Layered gadolinium-based nanoparticle as a novel delivery platform for microRNA therapeutics

LGdH nanoparticles were synthesized using a previously reported procedure with minor modifications [14]. Briefly, 5.75 ml of 1.5 M NaOH was added to 20 ml of 0.125 M GdCl₃·6H2O with vigorous stirring at room temperature to adjust the final pH to 6.5. The precipitate was heated at 80 °C for 2 h with vigorous stirring. The slurry of LGdH was centrifuged at 6000 rpm for 10 min and the supernatant was decanted. Nanoparticle pellet was washed thoroughly with deionized water and dispersed with a sonic dismembrator (Model 705, Fisher Scientific). The sizes of LGdH nanoparticles were determined by using the ellipse tool in the ImageJ program (Ver 1.48d, National Institute of Health).

A single-stranded miR-10b inhibitor, miRCURY LNA miRNA Inhibitor with the sequence of 5'-ACAAATTCGGTTCTACAGGGT-3' was used as AMO (Exiqon, Woburn, MA, USA). The equivalent oligodeoxynucleotide-based sequences in untagged and 5' Alexa Fluor 488-modified forms were used for characterization, confocal microscopy, and flow cytometry studies (Integrated DNA technologies). AMOs were loaded to LGdH nanoparticles via ion exchange. AMO and nanoparticle were mixed in deionized water (pH 7.0) at 1:1 mass ratio and vortexed immediately at the highest setting. The mixture was then shaken on a roller for 30 min, centrifuged, and resuspended in saline at room temperature prior to use.

The powdered LGdH and AMO-LGdH nanoparticle samples for x-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FTIR) were obtained by vacuum drying the sample at room temperature. The XRD analysis was conducted on a Rigaku Ultima IV Powder XRD system run at 38 kV and 38 mA using a x-ray generator with a cobalt tube. The search and match was done with JADE software (version 6.1.7601, Materials Data). The FTIR spectrophotometry data (Nicolet 8700, Thermo Scientific) was analyzed via OMNIC software (version 6.2, Thermo Electron Corporation). Zeta potential measurements were recorded at 25 °C in triplicates (Zetasizer Nano-ZS, Malvern Instruments). The average of zeta potential measurements is reported ± standard error of mean.

For the determination of AMO loading and entrapment efficiencies, amount of AMO in the supernatant was assessed via Beckman UV spectrophotometer. Calculations were done according to the following [36]:

$$\begin{eqnarray*}&&\begin{array}{lllllllllllllll} {\rm Loading}\;{\rm efficiency}(\%) \\ \quad =\frac{{\rm Mass}\;{\rm AMO}\;{\rm added}\;-\;{\rm Mass}\;{\rm AMO}\;{\rm in}\;{\rm supernatant}}{{\rm Mass}\;{\rm LGdH}\;{\rm nanoparticles}}\; \\ \quad \quad \times 100. \\ \end{array}\end{eqnarray*}$$

$$\begin{eqnarray*}&&\begin{array}{lllllllllllllll} {\rm Entrapment}\;{\rm efficiency}(\%) \\ \quad =\frac{{\rm Mass}\;{\rm AMO}\;{\rm added}\;-\;{\rm Mass}\;{\rm AMO}\;{\rm in}\;{\rm supernatant}}{{\rm Mass}\;{\rm AMO}\;{\rm added}} \\ \quad \quad \times 100. \\ \end{array}\end{eqnarray*}$$

Experiment was conducted in triplicates.

Human breast cancer cell lines MDA-MB-231 and MCF7 were obtained from American Type Culture Collection (ATCC) and cultured in Dulbecco's Modified Eagle Medium (Gibco) supplemented with 1% antibiotic-antimycotic (Invitrogen), 5 μg ml⁻¹ plasmocin prophylactic (Invivogen), and 10% fetal bovine serum (FBS; Gibco). Cells were kept in polystyrene flasks (Corning) in incubator with 5% CO₂ at 37 °C.

