Application of magnetic nanoparticles in adoptive cell therapy of cancer; training, guiding and imaging cells

ABSTRACT

Adoptive cell therapy (ACT) is on the horizon as a thrilling therapeutic plan for cancer. However, widespread application of ACT is often restricted by several challenges, including complexity of priming tumor-specific T cells and poor trafficking in solid tumors. The convergence of nanotechnology and cancer immunotherapy is coming of age and could address the limitations of ACT. Recent studies have provided evidence on the application of magnetic nanoparticles (MNPs) to generate smart immune cells and to bypass problems associated with conventional ACT. Herein, we review current progress in the application of MNPs to improve preparing, guiding and tracking immune cells in cancer ACT. Besides, we comment on the challenges ahead and strategies to optimize MNPs for clinical settings.

Plain language summary

Article highlights.

Background

• Adoptive cell therapy (ACT) harnesses cells of our immune system to treat cancer. However, widespread application of ACT is often restricted by several challenges, such as complexity of priming tumor-specific T cells and their poor trafficking in solid tumors.

• Magnetic nanoparticles (MNPs) are a class of nanoparticles featuring with magnetic properties, biocompatibility, low toxicity and easy synthesis. The most frequently used type of MNPs are superparamagnetic iron oxide nanoparticles (SPIONs), which are FDA-approved for clinical purposes and are potentially used in cancer treatment.

• The magnetic properties of SPIONs have rendered them a good target to be controlled and tracked under external magnetic field (EMF). This feature has been previously used in drug delivery in cancer, whereas MNP-bound agents are guided toward site of interest under a local EMF provided by external magnet.

MNPs for ACT of cancer

• It is hypothesized that MNP-labeled immune cells can be delivered to site of interest via applicating a local EMF near tumor site. However, results obtained from studies on T cells has shown that MNP-labeled T cells tend to accumulate in tumor draining lymph nodes, while MNP-labeled NK cells were efficiently delivered to tumor site.

• MNPs can also be used as artificial antigen-presenting cells (aAPCs) to activate and expand antigen-specific T cells prior to adoptive transfer.

• MNP labeling of immune cells enables us to track their biodistribution after adoptive transfer with magnetic resonance imaging (MRI) or magnetic particle imaging (MPI) devices.

1. Introduction

Adoptive cell therapy (ACT), also defined as cellular adoptive immunotherapy, is a vibrant field that harnesses cells of our immune system to treat cancer [1]. In ACT, immune cells, particularly T cells, are first extracted from the periphery blood of patients, next are genetically engineered and expanded ex-vivo and finally returned to the same donor or other patients [1]. Over the past two decades, the landscape of ACT has dramatically changed from infusion of tumor-infiltrating lymphocytes (TILs) toward more specific treatments including transfusion of T cell receptor (TCR)-T and chimeric antigen receptor (CAR)-T cells [2]. Among them, ACT with CAR-T cell products has gained Food and Drug Administration (FDA) approval for the treatment of several hematological malignancies [2,3]. Consequently, based on CAR technology, new classes of ACTs using other immune cells including macrophages and natural killer (NK) cells have been developed [4]. However, despite promising initial results in hematological malignancies, the application of ACT in solid tumors is still narrowed. Several characteristics of solid tumors, such as immunosuppressive tumor microenvironment (TME) and heterogenicity of tumor antigens reduce the trafficking of immune cells into the tumor site [5].

Nanomaterials enable immune cell products or immunomodulators to target tissues of interest while avoiding challenges associated with solid tumor microenvironment [6]. Due to their size and surface area-to-volume ratio, they have long been exploited to encapsulate drugs for their controlled delivery in cancer [6,7]. Magnetic nanoparticles (MNPs) are nanosized materials consisting of superparamagnetic and ferromagnetic elements including Iron, cobalt, nickel and their alloys. Among several types of MNPs, superparamagnetic Iron oxide nanoparticles (SPIONs) are the most frequently employed magnetic nanomaterials since they are biocompatible, safe and exhibit low cytotoxicity to living organisms [8]. The unique feature of MNPs to be manipulated and steered by an external magnetic force has encouraged the widespread application of these nanomaterials in drug delivery, photothermal and photodynamic therapies, magnetic hyperthermia and magnetic resonance imaging (MRI) in cancer [8]. Following the magnetic targeting principles, it is now possible to redirect MNP-loaded cells toward a site of interest by applying an external magnetic field (EMF). Currently, stem cells, dendritic cells and endothelial cells are being functionalized with SPIONs and are magnetically guided for applications in tissue regeneration and tumor vaccination [9]. Considering diagnosis, MNP-loaded cells can be tracked by MRI to visualize their biodistribution and to estimate their survival under pathophysiological conditions [10].

There is emerging research into how MNPs can be applied to improve ACT in malignancies. To name a few, MNPs conjugated with Major histocompatibility complex (MHC) molecules have been applied as nanoscale artificial antigen-presenting cells (aAPCs) to enhance ex-vivo expansion of T cells before adoptive transfer [11]. Moreover, the superparamagnetic features of MNPs are used to magnetically guide MNP-loaded T cells comparable to magnetic drug delivery [12]. In another strategy, loading NK cells with MNPs enabled their magnetic delivery toward solid tumors in a mouse model of lung small cell carcinoma [13]. Herein, we aim to provide a comprehensive review of current progress in the applications of MNPs in ACT. Furthermore, we will shed light on the challenges ahead and comment on strategies to address them.

