation of the MNPs in contact with the cell membrane, such that their heat dissipation is reduced [23]. This MNPs inactivation is triggered by them simply coming into contact with cells, independent of their core size, coating, the cellular environment, host cell line, or subcellular localization [23, 24, 27]. Therefore, new strategies must be implemented to obtain MNPs with more efficient heat release (preferentially due to Néel instead of Brownian relaxation), that also avoid aggregation, thereby resolving bottlenecks associated with MNP/cell interactions.
Chemokine receptor 9 (CCR9) is a seven‐transmembrane extracellular G protein coupled receptor with a low internalization rate that plays a crucial role in T cell development and tissue‐specific homing through binding to its specific ligand, CCL25 [28, 29]. CCR9 mRNA is weakly expressed in healthy tissues but according to the GENT2 database, CCR9 is found in several types of hematopoietic tumors. Specifically, >80% of patients with T‐lymphocyte tumors, 30% with splenic marginal zone lymphoma, and 20% with hairy cell leukemia and Hodgkin lymphoma express CCR9 mRNA strongly. Among nonhematopoietic tumors, nearly 70% of vaginal tumors and melanomas, as well as >25% of tumors in other organs (heart, bladder, small intestine, prostate, brain, and cervix) overexpress CCR9 mRNA. Consequently, CCR9 is considered an important therapeutic target [30, 31].
Here, we propose a strategy to prove that maintaining MNPs in close proximity to the cell membrane and in a less‐aggregates state could maximize heat generation in MHT in vitro. To this end, MNPs previously identified as excellent candidates for MHT due to their high SAR in aqueous media [32], were functionalized with the mouse monoclonal 92R antibody (mAb IgG2a), which specifically recognizes the N‐terminal domain of the human CCR9 receptor. CCR9 was selected as a model receptor because it is only internalized in vitro upon combined stimulations with IL‐2 and IL‐4 at low concentrations (ng/mL) [33]. Functionalizing with the 92R antibody enables the MNPs to specifically bind to CCR9 on the cell membrane of MOLT‐4 cells and remain there due to the receptor´s low internalization rate. This strategy prevents intracellular aggregation, immobilization, deactivation, and lysosomal degradation of MNPs, which would otherwise occur upon their internalization. In addition, this approach enables spatially controlled binding that helps to maintain the MNPs in a less‐aggregated state and at an optimal distance from the cell membrane, thereby enhancing heat generation. By keeping the MNPs in close proximity to the cell surface nonspecific aggregation is reduced, heating efficiency preserved, and overall MHT performance is improved. This proof‐of‐concept approach offers an innovative framework for modulating MNPs–cell to enhance magnetic hyperthermia applications and may represent an important first step toward the development of an optimized antitumor MHT combined with 92R‐based immunotherapy in vivo [29, 30, 34].
Results and Discussion
2
MNPs Synthesis, Dimercaptosuccinic Acid (DMSA) Coating, and Their Physico‐Chemical Characterization
2.1
IONPs were obtained by the microwave‐assisted polyol method following the protocol proposed previously [35]. Compared to other approaches like thermal decomposition, microwave‐assisted synthesis has many advantages when contemplating industrial synthesis for biomedical purposes: ease of handling, more effective heating that avoids thermal gradients, a faster procedure with lower energy demands, very reproducible production of monodispersed MNPs, high synthetic yields, scalable, and the possibility of establishing a continuous process [35, 36, 37]. The size and shape of the MNPs core obtained in this way was assessed by transmission electron microscopy (TEM), demonstrating that uniform 12 nm quasispherical MNPs were generated (Figure 1A). Moreover, the X‐ray diffraction (XRD) pattern suggested that magnetite/maghemite MNPs were obtained, with no secondary phases. All the peaks corresponded to an inverse spinel structure and the crystal size calculated from the broadening of the magnetite {220} crystalline plane was 13.6 nm according to the Scherrer equation (Figure 1B).
*MNPs and DMSA–MNPs characterization. (A) TEM image of 12 nm MNPs (scale bar = 50 nm, inset scale bar = 20 nm) and their size distribution. (B) XRD of MNPs: the black marks correspond to the magnetite patterns. (C) Magnetization curves of the MNPs at room temperature obtained in a VSM. (D) FTIR spectrum of the DMSA‐coated MNPs (light blue, main peaks highlighted), uncoated MNPs (gray), and free DMSA (red). (E) SAR of MNPs in water and DMEM + FBS 10% under an AMF of 530 kHz and 20 mT, before (gray) and after (light blue) DMSA coating (n = 5: n.s. = not significant, *p < 0.05, *p < 0.01). (F) Evolution of the SAR over time in DMEM+FBS (n = 3). (G) Evolution of the SAR at increasing glycerol concentrations (n = 3).
The superparamagnetic behavior of the MNPs was probed in a vibrating sample magnetometer (VSM: M R and H C ≈ 0), showing a M S ═ 84.5 emu/g_Fe_ (within the range for magnetite/maghemite NPs at room temperature, M S ═ 76–86 emu/g_Fe_ [38]: Figure 1C). Finally, the MNPs were coated with DMSA according to the protocol described elsewhere [39]. Infrared spectra confirmed the correct DMSA coating (Figure 1D), with: spectra bands at 750 and 400 cm^−1^ assigned to Fe‐O vibrations of magnetite; bands at 1040 cm^−1^ corresponding to C*─O and C─S groups; bands at 1385 and 1625 cm^−1^ corresponding to the symmetric and asymmetric vibrations of the DMSA carboxylic groups, respectively; and bands at 3400 and 2295–2950 cm^−1^ corresponding to the O─H of water and C─*H, respectively (not shown) [40]. A crucial observation is the disappearance of the free C═O stretching band of DMSA at ∼1700 cm^−1^, which shifts to the 1625–1385 cm^−1^ region. This shift evidences the deprotonation of the carboxylic acid groups to carboxylates (–COO^−^) and their subsequent coordination to the magnetite surface. Hydrodynamic diameters (HD) of 41 nm and a polydispersity index (PDI) ═ 0.16 were obtained for DMSA–MNPs by dynamic light scattering (DLS) in water at pH 7, indicating excellent colloidal dispersion. The surface charge was ‐35 mV, confirming the presence of the DMSA coating. When DMSA–MNPs were transferred to Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% Fetal Bovine Serum (DMEM + FBS 10%) their hydrodynamic diameter increased to 230 nm, although good colloidal stability was maintained (PDI = 0.24). Finally, the SAR of the MNPs was calculated in water and in DMEM + FBS 10%, before and after DMSA coating, and at an AMF frequency of 530 kHz and an intensity of 20 mT (Figure 1E). These AMF conditions were selected after screening different intensities and frequencies (Figure S1).
These values indicated the importance of the MNPs coating for their applications in biological media, avoiding significant aggregation of the magnetic cores due to interactions with proteins and other elements in the medium, consequently provoking a dramatic decrease in their heat release capacity. DMSA coating provided good colloidal stability to the MNPs in water and DMEM + FBS 10%, which retained 100% and 75% of their SAR relative to their initial values when dispersed in water and DMEM + FBS 10%, respectively. The slight decrease in the SAR in DMEM + FBS 10% could be explained by a small increase in the PDI, reflecting how some degree of aggregation enhanced the magnetic dipolar interactions between particles. However, the SAR remained stable for 15 days in DMEM + FBS 10% with little variation, demonstrating the MNP's stability over long periods of time (Figure 1F).
Finally, the contribution of Néel and Brownian relaxation to the total heat release was evaluated by measuring the SAR in increasing concentrations of glycerol, blocking Brownian relaxation by augmenting the viscosity of the medium. Consequently, 80% of the heat released was seen to be due to Neel relaxation, with a minor contribution of Brownian relaxation for these MNPs (Figure 1G). To summarize, 12 nm quasispherical magnetite/maghemite superparamagnetic MNPs synthesized in a microwave and covered with DMSA are excellent candidates to perform MHT due to the strong heat release in biological media over long periods of time. This release is triggered mainly by Néel relaxation, which would avoid problems related to the loss of heat efficiency due to blocking Brownian relaxation as a result of cell‐MNP interactions.
DMSA Coated MNPs Interactions with MOLT‐4 Cells, Establishing the Conditions for In Vitro MHT Experiments
2.2
The biocompatibility of DMSA–MNPs with MOLT‐4 cells was evaluated using the Presto Blue viability assay, exposing cells to increasing concentrations of the DMSA–MNPs (from 0 to 500 µg_Fe_/mL) for either 2 or 24 h. The MNPs proved to be relatively biocompatible, exhibiting high biocompatibility at doses below 250 µg_Fe_/mL for both time points tested, inducing <15% cell death. However, cytotoxicity increased to 25% after a 2‐h incubation with the highest dose (500 µg_Fe_/mL: Figure 2A,B). As shown previously, MNPs attach to the membrane of Jurkat lymphoid cells [23, 41] and so, the amount of DMSA–MNPs attached to MOLT‐4 cells was assessed after different periods of exposure to the MNPs. After exposing the cells to the MNPs, the excess MNPs that did not interact were removed by washing 3 times with PBS and the amount of Fe per cell was evaluated through inductively coupled plasma optical emission spectroscopy (ICP‐OES). The system reached saturation in a short time (2 h, 3.1 pg Fe/cell) and after even longer incubations the same amount of MNPs were attached to the cells (Figure 2C), albeit 3–5 times lower than the 3‐aminopropyl‐triethoxysilane (APS) coated MNPs attached to Jurkat T lymphocytes [23, 41]. Confocal and TEM images were used to confirm the MNPs were attached to the cell membrane, demonstrating that the MNPs were indeed bound to the MOLT‐4 cell membranes. Internalization of MNPs within these cells is highly exceptional (Figure 2D,E). This was described as typical behavior of lymphocyte cell lines, such as Jurkat cells [23]. Moreover, the amount of DMSA–MNPs attached to the MOLT‐4 cell membrane (around 3 pg/cell) was similar to that of Jurkat cells and of other tumor lines like U87MG or MDA‐MB‐231 cells with 12 nm MNPs synthesized by coprecipitation, even though these latter cell lines internalize the MNPs [42]. Finally, the DMSA–MNPS concentration necessary to reach the HT therapeutic window was studied, concluding that doses > 150 µg_Fe_/mL are required and that doses around 500 µg_Fe_/mL are suitable to reach macroscopic HT temperatures (Figure 2F). Hence, the conditions selected for in vitro MHT experiments were a 2 h exposure to 500 µg_Fe_/mL MNPs.