For preparation for TEM, MDA-MB-231 cells were seeded onto sterile plastic cover slips in a six-well plate 24 h before the treatment. On the day of the treatment, cells were treated with 200 μg ml⁻¹ LGdH nanoparticle and incubated for 1, 6, 24, or 48 h. The cells on cover slips were washed twice with PBS, and then fixed with 2% paraformaldehyde and 2.5% glutaraldehyde. The cells were then dehydrated with an ethanol and block stained with 1% uranyl acetate, infiltrated by Spurrs' resin, and polymerized in 65°–70 °C oven. Thin sections of 100 nm were cut with ultramicrotome (Leica EM UC6) and coated on a 400 mesh copper grid. The sections were stained with 2% Uranyl acetate and 1% lead citrate.

The images were taken using JEM 2100 transmission electron microscope at 120 kV (JEOL).

Viability of cells treated with 20, 50, 100, or 200 μg ml⁻¹ concentrations of LGdH nanoparticles or AMO-loaded LGdH nanoparticles was determined using the AlamarBlue Cell Viability Reagent (Invitrogen) as per manufacturer's instructions. Briefly, MDA-MB-231 cells were seeded at 4 × 103 cells per well on a 96-well assay plate (Corning) and allowed to attach for 24 h. On the next day, cells were treated and incubated for another 48 h. After the addition of AlamarBlue reagent, cells were incubated for 4 h. Absorbance was monitored at 590 nm using Optima 96-well plate reader (BMG LABTECH). Experiment was conducted in quadruplicates.

MDA-MB-231 cells were seeded onto six-well culture plates (10⁵ cells per well) and allowed to attach for 24 h. Media in all the wells were replaced with DMEM supplemented with 10% FBS without antibiotics prior to transfections. AMO-LGdH nanoparticle complexes were prepared at 18 μg: 18 μg weight ratios, and were added to each well to give a final LGdH concentration of 9 μg mL⁻¹. Transfections using Lipofectamine RNAiMAX (Life Technologies) were performed following the manufacturer's protocol. About one-half of AMO in AMO-LGdH nanoparticles was used for AMO-lipofectamine and AMO-only groups to compensate for the loading efficiencies. Cell transfections were complete in 24 h.

Cells were starved for 24 h in serum-free DMEM medium prior to the assay. Both migration (8 μm Mililcell cell culture inserts, Merck Millipore) and invasion assays (8 μm QCM™ High Sensitivity Non-cross-linked Collagen Invasion Assay, Merck Millipore) were conducted as per manufacturer's instructions. Briefly, after plating cells in inserts, cells were incubated for 6 h for the migration assay and 72 h for an invasion assay. Media supplemented with 10% FBS and 20% FBS was used as a chemo-attractant for the migration and invasion assays, respectively. For the migration assay, inserts were washed twice with PBS, fixed with 4% paraformaldehyde at room temperature, permeabilized with 100% methanol, and stained with crystal violet. For the invasion assay, staining reagent in the kit was used as per manufacturer's instructions. Cotton swabs were used to scrape off non-migrated cells from the top membrane prior to imaging under Olympus IX70 microscope. One center field views was taken per well using Olympus Plan 10X/0.25PhI or UPlanFl 20x/0.5 lenses, Zeiss Axiocam MRc camera, and AxioVision software (Version 4.4.0.0, Carl Zeiss). Both assays were conducted in triplicates. The numbers of migrated cells were counted with Metamorph software (64-bit Version 7.8.4.0; Molecular Devices).

MDA-MB-231 cells were seeded onto glass cover slips (Corning) 24 h prior to treatment in a six-well plate. The day after transfection, the media was removed and the wells were washed three times with warm phosphate buffered saline (PBS; Gibco, USA). Cells were fixed in warm 4% paraformaldehyde (Sigma Aldrich) in PBS for 20 min The cover slips were washed twice in PBS, and then mounted on glass slides using a mounting medium with DAPI (Life Technologies). Images were acquired on a Zeiss LSM-710 laser scanning microscope (Carl Zeiss) with 40 × 1.3 Oil DIC M27 using ZEN 2011 software (black edition 64-bit, version 7.0.4.0, Carl Zeiss). All treatment groups were prepared in triplicates.

