Mechanical and dynamic characteristics of encapsulated microbubbles coupled by magnetic nanoparticles as multifunctional imaging and drug delivery agents

Figure 1 illustrates the procedures in preparing SPIO-albumin-shelled perfluorocarbon MBs (referred to as SPIO-albumin MBs), which includes two major steps: synthesis of SPIOs (figure 1(a)) and assembly of SPIO-albumin MBs (figure 1(b)).

Zoom In Zoom Out Reset image size

Firstly, aqueous solution of FeCl₃•6H₂O and FeSO₄•7H₂O was made to serve as a source of iron. Under nitrogen atmosphere to prevent oxidation, concentrated NH₃•H₂O was added into the iron-containing solution until its pH reached 11.0. Then, oleic acid (OA; Sinopharm Chemical Reagent Co, Ltd, Shanghai, China) was added into the alkaline solution to get mono-disperse hydrophobic Fe₃O₄-OA nanoparticles. Via a surface double-exchange reaction between OA and meso-2, 3-Dimercaptosuccinic acid (DMSA; Sinopharm Chemical Reagent Co, Ltd, Shanghai, China), mono-disperse water-soluble SPIO (Fe₃O₄-DMSA) nanoparticles could be obtained and collected through a magnetic separation procedure. After a pH neutralization process, the SPIO solution was dialyzed through the dialysis membranes (Cellu-Sep T3, Membrane Filtration Products Inc, Seguin, Texas, USA) in pure water environment for 3 d to remove excess impurities. The final sample was filtrated through a 0.22 μm membrane and stored at 4 °C. In the second step, SPIO-albumin MBs were synthesized by using a sonicating method similar to that reported by Porter et al (1995). Briefly, 10% bovine serum albumin and 60% sucrose were mixed with a volume ratio of 1:1 in deionized water. A certain amount of SPIOs were then added into the mixed solution. The mixed solution was put in a bell-like glass chamber (9 cm radius and 20 cm height) saturated with perfluorocarbon (C₃F₈) overnight, then sonicated with an ultrasonic processor (VCX750, Sonics and Materials Inc, Newtown, CT, USA) for 2 min, while a syringe pump (LEGATO 270, KD Scientific Inc Holliston, MA, USA) was used to continuously inflate C₃F₈ gas. The ultrasonic processor worked at the burst mode with an on/off ratio of 3:1, a working frequency of 20 kHz and an intensity of 300 W. Then, SPIO-albumin MBs were centrifuged and then washed to eliminate free SPIOs. Finally, the upper layer of MB suspensions was collected for subsequent assessments. The detailed description of synthesis procedures can be found in the supplementary data (stacks.iop.org/PMB/59/226729/mmedia).

High-resolution 3D topography images were constructed using Multimode 8 atomic force microscopy (AFM) with a NanoScope V controller (Bruker Corporation, Billerica, MA, USA) (Vinogradova 2004). SCANASYST-AIR cantilevers with a nominal spring constant of 0.4 N m⁻¹ were used for imaging purpose. The cantilever has a pyramid tip with a side angle of 17.5° and a nominal radius of 2 nm. During the imaging processes, the cantilevers were made to oscillate vertically at 5% below their natural resonant frequency and move in a raster mode within a particular region of interest. The resonant frequency of the SCANASYST-AIR cantilever was about 70 kHz. Prepared microbubble samples were diluted 10 times, dropped onto a fresh mica plate and placed at room temperature for 30 min. All measurements were taken in a water environment on the mica plate. The scanning was performed within a square region (5  ×  5 μm) at a scanning rate of 0.5 Hz, with a pixel resolution of 256  ×  256. At least ten bubbles were examined for each sample. Through post-processing of raw data with NanoScope Analysis software (Version 1.40, Bruker Co, Billerica, MA, USA), 3D topography images were finally constructed for tested MBs.

