Characterization of functionalized AuNPs
Different surface-functionalized AuNPs were synthesized according to the method described in our previous paper.42-43 AuNPs were functionalized with four ligands; citrate, dextran-10, dextrin and chitosan. It is known that except citrate, all other ligands show low toxicity, high dispersibility and enhanced biocompatibility. Dextran-10 (dex-10) molecule is a branched molecule, whereas dextrin and chitosan molecules are linear in nature. All AuNPs types were well dispersed in water and showed the SPR peak at ∼528 nm (Figure S1). An additional shoulder at 600 nm for dex-10-AuNPs is either due to the formation of some aggregates or to formation of networks via the branched dex-10 molecules. TEM micrographs show that all types of AuNPs are spherical with an average diameter for citrate-, dex-10-, chitosan- and dextrin-AuNPs of 15.6 ± 1 nm, 13 ± 2 nm, 23.5 ± 6 nm and 6.4 ± 2.5 nm, respectively (Figure S2).
The average hydrodynamic diameter (dH) for citrate-, dex-10-, chitosan- and dextrin-AuNPs determined by dynamic light scattering (DLS) in water was ∼19.2, ∼60, ∼64 and ∼40 nm, respectively (Figure S3A). The overall increase in diameter compared to TEM data is due to the hydrated expansion of the molecules in aqueous state. The zeta potential of functionalized AuNPs was measured (Figure S3B) to confirm the surface-functionalization. With the exception of chitosan-AuNPs which are positively charged due to the -NH2 groups, the surface of the other types of AuNPs was negatively charged due to the presence of –COOH and –OH groups.
Time-dependent absorption behavior of AuNPs in protein poor- and protein rich cell culture media
To understand the interaction of AuNPs with components of CCM, we exposed different surface- functionalized AuNPs to CCM composed of RPMI supplemented with 10% FCS and without FCS at 37 °C for different time intervals. RPMI medium supplemented with 10% FCS is considered as protein rich CCM (=7 mg/ml proteins), whereas RPMI without FCS considered as protein poor CCM. Protein poor CCM contains only proteins at very low concentration (ng/mL) in the form of growth factors. However, both CCM contain a high amount of amino acids, carbohydrates, vitamins and salts which can cover the surface of AuNPs.
First, we characterized the stability of the AuNPs over time in both protein poor- and protein rich CCM using UV-Vis spectroscopy. Figure 1 shows the UV-Vis spectra of the various surface- functionalized AuNPs in protein poor media at 37 °C for different time intervals (0.5, 4, 15 and 24 h). It can be seen that the SPR band of citrate-AuNPs shifted from 528 nm to longer wavelength 720 nm and the intensity was completely diminished with increasing time. Therefore, in the absence of serum proteins, the SPR band of the citrate-AuNPs rapidly disappears, redshifts, and broadens, and a new band at 550 and 750 nm emerges, which indicates the formation of larger non-spherical aggregates.44-45 This is because citrate molecules are easily replaced allowing a spontaneous adsorption of proteins when they are exposed to high concentration of charged molecules in CCM due to its weak binding nature. Similarly, dex-10-AuNPs also shifts the SPR band from 528 nm to 560 nm with an additional band at 655 nm and suggests the formation of both spherical and non-spherical aggregates. However, no much change in the intensity of absorbance with time was observed in case of dex-10-AuNPs.
Although dex-10 molecules bind to the AuNPs strong enough compared to the citrate molecules, they also form aggregates due to branching nature of dex-10 molecules; which allows fewer binding sites on the AuNPs surface. In the case of chitosan-AuNPs, an extra band at 590 nm was observed along with the SPR band at528 nm until 4 h incubation, suggesting the formation of spherical aggregates along with stable AuNP-
bimolecular corona. After 15 h incubation, the SPR band becomes broad with a peak maximum at 545 nm, which might be due to the formation of both stable AuNP-biomolecular conjugates and spherical aggregates. In case of dextrin-AuNPs, there was no shift in the SPR band until 4 h incubation, which indicates the formation of stable AuNP-biomolecular corona. However, after 15 h and 24 h of incubation, there was a slight shift in the SPR band from 528 to 535 nm suggesting strong interactions of AuNPs with biomolecules by the exchange of new molecules on the AuNP surface.
