Design of Albumin-Coated Microbubbles Loaded With Polylactide Nanoparticles

Abstract

Objectives: A protocol was designed to produce albumin-coated microbubbles (MBs) loaded with functionalized polylactide (PLA) nanoparticles (NPs) for future drug delivery studies.

Methods: Microbubbles resulted from the sonication of 5% bovine serum albumin and 15% dextrose solution. Functionalized NPs were produced by mixing fluorescent PLA and PLA-polyethylene glycol-carboxylate conjugates. Nanoparticle-loaded MBs resulted from the covalent conjugation of functionalized NPs and MBs. Three NP/MB volume ratios (1/1, 1/10, and 1/100) and unloaded MBs were produced and compared. Statistical evaluations were based on quantitative analysis of 3 parameters at 4 time points (1, 4, 5, and 6 days post MB fabrication): MB diameter using a circle detection routine based on the Hough transform, MB number density using a hemocytometer, and NP-loading yield based on MB counts from fluorescence and light microscopic images. Loading capacity of the albumin-coated MBs was evaluated by fluorescence.

Results: Loaded MB sizes were stable over 6 days after production and were not significantly different from that of time-matched unloaded MBs. Number density evaluation showed that only 1/1 NP/MB volume ratio and unloaded MB number densities were stable over time, and that the 1/1 MB number density evaluated at each time point was not significantly different from that of unloaded MBs. The 1/10 and 1/100 NP/MB volume ratios had unstable number densities that were significantly different from that of unloaded MBs (P < .05). Fluorescence evaluation suggested that 1/1 MBs had a higher NP-loading yield than 1/10 and 1/100 MBs. Quantitative loading evaluation suggested that the 1/1 MBs had a loading capacity of 3700 NPs/MB.

Conclusions: A protocol was developed to load albumin MBs with functionalized PLA NPs for further drug delivery studies. The 1/1 NP/MB volume ratio appeared to be the most efficient to produce stable loaded MBs with a loading capacity of 3700 NPs/MB.

Keywords: fluorescence; functionalized nanoparticles; microbubbles; stability.

Figure 1

Schematic representation of the covalent linking of PLA NPs to the albumin MB shell via the carbodiimide technique: NPs are loaded to the MB shell through an amine bond between the carboxyl groups on the NP surface and the amine groups on the MB surface. EDC indicates 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride; and NHS, N-hydroxysuccinimide.

Figure 2

ImageJ-processed light microscopic image (A) and fluorescence microscopic image (B). Edges of the elements detected in both images were provided by ImageJ to allow the determination of number density.

Figure 3

Nanoparticle-loaded MB (volume ratios 1/1, 1/10, and 1/100) and unloaded MB (uMB) mean diameter as a function of time. Error bars represent the 95% CIs. D indicates day.

Figure 4

Nanoparticle-loaded MB (NP/MB volume ratios 1/1, 1/10, and 1/100) and unloaded MB (uMB) mean number density as a function of time. Error bars represent the 95% CIs. D indicates day.

Figure 5

Nanoparticle-loaded MB (NP/MB volume ratios 1/1, 1/10, and 1/100) NP-loading yield as a function of time. Error bars represent 95% CIs; 100% corresponds to the total number of MBs detected that showed fluorescence. D indicates day.

Figure 6

Postexcitation percentage curve for 1/1 NP-loaded MBs plotted against peak rarefactional pressure amplitude (PRPA) at 4.6 MHz. The solid curve represents the logistic fit, and the dashed curves represent the 95% CIs.

Figure 7

A, Fluorescence spectra of FITC with an excitation wavelength of 495 nm: the maximum emission wavelength was found at 517 nm. B, FITC standard curve: fluorescence intensity values at 517 nm as a function of FITC concentration.

Figure 8

Left (A) and right (B) 4T1 tumor tissue sections from a BALB/c mouse incubated with collagen IV and Cy5-labeled donkey anti-rabbit immunoglobulin G. Tumor sections were imaged at ×200 using FITC, 4′,6-diamidino-2-phenylindole, and Cy5 filters.