Flow Cytometry was conducted to determine the cell uptake of 5' Alexa Fluor 488-modified AMO-LGdH nanoparticles. The day after transfection, cells were trypsinized and centrifuged at 200 g for 5 mins. After removal of the supernatant, cells were washed three times with PBS. Cells were analyzed by a FACSCalibur flow cytometer (BD Biosciences, San Jose, CA) with a 15 mW air-cool argon-ion laser at an excitation frequency of 488 nm. Fluorescence signals were detected using a band-pass filter 530/30 nm. The BD FACSCalibur™ flow cytometer was automatically set up and compensated with BD Calibrite™ beads used with FACSComp™ software (BD Biosciences). Experiments were conducted in triplicate.

Cells were washed twice with PBS (Gibco) and whole cell lysates were harvested in lysis buffer (150 mM sodium chloride, 1% Triton X-100, 50 nM Tris, pH 8.0) using a cell scraper (Corning). Protease inhibitor cocktail (Sigma Aldrich) was added to the collection tube. Proteins were resolved by Mini-Protean TGX 4–20% gel (Biorad), transferred to PVDF membrane (Biorad), blocked in 5% skim milk in TBST, and blotted with antibodies for HOXD10 (1:1500, Abcam) and α-tubulin (1:2000, Abcam), and secondary antibody conjugated to HRP (1:2000, Abcam). Clarity Western ECL Substrate (Biorad) was used for film-based imaging.

Magnetic strength measurements were made with vibrating sample magnetometer (VSM 7400, Lakeshore Cryotronics) at room temperature. 3.0T Philips MRI using spin-echo sequence was used to acquire T1-weighted image (TE, 8 ms; TR, 800 ms).

Statistical significance for migration and invasion were determined by a one-way analysis of variance (ANOVA) followed by Dunnett's multiple comparison test with Graphpad Prism software (Prism 5 for Windows, Version 5.04, GraphPad Software). Statistical significance for cell toxicity was determined by a two-way analysis of variance (ANOVA) followed by Bonferroni multiple comparison tests. P values of <0.05 were considered to be statistically significant for all tests. Graphs produced by Graphpad Prism software are presented as means with error bars representing standard error of mean.

The protocol described in Materials and Methods produced homogenous and stable aqueous suspension of oval LGdH nanoparticles (figure 1(A)). The LGdH nanoparticles had an average major and minor diameters of 146 nm and 90 nm, respectively, within the range for effective endocytosis [29]. We tested several AMO to LGdH nanoparticle mass ratios to assess the loading and entrapment efficiencies. The loading efficiency was highest for the highest AMO to LGdH nanoparticle ratio tested (6:1), whereas the entrapment efficiency was highest for the lowest AMO to LGdH nanoparticle ratio tested (0.5:1) (figure 1(C)). At 1:1 AMO to LGdH nanoparticle mass ratio, the loading and entrapment efficiencies were 41.8% and 89.5%, respectively, and this ratio was selected for all the subsequent experiments. The apparent size and shape of LGdH nanoparticles did not change after the addition of AMO (AMO-LGdH; figure 1(B)), which indicated effective intercalation of AMOs into the interlayer spaces of LGdH nanoparticles [24]. Intercalation activity of AMOs into LGdH nanoparticles was further confirmed by the absence of agglomeration, which has reported to occur for LDHs subjected to suboptimal intercalation activity [37]. In line with previous observations for LDHs loaded with anionic biomolecules [17, 22, 24], a decrease in positive surface charge was observed for our AMO-loaded LGdH nanoparticles. The LGdH nanoparticles loaded with AMOs as measured by zeta potential had more neutral electrical characteristic in an aqueous solution with zeta potentials of +14.6 ± 0.1 mV compared to +45.6 ± 0.4 mV for pristine LGdH nanoparticles (figure 2).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Further characterization was conducted to confirm the association of AMOs with LGdH nanoparticles. The FTIR spectrum of the both samples showed absorption peaks that are characteristic of rare earth metal-based hydroxycarbonate nanoparticles with the vibration of the OH-group at 3525 cm⁻¹ and 3497 cm⁻¹ (figure 3) [38]. From the analysis of deconvolved FTIR spectra of the AMO-loaded LGdH nanoparticles, we were able to identify peaks that were absent in pure LGdH spectra that could specifically be attributed to nucleic acids (data not displayed). The bands at 1052 cm⁻¹ and 1223 cm⁻¹ could be assigned to C-O stretching vibration of deoxyribose or ribose residue and symmetric stretching vibration of PO₂⁻ group, respectively [39]. The characteristic bands were also present for in-plane base vibrations of thymine (1695 cm⁻¹), guanine (1588 cm⁻¹), and cytosine (1529 cm⁻¹), as well as ring vibrations of guanine and adenine (1482.5 cm⁻¹ and 1308 cm⁻¹) (figure 3) [40]. The FTIR result confirmed that AMO-LGdH nanoparticles consist of oligonucleotides and LGdH nanoparticles, and that the intrinsic structure of LGdH nanoparticles remains intact even after intercalation of AMOs into interlayer spaces.