*Study of the interaction between DMSA–MNPs and MOLT‐4 cells. (A) Presto Blue biocompatibility assays using increased doses (0–500 µgFe/mL) of DMSA–MNPs for 2 h or (B) 24 h (n = 3, p < 0.05). (C) Timing of Fe cell attachment at a DMSA–MNPS concentration 150 µgFe/mL. (D) Confocal and (E) electron microscopy images taken 2 h after incubation with MNPs (150 µgFe/mL). Scales bar = 20 and 2 µm, respectively. Insets in TEM images correspond to a more detailed area (scale bar = 100 nm). The black arrow in the inset indicates DMSA–MNPs. (F) Study of the temperature increase in an AMF (528 kHz, 20 mT) for 1 h using different DMSA–MNPS concentrations.
DMSA–MNPS Functionalization with Antibodies and Their Characterization
2.3
The functionalization of MNPs with antibodies to specifically recognize a desired target has been widely studied for various applications, such as the detection of SARS‐CoV‐2 [43] and targeting of tumor cells [44]. In contrast, in this study, the 92R antibody was selected for its specific recognition of the CCR9 receptor which is expressed by various tumor cell lines, including MOLT‐4 cells, with the primary aim of enable MNPs to specifically bind to CCR9 on the membrane of MOLT‐4 cells and remain localized there due the receptor´s low internalization rate, rather than for targeting purposes. The binding of 92R‐functionalized MNPs prevents intracellular aggregation, immobilization, deactivation, and lysosomal degradation of MNPs, which would otherwise occur upon their internalization. Furthermore, it allows for spatially controlled binding, helping to maintain the MNPs in a less‐aggregated state and at an optimal distance from the cell membrane, with the goal of enhancing heat generation. The 92R antibody (a mouse mAb, IgG2a) was attached to DMSA–MNPs through the activation of free DMSA carboxylic groups and via electrostatic interactions with the positive amino acids of the antibody's Fc, as proposed previously [45]. To ensure the specificity of the results obtained, an isotype control antibody (IC, mAb IgG2a) that does not recognize CCR9 was also used. Functionalization was carried out after optimizing several parameters (Figures S2 and S3). The antibody functionalization process was subsequently confirmed by Fourier Transform Infrared Spectroscopy (FTIR) analysis, revealing pronounced differences in the 1250–850 cm^−1^ region between DMSA–MNPs and antibody‐functionalized DMSA–MNPs. Antibody conjugation introduced additional amide III and C─N/N─H deformation bands, confirming successful functionalization and modification of the local chemical environment (Figure S3F). Therefore, depending on the functionalization performed, we obtained different DMSA–MNPS samples of distinct hydrodynamic diameter and ζ‐potentials in water (see Table 1). Increases in hydrodynamic diameter and PDI were observed as a result of DMSA‐groups of MNPs activation and subsequently blocking using Gly (DMSA–MNPs@Gly). Further enlargement was detected when, after the activation process, antibodies were attached (DMSA–MNPs@92R and DMSA–MNPs@). Antibody attachment also made the ζ‐potential of the MNPs less negative, likely due to the loss of the exposed free carboxylic groups of DMSA, now bound to positively charged antibody residues. To evaluate potential effects of protein corona formation, which can critically influence biological performance [46, 47], the hydrodynamic diameter was measured in DMEM supplemented with 10% FBS at different time points (Table S1). Results showed a slight increase in hydrodynamic diameter and PDI for antibody‐functionalized nanoparticles when exposed to DMEM + FBS, suggesting partial aggregation or interactions with serum proteins. Nevertheless, the nanoparticles maintained appropriate hydrodynamic diameters (approximately 150 nm) and good colloidal stability (PDI = 0.24) even after 72 h of incubation.
Similar amounts of both antibodies bound to the MNPs (12 µg Ab/mg MNPs) approximately 40% of the antibody added (Figure 3A). Regarding the heat‐release capacity, a decrease in the SAR was observed following antibody functionalization (Figure 3B), probably due to the enhanced shielding of the core, and the interactions between the MNPs surface and the environment, producing a loss of colloidal stability (see PDI values in Table 1). However, this decrease in the SAR was mild (>75% of the heat release capacity was maintained) and consequently, the capacity to raise the temperature up to therapeutic HT windows (39–45°C) using a DMSA–MNPs@antibody concentration of 500 µg_Fe_/mL confirmed their suitability for in vitro MHT experiments in DMEM + FBS 10%, despite potential protein corona interactions (Figure 3C–E). As control conditions, no MNPs were added before applying the AMF.
*MNPs antibody functionalization and characterization. (A) Analysis of DMSA–MNPs@Ab binding. Quantification of the Alexa488 Gαm IgG recovered in the supernatant after the DMSA–MNPs were functionalized for 2 h with the 92R or IC antibodies (30 µg antibody/mg MNP) and the Alexa488 Gαm IgG antibody bound to the MNPs relative to the total antibody added. (B) The SAR of the different functionalized MNPs: DMSA–MNPs (gray), DMSA–MNPs@92R (blue), and DMSA–MNPs@IC (yellow) at 528 kHz and 20 mT in water. (C) Temperature increases after AMF application for 1 h (528 kHz–20 mT) to different MNPs samples at 500 µgFe/mL in DMEM + FBS 10%. Controls (green) were carried out without adding MNPs. (D) Temperature curves of the different MNPs (c = 500 µgFe/mL) in DMEM + FBS 10% under AMFs (528 kHz–20 mT, 1 h). Controls (green) were carried out without the addition of the MNPs. (E) Timing of hydrodynamic diameter (HD) measurements in DMEM + 10% FBS over 72 h for DMSA–MNPs (dark gray), DMSA–MNPs@Gly (light gray), DMSA–MNPs@92R (dark blue), and DMSA–MNPs@IC (yellow), as well as in water for DMSA–MNPs@92R (light blue). (n = 5: n.s. = not significant, *p < 0.05, **p < 0.01, and **p < 0.001).
Antibody binding to the MNP's surface is challenging, mainly due to the large size of the antibodies and the risk that their binding might inactivate the antibody (as would occur, if it bound to the MNPs through its variable and not its constant region). This problem has been addressed in various studies, clarifying patterns that should be reproduced to ensure an adequate union [45]. After optimization, the 92R antibody and its isotype control were seen to have bound correctly to the MNPs surface due to the prior activation of the carboxyl groups of DMSA, which produces attraction to the amino groups of the antibody. Negative and positive charges are distributed evenly throughout the antibody structure, yet its asymmetric structure causes the frontal plane of the antibody to have the highest net charge. As such, binding of the antibody's constant region to the MNPs is favored, leaving the sites involved in antigen recognition close to the surface of the MNPs but not involved in its electrostatic binding to the MNP, thereby preserving the antibody's activity [45]. In the range studied, the amount of antibody bound to MNPs depends on the incubation time (highest after 2 h: Figure S3A) and not on the amount of antibody available (Figure S3B). However, an adequate orientation preserving its activity does depend on the amount of antibody used (Figure S3C). These observations complement other studies where exposure to excess antibody was counterproductive to obtaining well orientated bonding [45].
Specific Binding of DMSA–MNPs@92R to MOLT‐4 (CCR9+) Cells
2.4
A CCR9 knock‐out cell line was generated from the MOLT‐4 T lymphoblast cell line (MOLT‐4 CCR9– or MOLT‐4 CCR9 KO) using CRISPR/Cas9, with CCR9 expression analyzed and the cells sorted by flow cytometry (Figure S4A and B). Flow cytometry experiments were carried out on the MOLT‐4 (CCR9^+^) and MOLT‐4 CCR9 KO (CCR9–) cell lines and while MOLT‐4 cells were specifically recognized by the 92R free antibody and not by the IC free antibody, the MOLT‐4 CCR9 KO cells were not recognized by either antibody (Figure S4C and D). These observations were corroborated by confocal microscopy, where patches of CCR9 (detected by 92R binding) at the cellular membrane were evident in Figure S4E, with higher concentrations in certain membrane areas.
DMSA–MNPs@92R (500 µg_Fe_/mL) were incubated for 2 h with MOLT‐4 and MOLT‐4 CCR9 KO cells, and the cell‐MNP interactions were studied. The number of MNPs bound to the cells was evaluated by ICP‐OES, revealing more MNPs bound to the cells when they were functionalized with antibodies. This increase was observed regardless of whether antibody functionalization involved the 92R or its IC antibody, and irrespective of whether the cells studied were MOLT‐4 or MOLT‐4 CCR9 KO (Figure 4). This phenomenon could reflect the increase in hydrodynamic diameter and PDI of the functionalized MNPs, which would make them noticeably less colloidally stable and facilitate their interaction with the cell membrane.
*ICP‐OES study of the DMSA–MNPs (150 µgFe/mL) attached to cells after 2 h. The controls involved following the same procedure but without adding the MNPs (n = 3: n.s. = not significant, **p < 0.01 and **p < 0.001).
Despite the similar number of MNPs attached, flow cytometry and confocal microscopy assays were carried out to assess whether there were differences in how these MNPs interact with the cells, and specifically to determine whether the interaction is more specific between DMSA–MNPs@92R and CCR9^+^ cells. While flow cytometry showed that some MNPs were attached in a nonspecific manner to MOLT‐4 and MOLT‐4 CCR9 KO cells, the MOLT‐4 cells exposed to DMSA–MNPs@92R produced a more intense Alexa488 Gαm IgG fluorescence signal, suggesting that there was a specific union between DMSA–MNPs@92R and MOLT‐4 cells in addition to the nonspecific union observed (Figure 5A,B). Confocal microscopy images showed that, regardless of their functionalization, MNPs accumulated around the cell membrane without any internalization. However, when MOLT‐4 cells were exposed to DMSA–MNPs@92R, a higher amount of antibody was detected at the membrane compared to DMSA–MNPs@IC, indicating increased antibody presence due to specific 92R–CCR9 interaction, rather than a higher number of nanoparticles attached (Figure 5C). The colocalization patterns of MNPs with the 92R antibody (Figure 5D) indicated that not all MNPs were correctly functionalized, such that a heterogeneous population of MNPs exists in which some were functionalized with antibodies and others were not.