The stiffness of SPIO-albumin MBs was interrogated under the ScanAsyst-contact mode of MultiMode 8 AFM system, with a ramp rate of 0.5 Hz. Microbubble samples were diluted 100 times, dropped on the fresh mica plate and then mechanically interrogated by 40 nm Au/15 nm Ti coated tipless cantilevers (MLCT-O10, Bruker Corporation, Billerica, MA, USA) with nominal spring constants (k_c). Prior to measurements, spring constant calibration was performed on the mica plate. The spring constant k_c and deflection sensitivity were calibrated to be 0.0228 N m⁻¹ and 58.91 nm V⁻¹, respectively. At least 3 force-displacement curves (no permanent deformation) were acquired for each bubble to ensure experimental reproducibility. The bright-field microscopy images were taken firstly to ensure that the cantilever was parallel to the mica plate and aligned to MB central axis. The effective spring stiffness of individual microbubbles (k_m) was assessed by measuring the deflection error of AFM cantilever during compression (Sboros et al 2006, Chen et al 2013). During each measurement, the compression force (F) was applied to the tested microbubble through the cantilever and the deflection error-z curve was recorded as raw data. The deflection error, d = F/k_c, represented the deflection distance if the cantilever contacted at a hard surface, while the value of z corresponded to actually measured cantilever deflection distance. Thus, the effective deformation of the bubble was calculated as d_MB = z–d. The measured d versus. z curves were automatically converted to F versus. d_MB curves by using the NanoScope Analysis software and then the effective stiffness of individual microbubbles could be obtained by applying linear regression fit to the force-deformation (F versus. d_MB) curve.

To ascertain the ability of SPIO-albumin MBs for MRI contrast enhancement, T₁- and T₂-weighted MR images were captured. A 3 T MRI scanner (Magnetom Trio, Siemens, München, Germany) was used to conduct all MRI studies. A 12-channel head coil was used for radiofrequency (RF) transmission and reception. For each sample, 1 ml SPIO-albumin MBs taken from individual groups were diluted 15 times. Since the effective SPIO concentrations in these four tested groups were measured to be 0, 19.6, 114.7 and 292.0 μg ml⁻¹, the molar concentrations of iron (Fe) element in the diluted samples were calculated to be 0, 0.01688, 0.0985 and 0.252 mM, respectively. In addition to SPIO-albumin MB solutions, four SPIO solution samples with corresponding Fe molar concentrations were also prepared for comparison. Each sample solution was placed into a 5 ml centrifugal tube for MR scanning. The longitudinal and transverse relaxivities (i.e. r₁ and r₂) of SPIO-albumin microbubbles and SPIOs were assessed by acquiring series of T₁- and T₂-weighted images, respectively. The detailed parameters used for MRI assessments can be found in the supplementary data (stacks.iop.org/PMB/59/226729/mmedia).

A polyacrylamide gel phantom (10  ×  10  ×  10 cm³) was made to provide mimic tissue background in US image. A custom-made cubic chamber (3  ×  3 × 3 cm³) with thin polythene membranes acting as acoustic windows was embedded in the center of the gel phantom. A portable US imaging instrument (Terason T3000, Teratech Corporation, Burlington, MA, USA) with a curved abdomen probe (5C2A, 2–5 MHz working frequency) was used to acquire US images. Each measurement was repeated 5 times. The gray-scale image intensity was calculated in each ROI, using MATLAB software.

All VEGF₁₆₅ delivery experiments were performed in an acrylic tank filled with degassed water. The US exposure apparatus and passive cavitation detection (PCD) system is similar to that used by Zhang et al (2013). An arbitrary waveform generator supplied 1 MHz sinusoidal pulses at fixed 20-cycle pulse length and 250 Hz pulse repetition frequency (PRF), with varied acoustic peak negative amplitudes (p-) of 0 (sham), 0.05, 0.2, 0.5 or 1.0 MPa. The output signals from the waveform generator were amplified through an RF power amplifier with a fixed gain of 50 dB, which were used to drive a 1 MHz self-made focused source transducer. The focused transducer was comprised of an air-backed, 1.375 inch-diameter piezoelectric (PZT) disk (APC 880, APC International Ltd, Mackeyville, PA, USA) attached to an aluminum-focusing lens with a radius of 9.2 cm curvature and approximately 6.6 cm focal distance (−6 dB and focal width of ~8 mm). A plastic test tube of 10 mm diameter, 0.5 mm thick and 50 mm length filled with sample suspension (the liquid depth of the suspension was ~16 mm) was capped by a custom-built rubber stopper that was used as an acoustic absorber to minimize the effect of standing wave and then sealed with parafilm to minimize bacterial contamination. The test tube was aligned axially with the source transducer so that the center of the suspension was situated at the focal region.