Figure 2 shows the UV-Vis spectra of AuNPs in protein rich media at different time intervals (0.5, 4, 15 and 24 h). Except chitosan-AuNPs, all other AuNPs showed no shift in the SPR band up to 4 h incubation. Moreover, chitosan-AuNPs showed an additional band at 600 nm for 0.5 h and 4 h incubation. This might be due to the formation of some spherical aggregates by the strong electrostatic interaction between positively charged chitosan molecules and negatively charged proteins (e.g. serum albumin). However, after 15 h and 24 h incubation, a slight red shift in the SPR band was observed, which reveals strong interaction of all the AuNPs with biomolecules and formation of stable AuNP-biomolecular conjugates either by replacement or adsorption of biomolecules with time. It is important to consider that the aggregation of AuNPs is dependent on either direct contact between the surfaces of two AuNPs or weekly bounded ligands on the AuNPs surface and AuNP charge.
UV-Vis results suggest that the behavior of AuNPs changes with protein rich- and protein poor CCM. Dextrin-, and chitosan-AuNPs showed similar behavior in both protein rich- and protein poor CCM, whereas citrate- and dex-10-AuNPs change with media. We believe that AuNPs with weak ligands (citrate-, dex-10) undergo aggregation in protein poor media, whereas form stable conjugates in protein rich media. However, all the AuNPs changed their surface with time in both media as confirmed by shift in the SPR band with different incubation times. Similar behavior was observed for the magnetic nanoparticles, where an increased amount of serum proteins adsorbed on the NPs after 16 h of incubation.46
Time-dependent size distribution of AuNPs in protein poor- and protein rich cell culture media
To further understand the size and charge distribution of AuNPs in both protein poor- and rich media, we have carried out DLS and zeta potential measurements at different time incubations (0.5, 4, 15 and 24 h) at 37 °C. Figure 3 A and 3 B show the dH of various surface-functionalized AuNPs in protein poor- and rich media at different time intervals. In case of protein poor media, except dextrin-AuNPs, all other AuNPs showed two different sizes, which indicate the formation of spherical and non-spherical aggregates. Both citrate- and dex-10-AuNPs displayed a dH around 100 nm and 700-1000 nm; representing the formation of spherical and non-spherical aggregates. The variation in the dH over time suggests a change in adsorption behavior of biomolecules on the AuNPs surface. However, chitosan-AuNPs, showed dH around 100 nm and along with diameter around 1500 nm, which might be due to the formation of stable chitosan-AuNPs-biomolecular corona and non-spherical aggregates, respectively. Dextrin-AuNPs showed only a dH around 250 nm, which is due to the formation of stable dextrin-AuNPs-biomolecular conjugates. In case of protein rich medium shown in Figure 3 B; except chitosan-AuNPs, all other AuNPs showed only one dH around 40-100 nm, which indicate the formation of stable AuNP-biomolecular conjugates. Chitosan-AuNPs exhibit two peaks around 300 nm and 1500 nm which are altered with time, indicating the formation of stable and spherical aggregates. DLS data completely agrees with the UV-Vis spectra analyses.
To further understand the surface charge distribution of AuNPs in both protein poor- and rich media, we have studied the zeta potential of AuNPs after incubation at 37 °C for 24 h. Figure S4 shows the surface charge distribution of different surface-functionalized AuNPs in protein poor- and rich media at 37 °C after 24 h incubation. The surface charge of the AuNPs after incubation in both media showed increased negative values which also indicate the formation of a biomolecular corona.
Additionally, we have investigated the osmolarity of CCM in order to understand any change in the concentration of biomolecules by the addition of different surface-functionalized AuNPs with time. Generally, CCM is designed to have osmolarity between 260 and 320 milliosmoles (mOsm) to mimic the osmolarity of serum (290 mOsm/kg).47 Figure S5 shows the osmolarity of both protein poor- and rich CCM after incubation with the AuNPs for 0.5 and 24 h at 37 °C. Except dex-10-AuNPs, all other AuNPs maintained the osmolarity in both protein poor- and rich CCM. Dex-10-AuNPs in protein poor CCM exhibited increased osmolarity at shorter incubation time (0.5 h). However, after 24 h incubation there was no much change in the osmolarity. The increased osmolarity in protein poor CCM might be due to the change in the concentration of molecules by release of some dex-10 molecules. This data clearly shows that except dex-10-AuNPs, all the AuNPs have no influence on the concentrations of biomolecules in both CCM.
