Abstract
By complexing ferumoxides or superparamagnetic iron oxide (SPIO) to transfection agents (TAs), it is possible to magnetically label mammalian cells. There has been no systematic study comparing TAs complexed to SPIO as far as cell labeling efficiency and viability. This study investigates the toxicity and labeling efficiency at various doses of FEs complexed to different TAs in mammalian cells. Different classes of TAs were used, such as polycationic amines, dendrimers, and lipid-based agents. Cellular toxicity was measured using doses of TAs from 1 to 50 μg/mL in incubation media. Iron incorporation efficiency was measured by combining various amounts of FEs and different doses of TAs. Lipofectamine2000 showed toxicity at lowest dose (1 μg/mL), whereas FuGENE6 and low molecular weight poly-
Introduction
Different investigators have tried different techniques to incorporate dextran-coated superparamagnetic iron oxide (SPIO) nanoparticles into cells and to monitor the migration of locally implanted or intravenously administered magnetically labeled cells by cellular magnetic resonance imaging (MRI) [1–6]. Ferumoxides (FEs), an SPIO complexed to polycationic transfection agents (TAs) are used to magnetically label stem cells and other mammalian cells for cellular MRI [7–10]. TAs are macro-molecules possessing an electrostatic charge and are used for nonviral transfection of DNA into the nucleus [11–14]. There are various classes of TAs available including: polyamines [poly-
The purpose of this study was to determine the effect of different TAs either alone or in combination with SPIO at various concentrations on cell viability and function, and to optimize a dose ratio (SPIO/TA) that would facilitate the incorporation of SPIO into the cells.
Materials and Methods
Ferumoxides (Feridex IV, Berlex Laboratories, Wayne, NJ)
FEs are approximately 70–150 nm in size and are provided as a solution at an iron content of 11.2 mg/mL (11.2 of iron μg/μL). Different concentrations of FEs (μg/mL) were used to make FE/TA complexes.
Poly-L-lysine (Sigma, St. Louis, MO)
PLL is a polyamine group of TA. PLL having different MWs (0.5 to 388 kDa) were used. A stock solution was made from each MW at a concentration of 1 mg/mL in distilled water and kept frozen at −20°C until required.
Poly-L-omithine (Sigma)
PLO is also a polyamine TA and has a MW of 39 kDa. A stock solution was made at a concentration of 1 mg/mL in distilled water and kept frozen at −20°C until required.
Poly-L-arginine (Sigma)
PLA is also a polyamine TA and has a MW of 141.4 kDa. A stock solution was made at a concentration of 1 mg/mL in distilled water and kept frozen at — 20°C until required.
Superfect (Qiagen, Valencia, CA)
Superfect is a heat-activated dendrimer used as a TA. The stock solution contains 3 mg/mL of active reagents and was kept at 4°C.
Lipofectamine2000 (Invitrogen, Life Technology, Gaithersburg, MD)
Lipofectamine2000 is a lipid-based TA widely utilized for effective transfection of DNA into different cell lines. The stock solution contains 2 mg/mL of active reagents and was kept at 4°C.
FuGENE6 (Roche Diagnostic, Indianapolis, IN)
FuGENE6 is an advanced nonliposomal TA, which is used to transfect DNA to a variety of eukaryotic cells. The concentration of active reagent in FuGENE6 is not disclosed by the manufacturer, but the recommended dose is 1:2–3 [1 μL (1 μg/μL) of DNA to 2–3 μL of FuGENE6). Therefore, we assumed the stock solution contains at least 1 mg/mL of active reagents, and the doses were calculated accordingly.
All TAs used in this study were evaluated at various concentrations from 1 μg to 50 μg/mL in media. Of note, repeated thawing of the stock solutions of TA was not performed, nor is it recommended.
Measurement of Zeta Potential
To measure the zeta potential of the TAs and FEs [in millivolts (mV)], 2 mL working solutions of TAs and FEs were made at a concentration of 50 μg/mL and measured using a Zeta Potential analyzer (Brookhaven Instrument, New York). A total of 10 runs were collected, and results are reported as mean and standard deviation.
Cell Lines and Culture Conditions
Adherent cell lines [human cervical carcinoma (HeLa, NCI, Bethesda, MD), and human mesenchymal stem cells (MSCs, Cambrex, Baltimore, MD)], and cells grown in suspension [lymphoblastic cells LADMAC (ATCC, Manassas, VA)] were cultured and propagated according to the manufacturers' recommendations. To determine the toxic effects of different TAs at various concentrations, alone or complexed with various concentrations of FEs, cells were grown in either 24-well or 96-well plates. Results were averaged from at least three wells to determine differences from control cells without FE or TA. Adherent cells were grown in culture flasks until they were 80% confluent for histology, determination of labeling efficiency, and average iron content per cell.