A self-made 5 MHz single-element non-focused PZT transducer (a diameter of 11 mm and a bandwidth of 2–8 MHz), used for PCD measurements, was located perpendicularly to the 1 MHz focused transducer so that its natural focus was orthogonal to that of the source transducer. Co-focusing of the transmitter and receiver transducers and the in situ pressure calibration was performed by using the NTR needle hydrophone (TNU001A, NTR Systems, Inc, Seattle, WA, USA) with a 30 dB preamplifier (HPA30, NTR Systems, Inc, Seattle, WA, USA). The acoustic scattering from bubbles and acoustic emission from inertial cavitation in the focal volume was collected by the 5 MHz transducer and digitized by an oscilloscope. The recorded PCD waveforms were stored in a personal computer for subsequent signal processing using MATLAB. LabView program was used to control the waveform generator and the oscilloscope. The IC 'dose' (ICD) was quantified for each measured PCD signal to assess the cumulated IC energy over a specific US exposure duration. The detailed protocol used to quantify ICD can be found elsewhere (Tu et al 2006).

Human Embryonic Kidney (HEK) 293 T cells were maintained in Dulbecco's modified Eagle's medium (DMEM; Invitrogen Corp, Carlsbad, CA, USA) supplemented with 10% Fetal bovine serum (FBS; Invitrogen Corp, Carlsbad, CA, USA), 100 U ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin (Invitrogen Corp, Carlsbad, CA, USA) and incubated at 37 °C in a humidified atmosphere containing 5% CO₂. The cells were harvested using Trypsin-EDTA and resuspended in phosphate-buffered saline before US exposure experiments. In accordance with the method described in previous work (Zhang et al 2013), 8 μg of plasmid pEGFP-N1-VEGF₁₆₅ was rapidly added into branched polyethylenimine (bPEI; 25 KDa molecular weight; Sigma-Aldrich, St. Louis, MO, USA) solution to produce bPEI:VEGF₁₆₅ complexes in a pH 7.4 PBS buffer. The mixture was vortexed for 30s and incubated at room temperature for 30 min before adding to the cell suspension. The HEK 293 T cells were harvested using Trypsin-EDTA and resuspended in 0.5 ml PBS at a concentration of 2.0  ×  10⁶ cells ml⁻¹. The bPEI:VEGF₁₆₅ complexes and 1.0  ×  10⁷ MBs in 0.5 ml PBS were individually added to the cell suspension. Every sample was briefly vortexed before it was exposed to ultrasound for 20s. The negative control sample was 1.0  ×  10⁶ cells added into 1 ml PBS. Five replicates were performed for each treatment. Unless otherwise specified, after incubating the sonicated cells for 2 h at 37 °C and 5% CO₂, the cells were centrifuged at 900 rpm (300 g) for 5 min and collected. After washing twice with PBS, the cells were resuspended and cultured in a 6-well plate for another 48 h in DMEM supplemented with 10% FBS and 1% penicillin-streptomycin solution at 37 °C and 5% CO₂.

The treated HEK 293 cells were collected by centrifugation at 900 rpm for 5 min and washed twice in PBS. Then, the cells were resuspended, cultured in the 6-well plate and incubated for 48 h in DMEM supplemented with 10% FBS and 1% penicillin-streptomycin solution at 37 °C and 5% CO₂. After a 48 h incubation period, the supernatant from each sample was harvested and VEGF₁₆₅ protein levels were measured following the instructions provided in the human VEGF₁₆₅ Elisa kit (RandD Systems, Minneapolis, MN, USA).

Cell viability was measured using the Cell-Counting Kit-8 (CCK-8, Dojindo, Tokyo, Japan). The treated HEK 293 T cells were collected by centrifugation at 900 rpm for 5 min and washed twice in PBS. The cells were resuspended before being put into 96-well plates at a density of 3  ×  10³ cells well⁻¹. After a 24 h incubation period, CCK-8 solution was added to each well according to the manufacturer's instructions. Plates were incubated for 4 h and the absorbance was measured at 450 nm with a reference wavelength of 600 nm. Cell viability was expressed as the absorbance of the sample divided by that of the negative control.

The data were analyzed and reported as the mean ± standard deviation (SD) using Origin software (OriginLab Co, Northampton, MA, USA). The normality test was performed using Origin software based on Shapiro-wilk test; a t-test was also applied using Origin software to compare results, with p < 0.05 considered to be a statically significant difference.

