Loading erythrocytes with maghemite nanoparticles via osmotic pressure induced cell membrane pores

Loading Erythrocytes with Maghemite Nanoparticles via Osmotic Pressure Induced Cell Membrane Pores

Mounir Ibrahima,b, Leonard Weea, Martin Saundersb,c, Robert C. Woodwarda,b and Timothy G. St. Pierrea,b,*

aSchool of Physics, bCentre for Strategic Nanofabrication, and cCentre for Microscopy, Characterisation and Analysis, The University of Western Australia, Crawley, Western Australia 6009

*E-mail: [email protected]

Abstract. Encapsulating magnetic nanoparticles within red blood cells is one strategy for extending the lifetime of magnetic resonance imaging contrast agents in the bloodstream. Human red blood cells were incubated for 12 hours with iron oxide (γ-Fe2O3) nanoparticles with a broad range of particle and aggregate sizes (ranging from 10 to 600 nm) at different osmolarities ranging from 100 to 290 mOsm before being returned to an osmolarity of 300 mOsm. Concentrations of nanoparticles trapped within the cells were measured using transmission electron microscopy and iron-mapping by electron energy loss spectroscopy. An osmolarity of 200 mOsm was found to be the optimal condition for loading of the cells with nanoparticles. At this osmolarity, it was shown that the concentration of particles within the cells relative to the average concentration in the suspension is maximized. At 200 mOsm, the maximum size aggregate of particles that entered the cells was approximately 120 nm.

Keywords: Red blood cell; magnetic nanoparticles; contrast agent; magnetic resonance imaging PACS: 87.57.cj; 87.85.jf; 87.85.Rs; 87.16.Gj

INTRODUCTION

In clinical magnetic resonance imaging (MRI) studies, contrast agents are often used to enhance the contrast between different tissues by altering the local proton relaxation times 1. Recent developments in this area include the use of nanoparticles with large magnetic moments as contrast agents. Such nanoparticles exhibit a property known as superparamagnetism at room temperature when they are below a certain size that depends on the magneto-crystalline anisotropy of the particles. Their superparamagnetic properties have the advantage of high magnetic susceptibilities which result in large proton relaxivities when the particles are suspended in aqueous media. Magnetic particles are cleared from the blood stream via ingestion by macrophages. In order to extend the lifetime of magnetic particles in the blood stream, attempts have been made to encapsulate magnetic particles within red blood cells (RBCs) so that macrophage clearance is slowed. A convenient method for loading RBCs with magnetic nanoparticles involves inducing pores in the RBC membrane by varying the osmotic pressure. A key strategy for loading RBCs with magnetic particles is to incubate them in the presence of the particles under hypo-osmolar conditions 2, 3. The RBCs can then be re-sealed by bringing the osmolarity of the medium back up to physiological values. Here, we investigate the optimal conditions for loading RBCs by this technique and identify the limit of particle (or aggregate) size that can be incorporated through the pores.

METHODS AND MATERIALS

Maghemite (γ-Fe2O3) nanoparticles were obtained from Sigma Aldrich. Magnetometry measurements at 300 K were made using a Quantum Design MPMS-7 SQUID-based magnetic susceptometer. The particles were found to exhibit a wide range of aggregate sizes in aqueous suspension with larger aggregates falling out of suspension over time. In order to produce a stable suspension of particles, 50 mg of γ-Fe2O3 were mixed with 10 mL of distilled

375

Downloaded 16 Dec 2010 to 130.95.35.78. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions

water. The suspension was vigorously shaken for two minutes using a vortex mixer and was then sonicated for 15 minutes using an ultrasonic homogenizer (Biologics, Model 3000). The suspension was then centrifuged for 30 minutes at 3000 g. The supernatant was decanted and used as the stock γ-Fe2O3 suspension for the remaining steps. The size distribution of the nanoparticles in the stock suspension was measured using dynamic light scattering (DLS) (Zetasizer Nano-ZS, Malvern Instruments). Human blood (approx 8 mL) was collected into heparinised tubes and was centrifuged at 3000 g at 4°C for 10 minutes to separate the RBCs from the plasma. The RBCs were washed three times with phosphate buffered saline (PBS, 300 mOsm, pH 8). A stock suspension of RBCs at 300 mOsm was prepared by suspending approximately 2 mL of packed RBCs in 10 mL of PBS. The following protocol was used to incubate the RBCs in the presence of a stable suspension of γ-Fe2O3 nanoparticles at a given osmolarity. X mL (where X is 0.13, 0.4, 1.0, 2.0 or 4.0) of the γ-Fe2O3 suspension were added to Y mL (where Y is 4, 2, 2, 2 or 2) of the stock RBC suspension at 300 mOsm to produce a suspension of RBCs and magnetic nanoparticles at 290, 250, 200, 150 or 100 mOsm respectively. The RBCs were then incubated with γ-Fe2O3 nanoparticles at 4°C for 12 hours with gentle shaking using a vortex mixer. After the incubation period, the osmolarity of the suspension was brought back up to 300 mOsm by adding Z mL (where Z is 0.4, 1.2, 3.0, 6.0 or 12.0) of PBS at 400 mOsm to the suspension of cells and nanoparticles.

