Investigation of protein destabilization in monodisperse poly(lactide-co-glycolide) microparticles

Poly(lactide-co-glycolide) (PLG) microspheres are a powerful choice for controlled release of protein therapeutics. However, several properties of PLG microspheres, particularly hydrophobicity and acidic microenvironment, may have detrimental effects on encapsulated protein, causing structural destabilization, potential immunogenicity, and a loss of biological activity. Researchers have co-encapsulated a number of excipients to counteract PLG???s detrimental effects and protect encapsulated protein molecules, but it remains important to characterize the relationships that drive protein destabilization in PLG microspheres. A clear understanding of the fundamental mechanisms controlling protein stability and release will help us to better exploit the capabilities of PLG microspheres and to optimize microsphere formulations for more precise controlled release protein delivery. It is our overall hypothesis that PLG microsphere diameter controls the rate and degree of encapsulated protein destabilization. The degradation process of PLG microspheres and hindered diffusion of acidic by-products cause the formation of a pH gradient from the surface to the core of the particles, which autocatalyzes the PLG degradation reaction. This phenomenon drives the relationship between microsphere diameter and degradation, erosion, and release rates. Since acidic microenvironment development is intricately related to microsphere size, and acidity is a primary cause of PLG-encapsulated protein destabilization, we seek to correlate the degree and rate of protein destabilization to microsphere diameter. Our first step was to develop a method for studying protein destabilization in PLG microspheres. To observe the amount and structure of both released and encapsulated protein at each time step during an in vitro release study, we have developed a three-step extraction protocol to recover bovine serum albumin (BSA) from PLG microspheres. In this protocol, we first completely extract all water-soluble BSA entrapped within PLG microspheres, followed by any noncovalent then covalent protein aggregates. Regardless of modifications this protocol has proven to be less successful with a more sensitive protein therapeutic, polyclonal human immunoglobulin g (IgG), but our method will still allow us to investigate the stability of IgG released from PLG microspheres. Next, we investigated the relationships between PLG microsphere diameter and BSA stability and release. We first utilized the extraction protocol mentioned above and an extended in vitro release study to investigate BSA destabilization during the majority of the particles??? in vitro lifetime. The results demonstrate that microsphere diameter and initial PLG molecular weight both affect BSA destabilization and release rates. In particular, the microspheres made from 0.60 dL/g PLG underwent more autocatalytic degradation and faster overall BSA release than their 0.20 dL/g PLG counterparts, especially at larger particle diameters. Also, several BSA species were observed in the soluble entrapped protein and release supernatants: a dimer/trimer component, the BSA monomer, and a 55 kDa fragment. We next refined our scope with a second iteration of experiments involving higher BSA loading (10% wt BSA/wt PLG) and some larger PLG microspheres, including 30, 50, and 70 um diameter ranges. During these experiments, 55-, 40- and 25-kDa soluble BSA fragment species appeared over the course of time. The 70 um microspheres exhibited the earliest BSA fragmentation, followed by the 50 um then the 30 um. The 55 kDa BSA fragment prevailed, followed by the 40 kDa species; the 25 kDa fragment appeared the most/earliest in the 70 um spheres, then 50 um, and was not observed in the 30 um microsphere sample. These results describe a clear, direct relationship between 0.60 dL/g PLG particle diameter and the timing and degree of acid-induced encapsulated BSA fragmentation. We then chose to examine the destabilization and release properties of human polyclonal IgG in PLG microspheres. We first demonstrated the applicability of our characterization techniques by performing a preliminary encapsulation, release, and stability study of IgG in non-uniform PLG microspheres fabricated using a conventional homogenizer-based process. Our second experimental iteration utilized the same PLG molecular weight (0.60 dL/g i.v.), theoretical protein loading (10% w/w), and uniform microsphere diameter ranges as the secondary BSA study (again, 30, 50, and 70 um) in order to understand particle diameter-related IgG behavior and compare to our BSA findings. Several components, including the 150 kDa IgG monomer, a 250 kDa molecule, and dimers, were observed in the initial release supernatants from these studies; these non-monomeric species were more apparent in the larger PLG microspheres. In addition, the results of the Easy-Titer assay, which quantifies bioactive IgG, suggest that IgG remains reactive for a longer time in the larger PLG microspheres. Overall, both the aggregation and reactivity of released IgG correlated to PLG microsphere diameter, but the underlying causes are still unclear. In summary, this work relates PLG microspheres??? degradation and release process, microsphere diameter, and the destabilization mechanisms for two types of protein, BSA and IgG. It has become clear that PLG microsphere diameter is one of the key factors controlling the stability of encapsulated protein; this is especially obvious in the case of BSA