Transmission electron microscopy (TEM) images were taken for MB shells to verify that SPIOs had been successfully coupled to albumin-shelled MBs instead of leaving in the suspension as free residues. The albumin-shelled MB without loading SPIOs shows a uniformly clear surface (see figure 2(a)). However, it is evident in figure 2(b) that lots of nanometer-scale black dots with individual sizes around 8–12 nm exist on the bubble surface, which reveals SPIOs have been successfully integrated to MB shells. Energy dispersive x-ray (EDX) analysis was also applied at 15 kV to further confirm the successful embedding of SPIOs in these MBs. Compared to normal albumin-shelled MBs (figure 2(c)), the EDX spectrum of SPIO-albumin MBs (figure 2(d)) clearly demonstrates the presence of Fe elements in the assembly bubble system. Note that the encapsulated bubble cannot keep the spherical shape, which is explained by the gas inside the bubble being driven out during the fixing and dehydration processes of sample preparation for TEM. Additionally, since SPIO nanoparticles were distributed randomly in the mixed solution during the sonication process, magnetic nanoparticles observed in the TEM image are not uniformly distributed in the bubble shell.

Zoom In Zoom Out Reset image size

Typical 3D topography images of albumin-shelled and SPIO-albumin MBs are illustrated in figures 2(e) and 2(f). The structures shown here are not an ideal spherical shape, but a pyramidal shape, because the interrogating cantilever has a pyramid tip with a nominal radius of 2 nm. The AFM topography is still believed to be capable of revealing the real shape of the upper part of the bubble, since the tip size is much smaller than the bubble. As described in previous work, the roughness of MB top surface can be evaluated with the software (Sboros et al 2006). Meanwhile, the size of a particular MB can be estimated by fitting the MB topography with a sphere, although it might be underestimated because the adhesion with the hard mica plate may affect the sphericity at bubble bottom. In figure 2(e), the diameter of albumin-shelled bubble, it is estimated to be ~1.2 μm with a relatively smooth surface (root-mean-square roughness R_q = 168 nm, roughness average R_a = 144 nm). However, the SPIO-albumin MB (figure 2(f)) exhibits a rougher shell structure (R_q = 328 nm, R_a = 265 nm) with a larger diameter (~1.8 μm), which might result from the embedding of SPIOs onto the bubble shell.

The above magnetization assessment, TEM images and EDX spectra have proven that SPIOs could be successfully loaded to albumin-shelled perfluorocarbon MBs. Furthermore, AFM topography demonstrates that SPIO-albumin MBs have a relatively rougher surface and larger size than that of un-loaded albumin MBs, which indicates the loading of SPIOs might alter the MB shell structure.

In the following assessments, SPIO-albumin MB samples were categorized into four groups with varied SPIO concentration (as listed in table 1). In each group, a certain amount of SPIOs (e.g., 0, 0.2, 1.2, or 3.2 mg) were added into 10 ml albumin/sucrose solution (a volume ratio of 1:1) for SPIO-albumin MB preparation and then the effective amount of SPIOs loaded to albumin-shelled MBs was determined by using an Ultraviolet-Visible Spectrometer (UV3600, SHIMADZU CO, Tokyo, Japan). The detailed procedures can be found in the supplementary data (stacks.iop.org/PMB/59/226729/mmedia).

A proper size is a key factor in terms of MBs behaving as an effective imaging or therapy agent that can safely pass through pulmonary and vascular circulation system, while proving sensitive US and magnetic responses. Thus, MB size distributions and concentrations were measured under varied SPIO concentrations by using the single particle optical sensing (SPOS) technology; the detailed procedures can be found in the supplementary data (stacks.iop.org/PMB/59/226729/mmedia). For samples in these four groups, MB concentrations were measured to be about 1–2  ×  10⁸ bubbles ml⁻¹. An ascending trend can be clearly observed in figure 3 for MB mean diameter with the increasing SPIO concentration. At the highest SPIO concentration (viz, 292.0 μg ml⁻¹) tested here, the mean diameter of SPIO-albumin MBs is about 3.5 μm with 99% smaller than 8 μm, which indicates that these SPIO-albumin MBs are small enough to serve as clinic MB agents.