2. Magnetic nanoparticles

Among different types of nanoparticles, MNPs are frequently used in cancer treatment due to their unique physical and chemical properties, easy synthesis, easy surface modifications, low toxicity, magnetic properties and good biodegradability [14]. MNPs are made using a core made of oxide or metal alloy of nickel (Ni), cobalt (Co), iron (Fe), manganese (Mn) and gadolinium (Gd) [14,15]. However, Ni and Co are prone to oxidation and weak magnetic properties and additionally, can be toxic for biological systems [16,17]. On the other hand, magnetite (Fe₃O₄), hematite (α-Fe₂O₃) and maghemite (γ-Fe₂O₃) are less toxic and thus, are the most widely used type of magnetic cores for MNPs [14,18].

However, due to the hydrophobic surfaces with large surface area-to-volume ratios, MNPs are susceptible to agglomerate [19]. To overcome this, MNPs are coated with surface coating materials to prevent the accumulation of iron oxide nuclei and improve the dispersion of nanoparticles. Several materials are used to coat MNPs such as dextran, polyethylene glycol, liposomes, peptides, etc. [20]. Surface coating enables MNPs to be dispersed into homogenous ferrofluids and thus, will endow MNPs with improved stability [20]. It also creates suitable conditions for the binding of nanoparticles to drug molecules, binding of various ligands to the iron oxide core and limiting specific cellular interactions for lower cytotoxicity and higher biocompatibility [19].

3. Application of MNPs to produce therapeutic T-cell products

The several types of therapeutic applications of MNPs in adoptive immunotherapy are depicted in Figure 1. Herein, we discuss these applications with a focus on recent advances.

Figure 1.

Application of MNPs in ACT of cancer. (A) MNPs decorated with T cell stimuli can be used as aAPCs to activate T cells. Besides, aAPC-CTL complexes can be passed through a magnetic column to elute antigen-specific T cells and discard negative fraction. In another approach, coculturing CD8⁺ T cells with nano-aAPCs in a magnetic field can induce T cell receptor (TCR) clustering and promote activation of T cells. activated T cells will ultimately be expanded and primed for adoptive cell transfer. (B) MNP-bound T cells or NK cells can be magnetically guided toward site of interest by applying an external magnetic field. (C) MRI and MPI devices enable us to track MNP-labeled immune cells biodistribution after injection.

3.1. MNPs as artificial antigen-presenting cells

Ex-vivo expansion and activation of antigen-specific T cells is a crucial step to achieve a therapeutic amount of T cells before adoptive transfer. Traditionally, this is accomplished through stimulation of T cells by natural antigen-presenting cells (APCs) such as dendritic cells (DCs) [21]. However, this method has faced unavoidable challenges regarding the hurdles in isolating natural APCs, poor reproducibility and its time-consuming process [22]. One of the promising strategies to overcome these barriers is to design nanosized aAPCs mimicking the interface between T cells and natural APCs. Generally, these nanosized platforms are constructed by coupling T cell activating proteins with nanoparticle cores, such as carbon nanotubes and MNPs, among others. Several features of MNPs such as their controllable nature under magnetic fields and their capacity to be formulated in a defined size have rendered them a suitable nanoparticle to be used in aAPCs structure [23].

The first model of nanoscale aAPC was manufactured based on iron-dextran paramagnetic nanoparticles decorated with peptide-MHC (as activating signal) and anti-CD28 or B7.1-Ig (as co-stimulatory signals). These nano-aAPCs induced robust proliferation of antigen-specific T cells in vitro. Besides, generated cytotoxic T lymphocytes (CTLs) effectively inhibited tumor growth in a subcutaneous mouse melanoma model [24]. In another study, Zhang et al. developed a biomimetic nano-aAPC platform in which magnetic nanoclusters (MNCs) were coated with a leukocyte cell membrane. The cell membrane was decorated with T-cell stimuli, including peptide-loaded MHC-I and anti-CD28 as a co-stimulatory domain through Cu-free click reactions. Upon coculturing, the biomimetic nano-aAPC efficiently activated and expanded CD8⁺ T cells before adoptive transfer [25].

Moreover, paramagnetic-based nano-aAPCs can offer a strategy to both “enrich” and “expand” antigen-specific tumor cells upon an EMF powered by magnetic columns [11]. The motivation for enriching T cells before their expansion arises from the observation that adoptively transferred antigen-specific T cells are in compete with polyclonal co-transferred T cells for growth signals, restricting the bioavailability of tumor-specific T cells [26]. To this end, Percia et al. generated paramagnetic nano-aAPCs by coupling iron-dextran nanoparticles to MHC-Ig dimer as activating and anti-CD28 antibody as co-stimulatory signals. Generated nano-aAPCs were then incubated with Naive CD8⁺ T cells from a wild-type B6 mouse for the enrichment step and then passed through a magnetic column [11]. This allowed for the magnet-bound antigen-specific T cells to be eluted and cultured in vitro. Using this enrichment and expansion (E+E) method increased the proliferation of tumor-specific lymphocytes up to 3–10 folds after in-vitro culture and up to 3–4 folds after adoptive transfer in-vivo [11]. In another study, Ichikawa et al. utilized this E+E method to expand tumor-specific T cells derived from patients with metastatic melanoma. They demonstrated that paramagnetic nano-aAPCs bearing MHC class I molecules loaded with the Melanoma-associated antigen recognized by T cells-1 (MART-1) peptide effectively expanded MART-1 specific CD8⁺ T cells, which yielded superior expansion compared with mature DCs and CD3/CD28 Dynabeads [27].