*Flow cytometry and confocal microscopy analysis of the specific binding of DMSA–MNPs@Ab to different MOLT‐4 cell lines. (A) Flow cytometry data from MOLT‐4 (Top; blue edges) and MOLT‐4 CCR9 KO (bottom; red edges) cells incubated for 2 h with DMSA–MNPs (gray), DMSA–MNPs@92R (blue) or DMSA–MNPs@IC (yellow, 500 µgFe/mL). (B) Quantification of the geometric mean of the fluorescence intensity peak (n = 5, *p < 0.01). (C) Top panel: Confocal microscopy images resulting from z‐stacks of three central slices of MOLT‐4 cell lines incubated for 2 h with DMSA–MNPs, DMSA–MNPs@92R, and DMSA–MNPs@IC (500 µgFe/mL: scale bar = 20 µm). Upper right insets represent a magnification of single cells where the gray and green arrows mark the MNPs and antibodies, respectively. Bottom panel represents a z‐stack 3D reconstruction of 5 central slices of a 0.8 µm z‐step zooms of single cell. The nucleus is shown in blue (DAPI), the cell membrane in red (WGA‐555), the MNPS in gray (reflection), and the antibodies in green (Alexa488 Gαm IgG: scale bars = 5 µm). (D) Confocal microscopy images resulting from different image processing of the lower panel of Figure 5C to highlight the colocalization of MNPs and 92R. The nucleus is shown in gray (DAPI), MNPs in red (reflection), antibodies in green (Alexa488 Gαm IgG), and DMSA–MNPs@92R colocalization in yellow. The arrows indicate the intensity of colocalization: red, no colocalization; yellow, maximum colocalization. Controls were processed in the same manner but without MNPs.
The incubation of MOLT‐4 cells with DMSA–MNPs@92R demonstrated the stability of antibody functionalization under the culture conditions tested. Moreover, an enrichment of antibody‐functionalized MNPs was observed in the MNPs population, as the majority of membrane‐bound MNPs had the 92R antibody coupled. By contrast, when MOLT‐4 cells were exposed to DMSA–MNPs@IC, a smaller proportion of the antibody functionalized MNPs were bound to cells (see Figure 5C,D). Finally, the absence of specific binding of DMSA–MNPs@92R to MOLT‐4 CCR9 KO cells was confirmed by confocal microscopy (Figure S5).
Since the functionalization method is probabilistic and nonguided, the heterogeneity of the MNPs population could be explained by at least four possibilities (detailed in Figure S6). The first may be due to nonfunctionalized MNPs and the second to functionalized MNPs but with the antibody bound to the particle through both variable domains, masking the CCR9 recognition site, neither of which is desirable. The third option would be the binding of the (bivalent) antibody such that only one CCR9 recognition site is exposed and the last, is the ideal situation where antibody binding leaves both recognition sites exposed. The flow cytometry and confocal microscopy data demonstrated that the association of MNPs functionalized with antibodies to MOLT‐4 cells is enhanced specifically. After MOLT‐4 and MOLT‐4 CCR9 KO incubation with DMSA–MNPs@92R, the cells were further incubated with a biotinylated peptide containing the amino acid sequence of CCR9 specifically recognized by 92R (624B), which was detected with streptavidin‐Allophycocyanin (APC) and visualized by confocal microscopy (Figure S7). Image analysis demonstrated that the presence of an active antibody bound to MNPs was exclusively enhanced in MOLT‐4 cells where most MNPs attached to the cell membrane had an active antibody. Conversely, the MNPs bound to MOLT‐4 CCR9 KO cell membranes predominantly had an inactive antibody. These observations suggest that the specific interaction between DMSA–MNPs@92R and MOLT‐4 cells enhances the proportion of 92R‐functionalized MNPs attached to the cell membrane relative to their proportion in the initial MNPs sample (Figure S6), even though the total number of membrane‐bound MNPs did not differ significantly (Figure 4).
In summary, while the overall number of MNPs attached to the membrane of MOLT‐4 and MOLT‐4 CCR9 KO cells was similar, the proportion of these MNPs functionalized with active 92R antibodies relative to nonfunctionalized MNPs was significantly higher on MOLT‐4 cells. This confirms that, as 92R specifically recognizes CCR9, its presence on MNPs promotes the selective binding of these functionalized MNPs to cells, while reducing nonspecific NP attachment (Figure 5B). To gain deeper insight into the interaction between MNPs and cells, TEM analysis was performed (Figure 6A). The analysis of the nanoparticles clusters directly contacting with the cell membrane revealed distinct attachment patterns depending on the type of antibody used. In MOLT‐4 cells, 92R‐functionalized MNPs exhibited a significantly higher proportion of filamentous or chain‐like assemblies at the membrane compared with all other conditions. In contrast, isotype control‐functionalized MNPs displayed predominantly random or isotropic aggregates. This random distribution was also observed in MOLT‐4 CCR9 KO cells, regardless of the antibody used (Figure 6B and Figure S8).
*(A) TEM images of MOLT‐4 (upper panels) and MOLT‐4 CCR9 KO (bottom panels) cells incubated for 2 h with DMSA–MNPs (black; second column from the left), DMSA–MNPs@92R (blue; second column from the right), or DMSA–MNPs@IC (yellow; rightmost column) at a concentration of 500 µgFe/mL. Control conditions (cells without MNPs incubation) are shown in left column. Scale bar = 200 nm. Insets show detailed regions; scale bar = 40 nm. (B) Quantification of nanoparticle groups directly interacting with the cell membrane (based on TEM images) classified according to their spatial distribution as chain‐like assemblies (dark gray) or randomly aggregated assemblies (light gray) (n = 100–140). ***p < 0.001, **p < 0.01, p < 0.05, n.s. = not significant. Error bars indicate the standard error (SE). Distance‐to‐membrane analysis of (C) DMSA‐MNPs@92R and (D) DMSA‐MNPs@IC in MOLT‐4 cells.
Moreover, examination of the TEM images confirmed a notably higher accumulation of nanoparticles at the cell membrane when the MNPs were functionalized with antibodies, regardless of whether 92R or the isotype control was used. However, the spatial localization of the nanoparticles in close proximity to the cell membrane of MOLT‐4 cells differed between conditions. For DMSA‐MNPs@92R, most nanoparticles localized close to the membrane, with their number decreasing sharply beyond 200 nm (Figure 6C and Figure S8). Conversely, DMSA‐MNPs@IC displayed a uniform distribution, without an increased proportion near the cell membrane (Figure 6D and Figure S8). These findings suggest different binding mechanisms depending on the antibody. For 92R‐fucntionalized nanoparticles, the defined proximity to the membrane is consistent with antibody‐receptor interactions, whereas a nondefined distribution could reflect electrostatic interactions, where continuous binding and release of the nanoparticles prevents defined localization.
Collectively, these findings indicate that the specific interaction between 92R‐functionalized MNPs and CCR9 modulates the spatial organization of nanoparticles at the cell membrane. They also suggest that 92R functionalization may enhance the local retention of nanoparticles in a less aggregated disposition, potentially increasing their biological efficacy at the membrane interface.
A limitation of this study is the absence of real‐time analyses of nanoparticles binding to the cell membrane. However, previous reports have established that the kinetics of nanoparticle–cell interactions are strongly influences by the nature of the binding forces involved. Electrostatic interactions typically occur rapidly and nonspecifically, resulting in transient adsorption driven by surface charge density [48, 49]. In contrast, antibody–antigen recognition is characterized by high‐affinity, specific interactions with slower dissociation kinetics, leading to more stable and selective binding once equilibrium is reached [49]. Reported time scales suggest that interactions between antibody‐functionalized nanoparticles with their specific cellular antigen generally reach equilibrium well before 2 h under physiological conditions [50]. Therefore, the 2 h incubation period employed in this study is sufficient to ensure effective binding of DMSA–MNPs@92R to CCR9‐expressing cells. This supports the validity of the experimental observations and demonstrates that the ability of the antibody‐functionalized system to interact with the receptor is preserved despite protein corona formation under physiological conditions.
The combination of specific 92R binding to CCR9 and of nonspecific binding of MNPs to the cell surface would explain the higher number of DMSA–MNPs@92R interacting with the cell membrane of MOLT‐4 (CCR9^+^) compared to MOLT‐4 CCR9 KO cells. Microscopy confirmed that the specific interaction of DMSA–MNPs@92R with CCR9 at the surface of MOLT‐4 cells did not induce MNPs internalization, nor did the nonspecific interaction of DMSA–MNPs with the cell membrane, in line with our previous studies on the interaction of MNPs with T cells [23, 41]. Nevertheless, in other cell types MNPs can be internalized by different mechanisms [42, 46], and the functionalization of MNPs with the 92R antibody could affect the uptake and subcellular localization of the MNPs. To explore this possibility, we performed experiments on other CCR9^+^ cell lines that intrinsically internalize MNPs, such as HEK293T cells expressing recombinant human CCR9 (HEK293T–CCR9). These analyses showed that a non‐specific interaction between DMSA–MNPs@92R and HEK293T (CCR9–), and between DMSA–MNPs@IC and HEK293T–CCR9 cells occurred, whereas a specific interaction was only observed between HEK293T–CCR9 cells and DMSA–MNPs@92R (Figure S9).
Biological Effects of In Vitro MHT on CCR9+ Leukemia Cells
2.5
MHT produces a wide variety of biological effects at the cellular level, tumor cell death being the most desirable. These effects can be classified into different groups according to their main cause [21], several of which are triggered by the rise in temperature (hyperthermia), including: enhanced oxidative stress, membrane receptor inactivation and increased ion permeability, membrane instability and an increase in membrane fluidity, altered cytoskeletal organization, enhanced protein denaturation and aggregation of insoluble proteins in the nucleus (which in turn promote heat shock protein expression and centrosome damage), and ultimately, DNA damage [51, 52, 53]. By contrast, some effects are not related to increases in temperature, such as lysosome permeabilization or rupture due to mechanical rotation or MNPs vibration, and consequently an increase in cytoplasmic reactive oxidative species (ROS) and cathepsin D [54, 55, 56]. Another effect that is exclusively related to MNPs is the intensification of intracellular reactive oxygen species (ROS) formation through Fenton reactions [42, 57]. Finally, a secondary effect of MHT is immune system activation, reflected in the activation of dendritic and NK cells [58]. Furthermore, an increase in temperature at the cellular or subcellular level, as opposed to the macroscopic level, may be provoked by a local hot–spot effect of MNPs, whereby local heat release enhances all the aforementioned biological effects directly related to hyperthermia [59, 60]. Eventually these biological effects provoke cell death, which will be investigated as the primary indicator of MHT efficacy in tumor cells treated with MNPs.
