AgNPs induce mitochondrial apoptosis in HG-3 CLL cells
In order to get insights into the mechanisms involved in the cytotoxic effects of AgNPs, we measured different markers of apoptosis in HG-3 cells upon treatment. We found an increased expression of the pro-apoptotic Bax protein and unchanged levels of the anti-apoptotic Bcl-2. This led to a significant unbalance in the Bax/Bcl-2 ratio (1.49 ± 0.29-fold increase vs control set as 1; Fig. 1D), indicating a susceptibility of CLL cells to an intrinsic apoptosis pathway.
We also observed that HG-3 cells treated with AgNPs showed a decreased mitochondrial membrane potential registered as mean fluorescence intensity (MFI) by flow cytometry (0.53 ± 0.31-fold decrease; Fig. 1E), indicating that the increased Bax expression could result in the rupture of mitochondrial outer membrane. It is known that mitochondrial outer membrane permeabilization (MOMP) triggers a cascade of downstream caspase activation involved in the apoptotic process induced by AgNPs (14, 15). We found that HG-3 cells treated with AgNPs showed a significant increase in the levels of cleaved caspase-9 (2.23 ± 0.67-fold increase), caspase-3 (12.38 ± 19.19-fold increase) and caspase-7 (13.23 ± 15.29-fold increase) (Fig. 1F). As further support of apoptosis elicited by AgNPs, we detected increased levels of the cleaved PARP fragment (2.14 ± 1.11-fold increase vs control set as 1), a known target of caspase activation (Fig. 1F). Altogether, these data demonstrate that AgNPs induce CLL cell injury trough mitochondrial intrinsic apoptotic mechanism.
AgNPs modulate the expression of calcium channel in CLL cells
As several studies have reported that AgNPs modulate cellular calcium homeostasis (16), we first characterize Ca2+ homeostasis in CLL cells as a potential target for nanoparticles effects. We compared Ca2+ channel mRNA expression of primary CLL cells with that of healthy B cells (N = 8, for both). Specifically, we analyzed Ca2+ channels localized on cell surface, on mitochondria and endoplasmic reticulum (ER) membranes, or associated to the ER-mitochondria crosstalk (17–21). Figure 2B shows that CLL cells exhibited a significant up-regulation of KCNN4 (4.52 ± 1.73 vs 0.58 ± 0.2), MCU (0.76 ± 0.05 vs 0.56 ± 0.11), IP3R3 (1.74 ± 0.73 vs 0.62 ± 0.13) and ATP2A2 (2.12 ± 0.29 vs 1.17 ± 0.23), and a down-regulation of VDAC1 (0.35 ± 1.73 vs 0.51 ± 0.12). These results suggest that alterations in these regulators of Ca2+ homeostasis represent a hallmark of CLL cells, which might render them sensitive to Ca2+-targeting agents.
To further define Ca2+ homeostasis as vulnerabilities targeted by AgNPs, we measured mRNA levels of Ca2+ modulators in HG3 cells treated with AgNPs (1nM for 6h). As shown in Fig. 2A, q-PCR analysis revealed that AgNPs significantly up-regulated the expression of KCNN4, MCU and VDAC1 (1.37 ± 0.22, 1.82 ± 0.41 and 1.41 ± 0.3, respectively) whereas IP3R3 and ATP2A2 levels were down-regulated, compared to control set as 1 (0.65 ± 0.14 and 0.58 ± 0.23, respectively). These results indicate that AgNPs deregulate important channels involved in the regulation of Ca2+ homeostasis in CLL cells.
AgNPs increase Ca2+ influx and stimulate ROS production in HG-3 cells
To better define whether AgNPS target Ca2+ homeostasis in CLL, we measured intracellular Ca2+ concentration upon exposure of HG-3 cells to AgNPs or vehicle control (Ctrl). As shown in Fig. 3A, FURA-2-AM Ca2+ imaging assay demonstrated a significant increase of intracellular Ca2+ concentration in AgNPs-treated compared to Ctrl cells (0.380 ± 0.058 vs 0.11 ± 0.005). In the same experimental conditions, when replacing the Ca2+ from the external solution with magnesium, there was no modification in the Ca2+ levels in the presence of AgNPs with respect to Ctrl, suggesting that Ca2+ ions incorporated by HG-3 cells after AgNPs exposure derived from the extracellular environment (Fig. 3A). Similar results were obtained with the “Fluo-4 DirectTM Calcium Assay” performed by flow cytometry. In particular, the mean fluorescence intensity (MFI) baseline of the cells labeled with Fluo-4 was significantly increased after the addition of AgNPs when compared to Ctrl (2.74 ± 2.03-fold increase vs Ctrl set as 1) (Fig. 3B).
Based on the evidence that high cytosolic Ca2+ levels stimulate mitochondria to produce high amount of ROS (22), we investigated ROS levels in HG-3 cells treated with AgNPs at different time points ranging from 1 to 6 hours. Results revealed the highest ROS production after 2h-treatmemt (1.77 ± 0.49-fold increase vs Ctrl set as 1), time at which cell viability was significantly reduced to 51.27 ± 0.08% compared to Ctrl (Fig. 3C). These data suggest that AgNPs induce CLL cell cytotoxicity by increasing intracellular Ca2+ levels with the consequent ROS overproduction.