Zoom In Zoom Out Reset image size

Figure 4(a) shows typical force-deformation curves measured for SPIO-albumin MBs with a diameter of ~3 μm, under varied SPIO concentration. Upon contact with the MB, the cantilever starts to compress the bubble and deflect upwards. Elasticity measurements were performed by adopting a linear elastic assumption for MB deformation based on Hooke's Law. As shown in figure 4(a), with gradually increasing compression force, the force-deformation curves enter a linear region in which the gradient is associated with the effective MB stiffness (Vinogradova 2004, Chen et al 2013). By applying linear regression analyses to the measured force-deformation curves, MB stiffness is plotted as the function of SPIO concentration (figure 4(b)), which indicates that the stiffness of SPIO-albumin MBs decreases with increasing SPIO concentration. The size dependence of MB stiffness was also investigated at a fixed SPIO concentration (i.e. 292.0 μg ml⁻¹ for MBs in Group 4). Figure 4(c) illustrates typical force-deformation curves for MBs with different sizes. It is observed in figure 4(d) that, as the bubble diameter increases from 1 to 5 μm, the MB stiffness decreases from 0.148  ±  0.003 to 0.771  ±  0.001 N m⁻¹.

Zoom In Zoom Out Reset image size

It is well known that SPIOs can be used as effective MRI contrast agents. Thus, the magnetic responses of SPIO-albumin MBs were assessed firstly to evaluate their potentiality as a contrast agent for MRI. T₁- and T₂-weighted MR images were captured for MB samples with varied SPIO concentration. Typical T₁- and T₂-weighted images measured for SPIO-albumin MBs and SPIO solutions are illustrated in figures 5(a) and 5(b), respectively. With the increasing Fe concentration, the T₂-weighted images obviously turn darker for both SPIO-albumin MBs and SPIOs (figure 5(b)), while only moderate contrast change can be observed in T₁-weighted images (figure 5(a)), which suggests that the tested MBs might be more favorable for the contrast enhancement in T₂-weighted MRI due to the integration with SPIOs. The longitudinal and transverse relaxation rates (i.e. 1/T₁ and 1/T₂) are plotted as a function of Fe concentration in figures 5(c) and 5(d). The slope of linear regression fit gives the longitudinal and transverse relaxivities (i.e. r₁ and r₂) for the tested sample. The ratio r₂/r₁ determines whether a contrast agent is most suitable for enhancing MR contrast in T₁- or T₂-weighted image (Tromsdorf et al 2009, Kim et al 2011). According to the results shown in figures 5(d) and 5(c), the r₂ and the r₂/r₁ ratio calculated for the SPIO-albumin MBs are 193.62 S⁻¹ mM⁻¹ and 35.66, respectively, which are significantly enhanced compared to those of SPIOs (about 2 and 2.5 times, respectively). The enhancement of r₂ value might be attributed to the synergetic magnetism effect arising from the aggregation and stabilization of multiple SPIOs in MB shells (Berret et al 2006). The higher measured r₂/r₁ ratio suggests the current SPIO-albumin MBs may be superior negative MRI contrast agents beyond regular SPIOs.

Zoom In Zoom Out Reset image size

Moreover, US imaging was carried out for the gel phantom in the testing chamber to investigate the impact of the integration of SPIOs on the acoustic scattering capability of SPIO-albumin MBs, which is a dominant factor for their performance in US imaging. To provide significant and distinguishable ultrasound contrast enhancement effect, the MB samples in each group were diluted 30 times and then injected into the testing chamber. Typical US images captured for degassed water and MB solutions with varied SPIO concentrations are shown in figure 6. Compared to degassed water (figure 6(a)), hyperechoic regions can be clearly observed in the testing chamber by adding MB solutions with higher SPIO concentration (figures 6(b)–6(e)). For each sample, the gray-scale image intensity is calculated in the region of interest (ROI) using MATLAB software. The contrast-to-tissue ratio (CTR) is estimated as the ratio of averaged gray-scale intensities between MB and tissue (phantom) regions. According to five repeated measurements, the averaged CTR values are calculated to be −0.59  ±  0.001, 5.89  ±  0.70, 8.97  ±  0.80, 11.04  ±  0.72 and 12.844  ±  0.56 dB for the samples of degassed water and Group 1 to 4, respectively. A negative CTR is discerned for degassed water, which indicates that the tissue-mimicking phantom has stronger echo response than the water. By adding albumin-shelled MBs without SPIOs, significant contrast enhancement effect is achieved. It is noteworthy that, as the effective concentration of SPIOs loaded to MBs increases from 0 to 292.0 μg ml⁻¹, the corresponding CTR is further improved, which suggests that greater US echo intensity can be contributed by MBs integrated with more SPIOs. This will be discussed in detail in the section of discussion.