Suitability of MOLT‐4 for In Vitro MHT Experiments and Their Intrinsic Response to Heat
2.6
Cell susceptibility to heat and their response to thermal shock is an important issue that needs to be established before performing MHT experiments in vitro. MOLT‐4 cells were exposed to temperatures between 37 and 45°C for 1 h, and their subsequent viability was evaluated using the Presto Blue reagent up to 24 h later. These experiments confirmed that temperatures of 42°C did not significantly affect the 24 h survival of MOLT‐4 cells, whereas at 45°C, approximately 50% of cell death was observed within the first 2 h following heat exposure (Figure S10A and B). Hence these cells showed a difference in behavior over a narrow range of temperatures. Elsewhere, maintaining cells at 44°C for only 10 min produced 25% apoptosis, with the participation of caspase 8 and Fas, reflecting a clear dependence on temperature of cell death [61]. Moreover, clear overexpression of the Hspa1b gene, which encodes for the heat shock protein family A (Hsp70) member 1B [62], peaked at 2 h after a heat shock of 42°C, reverting to basal levels after 4 h (Figure S10C). However, after only 1 h at 45°C the expression of Hsp70 increase drastically for up to 24 h later, in conjunction with extensive cell death and the incapacity to control the heat shock (Figure S10C). Indeed, an upregulation in transcriptional activity that provokes an increase in Hsp70 protein that remains evident for up to 12 h following heat shock was reported previously [61].
MOLT‐4 Cell Death as a Result of MHT through DMSA–MNPs@92R
2.7
It has been demonstrated that DMSA–MNPs are promising candidates for MHT (Figure 1) and that functionalization with antibodies preserves their therapeutic potential by maintaining efficient heat release (Figure 3), while also enabling specific binding to CCR9^+^ cells (Figure 5). Thus, the effect of 1 h in vitro MHT induced by AMFs (528 kHz, 20 mT) was assessed through the viability of MOLT‐4 cells previously exposed to DMSA–MNPs@92R (500 µg_Fe_/mL) for 2 h (DMSA–MNPs@92R + AMF). In addition, several experimental controls were used to validate the results including a negative control of cells incubated at 37°C (Ctrl 37°C) (see “MHT experiments in vitro,” in “Experimental Section”).
By exposing the cells directly to AMFs (528 kHz, 20 mT for 1 h: Ctrl AMF), its effect on cell viability was assessed, and as control for conventional temperature increase, cells were maintained for 1 h at 42 and 45°C (Ctrl 42 and Ctrl 45°C). To evaluate the DMSA–MNP's biocompatibility, cells were incubated with DMSA–MNPs, DMSA–MNPs@92R, and DMSA–MNPs@IC for 2 h (500 µg_Fe_/mL: DMSA–MNPs, DMSA–MNPs@IC, and DMSA–MNPs@92R); and to demonstrate the crucial role of 92R on in vitro therapeutic efficacy, MOLT‐4 cells were incubated for 2 h with DMSA–MNPs, DMSA–MNPs@IC, and DMSA–MNPs@92R (500 µg_Fe_/mL) and then subjected to AMF (528 kHz, 20 mT) for 1 h (DMSA–MNPs + AMF, DMSA–MNPs@IC + AMF and DMSA–MNPs@92R + AMF). Finally, to confirm the specificity of the therapy in MOLT‐4 (CCR9^+^) cells, the experiments were repeated on MOLT‐4 CCR9 KO (CCR9^−^) cells. Temperature was monitored meticulously during the in vitro MHT experiments, and it increased to approximately 41°C due to the response of DMSA–MNPs@Ab to the AMF, reaching 44°C with DMSA–MNPs while it remained at 37°C in the AMF control. This corresponds to a temperature increase of 4.5–5.5°C for DMSA–MNPs@Ab + AMF, 7.5°C for DMSA–MNPs + AMF and 1.5°C for Ctrl‐AMF relative to the basal temperature of 35.5–36.0°C (Figure 3C,D).
MOLT‐4 and MOLT‐4 CCR9 KO cell death was evaluated 24 h after treatment using flow cytometry with the Annexin V–PI assay. First, only the application of AMFs (Ctrl AMF) did not induce cell death, maintaining a viability of approximately 92% that was similar to the negative control (Ctrl 37°C: Figure 7A). Regarding thermotolerance, incubation at 42°C for 1 h caused a slight increase in cell death (up to 12%: Ctrl 42°C), while incubation at 45°C for 1 h significantly increased cell mortality to approximately 65%, which may be indicative of cells progressing toward late apoptosis (Figure 7A and Figure S11 and Table S2). These results highlight the sensitivity of MOLT‐4 cells to relatively minor temperature variations within the HT therapeutic window.
*Cell death provoked by in vitro MHT and assessed by Annexin V—PI flow cytometry of (A) MOLT‐4 (CCR9+) and (B) MOLT‐4 CCR9 KO (CCR9–) exposed to DMSA–MNPs, DMSA–MNPs@IC, and DMSA–MNPs@92R (500 µgFe/mL) for 2 h. Stippled bars indicate AMFs applied (530 kHz–20 mT). Numbers in boxes indicate the temperature reached during the experiment. Hspa1b overexpression assessed by qPCR‐RT (C) 2 and (D) 24 h after AMF application (n = 3–5: *p < 0.05; *p < 0.01; d = Cohen's effect size).
The increase of toxicity using high doses of DMSA–MNPs was also confirmed by flow cytometry, as incubation at 500 µg_Fe_/mL for 2 h resulted in cell death exceeding 30% (DMSA–MNPs). However, antibody functionalization of the DMSA–MNPs improved cell biocompatibility, reducing cell death to approximately 15% at 500 µg_Fe_/mL for 2 h (Figure 7A). DMSA possesses inherent cytotoxicity due to its exposed carboxyl and sulfhydryl groups, yet blocking these reactive groups by antibody binding, chemical blocking or steric hindrance dampens their toxicity. Here, antibody functionalization followed by further glycine blocking effectively reduced MNPs cytotoxicity by half (DMSA–MNPs@92R and DMSA–MNPs@IC in Figure 7A and Figure S12). In addition to the intrinsic effects attributed to specific recognition, antibody functionalization of MNPs also increases their biocompatibility, making the MNPs safer for therapeutic applications. This functionalization limits the side effects associated with toxicity in areas of the body where MNPs might accumulate nonspecifically. This is particularly important when high doses of MNPs are required, as reported in the clinical trials conducted to date. Furthermore, this reduced toxicity enables higher MNPs doses to be administered, which could enable higher temperatures to be reached in certain regions of the body, despite a minor reduction in the system's capacity to generate heat due to functionalization.
When considering MNPs application followed by AMF application, treatment with DMSA–MNPs increased the temperature to 43–44°C (Figure 3D), resulting in approximately 55% cell death after 24 h, likely due to an increase in early and late apoptosis (DMSA–MNPs + AMF; Figure 7A and Figure S11). Thus, the observed cell death can be attributed to the macroscopic temperature increase, as conventional heating at 42°C (Ctrl 42°C) and 45°C (Ctrl 45°C) led to 12% and 65% cell death, respectively. The cell death (55%) and macroscopic temperatures (43–44°C) reached by DMSA–MNPs + AMF treatment fall within this range, supporting the conclusion that thermal effects can explain the observed cytotoxic effects. Conversely, treatment with DMSA–MNPs@IC increased the temperature only to 40–41°C (Figure 3D), producing approximately 45% cell death (DMSA–MNPs@IC + AMF: Figure 7A). This enhanced cytotoxicity suggests that magnetic hyperthermia contributes to cell death through mechanisms beyond a simple temperature increase, as conventional heating treatment at 42°C (Ctrl 42°C) induced only 12% cell death, representing a 33% lower effect despite reaching 1–2°C higher. Notably, cell death due to MHT exceeded 70% when DMSA–MNPs@92R were combined with AMFs (DMSA–MNPs@92R + AMF: Figure 7A). In this case, the temperature reached 40–41°C (similar to DMSA–MNPs@IC) but the observed cell death was approximately 25% higher. In fact, a substantial effect on cell death was observed in MOLT‐4 cells treated with DMSA–MNPs@92R + AMF, indicating a significantly increased efficacy compared to the DMSA–MNPs@IC + AMF treatment (Cohen´s d = 1.36). This suggests the therapeutic potential of specific DMSA–MNPs@92R binding to MOLT‐4 cells for efficient MHT. Moreover, the contribution of cell death mechanisms beyond a simple temperature increase was clearly evident as conventional macroscopic heating up to 42°C (Ctrl 42°C) resulted in 58% lower cell death despite reaching a temperature 1–2°C higher.
When MOLT‐4 CCR9 KO cells were employed, macroscopic temperature increases and MNPs biocompatibility resulted in similar outcomes to those observed for MOLT‐4 cells (Ctrl AMF, Ctrl 42°C, Ctrl 45°C, and DMSA–MNPs: Figure 7B). However, MOLT‐4 CCR9 KO cell death was approximately 45% when DMSA–MNPs@92R were used in combination with AMFs (DMSA–MNPs@92R + AMF: Figure 7B), about 25% lower than in MOLT‐4 (CCR9^+^), but similar to that observed in MOLT‐4 cells when DMSA–MNPs@IC were combined with AMFs. This suggests that MOLT‐4 CCR9 KO cell death is caused by MHT but that it is not enhanced by the specific interaction between MNPs and the CCR9 receptors.
These findings demonstrate that therapeutic efficacy in vitro was exclusively enhanced when DMSA–MNPs@92R were used in combination with AMFs, specifically in MOLT‐4 (CCR9^+^) cells, highlighting the critical role of the MNP–cell interaction in achieving successful MHT. Here, most MNPs specifically attached to the cell membrane in a less‐aggregated manner via 92R–CCR9 binding, facilitating efficient energy dissipation as heat. The cell death observed cannot be solely attributed to macroscopic temperature increases, as exposure to 45°C resulted in lower cell death than that provoked at 41–42°C when DMSA–MNPs@92R were exposed to AMFs. Therefore, cell death is likely driven by local temperature elevations on a microscopic or subcellular level, potentially enhanced by additional effects of the MNPs response to AMFs that would explain the biological effects observed, such as a local heating, also known as hot‐spot effect.