Albeit nano-aAPCs have several advantages such as enhanced biodistribution in tumor sites when compared with micron-sized aAPCs, however, their small size confines their contact with T cells and subsequent T cell activation [23]. One promising feature of paramagnetic nano-aAPCs is that they can be clustered on T cells under an EMF to create a micron length-scale of interaction between aAPCs and T cells, subsequently enhancing contact site and T cell activation [28]. In a study by Percia et al., incubation of CD8⁺ T cells with paramagnetic Iron-dextran-based nano-aAPCs in a 0.2 T EMF-induced nano-aAPCs aggregation and doubling of T cell receptor (TCR) cluster size, ultimately resulting in improved T cell expansion both in-vitro and after adoptive immunotherapy in-vivo [28].

3.2. Magnetic immune cell delivery

Despite the clinical success of ACT in hematological malignancies, delivering immune cells to solid tumors remains limited, owing to insufficient trafficking of immune cells into the tumor site and immunosuppressive microenvironment. Manipulating immune cells to localize in the region of interest will thus improve the effectiveness of ACT. The ability of MNPs to be steered under an EMF has been previously used to enrich magnet-bound drugs in tumor sites. Nevertheless, recent studies highlight that magnetic control can also be safely applied to several immune cells such as T cells and NK cells to drive their site-specific enrichment. This Magnetic targeting can be further achieved by loading immune cells with MNPs, particularly SPIONs. Studies on magnetic delivery of immune cells are summarized in Table 1.

Table 1.

Current pre-clinical and studies of magnetic delivery of immune cells.

Study type	Type of tumor	Type and amount of MNP	Type of immune cell transplanted	Magnetic field strength	outcome	Study	Ref.
In vitro	NA	25 and 50 µg/ml - SPION	T cells	0.2 T	SPION-loaded T cells were attractable by magnets but were cytotoxic for T cells.	Mühlberger et al.	[29]
In vitro	NA	75 µg/ml- SPIONcitrate	T cells	0.5 T	SPION-loaded T cells were attractable by magnets and had no cytotoxic effect on cells vitality.	Mühlberger et al.	[30]
In vivo	EG7-OVA cells	150 µg/ml APS-coated SPION	CD8+ T cells	1.45 T	Localized EMF improved retention of MNP-loaded T cells in tumor draining lymph node but reduced infiltration of APS-MNP loaded T cells in tumor site.	Sanz-Ortega et al.	[32]
In vivo	EG7-OVA cells	20 pg Fe/cell azide-modified leukocyte membrane-coated magnetic Fe3O4 nanoclusters	CTLs	NA	Localized EMF promoted retention of aAPC-CTL complexes in tumor site and correspondingly, delayed tumor growth	Zhang et al.	[25]
In vivo	GFP labeled RPMI8226 human B cell lymphoma	20 μg Fe/ml Cy5.5-Fe3O4/SiO2	NK cells	340 G/mm	Placing a magnet near the tumor enhanced tumor infiltration of MNP-loaded NK cells into tumor site by 17-fold.	Jang et al.	[34]
In vivo	Human alveolar adenocarcinoma cell line (A549)	50 μg/ml Fe3O4 coated with PDA	NK cells	NA	magnetic delivery increased intra-tumoral accumulation of MNP-labeled NK cells which was followed by increased apoptosis of cancer cells.	Wu et al.	[13]
In vitro	NA	150 µg/ml APS-coated SPION	NK cells	Magnet A: 1.35 T Magnet B: 1.45 T	Magnetic delivery successfully enhanced retention of MNP-loaded NK cells in an in-vitro capillary flow system.	Sanz-Ortega et al.	[36]
In vivo	metastatic prostate tumors	100 μg/ml SPIONs	macrophages	7 T	Targeting SPION-labeled macrophages using MRI significantly increased macrophages uptake into tumor site, which was accompanied by increased tumor reduction.	Muthana et al.	[39]

aAPC: Artificial antigen-presenting cell; APS: (3-aminopropyl) triethoxysilane; CTL: Cytotoxic T Lymphocyte; EMF: External magnetic field; MNP: Magnetic nanoparticle; NA: Not available; NK: Natural killer; PDA: Polydopamine; SPION: Superparamagnetic iron-oxide nanoparticle.