Cell Proliferative Activity and Viability
Proliferative activity and viability of FE/TA labeled or TA treated cells were evaluated using an MTT (3- [4,5–2-yl]-2,5-diphenyl tetrazolium bromide) assay (Roche Molecular Biochemicals, Indianapolis, IN) and Trypan blue dye exclusion test, respectively, as described previously [8,9]. The values of the MTT assays of labeled and TA treated cells were expressed as the percentage of corresponding control cells.
TA Toxicity to Cell Lines
To determine the toxicity of TA, 2 × 105 LADMAC cells were put into 96-well plates and then incubated with different doses of TAs at 37°C in 95%/5% air/CO2. MTT assays were performed at 24 and 72 hr. The results of the MTT assays were expressed as the percentage of the corresponding control cells.
Cell Labeling Efficiency with FE/TA Complexes
Specified number of HeLa cells were incubated in cultural flasks and grown until 80–90% confluent. FE/TA complexes with different FE/TA ratios were made and added to the cells for incubation overnight. The final concentration in media of FEs was 25 μg/mL with the concentrations of TAs at 1, 5, and 25 μg/mL. Following incubation, cells were washed twice with phosphate-buffered saline (PBS), and were evaluated for the presence of intracellular iron on histology with Prussian blue stains. Cells were also collected in NMR tube for measurement of intracellular iron concentration.
Effects of Altering the Ratio of FE/TA on Labeling Efficiency, Cell Proliferative Activity and Viability
To examine the effect of altering the ratio of FE/TA on labeling efficiency, cell proliferative activity and viability, we focused on labeling cells with the 388 kDa PLL. MSCs and HeLa were plated in either 24- or 96-well plates and were grown until they were 80% confluent with refreshing of old media. FEs at a concentration of 50 μg/mL were put into a mixing flask or tube containing respective complete medium and then different concentrations of PLL (388 kDa) were added. The solution was then mixed for 30–60 min. In addition, FEs at a concentration of 50–500 μg/mL were mixed with PLL (388 kDa) at a ratio of FE/PLL of 1 μg/mL:0.03 μg/mL (PLL = 1.5–15 μg/mL) and added to the solution to determine the effect of increasing amounts of FEs would have on cell viability. For both experiments, equal volumes of FE/PLL complexes were added to the existing media in the cell culture and incubated for 24 hr at 37°C in 95%/5% air/CO2. After incubation overnight, cells were washed twice, replaced with fresh media and MTT assay and Trypan blue dye exclusion test were performed the following day. The results of MTT assays were compared with that of control cells. The results of Trypan blue tests were expressed as percent dead cells of total cells. To determine the iron concentration, the labeling was performed in culture flasks, cells were collected in NMR tube and average iron concentration per cell was measured.
Histology
After incubation with FE/TA, cells were washed twice to remove excess FE/TA, trypsinized (adherent cells), and transferred to cytospin slides. Cells were fixed with 4% glutaraldehyde, washed, incubated for 30 min with 2% potassium ferrocyanide (Perl's reagent for Prussian blue [PB] staining) in 3–7% hydrochloric acid, washed again and counterstained with nuclear fast red. Representative labeled cells were subjected to standard transmission electron microscopy to see the distribution of FE/PLL complexes in the cells.
Measurement of Iron Content
The iron concentration of the labeled cells was assessed using NMR relaxometry. Briefly, cells were incubated with different FE/TA complexes and at certain time points, washed cell suspensions with known cell density were first dried at 110°C for overnight and then completely digested in a mixture (500 μL) of perchloric and nitric acid at a 3:1 ratio. The samples were digested for at least 3 hr at 60°C using a heating block. For these 500-μL samples, the NMR relaxation rates 1/T1 and 1/T2 (sec−1) were then measured at room temperature and 1.0 Tesla using a custom-designed variable field NMR relaxometer [8,20] and a Carr–Purcell–Meiboom–Gill (CPMG) pulse sequence. Iron concentration in each sample was calculated from a standard curve that was derived from calibration standards of ferrous chloride containing 0.01 to 10 mMol/L of iron in the same acid mixture. Iron concentration was expressed as an average picogram iron per cell.
Data Analysis
Data are expressed as mean ± SD from three to nine data samples unless stated otherwise. MTT based toxicity or proliferation data are expressed as percentage of the corresponding control unlabelled cells and significant test was performed by ANOVA (Statview 4.51, Abacus Concept, Berkeley, CA) followed by post hoc test (Fisher's PLSD). A p value of < .05 was considered significant.