Temperature measurements at the macroscopic scale for in vivo experiments can be performed using various approaches, such as infrared thermal imaging [63]. However, these techniques lack the spatial resolution required to accurately assess temperature variations at the local or cellular level. Temperature measurements employing molecular thermometers could help confirm that the mechanism triggered by the MNPs was a local hot‐spot effect. Thus, MNPs functionalized with fluorescent molecules were used to demonstrate that after the application of AMFs, the temperature at the MNPs surface reached 51°C, while no macroscopic temperature variation occurred [55]. Another approach is to quantify HSP70, which is overexpressed in response to thermal shock as a protective mechanism [64]. To evaluate the role of a local hot‐spot effect in the cell death observed, Hspa1b overexpression was quantified by real‐time qPCR‐RT. Hspa1b expression differed significantly when cells were exposed to AMFs in the presence of DMSA–MNPs or DMSA–MNPs@92R. In the presence of the former, a temperature increases to 43–44°C was induced and Hspa1b expression closely resembled that observed in cells exposed to 42°C. However, while a lower temperature was induced in the presence of DMSA–MNPs@92R (40–41°C), Hspa1b expression was similar to that of cells exposed to 45°C (Figure 7C). Nevertheless, the expression of Hspa1b 24 h after treatment indicated that exposing cells to a macroscopic temperature increase up to 45°C was the only condition in which Hspa1b mRNA expression was markedly upregulation (Figure 7D). Hence, the Hspa1b expression provoked by MHT not only depended on a macroscopic temperature rise, suggesting that alternative mechanisms induce cellular local heating, such as a hot‐spot effect.
To investigate whether oxidative stress could be responsible for the observed cell death, we evaluated ROS generation through MNPs incubation and magnetic hyperthermia treatment. The results indicated that cell death was not associated with oxidative stress, as DMSA–MNPs did not induce any significant increase in ROS levels. This outcome is likely due to the fact that the nanoparticles remain in the extracellular space without internalization, in contrast to what it observed in other cell lines such as MDA‐MB‐231, where DMSA–MNPs are internalized and ROS levels increased (Figure S13). Comparable findings have been reported in other cell lines exposed to DMSA‐coated iron oxide nanoparticles [42, 65].
To summarize, while the number of MNPs attached to MOLT‐4 or MOLT‐4 CCR9 KO cell membranes remained unchanged after incubation with DMSA–MNPs@92R or DMSA–MNPs@IC (Figure 4), the proportion of MNPs correctly functionalized with the 92R antibody was enriched. Hence, there is a specific increase in MNPs binding to CCR9^+^ cells through the CCR9–92R interaction (Figure 5), which results in greater MHT efficiency. Consequently, a 25% increase in MOLT‐4 cell (CCR9^+^) death was observed after exposure to DMSA–MNPs@92R and an AMF than when they were exposed to DMSA–MNPs@IC, and relative to MOLT‐4 CCR9 KO (CCR9–) cells incubated with DMSA–MNPs@92R (Figure 7A,B). The specific binding of DMSA–MNPs@92R to CCR9 influences the interaction between the MNPs and cells, potentially arranging the MNPs in a specific manner and distance from the cell membrane. As opposed to nonspecific binding, this specific MNP–cell interaction could enhance the efficiency of heat dissipation by preserving the effective anisotropy and facilitating Brownian relaxation, avoiding dipolar interactions between magnetic nuclei that could provoke MNPs aggregation. Although this specific binding does not produce a measurable increase in macroscopic temperature, it is likely to enhance local thermal damage to the cell membrane (see Figure 8).
Scheme of the local heat release during MHT. In vitro experiments on MOLT‐4 and MOLT‐4 CCR9 KO cells using a strategy based on MNPs functionalization with the 92R antibody. (A) DMSA–MNPs@92R/ MOLT‐4 (CCR9+) interaction: specific binding of the MNPs to cells promotes maximal local heat release. (B) DMSA–MNPs@92R/MOLT‐4 CCR9 KO (CCR9–) interaction: nonspecific binding between MNPs and cells resulting in dampened local heat release.
Specific MNP–cell membrane recognition has been shown to enhance MHT, even when heterogeneous MNPs populations are used in which not all MNPs are properly functionalized and bound in the same way to cells. For this reason, optimizing the functionalization and purification of the MNPs to obtain a homogeneous population of 100% correctly functionalized particles will enhance biocompatibility, reducing the toxicity associated with unblocked reactive groups of DMSA on nonfunctionalized MNPs. Simultaneously, the effectiveness of the therapy can be improved by enhancing specific MNPs binding to the cell membrane. When considering therapies involving intratumoral administration (as in the clinical trials carried out), the specific binding of CCR9–MNPs@92R would enhance their retention by the target tissue, preventing clearance and ensuring optimal MNPs concentrations in the tumor. This would enable MHT to be repeated over a longer period of time without the need for additional injections.
It is also worth highlighting that MNPs functionalization with a monoclonal antibody specific to the CCR9 cell membrane receptor could pave the way to develop alternative immunological therapies based on specific CCR9–92R interactions. First, 92R mediates complement‐dependent cytotoxicity in vitro, inducing MOLT4‐leukemia cell lysis due to complement fixation and antibody‐dependent cytotoxicity by NK cells. However, 92R functionalized MNPs (as demonstrated with the DMSA–MNPs@92R controls: Figure 7A and Figure S12) or 92R by itself does not directly induce cell apoptosis [29, 34]. Second, in mouse models the 92R antibody inhibits the growth of human CCR9^+^ xenograft tumors [29, 34]. Similarly, treatment with 92R produces a major decrease in MOLT‐4 cell number in a mouse model of orthotopic xenotransplants, demonstrating the specificity of the treatment through a lack of effect on the CCR9^−^ Jurkat leukemia cell line [30]. Similar results were obtained with nonhematopoietic CCR9^+^ tumors, producing smaller tumors of HEK293T cells (a human embryonic kidney cell line) transfected with recombinant CCR9 or lower weights of AsPC‐1 (human pancreas adenocarcinoma cell line, CCR9^+^) tumors from subcutaneous xenotransplants [30]. Finally, in the context of human patients, a negative correlation between the presence of CCR9 and patient survival was demonstrated for different hematopoietic and nonhematopoietic tumors [30]. Therefore, CCR9 has been proposed as a risk factor for tumor infiltration or invasive tumors, and the capacity of 92R to inhibit tumor infiltration would open the door to treatments in patients with advanced acute lymphoblastic leukemia that develop infiltrations in the CNS [30]. Taking this into account, it can be expected that the 92R antibody bound to MNPs may exert a therapeutic effect alone in vivo or in applied therapy contexts. Therefore, synergy between immunotherapy and MHT could occur, increasing the efficacy of anti‐tumor therapies and reducing the side effects associated with conventional treatments like systemic chemotherapy.
Finally, these results open the possibility of scaling up the proposed DMSA–MNPs@92R–CCR9 binding strategy as a therapy for other CCR9^+^ tumors. The functionalization of MNPs with other antibodies specific to receptors with a low‐internalization rate in different tumors could be explored, or with antibodies that promote specific localization of MNPs within selected sites of the cell. This approach may help prevent phenomena such as blocking Brownian relaxation, MNPs aggregation, lysosomal uptake or other events that dampen heat release.
Conclusions
3
The microwave‐assisted synthesis of 12 nm IONPs coated with DMSA (DMSA–MNPs) enabled us to demonstrate their potential as agents for MHT. These MNPs exhibit high SARs, which remain high over extended periods in biological media. Heat generation primarily occurs through Néel relaxation, further supporting their suitability for therapeutic applications. The functionalization of these MNPs with the 92R monoclonal antibody and with the isotype control antibody (DMSA–MNPs@92R and DMSA–MNPs@IC, respectively) maintained their heat‐release capacity above 75% in biological media. Moreover, a CCR9 knock‐out MOLT‐4 variant (MOLT‐4 CCR9 KO) was generated and specific binding of DMSA–MNPs@92R to MOLT‐4 cells (via the 92R–CCR9 interaction) was confirmed by flow cytometry and confocal microscopy, with MNPs specifically attached to the cell membrane, despite a certain degree of nonspecific cell membrane attachment of DMSA–MNPs@IC to MOLT‐4 cells and DMSA–MNPs@92R to MOLT‐4 CCR9 KO cells. This specific binding was essential to enhance in vitro cell death induced by MHT, with a 25% increase relative to DMSA–MNPs@IC and a 15% increase relative to nonfunctionalized DMSA–MNPs, despite reaching a lower macroscopic temperature increase (approximately 4°C lower). These findings highlight the critical role of controlled MNP–cell interactions in the effectiveness of MHT. Moreover, a local hot‐spot effect was proposed to trigger cell death, as suggested by the significant increase in cell death despite a smaller macroscopic temperature rise (up to 40–41°C) compared with direct exposure to 42°C. This effect was further supported by the observed overexpression of Hspa1b (HSP70 mRNA). Thus, the lower temperature reached, along with the enhanced local heat effects (as opposed those of macroscopic temperature rises), could represent an approach producing milder side effects on healthy cells surrounding the tumor.
The functionalization of DMSA–MNPs using the monoclonal 92R antibody (DMSA–MNPs@92R) represents a robust system with significant potential for MHT, offering several key advantages. These include: enhanced safety due to the increased biocompatibility resulting from the blocking of reactive DMSA groups; specificity due to the recognition of CCR9^+^ cells (e.g. the MOLT‐4 cell line); specific localization of MNPs on the CCR9^+^ cell membrane at an appropriate distance, minimizing aggregation and other undesirable events, and thus enhancing heat release; high affinity ensuring a strong bond between MNPs and the CCR9^+^ cells, thereby reducing clearance following clinical intratumoral injection; and synergy, combining immunotherapy and MHT for more effective therapies. Collectively, these advantages could position this system as a promising alternative to more aggressive and less specific treatments, such as systemic chemotherapy, paving the way for more personalized therapeutic approaches with potentially milder side effects, thereby enhancing treatment efficacy and patient safety.
Experimental Section
4
Materials
4.1
Iron(II) acetate (Fe(ac)2, 98%), diethylene glycol (DEG, 99%), nitric acid (HNO_3_, 65%), iron(III) nitrate (Fe(NO_3_)3, 98%), ethanol (99.5%), acetone (99.5%), meso‐2,3‐dimercaptosuccinic acid (DMSA, 98.0%), Potassium hydroxide (KOH, 85%), 4‐Morpholineethanesulfonic acid (MES, 99%), N‐(3‐Dimethylaminopropyl)‐N′‐ethylcarbodiimide (EDC, 97%), N‐Hydroxysuccinimide (NHS, 98%), Glycine (Gly, 99%) and poly‐L‐Lysine (poly‐Lys), Phosphate buffered saline (PBS: 150 mm NaCl, 10 mm Na_2_HPO_4_ [pH 7.4]), bovine serum albumin (BSA), PBS‐B (PBS with 0.5% BSA), 1% FBS and 0.1% sodium azide were all purchased from Sigma‐Aldrich. Goat F(ab´)2 Anti‐mouse IgG (H+L)‐PE (diluted 1:300, 0.25 mg/mL) was purchased from Southern Biotech (Cat. 1032‐09, Lot A2418‐X398B. Exp. 2020‐01). The mouse monoclonal 92R antibody and its isotype control (IC) were generated by our group, and the biotinylated 624B peptide was generated at the Proteomics facility at the CNB.
