Magnetic Drug Targeting Reduces the Chemotherapeutic Burden on Circulating Leukocytes

Magnetic drug targeting (MDT) improves the integrity of healthy tissues and cells during treatment with cytotoxic drugs. An anticancer drug is bound to superparamagnetic iron oxide nanoparticles (SPION), injected into the vascular supply of the tumor and directed into the tumor by means of an external magnetic field. In this study, we investigated the impact of SPION, mitoxantrone (MTO) and SPIONMTO on cell viability in vitro and the nonspecific uptake of MTO into circulating leukocytes in vivo. MDT was compared with conventional chemotherapy. MTO uptake and the impact on cell viability were assessed by flow cytometry in a Jurkat cell culture. In order to analyze MTO loading of circulating leukocytes in vivo, we treated tumor-bearing rabbits with MDT and conventional chemotherapy. In vitro experiments showed a dose-dependent MTO uptake and reduction in the viability and proliferation of Jurkat cells. MTO and SPIONMTO showed similar cytotoxic activity. Non-loaded SPION did not have any effect on cell viability in the concentrations tested. Compared with systemic administration in vivo, MDT employing SPIONMTO significantly decreased the chemotherapeutic load in circulating leukocytes. We demonstrated that MDT spares the immune system in comparison with conventional chemotherapy.


Introduction
Together with surgery and radiotherapy, chemotherapy is still a mainstay of cancer treatment today. Its main disadvantage lies in the severe systemic adverse effects inherent in this mode of therapy. In order to kill tumor cells directly, anticancer drugs are usually given at the maximum dose tolerated by the patient [1], which means that the dose necessary for complete tumor eradication is barely achieved. Targeted drug delivery to the tumor may overcome this dilemma [2], and nanotechnology provides new possibilities in this rapidly growing medical field, e.g., in otorhinolaryngology [3]. Magnetic drug targeting (MDT), a new and innovative approach, has been developed to concentrate anticancer drugs in the tumor and reduce leakage into healthy tissues and cells. Superparamagnetic iron oxide nanoparticles (SPION) are functionalized with a chemotherapeutic agent, injected into the vascular supply of the tumor, and directed into the tumor itself by means of an external magnetic field ( Figure 1) [4].
Systemic distribution and adverse drug reactions are thereby decreased, at the same time as the local impact on the tumor is increased. Lower doses of the anticancer drug can be administered, as enhancement rates are 26 times higher with MDT than with conventional systemic chemotherapy [5]. Previous animal studies with an experimental VX-2 squamous cell carcinoma in rabbits successfully demonstrated MDT-induced tumor remission using nanoparticle-coupled mitoxantrone (MTO) [6]. MTO acts against a variety of hematologic and solid tumors by impairing DNA replication, transcription and repair [7], and may cause adverse effects such as nausea, fever, anemia, immunosuppression, and cardiotoxicity [8].
The immune system plays a key role in carcinogenesis and cancer therapy. The initiation, growth and metastasis of a tumor require "cancer immunoediting" [9]. In addition, more and more attention is being paid to the immunological aspects of cancer therapy: antibodies [10], vaccines [11], and the induction of anti-tumor immunity during radiotherapy and chemotherapy [12]. This pilot study aimed to quantify MTO in peripheral blood leukocytes (PBL) using its inherent fluorescent properties, since PBL are crucial components of the immune system. We observed that targeted delivery of MTO by MDT reduces the cytostatic burden on circulating leukocytes. The solution is injected into the vascular supply of the tumor and accumulated at the intended site by means of an external magnetic field.