Results
Zeta Potentials
Table 1 is a summary of the results of the zeta potentials for FEs and the TAs used. There were wide variations in the zeta potential of TAs from + 2 to 88 mV, whereas FEs have a negative zeta potential of −32 mV. Polycations containing lysine had higher zeta potentials but no MW dependency was observed.
Toxicity of TAs
Figure 1 contains the results of the MTT analysis for LADMAC cells incubated in the presence of various concentrations of TAs for 24 hr (Figure 1A) and 72 hr (Figure 2B). PLL, which has been used as a transfection as well as a cell adhesion agent, showed concentration and MW-dependent cell toxicity when compared with cells grown in media alone (control). All MW of PLL demonstrated a significant decrease in cell proliferative activity starting at 5 μg/mL dose at 24 hr; however, at 72 hr PLL 0.5–2 kDa demonstrated no loss of cellular proliferative activity even at the highest dose (50 μg/mL) as compared to controls. PLL 23 kDa and PLL 388 kDa showed significant loss in proliferative activity compared with controls at 72 hr at concentrations of 25 and 50 μg/mL, respectively. Cells incubated with PLO and PIA at MWs of 39 and 141 kDa, respectively, exhibited a significant decrease in proliferation at a concentration of 1 μg/mL at 24 hr and at 10 μg/mL at 72 hr compared with controls. Superfect and Lipofectamine 2000 demonstrated significant decrease in cell proliferation activity at concentrations starting at 5 μg/mL at both 24 and 72 hr. FuGENE6 demonstrated no effect of cell proliferative activity at the highest concentration tested (50 μg/mL) at both 24 and 72 hr.
Cell Iron Content with Different FE/TA Ratios
Table 1 contains a summary of the iron content per cell for the various combinations of FE/TA ratios and notes significant differences among the TAs and different doses. FE/PLO resulted in the highest incorporation of iron in the LADMAC cells at 9.92 ± 0.45 pg/cell when used 25:1 μg/mL (FE vs. TA), and FEs alone showed the lowest at 1.08 pg/cell. With the rest of the combinations of FE/TA, iron incorporation in the cells was in between this range following overnight incubation. At a ratio of FE/TA of 25:5 μg/mL, superfect resulted in the 15–18 ± 0.87 pg of iron/cell and the low MW PLL had the lowest incorporation of iron in cells at 1.26 ± 0.05 pg/cell. Of note when FEs were complexed to PLO or PLA at a ratio of 25:5 μg/mL, there was a significant (p ≤ .001) decreased iron incorporation in cells.
Details of Zeta Potentials of FEs and the TAs as well as the Iron Incorporation into HeLa Cells with Different Doses of the TAs
There are significant differences (p ≤ .001) among the TAs.
Significant differences between 1 and 5 μg/mL of corresponding TAs.
The dose effects of different transfection agents alone at 24 hr (A) and 72 hr (B) after incubation. Data are expressed as mean ± standard deviation of percent corresponding control. Note the recovery of proliferative activity at 72 hr for the polycations only. *Significant differences with the corresponding control.
Iron Incorporation and Toxicity of PLL at Constant Concentration of FEs
The toxicity profile and the ability to magnetically label MSCs change significantly (p ≤ .001) for PLL at an MW of 388 kDa when the TA is complexed to FEs at an FE concentration of 25 μg/mL compared with control cells (Figure 2). PLL 388 kDa at a dose of 7.5 μg/mL showed no loss of cell proliferative activity (MTT assay) (Figure 2B) compared to control, however, 20% of the cells were dead on Trypan blue analysis compared with <5% for control media alone (p ≤ .001). At a ratio of 25:7.5 μg/mL of FE/PLL (MW = 388 kDa), iron incorporation was very high compared with the other ratios used. When PLL was in excess of FEs (ratio of FE/PLL 25:30 μg/mL), there was a significant effect on cell proliferation and viability with almost complete cell death compared to control.