Mechanical behavior of HL60 cells after treatment with AuNPs at different time intervals
We next studied the mechanical properties of HL60 cells after interaction with different surface-functionalized AuNPs at different incubation times. Cell elasticity can be used to express the resistance of the cell to an externally induced deformation. It is usually referred to as the Young’s modulus (E), which is the ratio between the applied mechanical stress and the resulting strain.48 In general, mechanical properties of a cell are related to the structure of the cytoskeleton. Alterations in mechanical properties can reflect changes in the cytoskeleton and by that in function. For example, the actin microfilaments are well organized in healthy cells compared to the unhealthy cells with a relatively high degree of stiffness, resulting in a larger Young’s modulus.49 However, these microfilaments are not well organized or are less observed in cancerous cells; hence the cells are softer and more flexible.
Typically, cell stiffness is measured through indentation experiments using atomic force microscopy (AFM).28 AFM is a state-of-the-art surface sensitive method that has recently been used for understanding the nanoparticle-to-nanoparticle and cell interactions in physiologic fluids.50-51 Though AFM is a widely used technique to study the cell mechanical properties it is hard to characterize suspended cells.
One of the recent advancements in studying the biomechanics of cells is RT-DC, a technique which facilitates high throughput characterization of cells in suspension. RT-DC being a microfluidic technique deforms the cells hydrodynamically and enables analysis in real-time at high speed (1000 cells s-1).
Figure 4 A gives a schematic representation of a RT-DC setup while Figure 4 B and 4 C highlight typical scatter plots for deformation and Young’s modulus vs. cell size, respectively. Figure 4 D and 4 E shows the time dependent alterations of cell deformation and Young’s modulus for HL60 cells in presence of different surface-functionalized AuNPs, respectively. Monitoring the mean cell deformation in protein poor medium over time for three experimental replicates where each measurement consists of several thousand single cell measurements, only a significant decrease in deformation has been observed for dex-10-AuNPs (p=0.04) after 0.5 h exposure. In contrast, an increase in the mean Young’s modulus was observed for the dex-10-, citrate- and chitosan-AuNPs at a short incubation time (0.5 h) but not at later time points of incubation (4, 15 and 24 h).
We presume that at 0.5 h incubation, except dextrin-AuNPs all other AuNPs form aggregates in the protein poor medium (see Figure 1) which may interrupt the stability of actin filaments and results in stiffening of HL60 cells. With increasing time, AuNPs form stable and spherical aggregates by adsorbing various biomolecules on their surface and have little influence on the cytoskeleton and correspondingly on cell stiffness.
We have further increased the AuNP concentration (50 nM) and studied the Young’s modulus at two different time points (0.5 and 24 h) (Figure S6). At 0.5 h incubation, all the AuNPs showed insignificant changes in cell deformation as well as Young’s modulus except dex-10-AuNPs. In contrast, we observed a significant decrease in deformation and an increase in cell stiffness in presence of all AuNPs after 24 h. The decrease in the Young’s modulus at 0.5 h incubation and increase after 24 h for dex-10-AuNPs might be either due to change in the adsorption behavior of biomolecules on the NPs surface or uptake or aggregation behavior of the NPs. These results indicate that cell stiffness in the presence of AuNPs changes with AuNP concentration and time. We believe that lower concentration of the AuNPs can be used for drug delivery applications as they very little affect the mechanical behavior of HL60 cells.
To understand the effect of FCS, similarly, we have studied the mechanical properties of HL60 cells after 0.5 h and 24 h exposure in protein rich media (Figure 5). We found that except dextrin-AuNPs, all other AuNPs showed a significant decrease in deformation at 0.5 h, whereas no such changes were observed at 24 h. However, only citrate-AuNPs showed increased cell stiffness at 0.5 h, but no such effect was observed at 24 h. Interestingly, after 24 h all AuNPs showed no significant changes in cell deformation and cell stiffness.