Zoom In Zoom Out Reset image size

In order to further confirm the association between LGdH nanoparticles and AMOs, we employed XRD to compare the crystal structures (figure 4). The XRD peaks for both samples were ascribed to hydrotalcite (JCPDS card 22–700), the magnesium and aluminum hydroxycarbonate that has a structure of layered double hydroxide. The basal spacing of pristine LGdH nanoparticles was calculated to be 0.893 nm, and this spacing did not expand with the addition of AMOs. The structural information based on XRD patterns indicated that AMO-LGdH nanoparticles retained the crystal structure of the pristine LGdH nanoparticles, as the only difference in the XRD patterns was the overall decrease in the diffraction intensity. This suggested that that AMOs were intercalated into the interlayer spaces without disturbing the basal spacing of pristine LGdH nanoparticles. The association of AMOs with LGdH nanoparticles without expansion was an expected outcome as the dimensions of a single-stranded AMO molecule does not necessitate an expansion of LGdH nanoparticles' interlayer space in order to intercalate [41, 42]. Overall, our characterization data demonstrated that AMOs associate with LGdH nanoparticles by intercalating within the interlayer regions of LGdH nanoparticles. Our data, together with the knowledge on small nucleic acids intercalated into interlayer spaces of LDHs [17], confirmed successful intercalation through electrostatic interactions between the negatively charged phosphate groups of the AMOs with the positively charged gadolinium hydroxide layers of LGdH nanoparticles.

Zoom In Zoom Out Reset image size

The cell toxicity of LGdH and AMO-LGdH nanoparticles was tested for a range of LGdH nanoparticle concentrations (or equivalent LGdH nanoparticle concentrations for AMO-LGdH group) in MDA-MB-231 cells (figure 5). The maximum concentration and the treatment duration tested were ten times and two times greater, respectively, compared to the parameters used for all the subsequent studies. As previously reported for the LDH nanoparticles [26, 27], LGdH nanoparticles did not cause significant cell toxicity for concentrations up to 200 μg ml. In addition, viability of cells treated with AMO-LGdH nanoparticles were not significantly different from cells treated with pristine LGdH nanoparticles for all the concentrations tested. This result confirmed that LGdH nanoparticles are biocompatible and that miR-10b antagonism alone does not affect the viability of the metastatic cell line [12].

Zoom In Zoom Out Reset image size

To characterize LGdH nanoparticles as a delivery platform, we tested whether cancer cells can readily take up LGdH nanoparticles at the specified time points. Using TEM, we observed cytosolic localization of nanoparticles in MCF7 cells at both early and later time points. At one-hour time point, LGdH nanoparticles were primarily at the membrane interface with minimal cytosolic localizations (figure 6(A)). After six hours of incubation with LGdH nanoparticles, considerable amount of LGdH nanoparticles were internalized in the endosomal compartments (figure 6(B)). After a day of incubation, some LGdH nanoparticles were degraded inside the endosomal compartments, as evidenced by the decrease in densities of LGdH nanoparticles in the endosomes compared to extracellular nanoparticles (figure 6(C)). The empty endosomal structures as well as LGdH nanoparticle-containing endosomes were visible on a second day (figure 6(D)).