We then performed Transmission Electron Microscopy (TEM) analysis in HG-3 cells after 2h-treatment with AgNPs and showed that AgNPs localize in close proximity to mitochondria (Fig. 3D). These results confirm the cellular uptake of AgNPs and suggest that AgNPs might directly interfere with mitochondrial functions through a physical interaction with mitochondria structures, in addition to Ca2+-mediated mechanisms.
AgNPs potentiate the cytotoxic activity of agents targeting Ca 2+ homeostasis and mitochondria functions in HG-3 cells
We tested the combined effects of AgNPs and selected drugs with anti-CLL activity and the capacity to target mitochondrial integrity and/or Ca2+ homeostasis on the viability of HG-3 cells.
-
We used the selective Bcl-2 inhibitor Venetoclax based on its role in inducing MOMP (23). Figure 4A shows that combined treatment with AgNPs and Venetoclax potentiated the cytoxicity of each single agent, resulting in a greater reduction in cell viability compared to each single drug. Specifically, cell viability, which was reduced to 70.08 ± 32.97% by Venetoclax and to 28.84 ± 19.79% by AgNPs, as compared to controls set to 100%, was further lowered to 8.98 ± 13.97% by drug combination. To determine whether the AgNPs/Venetoclax combination was synergic or additive, we conducted cytotoxicity tests at several drug concentrations and used the Chou-Talalay model. Combination Index (CI) plot confirmed the synergistic effect of drug combination with 1nM AgNPs + 2nM Venetoclax and 2nM AgNPs + 4nM Venetoclax (Fig. 4A, right).
-
We tested the BTK inhibitor Ibrutinib and AgNPs combination, since Ibrutinib sensitizes cancer cells to ROS inductor agents (24). As shown in Fig. 4B, AgNPs also potentiated the cytotoxic activity of Ibrutinib. Indeed, viability of HG-3 cells, which was decreased to 57.77 ± 6% after Ibrutinib treatment, and to 63.87 ± 3.67% by AgNPs as compared to controls set to 100%, was further reduced to 25.79 ± 11.3% by drug combination. The Chou-Talalay analysis revealed synergism between AgNPs and Ibrutinib (Fig. 4B, right).
-
We combined AgNPs and Bepridil which perturbs Ca2+ homeostasis in CLL cells, representing a potential option for CLL treatment (25). Figure 4C shows that HG-3 cell viability, which was reduced to 51.59 ± 19.58% by Bepridil alone, and to 30.25 ± 23.59% by AgNPs, as compared to controls (Ctrl) set to 100%, was further lowered to 7.32 ± 10.64% by the combination. The combination of AgNPs with Bepridil resulted synergistic, as assessed by the Chou-Talalay method (Fig. 4C, right).
AgNPs conjugated with Rituximab display targeting capability and in vivo anti-leukemic activity
The greatest limitation of nanotechnologies applications in clinical translation is due to the non-specific distribution in vivo (26). In order to develop a targeted therapy mediated by ligand-receptor specific affinity, we coated AgNPs with the anti-CD20 antibody Rituximab (AgNPs@Rituximab) and tested their anti-leukemic activity, by An V/PI assay, against HG-3 cells after 24h of in vitro treatment.
Results in Fig. 5A showed that AgNPs alone reduced cell viability to 54.42 ± 4.07% compared to 95.33 ± 1.41% of the control, in keeping with data shown in Fig. 1C, whereas Rituximab alone did not affect cell viability (95.37 ± 0.97%). When uncoated AgNPs and Rituximab were administered simultaneously, we observed a reduction in cell viability similar to that induced by AgNPs alone. Notably, when cells were treated with AgNPs@Rituximab, cell viability was drastically lowered to 27.88 ± 3.07%, indicating that the conjugation with Rituximab potentiates the anti-leukemic effect of AgNPs (Fig. 5A).
With the aim to demonstrate the advantages obtained from the conjugation of AgNPs with Rituximab, TEM imaging was used to evaluate targeting capability of AgNPs@Rituximab towards HG-3 cells. As shown in Fig. 5B, we demonstrated the adhesion of AgNPs@Rituximab with HG-3 cell membrane after 0.5h-treatment. In addition, we observed a tendency of cell membrane invagination that documented the first step toward internalization of AgNPs. After 2h-treatment, cell membrane was markedly coated with AgNPs@Rituximab, which were also found in the cytoplasm suggesting their cellular uptake through cell membrane, probably within endocytic-like structures. AgNPs appeared localized in the cellular matrix after 12h-exposure in close proximity to the nucleus (Fig. 5B). These data documented the success of AgNPs@Rituximab multicomplex manufacturing and, importantly, their specificity in targeting CLL cells.
Finally, with the aim to evaluate in vivo the efficacy of AgNPs@Rituximab, we performed experiments in xenograft models of CLL, obtained by transplanting HG-3 cells in NSG mice. Mice were divided into four groups and treated at day 4, 8, 12 and 16 from the transplantation (day 0) with vehicle, AgNPs, Rituximab and AgNPs@Rituximab, respectively (Fig. 5C). AgNPs@Rituximab treatment significantly improve survival of transplanted mouse compared to Rituximab and vehicle mouse cohorts, with a median survival of 38, 27 and 21 days, respectively. Conversely, administration of AgNPs alone had only a modest effect on median survival as compared to control mice (23 days vs 21 days), suggesting that conjugation with Rituximab is important for increasing AgNPs specificity towards leukemic cells in vivo (Fig. 5D).