3.2.1. T cell delivery

In an early study, Mühlberger et al. [29] demonstrated that T cells incorporated with high-concentration SPIONs (25 and 50 µg/ml) were attractable by magnets, but were not suitable for clinical setting due to toxicity concerns. Coating particles with bovine or human serum albumin to prevent cytotoxicity, led to either immune reactions or decreased particle uptake by T cells, respectively. This motivated later works to use citrate-coated SPION (SPION^citrate) which is both cyto-immunocompatible and has adequate cellular loading [30]. Literature has demonstrated that SPION^citrate does not tend to aggregate in blood and cause thromboembolic events [30]. Besides, these nanoparticles can be safely incorporated into T cells without affecting their cellular functionality [12]. T cells incubated with SPIONs retained proliferation, cytokine secretion and cytotoxicity against cancer cells after antigen-specific activation by the TCR [12]. Results obtained from in vitro studies demonstrated that loading T cells with an optimum MNP concentration of 75 µg Fe/ml was sufficient enough for the magnetic attraction of cells by external magnets [12]. To evaluate the applicability of EMF-mediated accumulation of T cells in vivo, Sanz-Ortega et al. [31] loaded CD8⁺ T cells with 150 µg/ml 3-aminopropyl-triethoxysilane (APS) coated MNPs and inoculated them intravenously into mice model. Collection of different lymphoid organs of mice indicated that MNP loading itself promoted homing of T cells in lymph nodes even in the absence of an external EMF. Besides, with the application of a localized EMF, the accumulation of MNP-loaded T cells was enhanced in local lymph nodes. Nonetheless, a later study by this group showed that albeit localized EMF can improve the accumulation of APS-MNP-loaded T cells in tumor-draining lymph nodes, However, exposure to EMF reduced the infiltration of APS-MNP loaded T cells in tumor site and resulted in a smaller tumor reduction rate [32]. The authors explained this might be due to the fact that loading T cells with MNPs and the EMF decreases the velocity of the T cells and thus might prolong the attachment of T cells in lymph vasculatures. This suggests that although guiding T cells using EMF is less efficient in terms of reducing tumor size, but might have the potential to prevent tumor metastasis through lymph nodes. This hypothesis requires further investigation.

As aforementioned, biomimetic magnetosomes as aAPCs are effective in terms of T cell expansion and activation in vitro [25]. Of interest, when Zhang et al. were evaluating the in-vivo fate of the expanded CTLs, they found that most of the magnetic aAPCs remained on the CTL surface during sample preparation, which might be attributed to the high affinity of aAPCs with CTLs. This phenomenon endowed CTL-aAPC complexes to be guided under an EMF and tracked visually using MRI. Reinfused CTL-aAPC complexes were delivered sufficiently to the target site by magnetic control and correspondingly, delayed tumor growth with few side effects [25].

3.2.2. NK cell delivery

Aside from T cells, NK cells constitute an important domain of lymphocytes essential for recognizing and eliminating certain tumor types. Clinical research has reported that enhanced tumor infiltration of NK cells correlates with a better prognosis of cancer, bringing NK cells as a promising cell type to be applied in ACT [33]. Ex-vivo manipulation of NK cells using MNPs is being investigated to magnetically drive their infiltration into the tumor site. In 2012, Jang et al. [34] established an immune-cell delivery system by loading human NK cells (NK-92MI) with Cy5.5-Fe₃O₄/SiO₂ MNPs and exposing generated NK cells to EMFs after adoptive transfer in a mouse model. Fluorescence-activated cell Sorting (FACS) analysis revealed that placing a magnet near the tumor enhanced the infiltration of MNP-loaded NK cells in the tumor site by 17-fold. In this study, NK cells were transfected with an MNP concentration of 20 μg Fe/ml, in which the highest killing activity of NK cells was observed in vitro. Of note, this concentration is remarkably lower than what is conventionally used for MRI monitoring and magnetic drug delivery, highlighting its promising advantage in reducing nanoparticle chemical toxicity. In another study, Wu et al. [13] synthesized Fe₃O₄@polydopamine (PDA) labeled NK cells with various concentrations of MNPs ranging from 25–100 μg/ml. They demonstrated that magnetic delivery of Fe3O4@PDA-loaded NK cells can significantly inhibit tumor growth in xenograft mice models bearing lung cancer cells.

However, it has been shown that NK cells are highly resistant to transfection by conventional methods, supporting that loading NK cells with MNPs could be challenging in primary cells [35]. To overcome this, Sanz-Ortega et al. [36] generated magnetic NK cells by attaching (3-aminopropyl) triethoxysilane (APS)-MNPs to the cell surface. They demonstrated that the APS-MNP-bound NK cells preserve the biological functions of primary NK cells and can also be successfully retained at the site of interest by applying an EMF.

3.2.3. Delivering immune cells using magnetic resonance targeting

Despite promising initial results, using external magnets to guide MNP-labeled cells can face limitations. External permanent magnets can only be applied to accumulate MNP-labeled cells in tissues near the body surface. Meanwhile, using implanted magnets to achieve magnetic field gradients in deeper tissues requires invasive procedures. An alternative approach for the magnetic delivery of immune cells is magnetic resonance targeting (MRT). MRT uses the innate magnetic field in all MRI devices to guide ferromagnetic-containing objects in deep tissues of the body [37]. This method is especially appealing owing to its high strength and spatiotemporal control over magnetic networks provided by MRI systems.

Previous studies have demonstrated the applicability of the MRI system to guide MNP-loaded monocytes in a vascular model [38]. Moreover, in 2015, Muthana et al. [39] used an MRI scanner to drive the accumulation of SPION-conjugated macrophages to both primary and metastatic tumor sites in an animal model bearing metastatic prostate tumors. In detail, after intravenous administration of 3 million MNP-labeled macrophages, they placed mice into a 7 T MRI scanner to provide magnetic gradients over target regions. They demonstrated that using the MRT method significantly increased SPION-loaded macrophage uptake in primary and metastatic tumor sites, which was followed by increased tumor reduction compared with control groups. Nevertheless, these encourage future research to explore the efficacy of the MRT method in redirecting other immune cells such as NK and T cells toward tumor sites, as well as assessing its safety and applicability in clinical settings.