MNPs Synthesis
4.2
The synthesis of 12 nm quasispherical IONPs was carried out by the polyol method assisted by microwaves (Monowave 300, Anton Paar) and based on the protocol proposed previously [35]. The microwave apparatus operates at a frequency of 2.45 GHz and is equipped with magnetic stirring, temperature control by optic fiber and a manometer. Briefly, 300 mg of Fe(ac)2 (1.72 mmol), 18.3 mL of DEG (0.33 mmol), and 0.7 mL of distilled H_2_O (0.04 mmol) was mixed in a G30 vial, homogenized by vortex and ultrasound (Elmasonic S30, Elma Schmidbauer GmbH, Singen, Germany), and introduced into the microwave. For synthesis, a constant stirring of 600 rpm was selected and the temperature rose 170°C over a 3.75°C/min heating ramp, retaining the reaction at 170°C for 2 h.
Finally, the sample was cooled at room temperature, washed 5 times with ethanol and the MNPs were recovered by centrifugation for 20 min at 7500 rpm (Sigma Centrifuge 4−15, rotor 12165‐H), obtaining around 100 mg of MNPs from each vial. The MNPs were then acid treated [66], adding 8.6 mL of HNO_3_ (2 m) and stirring constantly at 600 rpm for 15 min. The HNO_3_ was removed by magnetic decantation, and 2.14 mL of Fe(NO_3_)3 (1 m) and 3.71 mL of distilled H_2_O was added. The mixture was boiled for 30 min with constant stirring at 600 rpm and after cooling at room temperature, the solvent was removed by magnetic decantation and 8.6 mL of HNO_3_ (2 m) was added, stirring again for 15 min at 600 rpm. After removing the acid and washing the sample 3 times with acetone, the MNPs were recovered by magnetic decantation and finally, traces of acetone were eliminated in a rotavapor and the sample was resuspended in distilled H_2_O.
MNPs Core Characterization
4.3
The particle size, shape and distribution were determined by TEM. Images were captured on a 100 keV JEOL‐JEM 1010 microscope equipped with a Gatan Orius 200 SC digital camera working at an acceleration voltage of 60 kV. The mean particle size and size distribution were evaluated by measuring the largest internal dimension of at least 300 particles with the ImageJ software (NIH, USA), followed by data fitting to a log‐normal distribution. The crystal structure of the particles was identified by XRD performed using a Bruker D8 ADVANCE diffractometer with Cu Kα radiation, scanning between 2θ values ranging from 10° to 70°. The crystal size of MNPs was calculated using the Scherrer equation
The iron concentration of the magnetic colloids was measured by elemental ICP−OES (Plasma Emission Spectrometer ICP PerkinElmer mod. OPTIMA 2100 DV, PerkinElmer, Waltham, MA). The samples (20 µL) were digested overnight in 1 mL of HCl at 90°C and diluted to a volume of 20 mL with Milli‐Q water. Magnetic behavior was analyzed at room temperature as a function of the magnetic field (±2.5 T) using a vibrating sample magnetometer (VSM: MLVSM9 MagLab 9T).
MNPs Surface Coating with DMSA
4.4
DMSA–MNPs were obtained following a previously reported procedure [39]. and based on earlier studies [67]. Briefly, DMSA (5 mg, 0.027 mmol) was added to 10 mL of MNPs solution (4.3 mg/mL in water [pH 3]) and stirred for 10 min. The pH was then raised to 10.5 with KOH and the sample was homogenized by ultrasound for a further 15 min. To remove the unreacted DMSA and other impurities, the sample was dialyzed for 3 days in distilled water through a cellulose membrane tubing (typical molecular weight cut‐off = 14,000 Da), with periodic water changes. Finally, the solution was adjusted to pH 7, concentrated using Amicon filter tubes (MWCO 10 kDa) and filtered through a 0.22 µm pore size membrane.
DMSA–MNPs Characterization
4.5
Colloidal characterization was performed by dynamic light scattering (DLS) on a ZetaSizer Nano ZS (Malvern). The hydrodynamic diameter (in intensity) and ζ‐potential (ζ‐pot) at pH 7 were obtained from three different measurements at 25°C. Infrared spectra were recorded in KBr pellets (2% wt of the sample) on a Bruker Vertex 70 V spectrophotometer.
DMSA–MNPs Antibody Functionalization and Binding Verification
4.6
DMSA–MNPs were functionalized with the 92R monoclonal antibody and its IC, adapting an earlier protocol [45]. First, DMSA–MNPs (10 mg_Fe_) were washed in MES buffer (10 mm, pH 6.1) and concentrated to a final concentration of 10 mg_Fe_/mL by centrifugation (1800 rpm, 10 min) with Amicon filter tubes (MWCO 100 kDa). Then, 0.5 mL of EDC (1.92 mg/mL) and 0.5 mL of NHS (1.72 mg/mL) were added (0.096 mg EDC/mg_Fe_ MNPs, 0.086 mg NHS/mg_Fe_ MNPs), and the sample was maintained for 15 min in a rotating wheel at room temperature. The activated MNPs obtained were then washed 3 times with MES buffer (10 mm, pH 5.4) and concentrated to 1 mL. Subsequently, 1 mL of antibody (288 µg/mL) was added (28.8 µg antibody/mg_Fe_ MNPs) and the mixture was maintained in an orbital shaker (500 rpm) at 37°C for 2 h. After incubating with the antibodies, the active sites of the MNPs were blocked in 2 mL of Gly (50 mm) for 16 h, after which excess Gly and other impurities were eliminated by washing in MES buffer (10 mm, pH 6.1), recovering the MNPs by centrifugation (1800 rpm, 10 min) in Amicon filter tubes (MWCO 100 kDa). Finally, the sample was filtered using a 0.45 µm pore size membrane in a laminar flow hood to ensure sterility. The attachment of antibodies to activated DMSA–MNPs was assessed using a secondary Alexa 488 conjugated goat anti mouse IgG (H + L) antibody (Gαm IgG Alexa488), which recognizes both the 92R and IC antibodies. Nonactivated DMSA–MNPs or glycine blocked MNPs (DMSA–MNPs@Gly) did not bind to the secondary antibody, while this antibody did bind to activated DMSA–MNPs (DMSA–MNPs@92R and DMSA–MNPs@IC).To evaluate MNPs antibody binding, MNPs (100 µg_Fe_) were dispersed in 100 µL of an Alexa 488 conjugated goat anti‐mouse antibody (Alexa488 Gαm IgG (H+L): Thermo Fisher) at 10 µg/mL in PBS‐B and incubated for 1 h. The sample was then washed and recovered by centrifugation (10,000 rpm, 12 min) several times. The presence of Alexa488 Gαm IgG in the supernatant and pellet was detected through fluorescence on a Spectramax iD3 fluorimeter (λ excitation = 485; λ emission = 525 nm) and quantified through prior elaboration of a straight calibration line between 0 and 20 µg/mL.
Moreover, the in silico MNP–92R binding was assayed incubating for 40 min on a rotating wheel with the 624B peptide (0.25 µg/mL in PBS‐B), a biotinylated peptide that specifically binds to the 92R antibody. The sample was then washed several times, recovered by centrifugation (10,000 rpm, 12 min), and Streptavidin‐APC (0.5 µg/mL: ThermoFisher) was added and incubated for 40 min on a rotating wheel. The sample was then washed and recovered by centrifugation (10,000 rpm, 12 min), and APC fluorescence was measured on a Spectramax iD3 fluorimeter (λ excitation = 630; λ emission = 670 nm).
MNPs Heating Efficiency (SAR)
4.7
The SAR of MNPs in water under an AMF was analyzed in Eppendorf tubes containing 0.25 mL of the sample (1 mg_Fe_/mL). The equipment used was a MagnetTherm (NanoTherics) equipped with a water recirculation‐based cooler, a direct current power supply, a wave‐generating device, an oscilloscope, an electromagnetic coil, interchangeable capacitors (6.2–200 nF), an optical fiber for measuring temperature (range −40–200°C) and software to control the different parameters. The eppendorf tube was located inside a polystyrene cylinder cavity, which guaranteed a fixed position at the center of the magnetic coil and thermal isolation. The temperature change was measured as a function of time (dT/dt), and the linear slope between the initial 10 and 40 s was used to evaluate the heating efficiency in terms of the SAR as power dissipation per unit mass of an element (W/g), using the following formula
where *C_v_
- is the specific heat capacity of water (4185 J/kg·K) and m Fe is the iron content per unit mass of the MNPs.
MOLT‐4 CCR9–Knock‐Out (MOLT‐4 CCR9 KO)
4.8
The MOLT‐4 CCR9–knock‐out (MOLT‐4 CCR9 KO) was generated using a strategy described previously based on CRISPR/Cas9 [68]. Gene editing and cell electroporation were carried out at the Molecular Cytogenetics Unit of the Spanish National Cancer Research Center (CNIO). Briefly, single guide RNAs (sgRNAs) were designed using the Benchling CRISPR Guide RNA Design Tool (https://www.benchling.com). Specific sgRNAs were generated to target the CCR9 gene (ENSG00000173585, exon 3) and sgAAVS1 was used as a control guide: sgCCR9.1, GTACTGGCTCGTGTTCATCG; sgCCR9.2, CCTAACATGGCTGATGACTA; and sgAAVS1, CCTCTAAGGTTTGCTTACGA. Ribonucleoprotein complexes were formed over at least 10 min at room temperature using in vitro synthesized guide sequences and the Cas9 protein (Integrated DNA Technologies, IDT). MOLT‐4 cells were electroporated with 3 × 10 ms pulses at 1350 V using the Neon Transfection System (Thermo Fisher Scientific) and after electroporation, the cells were maintained in culture without antibiotics for 18 h. The loss of CCR9 expression at the MOLT‐4 cell surface was analyzed by flow cytometry at 18 and 187 h after electroporation. The cell population previously treated with sgCCR9 was cultured for 2 weeks and then selected by flow cytometry (Cell Sorter Beckman Coulter Moflow XDP). The population of MOLT‐4 CCR9^−^ cells was collected and cultured, and then subjected to limiting dilution cloning to obtain a stable CCR9 knock‐out MOLT‐4 cell line (MOLT‐4 CCR9 KO). Finally, flow cytometry was carried out as follow. Cells were preincubated in PBS with 1% goat serum for 10 min at 4°C to block nonspecific binding and the cells were then incubated with the 92R anti‐CCR9 monoclonal antibody or an IC mouse IgG2a for 45 min at 4°C. The cells were then washed with PBS and subsequently labeled with a Goat F(ab’)2 anti‐mouse IgG (H+L)‐PE antibody (30 min, 4°C, in the dark). The samples were finally analyzed on a Beckman Coulter Cytomics FC500 cytometer and the data were processed with the Kaluza analysis software (Beckman Coulter).