The resulting cell suspensions were prepared for transmission electron microscopy (TEM) in the following way. The suspensions were centrifuged and the cell pellet was re-suspended in PBS at a volume ratio of 1:1. The suspension was then fixed with glutaraldehyde at a volume ratio of 1:20. After fixation, the cells were serially dehydrated for 40 seconds each time in consecutive solutions of 50, 70, 95 and 100% ethanol. The sample was dried by washing twice in 100% acetone. After dehydration, the sample was infiltrated with Spurrs resin for 3 minutes each in consecutive solutions of acetone and Spurrs resin at volume ratios of 3:1, 1:2 and 1:3, respectively. The sample was bonded in 100% Spurrs resin followed by polymerization overnight at 70°C. An ultramicrotome was used to cut 100-nm thick sections from the resin block. The sections were then mounted on carbon-coated copper grids. A series of control samples was prepared using the above protocol using distilled water in place of the stock γ-Fe2O3 suspension. TEM (JEOL-2100, Philips) was used to record bright-field images of these sections. Energy-filtered TEM (EFTEM) was conducted using an electron energy-filter system (Gatan, Tridiem). Elemental maps for iron were acquired by the conventional three-window method 4 using the Fe M-edge. Background (pre-edge) images were acquired at 45 eV and 50 eV, with the signal (post-edge) image acquired at 59 eV. All images were acquired with an energy window (slit width) of 5 eV. These conditions were optimized to provide good signal-to-noise ratio with suitable background removal. All TEM images were acquired at 120 keV at room temperature. Particle size distributions within the RBCs and outside the RBCs were measured by analysing at least 20 iron maps, each with a field of view of 5 × 5 μm. Freeware ImageJ (NIH) was used to measure the intracellular and extracellular areas in each field of view, as well as the intracellular and extracellular size distribution for particles/aggregates specifically containing iron. Each particle/aggregate was fitted with an ellipse and its effective size d was calculated as 2√( ab), where a and b are the semi-minor and semi-major axes of the fitted ellipse, respectively. The volume of each particle or cluster was estimated as (1/6)πd3. Concentrations of γ-Fe2O3 trapped within the cells were then estimated from the ratio of total cell to nanoparticle volume in the imaged 100-nm thick sections by assuming that each particle/aggregate had the density of bulk γ-Fe2O3. In order to determine the iron concentration in the stock γ-Fe2O3 suspension, 1 mL of stock γ-Fe2O3 was added to 10 mL of concentrated HNO3 and heated at 95°C to reduce the volume to 1 mL in total. The digestion was then made up to 30 mL with dilute (1%) HNO3. The iron concentration in the digested solution was measured by inductively coupled plasma mass spectrometry (ICP-MS).

RESULTS

Figure 1 shows a TEM image of γ-Fe2O3 nanoparticle aggregates prepared directly from a suspension in distilled water by casting them on the TEM grid. This figure demonstrates that the majority of particles are roughly spherical with the smallest particles being approximately 10 nm in size. DLS measurements on the suspensions indicated particle sizes ranging from 30 to 300 nm demonstrating that stable aggregates were present in the aqueous suspension as well as being observed in the dry state under TEM. A magnetic hysteresis curve of the γ-Fe2O3 nanoparticles at 300 K is presented in Figure 2. The magnetometry measurements show that the γ-Fe2O3 nanoparticles mainly exhibit superparamagnetic behavior but with some remanent magnetization at zero field (Figure 2(b)). The saturation magnetization for the γ-Fe2O3 is 66 emu/g, a value slightly lower than the published value for bulk γ-Fe2O3 (76 emu/g) 5. Figure 3(a) is a TEM micrograph of the γ-Fe2O3 nanoparticles embedded in resin showing the presence of a large aggregate. Figure 3(b) shows the EELS iron map corresponding to the same region imaged in Figure 3(a). Figure 4(a) shows a TEM image of an unstained section through a red blood cell after