3.3. Tracking adoptive cell therapy using MRI

One of the requirements for developing safe cell therapies with maximal efficacy is non-invasive imaging means to visualize the biodistribution of cells after administration. This enables us to ascertain if the cells are homed in the tumor site and assist in regulating the cell dose. Traditionally, radionuclide imaging such as single photon emission computed tomography (SPECT) and positron emission tomography (PET) have been used for imaging immune cells [40]. However, cytotoxicity concerns along with considerable associated costs have narrowed the clinical application of radionuclide-based imaging [41]. Furthermore, the radionuclide-based method should be combined with other modalities such as computed tomography (CT) scan or MRI to depict anatomical biodistribution [41]. These limitations have provided rationales for the development of alternative approaches, particularly, MRI-based techniques. MRI is a safe method that provides high-resolution anatomical details without ionizing radiation.

SPIONs have been commonly used as MRI contrast agent for imaging of cells without affecting their viability. In an early study conducted by Kircher et al., ovalbumin (OVA)-specific CD8⁺ T cells were labeled with a highly derivatized cross-linked iron oxide nanoparticle (CLIO-HD). labeling T cells enabled MRI tracking of their tumor accumulation after administration into mice model bearing B16-OVA melanoma [42]. Later studies have also labeled CAR-T cells with MNPs to monitor their biodistribution. Bhatnagar et al. loaded CD19-specific CAR-T cells with MRI contrast agent SPION-copper-64 (SPION-⁶⁴Cu) and successfully detected their accumulation in an in-vitro CD19⁺ lymphoma model [43]. Xie and collaborators loaded interleukin 13 receptor alpha 2 (IL13Rα2) and the EGFRvIII targeting CAR-T cells with ultra-small superparamagnetic iron oxide nanoparticles (USPIOs) at a concentration of 37.5 Fe μg/ml and monitored their dynamic infiltration in glioblastoma. After administration, images obtained from 7.0-T MRI demonstrated enhanced spot-like signals within 3 to 14 days in tumor parenchyma, indicating the presence of CAR-T cells, which was further validated by pathological staining. Of note, the therapeutic efficacy correlated positively with increasing signals [44]. Therefore, USPION-enhanced MRI enabled both tracking the labeled CAR-T cells and assessing their therapeutic outcomes. In another study, Kiru et al. used ferumoxytol (an FDA–approved iron supplement) to label and track CAR-T cells in vivo. In this regard, they used the mechanoporation technique to label anti-B7-H3 CAR-T cells with ferumoxytol. 24 hours post-injection, MRI imaging demonstrated a dose-dependent enhancement of T1-weighted MRI signals in the spleen and at a minimal level in the tumor site, while tumor infiltration of MNP-labeled CAR T cells was significantly increased by day 3 [10].

The first human study in this field was conducted by Singla et al. in 2020, in which a cohort of patients with solid tumors received SPION-⁶⁴Cu labeled CAR-T cells to track their biodistribution. Signals received from dual PET-MR showed a transient accumulation of labeled cells in the lung's posterior basal segments and an intense uptake by the spleen, liver and bone marrow, suggesting the margination of labeled cells to the reticuloendothelial system [45]. Minimal amounts of labeled cells detected in the brain and heart support their safety upon administration.

In addition to CAR-T cells, a considerable volume of in-vivo studies has been published on MRI-based tracking of NK cells toward Liver solid tumors [46–48]. Surveys such as those carried out by Su et al. demonstrated that heparin-protamine-ferumoxytol (HPF) nanocomplexes allowed MRI monitoring of HPF-labeled NK cells after transcatheter intra-hepatic arterial local delivery of NK cells. Of interest, they reported that serial MRI monitoring of NK cell accumulation 24h after injection can represent an early biomarker for anticipation of longitudinal response [46].

3.4. Tracking adoptive cell therapy using MPI

Magnetic particle imaging (MPI) is a non-ionizing tomographic imaging technique that detects the biodistribution of MNP tracers. MPI system was first established in 2005 and is rapidly advancing toward clinical settings. Unlike MRI, MPI visualizes only nanoparticles and does not provide background signals arising from diamagnetic tissues in the neighborhood [49]. Furthermore, MPI does not develop artifacts obtained from hypointense MNPs visualized with MRI. Indeed, signals from MPI are depth-independent and have no penetration limitations. Images provided by MPI can be combined with other modalities, such as CT scan and MRI to depict anatomical distribution of nanoparticles. To date, several studies have used the MPI technique for cell-tracking studies on stem cells. However, studies on MPI imaging of immune cells labeled with SPIONs are sparse. In a mouse bearing malignant glioma tumor cells, MPI allowed the detection of ferucarbotran-labeled CD8⁺ T cells at the tumor site after intravenous or intracerebroventricular administration [50]. In another study using the osteosarcoma model, MPI imaging was used to confirm the retention of ferumoxytol-loaded CAR-T cells in the tumor site after intravenous administration [10]. Nevertheless, further studies translating MPI applications toward clinical settings are required.