Zoom In Zoom Out Reset image size

The internalization of 5'-AlexaFluor488-tagged AMO-LGdH nanoparticles was visualized via confocal imaging and quantified via flow cytometry. Punctate localizations were observed once AMO-LGdH nanoparticles were internalized into MDA-MB-231 cells, confirming the endosomal compartment localization as observed under TEM (figure 7). Similar punctuate appearance was seen with AlexaFluor-488 tagged AMO transfected via Lipofectamine (AMO-Lipo) method. Fluorescence signal was detected in almost all cells for AMO-LGdH and AMO-Lipo groups, whereas no apparent signal was detected in cells treated with AlexaFluor488-AMOs without a vehicle (AMO-Only). Flow cytometry demonstrated that transfection was efficient for both the AMO-Lipo and AMO-LGdH treatment groups with more than 99% of cells treated being positive with fluorescence signal (figure 8). This transfection efficiency is in line with a previously reported value for short oligodeoxynucleotide delivered by LDH nanoparticle [24]. Despite the similar efficiencies in transfection, AMO-LGdH group had a broader distribution compared to the AMO-Lipo group indicating a greater variability in the number of AMO-LGdH taken up per cell. The absence of fluorescence signal in AMO-only group confirmed the results of our confocal imaging results.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

It was previously reported that LDH nanoparticles carry biological molecules in the interlayer spaces for subsequent release in acidic buffer below pH 2 or in buffer of sodium chloride concentrations similar to that of bodily fluid without structural compromises to the cargo molecules [43]. Although our characterization studies revealed that many AMOs are likely to be absorbed onto the surfaces than intercalated between the layers, we tested whether LGdH nanoparticles can deliver and confer protection for the AMOs in vitro. For this, we used an AMO against miR-10b, a miRNA whose expression is known to correlate with metastatic potential of the cancer cells [12, 44]. We specifically tested changes in the migration and invasion properties, which are two separate steps in early metastasis (figure 9) [45]. We observed that AMO against miR-10b primarily affected invasive properties of MDA-MB-231 cells, with no significant decrease in the motility of MDA-MB-231 cells. AMO against miR-10b significantly reduced the invasive properties of the cell when transfected by LGdH nanoparticles. No significant difference was observed between the AMO-Lipo and AMO-LGdH nanoparticle groups. Our result is in line with previously reported result in which miR-10b inhibition specifically affected invasion but not migration in MDA-MB-231 cells [12].

Zoom In Zoom Out Reset image size

We further tested the integrity of AMOs delivered by LGdH nanoparticles by checking the protein expression of miR-10b downstream effector (figure 10). Although the role of miR-10b has been implicated in numerous molecular pathways, the pathway leading to the metastatic program in breast cancer has been shown to include a pro-metastatic transcription factor Twist [12]. Twist directly induces the expression of miR-10b, which in turn indirectly targets homeobox D10 (HOXD1O) mRNA and inhibits its translation [12]. Inhibition of HOXD10 expression results in the increase of Ras homolog gene family, member C (RHOC) protein activity, a positive regulator of the metastic program in cells [46, 47]. With LGdH nanoparticles, we successfully delivered AMO against miR-10b and obtained HOXD10 expression similar to that for AMO-Lipo group (figure 10). HOXD10 expression was absent in the AMO-only group. This indicated that LGdH nanoparticles deliver functional AMOs in vitro by conferring some degree of protection in addition to promoting efficient cancer cell uptake.

Zoom In Zoom Out Reset image size

To confirm that LGdH nanoparticles can be tracked with MR imaging, we obtained magnetization graph for LGdH nanoparticles (figure 11(A)) and assessed whether an effective T₁-weighted MRI image could be taken in 3.0 T clinical MR scanner (figure 11(B)). The LGdH nanoparticles exhibited paramagnetic property as demonstrated by a linear relationship between magnetization and applied field. Concentration-dependent T₁ signal intensity was seen for LGdH nanoparticles suspended in deionized water for concentration up to 5 mM, with no further contrast enhancement occurring between 5 mM and 10 mM. The MR imaging result demonstrated that the gadolinium-containing LGdH nanoparticles exhibit good T₁ relaxivity and can be an effective T₁ MRI contrast agent at concentrations below 5 mM.

Zoom In Zoom Out Reset image size