Increasing Concentration of FEs at Constant FE/PLL (388 kDa) Ratio
To determine the dose-limiting toxicity of FEs to cells, increasing amounts of FE/PLL were added to the cells at a fixed ratio of FE/PLL of 1:0.03 μg/mL and incubated overnight. Cells were collected for viability, iron concentration measurement, and histology at specific time intervals. There was a linear increase in cellular uptake of iron with increasing concentration of FEs in media. However, cell viability was compromised in MSCs at a dose of 100 μg/mL of FEs (Figure 3). The mean iron content per cell was highest when MSC or HeLa cells were incubated with an FE/PLL ratio of 250:7.5 μg/mL. However, approximately 40% of MSCs were dead based on Trypan blue dye exclusion. HeLa cells incubated at the highest ratio of FE/PLL (250:7.5 μg/mL) demonstrated no significant increase in percentage of dead cells when compared to control unlabeled cells. At an FE/PLL ratio of 250:7.5 μg/mL, all cells completely contained Prussian blue positive endosomes (Figure 4). Representative transmission electron microscopy images of human MSCs (i.e., adherent cells) showed endosomal incorporation of the FE/PLL complexes by overnight incubation (Figure 4E), although at 1 hr complexes were attached at the surface of the cells. No definite uptake into the nucleus was observed and after washing the complexes were not seen attached to the surface of the cells by overnight.
The effect of increasing amount of PLL (388 kDa) to a fixed dose of FEs (25 μg/mL) on the iron incorporation (A), proliferative activity (B), and viability (C) in mesenchymal stem cells. Note less iron incorporation with 100% cell death at a dose ratio of 25:30 μg/mL for PLL 388 kDa. *Significant differences compared to unlabeled cells.
Discussion
Polycationic amines (i.e., PLL, PLO, and PLA) are not commonly used DNA TAs primarily because of the variable transfection efficiency and undesirable endosomal capture of DNA, and variation in MW, size, and polymerization [21,22]. PLL has been used to effectively treat glass slides, cover slips, and culture flasks for better attachment or adherence of cells [23,24], but in general, polyamines are toxic to cells [18,25]. Combining excessive amounts of polycations with cells results in pore formation in the membrane and an intracellular ionic imbalance that can ultimately lead to cell death [18,26,27]. This study clearly shows that there is a concentration and MW dependency for cell toxicity and viability for polyamine TAs alone. PLL (which is used in combination with dextran-coated SPIO to magnetically label cells [7–9,20]), when incubated alone with LADMAC cells, showed significant decrease in cell proliferation with a concentration of 5 pg/mL media at 24 hours when compared with control cells. However, the initial decrease in MTT activity compared with control cells observed with the PLL examined, recovered with longer incubation times up to a maximum concentration of PLL 25 pg/mL for MWs 0.5–388 kDa. The near normalization observed on the MTT assay at 72 hr for the LADMAC cells at PLL concentrations below 25 pg/mL might be due to metabolic deactivation of PLL by endosomal/lysosomal enzymes [28]. The toxicity profile observed in the current study is similar to previous reports [18,28]. Although free PLL (i.e., 23 and 388 kDa) resulted in a significant decrease proliferation at a dose above 1 pg/mL at 24 hr, when it is complexed with FEs there is a clear shift in the maximal tolerated concentration by the cells before significant cell death (i.e., greater than unlabeled cells of approximately 5%) occurs. Complexing PLL with FEs at the ratios used in this study results in no free PLL in the media and therefore does not affect cell viability or proliferation. When PLL is added in 10-fold or more excess to the amount needed to complex with FEs (FE/PLL >25:7.5 pg/mL ratio) (Figure 2), there is an increase in the iron content in cells accompanied by an increase in cell death (see below). We observed that these FE/polycation complexes tend to make larger complexes in the incubation media over time (data not shown) if the balance between the FEs and polycations is not maintained and if there is no serum present in the incubation media. The ratio used in this study resulted in the complexes having an increase in the zeta potential over FEs alone, and based on the stability of the zeta potentials for FE/TA complexes, there was no evidence of dissociation of the complex with time [11]. Furthermore, by vigorously pipetting or placing sample on vortex for 15–30 sec breaks down any clumps of the FE/TA complex that may form over time to nanosized particles.
The effect of increasing amount of FEs with a fixed ratio of PLL (1:0.03 w/w) on the iron incorporation (A, C) and viability (B, D) in MSCs (A, B) and HeLa cells (C, D). Comparing the HeLa cells, MSCs showed more iron incorporation and loss of cell viability with a dose of 100:3. No loss of cell viability is seen with a dose of 250:7.5 for HeLa cells
The other macromolecular polycations evaluated (e.g., PLO and PLA) had a significant decrease in proliferative activity for LADMAC cells at 24 hr even with the lowest dose of TA used (1 μg/mL). However, cells also showed recovery of the MTT assay similar to unlabeled cells by 72 hr at concentrations of 25 μg/mL similar to what was observed with PLL. This finding is consistent with the hypothesis that cells may be able to deactivate macromolecular polycations in the endosomes/lysosomes with longer incubation times.