4. Strategies to optimize MNP-based cell therapy

Targeting immune cells requires an appropriate design of nanoparticles to maximize cellular uptake and mitigate undesired reactions. Moreover, nanoparticles and their bound objects can be phagocyted by macrophages in organs such as the spleen, liver and lungs [51]. This phenomenon can further prevent MNP-loaded immune cells from reaching a sufficient concentration in the tumor site. Herein, we aim to emphasize the importance of the shape, size and surface modifications of MNPs and their tuning as a potent strategy to maximize MNP-based ACT. Further, we will discuss the potential of magnetic gold nanoparticles to improve immune cell therapy.

4.1. Facilitating particle uptake

As mentioned, lymphocytes and NK cells tend to be resistant to transfection. Therefore, when MNPs are cocultured with these cells, nanoparticles are most likely attached to the cell surface through electrostatic interactions rather than being accumulated intracellularly [31]. However, internalization of dextran-coated MNPs is important since dextran on the cell exterior can bind to antidextran antibodies and provoke antibody-mediated cytotoxicity [52]. Nevertheless, one of the promising strategies to favor the endocytosis of MNPs is to endow them with membrane translocating signals (MTSs). Several effectors can be utilized for such purposes, such as HIV-Tat, transferrin and antibodies that bind to their specific receptors on cell membranes [53]. Another strategy is to exploit cationic agents that create highly charged complexes with MNPs. However, this approach requires long incubation periods [52].

4.2. Improving the bioavailability of MNPs in the tumor site (Figure 2)

Figure 2.

Important determinants of magnetic nanoparticles functionality.

4.2.1. Effect of nanoparticle shape

As mentioned above, the engulfment of MNPs by phagocytes can lead to the elimination of MNP-labeled immune cells. In this regard, nanoparticle shape plays a key role in its tendency to be engulfed by phagocytes [51]. It has been shown that non-spherical particles such as discs, chains and rods exhibit lower phagocytosis and have a higher propensity to escape accumulation in Reticuloendothelial systems (RES) such as the liver. This so-called phenomenon can thus improve the deliverance of MNPs toward the tumor site, rather than being trapped in RES [54]. In addition, the physical characteristics of the particles can also affect their dispersion once they are administrated intravenously. Compared with spherical MNPs, non-spherical MNPs tend to deflect across the blood streamline toward the vessel wall and thus, have more tendency to marginate tumor site endothelium [55].

4.2.2. Effect of nanoparticle size

The size of MNPs plays a crucial role in pharmacokinetics, toxicity, blood circulation half-life, biodistribution, tumor penetration and their interaction with biological systems. The size of nanoparticles injected into the vein should be small enough not to block the blood capillaries and at the same time be able to pass through the pores of the blood vessels and diffuse into the tissues [54]. Spherical particles can vary in size. It has been demonstrated that Nanoparticles ranging between 100–200 nm can efficiently extravasate through the vascular endothelium of the tumor site and escape filtration via the liver and spleen [14]. As size increases above 150 nm, MNPs are prone to be captured by the liver and spleen. However, smaller nanoparticles less than 5 nm are filtered out by the kidneys [14].

Moreover, the size of MNPs must be small enough not to cause needle clogging. Considering this, MNPs with a size of 10–20 nm have a great ability for injection [54]. However, their small magnetic volume limits the magneto-mechanical effects. To solve this problem, clustered SPIONs are used, which are maintained by ligands and have a size of 100 nm to several micrometers [54].

4.2.3. Effect of nanoparticles surface charge & modifications

The surface charge of magnetic nanoparticles is due to the presence of different functional groups on their surface. Surface charge can affect cellular uptake, biodistribution, circulation time of nanoparticles in blood and interaction with local macrophages of RES [14]. Generally, due to the negative surface charge of cell membranes, MNPs with positive surface charge can be internalized into cells more readily than nanoparticles with negative or neutral surface charge [56]. This could therefore facilitate cellular uptake of MNPs in the manufacturing process of MNP-loaded immune cells. on the other hand, research has demonstrated that nanoparticles in the range of 20–150 nm with a positive or negative charge are more susceptible to separation by macrophages in the lung, liver and spleen. In contrast, nanoparticles ranging between 20–150 nm with a neutral charge are shown to have a longer circulation life and less accumulation in these organs [57].

Unmodified IONPs usually show higher toxicity, requiring surface modification of IONPs as an essential strategy for clinical applications. Surface coating can reduce specific cellular interactions to provide MNPs with lower cytotoxicity. In addition, surface modification protects against iron oxide nucleation and improves the intracellular dispersion of MNPs [19]. Proteins, liposomes, polymers and inorganic materials are the most prevalent coating materials used for MNPs. A proper coating material for modification of MNP should have a high affinity for particles and not stimulate the immune system at the same time.

4.2.4. Magnetic gold nanoparticles

Another kind of nanoparticle employed in biomedicine is gold nanoparticles or Au NPs. They are referred to in the literature as biocompatible drug delivery nanosystems that are also utilized as biosensors and for hyperthermia [58]. In the bulk state, metals such as Au and Ag are diamagnetic, whereas they exhibit magnetic properties at the nanoscale [59]. Previous studies in this field have exploited gold-coated iron-oxide nanoparticles for magnetic drug delivery of anti-cancer agents such as doxorubicin (DOX) [60]. It has been shown that using gold shell-iron core MNPs such as Fe3O4@Au NPs can allow magnetic delivery of drug while decreasing cytotoxicity of Fe₃O₄ NPs [61,62]. However, no study has studied the implication of gold nanoparticles in magnetic immune cell delivery, so far. Nevertheless, these findings encourage future studies to explore the efficacy and safety of gold nanoparticles for the delivery of immune cells.