Zoom In Zoom Out Reset image size

Beyond a dual-modality imaging probe, the proposed SPIO-albumin MBs were also designed as a therapeutic agent. Thus, the enhancement effect of SPIO-albumin MBs on US-facilitated VEGF₁₆₅ transfection effect was also evaluated in vitro under varied SPIO concentrations. In the experiments, the tested samples were divided into five series: (1) HEK 293 T cells mixed with bPEI:VEGF₁₆₅ complexes only; and (2)–(5) cells + bPEI:VEGF₁₆₅ + MBs in Group 1–4 (i.e. SPIO-concentration = 0, 19.6, 114.7 and 292.0 μg ml⁻¹, respectively). 1 MHz US exposures were applied to the tested samples with varying acoustic driving pressures (p-).

In order to examine the capability of SPIO-albumin MBs on US-facilitated gene transfection, VEGF₁₆₅ protein secreted into the supernatant of treated cell suspensions was quantitatively measured using ELISA assay and the cell viability was evaluated by CCK-8 assessments. Meanwhile, MB IC energy accumulated during US exposure period was quantified as ICD based on PCD measurements, since MB IC activities have been claimed to play an important role in US-facilitated gene/drug delivery through sonoporation (Miller et al 2002, Van Wamel et al 2006, Zhang et al 2013). The measured ICD, cell viability and VEGF₁₆₅ transfection efficiency are plotted in figure 7, as a function of p-. It is clearly shown in figure 7(a) that ICD generally increases with the increasing p- and the addition of MBs with higher SPIO concentrations. Cell viability can be expressed as a percentage of the number of viable cells in the experimental samples divided by the number of viable cells in the control samples. As shown in figure 7(b), if sham treatments (i.e. p- = 0) are applied, about a 20% reduction of cell viability might be attributed to the cytotoxicity of bPEI without adding MBs (Fischer et al 1999). However, a slight decline of cell viability (less than 10%) can be observed by adding MBs with increasing SPIO concentration (i.e. from 19.6 to 292.0 μg ml⁻¹). After US exposures, cell viability can be significantly lowered with the increasing p- and SPIO concentration (a maximum reduction of 56.5% is observed under current experiment conditions). Figure 7(c) shows VEGF₁₆₅ transfection efficiency measured for individual samples. For the sham treatments (p- = 0), VEGF₁₆₅ transfection is mainly facilitated by bPEI, thus there is no obvious enhancement observed with the addition of MBs. For cells mixed with bPEI:VEGF₁₆₅ complexes only, US exposures might slightly benefit bPEI:VEGF₁₆₅ transfection without MB IC activities. As albumin-shelled MBs are added into the suspension (viz, cells mixed with bPEI:VEGF₁₆₅ complexes with MBs in Group 1), the DNA transfection efficiency can be statistically raised, compared with the samples mixed with bPEI:VEGF₁₆₅ only. When cells were mixed with bPEI:VEGF₁₆₅ complexes and MBs in Group 2, the coating of SPIOs on MB shells (effective SPIO concentration = 19.6 μg ml⁻¹) significantly improves US-facilitated VEGF₁₆₅ transfection. However, as the concentration of SPIOs in MB shells increases to 114.7 μg ml⁻¹ (viz, cells mixed with bPEI:VEGF₁₆₅ complexes with MBs in Group 3), the transfection efficiency initially increases with the increasing p-, then reaches a peak level of 613.1  ±  46.02 pg ml⁻¹ at p- = 0.5 MPa. As p- is continuously raised to 1.0 MPa, the measured VEGF₁₆₅ concentration, in turn, drops to 542.9  ±  17.26 pg ml⁻¹. When SPIO concentration further increases to 292.0 μg ml⁻¹ (viz, cells mixed with bPEI:VEGF₁₆₅ complexes with MBs in Group 4), the measured VEGF₁₆₅ concentration will increase first, then decrease with the increasing p-, although the peak value appears even earlier at p- = 0.05 MPa. Thus, it is obvious that, compared with cell viability and ICD, much more complicated variation happens in VEGF transfection efficiency assessments.

Zoom In Zoom Out Reset image size