As compared to polyamines, heat-activated dendrimers (superfect) and liposomal TA (e.g., Lipofectamine2000) had a significantly greater effect on cell proliferative activity compared to control cells at low concentrations (>1 μg/mL) at 24 hr, without recovery at 72 hr (Figure 1B). The inability to recover proliferative activity at 72 hr may be due to the inability of the lysosomal environment to degrade these TAs. The results from this study on cellular toxicity of superfect and Lipofectamine2000 are similar to what has been reported in the literature [29].
In contrast to other SPIO complexes being used for cell labeling [4,5], our previous studies, as well as this study, proved the endosomal incorporation of FE/TA complexes and none of the complexes was seen inside the nucleus (Figure 4). Incorporation into the nucleus would damage the DNA by forming reactive oxygen species (ROS) [30,31]. Our previous studies showed an echo time dependence of the R2 relaxation properties of FE/TA complexes when used for labeling indicative of internalized FE/TA complexes in different cell lines [7,20]. The average iron content in cells following incubation with FEs complexed to various TAs did not correlate with the TA zeta potential or MW. This observation was true for all the polyamines as well as for the lipid-based transfection agents (e.g., Lipofectamine2000 and FuGENE6). Labeling efficiency based on measured average iron per cell was high for the FE/superfect complex despite the large difference in zeta potentials between the two macro molecules (see Table 1). The increased labeling efficiency may be due to the heat-activated dendrimer's ability to form a complex with the FEs nanoparticle at optimal size and surface charge distribution (i.e., proper titration of FEs to TAs for optimum iron incorporation to the cells) for endocytosis [29]. FuGENE6, an advanced lipid-based TA, had no toxicity up to concentrations of 50 μg/mL in media when incubated alone with cells, however, was inefficient at complexing with FEs for magnetically labeling cells (Table 1). These results are similar to the finding reported by Hoehn et al. [10] who reported an approximate 70% labeling efficiency of neural stem cells when FuGENE6 was complexed to dextran-coated SPIO. In the current study, PLL (388 kDa) and superfect demonstrated similar iron incorporation into the cells when the same incubation conditions were used. Despite superfect's effect on cell viability when incubated with cells alone for 24 and 72 hr, the heat-activated dendrimer was effective at complexing with the SPIO nanoparticles and therefore could be a reasonable, albeit more expensive, alternative to PLL (388 kDa) for magnetically labeling cells. The results from the current study indicate that toxicity and the amount of iron content per cell depend on the ratio of FE to TA as well as the total amount of iron in the media. We observed a significant decrease in MSC viability when the average iron content per cell was greater than 100 pg/cell when using FE/PLL at high concentrations in the media (Figure 3). The increase in stem cell death due to high concentrations of iron oxide in the endosomes may be due to acid digestion at lysosomal pH, releasing iron (Fe2+) into the cytoplasm. This would facilitate the possibility of a Harber-Weiss reaction with excessive free radical formation or ROS, thereby causing damage to DNA and cell death [30,31]. Although it appears that human MSCs are sensitive to high concentrations of intracellular iron in endosomes, a similar toxicity profile was not observed with malignant HeLa cells at iron concentrations of >40 pg/cell. The relative lack of toxicity observed when labeling HeLa cells with FE/PLL may be due to either the proliferative activity of the cells or less free radical formation and toxicity to the cells or that malignant HeLa cells are able to utilize free ionic iron in metabolic pathways at a faster rate and therefore results in less toxicity to the cell. It has been shown that malignant cells have the ability to neutralize the effect of free radical damage to DNA in the nucleus as compared to normal cells [32–34]. Therefore, this cellular defense mechanism may be responsible for limited toxicity and higher iron content observed with HeLa cells compared to human MSCs when grown in culture.
The endosomal iron in HeLa cells with different doses of FEs and PLL. (A) 25:0.75, (B) 50:1.5, and (C) 250:7.5. Representative transmission electron microscopic images of MSCs incubated with 25 μg of FEs and 0.75 μg of PLL at 1 hr × 25,000 (D) and overnight × 15,000 (E). Note the attachment of FE-PLL complexes at the surface of the cell by 1 hr and endosomal incorporation (arrows) by overnight. No definite nuclear uptake is seen.
In conclusion, FEs complexed to either high MW PLL or superfect would be an appropriate choice for magnetic cell labeling based upon toxicity profile and iron content per cell. We would recommend using a ratio of FEs to PLL between 1:0.03 μg/mL and 1:0.05 μg/mL as a starting point for magnetically labeling cells that could be detected in vivo by cellular MRI.
Footnotes
Acknowledgments
We thankfully acknowledge the generous help of Dr. S. Cheng and Miss Virginia Tanner of EM facility at NINDS.