5. Challenges & concerns

Although labeling immune cells with MNPs can yield them with exciting features, our current perception of their impact on living organisms is not comprehensive, including the state and the fate of MNPs in-vivo, how they interrelate with immune cell components and whether they can provoke immune responses in the body [63]. In the upcoming section, we will discuss some of the challenges ahead to reach the goal of manufacturing clinical-grade MNPs.

5.1. Stability of MNPs in cell labeling milieu

Dispersion of nanoparticles in the cell labeling environment plays a key role in how cells react to MNPs. Immune cells respond in different manners whether MNPs are distributed in suspension form or are aggregated in the biological medium. MNP aggregates can be cytotoxic to immune cells whereas the same MNPs dispersed in suspension are shown to have no detrimental effects [64]. Additionally, the characteristic of the MNP surface can be changed in the culture medium, due to the absorption of different proteins and macromolecules present in the medium [65,66]. Nevertheless, the MNPs in the culture medium are not the same as the original MNPs designed by the chemist. To address MNP aggregation and avoid absorption of macromolecules on the MNP surface, cell labeling can be carried out in a serum-free culture supplemented with free citrate [63].

5.2. Functionality of MNPs in physiological conditions

When translating nanoparticles into clinical settings, complex biological features of the in-vivo environment such as proteins and electrolytes can render MNPs to aggregate and, therefore fail to reach the action site [67]. When employed as drug transporters, nanoparticles are subjected to a physiological media that is rich in proteins and salt. The stability of nanoparticles is impacted by both of these variables. Proteins are adsorbed on the surface of nanoparticles and alter their size and surface charge, whereas high salt concentration decreases electrostatic repulsion between nanoparticles, causing them to aggregate [68]. One proposed option is to enhance the repulsion between nanoparticles, which improves colloidal dispersion. However, chemical stability in the biological environment is compromised in the case of nanoparticles utilized in biomedical therapy. For instance, many nanoparticles can aggregate under the acidic (pH = 6–7) conditions seen in cancer cells. However, stable nanoparticles may also be produced at neutral or acidic pH using existing techniques. In this regard, Shakiba et al. [69] improved the interfacial characteristics of AuNPs by forming a self-assembled monolayer, which allowed the particles to be manufactured with great stability in acidic environments. The development of an aqueous layer around the particles leads to increased stability and has facilitated the dispersion of particles in the biological medium.

Another problem arises from the formation of nanoparticle-crown protein (NP-PC) complexes. In an ideal situation, nanoparticles travel through the bloodstream to reach the target of interest. However, once injected into a biological environment, they interact with biological fluid components such as proteins and convert into a nanoparticle that proteins are attached to their surface in a crown shape form, called NP-PC complexes [70]. The development of protein crowns not only affects the toxicity of MNPs but can also influence immune-related responses provoked by them. The possibility for the particles to promote immunogenicity by exposing protein epitopes in an abnormal conformation on their surface was postulated as the cytotoxic mechanism driving this behavior [71]. For example, it was observed that the presence of poly(acrylic) acid on the surface of gold NPs triggered the unfolding of adsorbed fibrinogen, which, in turn, interacted with the leukocyte receptor MAC-1, producing an inflammatory response [72]. PCs has also the ability to increase immune responses against nanoparticles. In a study conducted by Yan et al., it has been demonstrated that the presence of albumin in the surface of poly-(methacrylic acid) nanoporous polymer particles increased the uptake of NP-PC complexes by macrophages, accompanied by increased secretion of inflammatory cytokines [73]. It has been demonstrated that altering the surface of nanoparticles with long-chain polymers, such as polyethylene glycol (PEG), reduces the amount of non-specific protein absorbed onto the nanoparticle surface [74]. Thus, developing stealth MNPs coated by PEG might be a potent strategy to evade macrophage-mediated clearance and undesired immune responses against MNPs.

5.3. Route of administration

MNPs can be delivered to the tumor site either by systematic administration (i.e. intravenously or intra-peritoneal injection) or directly into the tumor site. MNPs can be injected locally using hypodermic needles or catheters. The direct injection approach involves injecting magnetic fluid containing a certain concentration of MNPs straight into the malignant tissue. Since it doesn't require any further assistance for particle localization, this approach is the simplest. Therefore, in in-vivo applications like clinical trials, the direct injection technique is the most often used strategy [75]. The controllability of MNP concentration at the tumor site is another benefit of direct injection. However, this approach works best when the tumor location is easily accessible. Furthermore, MNPs have a propensity to aggregate after injection, which adversely affects their homogeneity within the tumor [76]. Nevertheless, this strategy is limited to several types of malignancies. In systematic delivery, MNPs are administered intravenously or via intra-peritoneal injection. MNP concentrations rise in the circulation due to increased permeability and retention (EPR) effects of the cancer environment [77]. The vascular architecture of tumors differs from that of healthy organs. Tumor tissue's damaged vasculature can help tumor cells develop by providing them with nutrients and oxygen, but it can also make it easier for macromolecules and nanoparticles to enter the tumor site. As a result, nanoparticles are transported from the bloodstream to the cancer location by intravenous injection. NPs that have been deposited in a tumor have a long retention period. The longer retention period is probably caused by MNPs' easier passage through the vasculature. MNPs' prolonged retention period and simple conveyance help to keep the bloodstream stable and avoid agglomeration [78].