376


incubation with the γ-Fe2O3 nanoparticles at 200 mOsm with its corresponding iron map image in Figure 4(b). The image indicates that nanoparticles can be found inside the cell. In comparison, a TEM and iron-map image of an unstained section through a control RBC are shown in Figure 5. While electron dense particles can be seen in the TEM image, the iron map indicates the absence of γ-Fe2O3 nanoparticles. Similar TEM and iron-map images of RBCs loaded at 150 and 250 mOsm showed low uptake of γ-Fe2O3 into the RBCs. No iron-containing particles were found inside the RBCs when loaded at 290 mOsm (see Figure 6). The concentrations of γ-Fe2O3 iron found inside the cells determined from analysis of the TEM iron map images together with the expected average concentrations of γ-Fe2O3 iron in the final cell suspensions determined from the ICP measured concentration of iron in the stock nanoparticle suspension are shown for each preparation in Figure 7. The concentration of γ-Fe2O3 iron in the cells incubated at 290 mOsm was found to be zero. Cells incubated at 250 mOsm appeared to have intracellular γ-Fe2O3 iron concentrations slightly less than the predicted average γ-Fe2O3 iron concentration. Cells incubated at 200 mOsm were found to have intracellular γ-Fe2O3 iron concentrations somewhat higher than the predicted average γ-Fe2O3 iron concentration. For the cells incubated at 150 mOsm, only a few were found intact. Although the intracellular γ-Fe2O3 iron concentration was found to be less than the predicted average γ-Fe2O3 iron concentration, the error on the measurement could be somewhat larger than that calculated because of the small number of cells measureable. At 100 mOsm, no intact cells were found and hence no intracellular γ-Fe2O3 iron concentration is reported.

The distributions of particle/aggregate sizes found inside and outside of the cells is shown in Figure 8. The largest clusters found inside the cell were approximately 120 nm (found in the cells incubated at 200 mOsm).

FIGURE 1. TEM of γ-Fe2O3 nanoparticles cast onto grid from aqueous suspension.

DISCUSSION

Previous studies of the incorporation of magnetic nanoparticles into RBCs have shown that it is possible to load particles into the cells by inducing pore formation through osmotic pressure 2, 3. This study aimed to find the optimum osmolarity for loading and the limits on the particle sizes that can be incorporated within the cells.

While the fundamental particle sizes of the γ-Fe2O3 used in this study were generally less than 100 nm, both the TEM and DLS indicated that particles were mostly found in aggregates ranging up to a few hundred nanometers in size. The suppressed saturation magnetization compared with the literature value for bulk γ-Fe2O3 5 is consistent with the TEM observation that the fundamental particle sizes range down to approximately 10 nm. The mixture of superparamagnetic and magnetically blocked behavior also indicates a large range of particle sizes.

The observation that no γ-Fe2O3 particles were found within cells incubated with nanoparticles at 290 mOsm (Figures 6, 7, and 8) while γ-Fe2O3 particles were found within cells incubated at lower osmolarities (Figures 4, 7, and 8) strongly suggests that the mechanism by which particles have entered the cells is through osmotic pressure induced pores in the cell membrane.

377


(a) (b)

FIGURE 2. (a) Magnetization vs applied magnetic field for the γ-Fe2O3 nanoparticles and (b) magnetization vs applied field

behavior close to zero field. .

(a) (b)

FIGURE 3. (a) Transmission electron micrograph of γ-Fe2O3 nanoparticles embedded in resin and (b) iron map corresponding

to region imaged in (a) generated from electron energy loss imaging.

(a) (b)

FIGURE 4. (a) TEM of unstained human red blood cell after incubation with γ-Fe2O3 nanoparticles at 200 mOsm and (b) iron

map corresponding to region imaged in (a) generated from electron energy loss imaging.

378


(a) (b)

FIGURE 5. (a) TEM of unstained control human red blood cell after incubation at 200 mOsm with no γ-Fe2O3 nanoparticles and (b) iron map corresponding to region imaged in (a) generated from electron energy loss imaging.

(a) (b)

FIGURE 6. (a) TEM of unstained human red blood cell after incubation with γ-Fe2O3 nanoparticles at 290 mOsm and (b) iron map corresponding to region imaged in (a) generated from electron energy loss imaging.