Cell Culture
4.9
Both the MOLT‐4 and MOLT‐4 CCR9 KO cell lines were cultured in High glucose DMEM supplemented with 10% FBS, 1% nonessential amino acids, 2 mm l‐Glutamine, 1 mm Sodium Pyruvate, 100 IU/mL Penicillin and 100 mg/mL streptomycin. The cell cultures were maintained between 3.5 × 10^5^ and 2.0 × 10^6^ cells/mL at 37°C in an atmosphere of 5% CO_2_ and at 90% relative humidity.
TEM Imaging of the Cells
4.10
To confirm the subcellular distribution and attachment of MNPs, cells were seeded in petri dishes for 24 h and then treated for 2 h with the different MNPs samples (500 µg_Fe_/mL). After treatment, the free MNPs were removed by washing with PBS and the cells were then fixed for 2 h at room temperature with 2% glutaraldehyde and 1% tannic acid diluted in 0.4 m HEPES [pH 7.2]. After fixation, the cells were recovered by centrifugation and resuspended in HEPES buffer, postfixed at 4°C in 1% osmium tetroxide (1 h) and 2% uranyl acetate (30 min), and then dehydrated in a series of acetone solutions before gradually infiltrating them with Epon resin. The resin was allowed to polymerize (60°C, 48 h) and ultrathin sections (60–70 nm) were obtained with a diamond knife mounted on a Leica EM UC6 ultramicrotome. The sections were finally attached to a formvar/carbon‐coated gold grid and visualized on a JEOL 1400Flash transmission electron microscope coupled to a Gatan OneView Camera.
TEM Image Analysis
4.11
Groups of nanoparticles in direct contact with the cell membrane were first identified in the TEM images. A “group” was defined as any set of three or more nanoparticles located in close contact with each other and simultaneously contacting the same region of the plasma membrane. Each group was then categorized according to its spatial disposition as either “like‐chain assemblies” (nanoparticles arranged in a linear or quasilinear sequence) or “random assemblies” (nanoparticles lacking an apparent ordered arrangement and often observed even tightly packed together). Representative examples of both categories are shown in Figure S8A. The proportion of each disposition was calculated for the different experimental conditions (DMSA–MNPs@92R and DMSA–MNPs@IC in MOLT‐4 and MOLT‐4 CCR9 KO cell lines). The standar error (SE) of each proportion was calculated using the formula
where “p” is the proportion of assemblies in a given category and “n” is the total number of assemblies analyzed (n = 100–140). Statistical comparisons between conditions were performed using a Z‐test to evaluate significant differences in the distribution of nanoparticle assemblies. The significance level was set at α = 0.05 (see “Statistical Analysis”).
Nanoparticle distribution from the plasma membrane toward the extracellular space was quantified using ImageJ. For each image, a freehand line (700 nm width) was drawn along the plasma membrane and straightened to generate a rectangular projection with cytosol on the right and nanoparticle‐containing extracellular space on the left. Images were normalized to a 0–1 intensity scale using nanoparticle signal and the mean intensity of a nanoparticle‐free reference region. Intensity‐versus‐distance profiles were obtained by vertically averaging over a region of interest restricted to pixels from the original image. Profiles from 10 to 12 images were aligned by identifying the membrane as the point of maximal intensity rise and shifting this position to x = 0. Because nanoparticles appear dark, higher nanoparticle density corresponds to lower normalized intensity values (<1).
MNPs Biocompatibility (Presto Blue Assay)
4.12
MOLT‐4 cells were seeded in 96‐well plates at a density of 5 × 10^5^ cells/mL and maintained under normal culture conditions for 24 h. They were exposed to increasing concentrations of DMSA–MNPs (0–500 µg_Fe_/mL) for 2 and 24 h, and at the corresponding time a 1:10 dilution of Presto Blue (ThermoFisher) in culture medium was added. After a 4 h incubation, the fluorescence emitted by the cells was measured on a SpectraMax iD3 fluorimeter (λ excitation = 555, λ emission = 595 nm).
Cellular Iron Concentration Measured by ICP‐OES
4.13
The iron content of the cells was measured at different times by ICP‐OES. Cells were cultured in 12‐well plates at a density of 5 × 10^5^ cells/mL and maintained under normal culture conditions for 24 h. The cells were then incubated for 0–24 h with DMSA–MNPs (150 µg_Fe_/mL) and after washing three times with PBS, they were stained with Trypan Blue and counted in a Neubauer Chamber. The cell suspension was centrifuged at 1200 rpm for 5 min and the cell pellet was digested for 1 h at 90°C in 1 mL of HNO_3_ (65%). After cooling the tubes to RT, the volume was made up to 10 mL with distilled water and the iron content was determined by ICP‐OES (Plasma Emission Spectrometer ICP PerkinElmer mod. OPTIMA 2100 DV, PerkinElmer, Waltham, MA). Finally, the iron concentration was expressed relative to the number of cells in each sample.
Confocal Microscopy Study of MNPs Cellular Location
4.14
Cells were incubated at a density of 5 × 10^5^ cells/mL for 2 h on poly‐Lysine coated 12 mm diameter crystal disks. After 24 h, the cells were exposed to DMSA–MNPs (500 µg_Fe_/mL) and then stained for 5 min at 37°C with 100 µL of Alexa555 WGA (Thermo Fisher Scientific) diluted 1:2000 in DMEM. The cells were then fixed for 15 min at room temperature in 100 µL paraformaldehyde (4%, PFA) in PBS, then stained for 10 min at room temperature with 100 µL DAPI (Sigma‐Merck) diluted 1:500 in PBS and finally, mounted to slides with Fluoromont‐G (Thermo Fisher Scientific). Between each staining step the cells were washed 3 times with 100 µL PBS. Finally, images were obtained on a TCS SP5 multispectral confocal system from Leica using a HC PL APO CS2 oil objective 63X/1.40.
In Vitro Study of the CCR9 Binding Specificity of 92R by Flow Cytometry
4.15
This study was performed in MOLT‐4 and MOLT‐4 CCR9 KO cells using 92R and IC antibodies, both free and bound to DMSA–MNPs. The cells were seeded in 96‐well plates at a density of 5 × 10^5^ cells/mL and maintained under normal culture conditions for 24 h. Free antibodies (10 µg/mL) or DMSA–MNPs@Antibodies (500 µg_Fe_/mL) were the added for 2 h and after washing 3 times with PBS, the cells were recovered by centrifugation (1800 rpm for 2 min) and Alexa488 Gαm IgG (10 µg/mL) was added to the cells for 40 min. After washing 3 times with PBS and recovery by centrifugation (1800 rpm for 2 min), the fluorescence of the different samples was finally evaluated by flow cytometry (Bekman Coulter Cytomics FC500 cytometer) gating the living cells using Propidium Iodide (PI, FL3–). Hence, the final value is the geometric mean of the FL1 peak of the FL3– population, processing the data with FlowJo analysis software.
In Vitro Study of the 92R–CCR9 Binding Specificity by Confocal Microscopy
4.16
This study was performed on MOLT‐4 and MOLT‐4 CCR9 KO cells using 92R and IC antibodies, both free and bound to DMSA–MNPs. The experimental procedure was similar to the confocal microscopy study of MNPs cellular location described above but incubating the cells for 40 min with Alexa488 Gαm IgG (10 µg/mL) before staining, and with the corresponding washes with PBS.
In Vitro MHT Experiments
4.17
Cells were incubated at a density of 5 × 10^5^ cells/mL in 35 mm diameter Petri dishes and maintained under normal culture conditions for 24 h. The different experimental conditions used were: negative control, cells incubated under normal conditions (Ctrl 37°C); heating control, cells were incubated at 42 and 45°C for 1 h (Ctrl 42 and Ctrl 45°C); for the different controls based on MNPs, cells were incubated for 2 h at 37°C with the different MNPs samples (500 µg_Fe_/mL) and then maintained for 1 h more with DMSA–MNPs, DMSA–MNPs@92R, or DMSA–MNPs@IC; as a control for magnetic field application, cells were exposed to AMFs (530 kHz, 20 mT) for 1 h (Ctrl‐AMF). Finally, for the in vitro MHT experiments, cells were incubated at 37°C with the MNPs (500 µg_Fe_/mL) and exposed to the AMF (530 kHz, 20 mT) for 1 h: DMSA–MNPs + AMF, DMSA–MNPs@92R + AMF, and DMSA–MNPs@IC + AMF. Schematic representation is shown in Figure 9.
(A) Schematic representation of the workflow followed for in vitro MHT experiments on MOLT‐4 and MOLT‐4 CCR9 KO cells. Cells were exposed to MNPs (500 µgFe/mL) for 2 h and an AMF (530 kHz, 20 mT) was applied for 1 h. After a further 24 h cell death was evaluated by annexin V staining and flow cytometry. (B) Panel showing the different controls used for the in vitro MHT experiments, considering the temperature of the system, the application of AMFs and the MNPs used.
Cell Death Evaluated by Annexin V–PI Flow Cytometry
4.18
The cells (2.5 × 10^5^ cells) were recovered by centrifugation (1500 rpm, 5 min) 24 h after the in vitro MHT experiments and resuspended in 100 µL of 1:10 Annexin V buffer in PBS. Annexin V‐FITC (7.5 µL) was added to each sample (Annexin V‐FITC Dead Cell Apoptosis Kits for Flow Cytometry: Thermo Fisher) and incubated for 10 min at 4°C, and 5 µL of PI was added just before introducing the tube into the cytometer (Bekman Coulter Cytomics FC500 cytometer). FITC and PI were measured in FL1 and FL3, respectively, and the results were interpreted as: live cells, FL1– and FL3–; early apoptotic cells, FL1^+^ and FL3–; late apoptotic cells, FL1^+^ and FL3^+^; and necrotic cells, FL1– and FL3^+^.