Viability of HL60 cells after treatment with AuNPs at different time intervals
To understand the influence of mechanical stress on the cell viability, we investigated different surface-functionalized AuNPs incubated with HL60 cells in media with and without FCS protein at different time intervals. Figure 6A shows the time dependent cell viability of HL60 cells after treatment with different surface-functionalized AuNPs in media without FCS. Stronger luminescence signal indicates higher amounts of vital cells. AuNPs without FCS can avoid any unspecific interactions with the cells. Except dex-10-AuNPs, all other studied AuNPs maintained a significant cell viability until 15 h, while a slight decrease was found after 24 h. The decreased cell viability for dex-10-AuNPs might be due to the formation of non-spherical aggregates which are more toxic to the HL60 cells. However, there was no change in the Young’s modulus after 4 h incubation at lower concentration.
We have increased AuNP concentration (50 nM) to understand the effect of concentration on cell viability in protein poor medium at different time intervals (Figure S7). Except dextrin-AuNPs, all other types of surface-functionalized AuNPs showed decreased cell viability at all time points. The increased cell viability for dextrin-AuNPs compared to the other AuNPs is due to the formation of stable bioconjugates which are less toxic to the HL60 cells than spherical and non-spherical aggregates. Both citrate- and dex-10-AuNPs showed decreased cell viability until 15 h, but a slight increase in cell viability for dex-10-AuNPs was observed for 24 h. The decrease in cell viability might be due to the formation of spherical and non-spherical aggregates. However, chitosan-AuNPs showed increased viability with time (max at 15 h and 24 h). The decrease in cell viability for chitosan-AuNPs at shorter incubation (0.5 h) might be due to the formation of spherical aggregates. However, all studied AuNPs at higher concentration exhibited an increase in the Young’s modulus after 24 h incubation.
Figure 6 B shows the cell viability after treatment with different surface-functionalized AuNPs in media with FCS, at two different time points (0.5 and 24 h). Interestingly, both dex-10, chitosan-AuNPs showed a substantial decrease in viability at both time points, whereas dextrin-AuNPs showed increased viability at both time points. Citrate-AuNPs showed a slight decrease in cell viability at 24 h. However, no significant change was observed in the Young’s modulus after 24 h incubation in protein rich CCM.
Among all AuNPs, dextrin-AuNPs showed increased viability in both protein poor- and rich medium. Dex-10-AuNPs showed decreased viability in both protein poor- and rich medium. It is quite interesting that chitosan-AuNPs showed increased cell viability in protein poor medium than protein rich medium. Instead, citrate-AuNPs showed increased cell viability in protein rich medium compared to protein poor medium. Our results suggest that there is no correlation between cell stiffness and cell viability. Increase in the cell stiffness might be due to the rearrangement of actin and tubulin microfilaments by the AuNPs, whereas decrease in cell viability might be due to the formation of aggregates and their uptake.
Release of ROS upon interaction with AuNP
Finally, we have investigated the release of reactive oxygen species (ROS) in HL60 cells incubated with AuNPs in protein rich CCM. ROS are chemically reactive molecules, and can promote cell proliferation and differentiation at a moderate level. However, an elevated level of ROS can induce oxidative stress, resulting in severe damage to the DNA, protein, and cells.52-53 It was previously reported that upon interaction with NPs, intracellular ROS production may increase and by interfering with cellular organelles can cause DNA/RNA breakage, membrane destruction and increased toxicity and eventually cellular death.54
Mitochondria are important source of ROS in a cell. MitoSOX-red was used as a mitochondrial ROS (superoxide) indicator to investigate whether NPs stimulate excess of ROS release. Figure 7 shows the release of ROS in HL60 cells induced by different surface-functionalized AuNPs. We observed that among all the AuNPs, dex-10-AuNPs showed a significant increase (p=0.00191**) in the intracellular ROS levels. Citrate-AuNPs induced moderate increase in the ROS levels, whereas chitosan-, dextrin-AuNPs showed almost no influence on the ROS levels. Even though some reports suggest that chitosan triggers ROS induction in cancer cells, we did not observe any difference in the ROS levels in HL60 cells; this might be due to lower concentration of AuNPs.55