Analysis of SPION by Flow Cytometry
SPION can be detected by flow cytometry, as shown in Figure 2A. We found a clear distinction from Jurkat cells in size and granularity. Forward and side scatter analysis (FSC/SSC blots) showed SPION as a cloud of events two orders of magnitudes smaller and less granular than the cells. This difference allowed us to gate cells without contaminating SPION.
Since MTO is inherently fluorescent, it is possible to detect and quantify MTO bound to SPION (SPION MTO ) by flow cytometry. The flow cytometer uses separate fluorescence channels to detect emitted light: MTO was excited with 638 nm and the fluorescent emission was recorded with a 725 ± 10 nm band pass detector in channel 7 . Binding MTO to SPION resulted in a moderate increase in the SPION FSC and SSC signals ( Figure 2B

Uptake of MTO and SPION MTO by Jurkat Cells
To gain information about the in vitro uptake of MTO in cell culture, we treated Jurkat cells with MTO and SPION MTO in various concentrations and used flow cytometry to examine viable and dead Jurkat cells (Figure 3). Cells were classified as viable or dead, according to their morphological properties. SPION-treated cells, both viable and dead, exhibited only low autofluorescence. When we incubated viable cells with various concentrations of pure MTO, we found a dose-dependent increase in MFI . Whereas dead cells displayed a flat MFI signal over time, fluorescence increased in viable cells for at least 24 h. Most importantly, there was no relevant difference between cells incubated with pure MTO and those with SPION MTO . The MFI signals were strongly dependent on the MTO concentration in both cases, demonstrating that MTO and SPION MTO display similar cytotoxicity in vitro.

Apoptosis and Necrosis of Jurkat Cells Treated with SPION, MTO and SPION MTO
In viable cells, phosphatidylserine (PS) is found mainly in the inner leaflet of the plasma membrane. During apoptosis, PS is exposed on the cell surface. Redistribution of PS in apoptosis and the loss of plasma membrane integrity in necrosis were monitored by annexinV-FITC/propidium iodide (AxV/PI) staining. We

MDT Reduced the MTO Load of Circulating Leukocytes in Vivo
Exposing healthy tissues and cells to anticancer drugs is a crucial problem with conventional chemotherapy and causes multiple adverse reactions (e.g., leukopenia). To study this effect, we investigated MTO contamination in PBL during conventional chemotherapy and MDT in an in vivo rabbit tumor model. Using flow cytometry, we distinguished granulocytes, monocytes, and lymphocytes by their morphological features and quantified the MTO content as MFI [FL-7] ( Figure 6A). As shown previously [5], MDT with a 20% therapeutic dose (TD 20 ) of SPION MTO reduces the chemotherapeutic burden on healthy tissue and enhances MTO in the tumor (Table 1, Figure 6B).
We detected increasing MTO concentrations in all leukocytes immediately after starting the infusion. MTO concentrations reached their peak after 10 min. MTO loading started to decline as soon as the MTO infusion finished (after 20 min). Comparing conventional chemotherapy in rabbits given a 100% therapeutic dose (TD 100 ) of MTO (n = 4), TD 10 MTO (n = 2), or Ringer's solution (n = 2) intravenously, we found that MTO uptake in PBL clearly depended on the systemic dose administered. Granulocytes, monocytes, and lymphocytes displayed similar MTO profiles ( Figure 6C).
Next, we studied the uptake of SPION MTO into PBL. We compared the MTO content of PBL after intravenous injection of TD 10 MTO with that found after intra-arterial injection of TD 10 SPION MTO . PBL had a reduced MTO load after SPION MTO , but the difference was not statistically significant ( Figure 6D).
An important step in MDT is focusing SPION MTO on the target area with an external magnetic field. After injecting TD 10 SPION MTO into an artery directly supplying the tumor, MDT was applied to one group of rabbits (MDT), while the second group was not exposed to the magnetic field. As shown in Figure 6E, MTO uptake by PBL decreased even further. Compared with conventional chemotherapy, MDT resulted in lower MTO loading of circulating PBL-A difference that is statistically significant (p = 0.014, paired Student's t-test).  Figure 6F summarizes the mean MTO loads of PBL in the different groups of rabbits. We compared TD 100 MTO i.v. (n = 4), TD 10 MTO i.v. (n = 2), TD 10 SPION MTO i.a. (n = 3), TD 10 SPION MTO i.a. MDT (n = 2) and controls treated with Ringer's solution (n = 2). We demonstrated that PBL take up MTO in a dose-dependent manner, and that coupling MTO to SPION decreases the (unwanted) MTO load in PBLs, especially when MDT is applied. In summary, MDT significantly reduced the MTO burden of healthy PBL in comparison with conventional chemotherapy.