Although systematic administration of MNPs resulted in a more homogeneous dispersion of particles at the tumor site than intra-tumoral injection, it is challenging to provide the necessary dose of MNPs for successful therapy. In addition, MNPs may accumulate in certain healthy tissues. Using an active delivery strategy, MNPs can be more effectively localized to the tumor location. In this manner, a targeted agent that attaches to cell receptors modifies the surface of MNPs. These targeting agents can be receptor ligands [79], peptides [80], aptamers such as HER2 [81] and monoclonal antibodies. Ross et al. developed Herceptin (commercially sold as Trastuzumab) conjugated IONPs to control drug release for the treatment of breast cancer and reported that a combination of Herceptin and MNPs dramatically increased their killing capacity against breast cancer cells [82].

5.4. Toxicity

Cytotoxicity of IONPs varies from transient and acute toxicity to insignificant changes in vivo [20]. Cytotoxicity caused by IONPs depends on several factors, including size, shape, concentration, surface charge, type of coating, surface functional groups of IONPs, dose of exposed cells and exposure time [83]. Some of the negative side effects that magnetic nanoparticles might induce include oxidative stress, embryotoxicity, genotoxicity, mutagenicity and vascular embolism (Figure 3) [84]. Studies have shown that SPIONs can enter cells through passive diffusion and endocytosis. When entering the cells, MNPs interact with the nucleus and mitochondria and therefore can cause changes in gene expression and levels of oxidative stress, respectively. These particles can cause swelling and degeneration of mitochondria, loss of mitochondrial cristae, elevated levels of mitochondrial reactive oxygen species (mtROS) and calcium ions, reduced mitochondrial membrane potential (MMP) and decreased ATP levels due to oxidative damage [85–87]. Besides, the high amount of free iron ions in the tissue can cause an imbalance in homeostasis and create abnormal cellular responses such as osmotic damage, cytotoxicity, epigenetic events, DNA damage and inflammatory disorders that can lead to cancer or damage in later generations [88]. Exposure to MNPs is mainly associated with toxic effects such as excessive production of ROS that can be the source of mitochondrial dysfunction [89]. However, among different MNPs, IONPs were demonstrated to be the safest, with reported cytotoxicity at concentrations of 100 μg/ml or higher.

Figure 3.

Toxicity of magnetic nanoparticles.

IONPs can reportedly deposit in different tissues but are less likely to cause cytotoxic changes in vital organs. After a year of magnetic hyperthermia therapy in patients with prostate cancer, although MNP deposits were found to be present in the prostate, no systemic toxicity was identified [90]. In animal models, the median lethal dose (LD₅₀) of dextran-coated iron nanoparticles is found to be high at levels of 400 mg/kg, with cytotoxicity on lymphocytes, neutrophils and peritoneal cells [91]. Moreover, in a study conducted by Yu et al., skin irritation, telangiectasia and leucocyte infiltration detected after subcutaneous injection of dextran-coated MNPs almost subsided 72 hours later [92].

6. Conclusion

In summary, the application of MNPs in ACT of cancer represents a promising avenue to overcome the challenges of conventional methods in several aspects of this field. MNPs can be harnessed to expand and enrich tumor-specific T cells before administration [11], redirect immune cells toward the tumor site [31,36] and of note, track their biodistribution after injection [10]. The evolution of MNP-loaded immune cells brings clinicians one step closer to using MRI devices to drive the accumulation of immune cells in sites of interest, even in the deeper tissues of the body [39]. The benefits of MNPs, such as their biocompatibility and low toxicity profile have made it possible to prime them for several pre-clinical and clinical studies. However, studies on labeling immune cells with MNPs are sparse. Nevertheless, it is in earnest to use more diverse animal models to bridge the gap between recent findings and their future application in the human species.

7. Future perspective

In the last decade, substantial efforts have been devoted toward utilizing MNPs for anti-cancer treatment. Recent research in this field has now encouraged the use of MNPs to improve immune-cell therapy of cancer. However, taking into account that most of the studies are conducted on animal models, more in-depth information is required to bridge the gap between current findings and human species. Albeit MNPs are accepted as biocompatible materials, their interaction with biological components might differ under several pathologic conditions such as cancer. In this regard, the following items should be taken into consideration for future studies: understanding the most ideal coating material for iron oxide MNPs, taking into account that these particles tend to accumulate on cell surface and thus, are prone to provoke unintentional immune responses against injected cells. determining a safe dose for clinical settings in which the magnetic properties are still preserved. designing strategies to avoid retention of MNP-loaded immune cells in tumor draining lymph nodes. We believe that in the early future MNPs not only might take place in standard ACT of solid tumors, but to assay patient's response to therapeutic plan. Additionally, the evolution of MNP-labeled immune cells can bring clinicians one-step closer to harness the magnetic field of MRI devices to redirect MNP-labeled immune cells toward deep tissue tumors.

Author contributions

V Mohammadi and K Esmaeilzadeh contributed to data gathering, writing the primary manuscript and designing the figures. V Mohammadi contributed to designing tables and grammatical and scientific revising of the manuscript. A Esmaeilzadeh contributed to the hypothesis, correspondence and revising of the final manuscript before submission. All the authors viewed and confirmed the final manuscript before submission.

Financial disclosure

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Competing interests disclosure

The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.