FIGURE 7. Estimations of concentrations of γ-Fe2O3 Fe inside cells determined from TEM together with predicted average concentration of γ-Fe2O3 Fe in final cell suspension determined from ICP measurement of stock γ-Fe2O3 suspension. Error bars

on measurements inside cells are derived from the standard deviation of values determined from the twenty fields of view analysed. Error bars on the average Fe concentration values are determined from the analytical error on the ICP measurement

method

379


FIGURE 8. Distribution of particle/aggregate sizes inside (black) and outside (gray) cells after incubation with γ-Fe2O3 nanoparticles at the indicated osmolarity.

Figures 7 and 8 suggest that the optimum incubation condition for loading RBCs with nanoparticles is at 200 mOsm. At 200 mOsm, the ratio of the γ-Fe2O3 iron concentration within the cells to the average concentration in the cell suspension is greatest (Figure 7). The data suggest that the concentration within the cells is greater than that outside the cells. While there could be a trapping of particles within cells while the pores are still open, resulting in a greater concentration inside than out, the most likely explanation of this observation is that there is a systematic error in our calculation of intracellular concentrations. There may be a systematic error in the section thickness. Also, our calculations have assumed that the aggregates of nanoparticles have the density of γ-Fe2O3, whereas the aggregates must have a density somewhat less than that for γ-Fe2O3 because of the interstitial spaces between the nanoparticles. Hence it is unlikely that the intracellular concentration exceeds the extracellular concentration in the 200 mOsm preparation. Nevertheless, the error on the density estimate is likely to be systematic across the different loading conditions and as such the data strongly indicate that there is a more efficient loading of nanoparticles at 200 mOsm compared with the other incubation conditions. Figure 8 also suggests a more efficient loading of particles since a wider range of aggregate sizes is observed inside the cells. The increased range of aggregate sizes observed

380


in cells incubated at 200 mOsm compared with those incubated at 250 mOsm may indicate larger pores forming under the greater osmotic pressure. The cut-off of approximately 120-nm for the aggregate size distribution for cells incubated at 200 mOsm suggests that the pore sizes range up to at least this size. Interestingly, it has been found that the maximum pore sizes induced in RBC membranes by electroporation are also approximately 120 nm 6. The aggregate size distribution found in the cells incubated at 150 mOsm is not as broad as that found for the cells incubated at 200 mOsm. This observation may reflect the fact that very few cells survived the incubation process at 150 mOsm. Possibly only those with smaller pore sizes survived. Alternatively, the very low number of cells observed may result in the observed distribution being biased towards the individual cells measured because of a lack of statistical averaging.

CONCLUSION

Magnetic nanoparticles can be loaded into RBCs by temporarily incubating the cells with the nanoparticles under reduced osmolarity conditions. The optimum osmolarity for particle loading is close to 200 mOsm, a condition at which the greatest fraction of particles in the incubating medium enter the cells. The maximum size of aggregate that enters the cells under these conditions is approximately 120 nm. Osmolarities of 150 mOsm or lower cause excessive cell destruction. RBCs loaded with γ-Fe2O3 nanoparticles by this method are potential candidates for MRI contrast agents with an extended residence time in the blood stream.

ACKNOWLEDGMENTS

We thank John Murphy for assistance with resin embedding and sectioning of samples, Greg Black for assistance with ICP measurements, and Jessie Prestage for venepuncture and blood collection. This research was partly supported under Australian Research Council's Discovery Projects funding scheme (project number DP0985848).

REFERENCES

1. P. F. Renshaw, C. S. Owen, A. E. Evans and J. S. Leigh, Jr., Magn. Reson. Imaging 4, 351-357 (1986). 2. A. Antonelli, C. Sfara, L. Mosca, E. Manuali and M. Magnani, J Nanosci. Nanotechno. 8, 2270-2278 (2008). 3. M. Brähler, R. Georgieva, N. Buske, A. Muller, S. Muller, J. Pinkernelle, U. Teichgraber, A. Voigt and H. Baumler, Nano

Lett. 6, 2505-2509 (2006). 4. R. Brydson, Electron Energy Loss Spectroscopy. (Telos Springer-Verlag, Emeryville, CA, 2001). 5. A. Millan, A. Urtizberea, N. J. O. Silva, F. Palacio, V. S. Amaral, E. Snoeck and V. Serin, J. Magn. Mag. Mater. 312, L5-

L9 (2007). 6. D. C. Chang and T. S. Reese, Biophysical Journal 58, 1-12 (1990).

381


Loading erythrocytes with maghemite nanoparticles via osmotic pressure induced cell membrane pores

Documents