Tumor Model
An experimental VX-2 tumor was implanted subcutaneously into the left hind limb of female New Zealand white rabbits (Charles River Laboratories, Sulzfeld, Germany), aged 18-22 weeks and weighing 3.1-3.9 kg. After four to six weeks, the tumor had grown to a size that made intervention necessary.

Nanoparticles
The superparamagnetic iron oxide nanoparticles (SPION), composed of maghemite (Fe 2 O 3 ) and magnetite (Fe 3 O 4 ), were produced by SEON. Alkaline precipitation was followed by self-assembling functionalization with lauric acid and MTO [15]. The hydrodynamic diameter was set at approximately 100 nm using dynamic light scattering. MTO was bound to SPION (SPION MTO ) in an aqueous solution (ferrofluid). To analyze the MTO-content of the ferrofluid, MTO was extracted from 100 µL ferrofluid by the addition of 900 µL 1N HCl and incubation in a Thermomixer ® comfort (Eppendorf AG, Hamburg, Germany) at 600 rpm and room temperature for 1 h. After centrifuging with a Minispin ® plus (Eppendorf AG, Hamburg, Germany) at 14,500 rpm for 10 min, the supernatant was measured with HPLC, using a Waters Alliance model consisting of a separation module (2695 series), a dual wavelength absorbance detector (2487 series), and a 3.0 × 100 mm X-Bridge phenyl column (Waters, Germany). The mobile phase consisted of formiate buffer (pH 3.0) and methanol (80:20 v/v) [16].

Mode of Administration
Therapy was given as a single treatment cycle. We administered conventional chemotherapy intravenously into an ear vein, while MDT required intra-arterial access. For this purpose, the rabbits were taken to an animal operation theatre and the femoral artery prepared under sterile conditions. Pure MTO and drug-loaded SPIONs, as well as Ringer's solution, were injected via a 24 G i.v. catheter (B. Braun, Melsungen, Germany) over 20 min. In one group, an electromagnet (Siemens Healthcare, Erlangen, Germany) with a magnetic field gradient of 72 T/m located directly beneath the pole tip [17] attracted the drug-loaded nanoparticles to the tumor during the 20 min of drug administration and for a further 15 min afterwards. Blood was taken before administration (−60 min), during the infusion (−10 min, 0 min, +10 min), and afterwards (+20 min, +30 min, +40 min, +50 min, +60 min, +120 min). For whole blood analyses, the erythrocytes in rabbit EDTA-blood were subjected to hypotonic lysis twice before flow cytometry.

Cells and Culture Conditions
To investigate the effects of SPION, MTO and SPION MTO on mammalian cells, we used the non-adherent human T cell leukemia cell line Jurkat (ACC 282, DSMZ, Braunschweig, Germany). Cell culture was performed at 37 °C and 5.5% CO 2 in R10 medium (RPMI 1640 medium supplemented with 10% FCS, 1% glutamine, 1% penicillin-streptomycin (all from Life Technologies GmbH, Darmstadt, Germany), and 1% HEPES (10 mM, pH 7.2) (Merck KGaA, Darmstadt, Germany)). Cells were counted and adjusted to a density of 0.5 × 10 6 cells/mL in R10 medium for these experiments.

Measurement of Cellular Morphology
Dying cells change in morphology, and this can be detected by scatter alterations in flow cytometry [18]. Apoptosis is indicated by cell shrinkage, increased cytoplasmic granularity and vacuolization. Apoptotic cells are therefore detected as a population with decreased FSC and increased SSC [19].

Flow Cytometry
Flow cytometry was performed with a Gallios cytofluorometer™ (Beckman Coulter, Fullerton, CA, USA). The excitation wavelength for FITC and PI was 488 nm; FITC fluorescence was recorded by the FL-1 sensor (525/38 nm band pass), PI fluorescence by the FL-3 sensor (620/30 nm band pass). MTO fluorescence was excited at 638 nm and recorded by the FL-7 sensor (725/20 nm band pass). Electronic compensation was used to eliminate any fluorescence bleed-through. Data were analyzed with Kaluza™ software (Beckman Coulter, Fullerton, CA, USA).