Hspa1 Overexpression Quantified by Real Time qPCR‐RT
4.19
Cells from in vitro MHT experiments were recovered by centrifugation (1500 rpm, 5 min) and their RNA was extracted using the High Pure RNA Isolation Kit (Roche). The RNA recovered was quantified using NanoDrop and cDNA was synthesized with the High‐Capacity cDNA Reverse Transcriptase kit (Applied Biosystems, Thermo Fisher). The cDNA obtained was used to perform qPCR with the Power SYBR Green PCR mix (Applied Biosystems, Thermo Fisher) and using specific primers for Hspa1b and for BETACC (β‐Actin gene) as a housekeeping gene (Table S3: Sigma‐Aldrich). The qPCR‐RT data were analyzed following the 2 ΔΔCt method.
ROS Generation Quantified by Flow Cytometry
4.20
Cells from in vitro MHT experiments were collected by centrifugation (1500 rpm, 5 min), resuspended in 400 µL of a 1:400 dilution of dihydrorhodamine 123 (DHR; from Sigma‐Aldrich) prepared in DMEM + FBS 10%, and incubated for 30 min at 37°C in an atmosphere of 5% CO_2_ and at 90% relative humidity. After incubation, cells were centrifugated again (1500 rpm, 5 min) and resuspended in 300 µL of PBS‐B. PI was added immediately before acquisition on cytometer (Bekman Coulter Cytomics FC500 cytometer). DHR and PI fluorescence were detected in FL1 and FL3, respectively. Data were analyzed by gating live cells using (PI–, FL3– population). ROS levels were expressed as the geometric mean of the FL1 peak (DHR^+^) within the PI–population. Analyses were performed using FlowJo analysis software.
Statistical Analysis
4.21
The data shown correspond to the means (±SD) of at least three independent experiments. A Welch's t‐test and one or two factor ANOVA, plus multiple comparisons contrasted with a Dunnet test were performed to study the difference between the treatment data and that of the controls. The significance level used was α = 0.05 and results were considered significantly different when p < 0.05 (), very significant at p < 0.01 (), and highly significant at p < 0.001 ()
Cohen's d was used to assess differences between groups in cases where the t‐test did not reach statistical significance, but results approached the threshold, revealing a large effect size that suggests a relevant trend (Cohen's d threshold: d (0.01) = very small, d (0.2) = small, d (0.5) = medium, d (0.8) = large, d (1.2) = very large, and d (2.0) = huge) [69]. This highlights the practical relevance of the observed effect and supports the conclusion that the differences identified were not negligible
Finally, Z‐test was used for proportion comparisons, selected based on their theoretical appropriateness for binary categorical data and large sample sizes. Two distinct Z‐test approaches were employed:
For between‐condition comparisons (DMSA– NPs@92R + MOLT‐4 vs other conditions), two‐sample Z‐tests for proportions were conducted to test the null hypothesis that the proportion of “like‐chain assemblies” in the experimental condition equaled that in each comparison condition (H_0_: p 1 = p 2). The test statistic was calculated as
where p^=x1+x2n1+n2 represents the pooled proportion.
For within‐condition analyses, one‐sample Z‐tests for a proportion were used to test whether the proportion of like‐chain assemblies in each condition significantly differed from the theoretical value of 0.5 (H_0_: p = 0.5). The test statistic was computed as
Raw Data can be Consulted in Tables S4 and S5
4.22
The “Supporting Information” includes details of the MNPs heat release under different AMFs conditions. The optimization of the MNPs antibody functionalization, FTIR characterization, and representation of the resulting MNPs populations. The generation of the CCR9 knock‐out MOLT‐4 variant (MOLT‐4 CCR9 KO), and validation assays by confocal microscopy and flow cytometry. Confocal images of the interactions DMSA–MNPs@Antibodies with MOLT‐4 CCR9 KO. Confocal images of DMSA–MNPs@92R activity in MOLT‐4 and MOLT‐4 CCR9 KO cells. TEM images and analysis of the disposition of the nanoaprticles at the cell membrane. Interaction of DMSA–MNPs@Antibodies with HEK293T (CCR9–) and HEK293T–CCR9 (CCR9^+^) cells. Cells response to heat exposition. Analysis of early apoptosis, late apoptosis and necrosis in the in vitro MHT experiments. Biocompatibility of DMSA–MNPs@Antibodies by flow cytometry; and detailed data from Annexin‐PI flow cytometry assays. Flow cytometry analysis of ROS generation. Time‐dependent evolution of the HD and PDI of the different DMSA–MNPs samples in DMEM + 10% FBS. Primer sequences used for Hspa1b overexpression and sequence of the 624B biotinylated peptide are available.
Funding
D.E.‐B. is a predoctoral fellow at the Molecular Biosciences Program of the Universidad Autónoma de Madrid supported by the Spanish Ministry of Science and Innovation through a FPI Contract (PRE2018‐084189). This work was supported in part by the Spanish Ministry of Science and Innovation through Grant SAF‐2017‐82223‐R (to D.F.B., funded by MCIN/AEI/10.13039/501100011033 and by the ERDF a way of making Europe), PID2019‐105404RB‐I00 (to L.K., funded by MCIN/AEI/10.13039/501100011033), PID‐2020‐112685RB‐I00 (to D.F.B., funded by MCIN/AEI/10.13039/501100011033), PID2023‐150170OB‐I00 (to M.P.M., funded by MCIN/AEI/10.13039/501100011033), and PID‐2023‐146212OB‐I00 (to D.F.B, funded by MCIN/AEI/10.13039/501100011033 and by the ERDF, EU). This work was also supported by the CSIC Grant PIE‐202320E096 (to L.K.). T.S.v.Z. acknowledges the Severo Ochoa Program for Centres of Excellence, Grant CEX2023‐001286‐S funded by MICIU/AEI/10.13039/501100011033. The Centro Nacional de Biotecnología (CNB‐CSIC) a Severo Ochoa Center of Excellence (CEX2023‐001386‐S) funded by MICIU/AEI /10.13039/501100011033. The groups of D.F.B. and M.P.M. are part of the “Nanotechnology in Translational Hyperthermia” Network supported by the Spanish Ministry of Science and Innovation (HIPERNANO, RED2018‐102626‐T, funded by MCIN/AEI/10.13039/501100011033). This research work was performed in the framework of the Nanomedicine CSIC HUB and the CSIC Cancer HUB.
Conflicts of Interest
The authors have no competing financial interests to declare, except for the fact that the 92R mAb was generated by the laboratory of L. Kremer and is protected by the CSIC (US Patent US9884915B2; EP3074421A1).
Author Contributions
D.E.‐B. contributed to conceptualization, data curation, formal analysis, investigation, methodology, writing of the original draft, writing, reviewing, and editing. I.C.‐G. performed the investigation, data curation, investigation, methodology, writing, reviewing, and editing. T.S.v.Z. contributed to data curation, methodology, and validation. M.P.M. contributed to conceptualization, funding acquisition, investigation, methodology, supervision, writing, reviewing, and editing. L.K. contributed to conceptualization, funding acquisition, investigation, methodology, supervision, writing, reviewing, and editing. D.F.B. contributed to conceptualization, formal analysis, funding acquisition, investigation, methodology, project administrator, resources, supervision, validation, visualization, writing of the original draft, writing, reviewing, and editing. All authors contributed to the preparation of this manuscript and have given their approval to the final version of the manuscript submitted for publication.
Supporting information
Supporting File: adhm70608‐sup‐0001‐SuppMat.pdf.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1E. M. Scutigliani , Y. Liang , H. Crezee , R. Kanaar , and P. M. Krawczyk , “Modulating the Heat Stress Response to Improve Hyperthermia‐Based Anticancer Treatments,” Cancers 13, no. 6 (2021): 1243.33808973 10.3390/cancers 13061243 PMC 8001574 · doi ↗ · pubmed ↗
- 2N. van den Tempel , M. R. Horsman , and R. Kanaar , “Improving Efficacy of Hyperthermia in Oncology by Exploiting Biological Mechanisms,” International Journal of Hyperthermia 32, no. 4 (2016): 446–454.27086587 10.3109/02656736.2016.1157216 · doi ↗ · pubmed ↗
- 3W. Rao , Z.‐S. Deng , and J. Liu , “A Review of Hyperthermia Combined with Radiotherapy/Chemotherapy on Malignant Tumors,” Critical Reviews™ in Biomedical Engineering 38, no. 1 (2010): 101–116.10.1615/critrevbiomedeng.v 38.i 1.8021175406 · doi ↗ · pubmed ↗
- 4P. Wust , B. Hildebrandt , G. Sreenivasa , et al., “Hyperthermia in Combined Treatment of Cancer,” The Lancet Oncology 3, no. 8 (2002): 487–497.12147435 10.1016/s 1470-2045(02)00818-5 · doi ↗ · pubmed ↗
- 5J. Mahmood , H. D. Shukla , S. Soman , et al., “Radiotherapy, and Hyperthermia: A Combined Therapeutic Approach in Pancreatic Cancer Treatment,” Cancers (Basel) 10, no. 12 (2018): 469.30486519 10.3390/cancers 10120469 PMC 6316720 · doi ↗ · pubmed ↗
- 6J. van der Zee , Z. Vujaskovic , M. Kondo , and T. Sugahara , “The Kadota Fund International Forum 2004–Clinical Group Consensus*,” International Journal of Hyperthermia 24, no. 2 (2008): 111–122.18283588 10.1080/02656730801895058 PMC 2759185 · doi ↗ · pubmed ↗
- 7P. Wust , H. Riess , B. Hildebrandt , et al., “Feasibility and Analysis of Thermal Parameters for the Whole‐Bodyhyperthermia System IRATHERM‐2000,” International Journal of Hyperthermia 16, no. 4 (2000): 325–339.10949129 10.1080/02656730050074096 · doi ↗ · pubmed ↗
- 8B. Hildebrandt , J. Dräger , T. Kerner , et al., “Whole‐Body Hyperthermia in the Scope of von Ardenne's Systemic Cancer Multistep Therapy (s CMT) Combined with Chemotherapy in Patients with Metastatic Colorectal Cancer: A Phase I/II Study,” International Journal of Hyperthermia 20, no. 3 (2004): 317–333.15204528 10.1080/02656730310001637316 · doi ↗ · pubmed ↗