Conclusions
MDT reduces the chemotherapeutic burden of circulating leukocytes. In our experiments, we used SPION MTO with a high magnetic field gradient of 72 T/m, and compared this approach with conventional chemotherapy. As previously demonstrated using MDT (TD 20 SPION MTO ), we decreased the chemotherapeutic burden on healthy tissue and enhanced MTO in the tumor (Table 1, Figure 6B) [5].
Flow cytometry is the most appropriate technique for precise analysis at the single cell level. MTO and the closely related anthracycline antibiotics, daunorubicin, doxorubicin and idarubicin (data not shown) are clearly detectable by this method because of their inherent fluorescence [21]. Over the years, the technique has been improved [22]. In this study, we used flow cytometry to monitor the incorporation of MTO into non-adherent cells in cell cultures in vitro as well as into blood cells in vivo.
Using MFI , it was also possible to determine the MTO loading efficiency of SPION. Although flow cytometry cannot detect a single SPION because of its small size, we established a measurement strategy to quantify MTO loading under defined conditions. Depending on the amount of MTO bound to the particles, SPION had a hydrodynamic diameter of approximately 100 nm, as determined by dynamic light scattering. Under defined conditions of SPION content, dilution, and flow rate, quantifying MTO bound to SPION in this way can compete with the established HPLC method used to determine loading efficiency [16].
Using flow cytometry, we found a dose-dependent effect of MTO and SPION MTO on the proliferation and viability of non-adherent human Jurkat cells in vitro. The two compounds had very similar effects, confirming equal cytotoxic activity of these substances if they are not differentially routed by MDT. This finding confirmed earlier results of real-time cell analysis and lactate dehydrogenase release in adherent cancer cells [23]. As shown in previous studies, low and intermediate concentrations of non-loaded SPIONs did not affect cell viability [23]. In contrast to earlier experiments performed with MCF-7 cells (human breast carcinoma) [23], much lower MTO concentrations induced Jurkat cell death. This confirms the high lymphoblast susceptibility to MTO. AxV/PI staining identified MTO-induced cell death as early apoptosis or the later response of secondary necrosis.
Preservation of the immune system during cancer therapy is extremely important and the cytotoxic load in PBL can be used as a surrogate marker for the immune system burden. Unwanted cytotoxic drug uptake by PBL is a major obstacle in MTO dose escalation. Ballestrero therefore developed a treatment strategy compatible with the rescue of blood progenitor cells. He achieved an acceptable hematologic toxicity while escalating the dose of MTO up to 90 mg/m 2 [8].
During the past few decades, evidence emerged that low-dose chemotherapy has immunomodulatory or immunostimulatory effects [24]. Interestingly enough, it was the anthracyclines in particular which were reported to induce a specific anticancer immune response [25][26][27]. After the injection of mitoxantrone-treated B16-F1 cells into mice, their splenocytes displayed an increased cytolytic effect on B16-F1 cells [28].
Nanoparticles reportedly influence the immune system, causing either immunostimulation or immunosuppression [29]. CoCr nanoparticles, for example, have been found to inhibit T-cell proliferation, which may explain the fact that patients with metal-on-metal implants have reduced CD8 + T-cell populations [30]. Like chemotherapeutic agents, Fe 3 O 4 (magnetite) magnetic nanoparticles, especially in low-doses, were found to affect immune function in mice [31]. In our in vivo and in vitro settings, SPION displayed no detectable cytotoxicity.
In comparison with conventional routes of administration, spillover of the chemotherapeutic agent MTO into healthy circulating PBL is significantly reduced by MDT-based treatment regimens. MDT reduces the major adverse effects of chemotherapy and preserves PBL, taken as a surrogate population of the immune system. We conclude that MDT saves drug and spares the immune system from cytotoxicity.