(W1130-06-31) Evaluating Strategies to Minimize Leakage of Protein from Ethylcellulose Microcapsules Using Bovine Serum Albumin as Model Protein

Purpose: Phenylalanine Ammonia Lyase (PAL) is an enzyme that metabolizes Phe to trans cinnamic acid and ammonia. We propose to encapsulate phenylalanine ammonia lyase (PAL) in non-digestible ethylcellulose (EC) microcapsules as a potential treatment for PKU. However, there is the extensive leakage of PAL out of the microcapsules and leaked PAL will immediately lose its activity in the gastrointestinal tract. The overall objective of our work was to devise formulation strategies to reduce leakage of PAL from spray dried EC microcapsules. We hypothesized that a lower molecular weight grade of EC would exhibit greater chain mobility that would enable more complete encapsulation of the protein and reduce leakage. Our first specific aim was to determine the effect of EC molecular weight on microcapsule size, yield, encapsulation efficiency (EE), and protein leakage. We next hypothesized that reduction in the viscosity of the polymer solution would result in more complete protein encapsulation and reduced protein leakage. In our second specific aim, we used varying volumes of ethanol to reduce the viscosity of EC solutions and manufactured microcapsules using these polymer solutions and characterized them as described above. In our studies, bovine serum albumin (BSA) was used as the model protein for the more expensive PAL.

Methods: Microcapsules were manufactured by dispersing micronized BSA particles in 4 % w/v solutions of either 7 cps grade EC or 20 cps grade EC in dichloromethane (DCM). The suspension was spray dried using a Buchi Model B-290 spray drier. Microcapsule sizes of each batch were determined by a Malvern Mastersizer 3000. EE of BSA was determined by dissolving aliquots of microcapsules in DCM: ethanol (1:1) solution to dissolve the polymer shell and simultaneously precipitate the encapsulated protein. The protein pellet was resolubilized in an alkaline SDS solution and protein content determined by the Lowry assay. Protein leakage was determined by suspending aliquots of microcapsules in water and incubating them in a rotating shaker at 37°C for 15 minutes, 1 hour, and 3 hours. At each time point, samples of the release medium were taken, released protein precipitated with trichloroacetic acid, the protein pellet redissolved in alkaline SDS solution and protein content determined by the Lowry assay. 4% w/v solutions of 20 cps EC were made in: 100% DCM, 90:10 DCM:ethanol, and 80:20 DCM:ethanol. Viscosity profiles of these polymer solutions were determined using a Brookfield DV III+ rheometer over a range of rotation speeds. Subsequently, BSA microcapsules were manufactured using these EC solutions as described above. Size, yield, EE, and protein leakage from these batches of microcapsules were determined as described above.

Results: Yield was not affected by the molecular weight of EC. Particle sizes of the batch made with 7cps EC had a d50 of 6.7 microns and were larger than the 4.4 micron d50 of the 20 cps batch. EE of BSA in the 7cps batch was 118.6 ± 4.2 % (mean ± S.D.) and higher than the 60.8 + 8.0 % EE of the 20 cps batch. The burst release of BSA measured at 15 minutes of 90.9 + 68% in the 7 cps batch was higher than the burst release of 55.3 + 16 % in the 20 cps batch. However, cumulative protein leakage over 3 hours was 100 % in both batches. At a spindle speed of 50 rpm, the viscosity of 4 % w/v EC in 100 % DCM was 16.2 + 0.7 cps while the viscosities in 90:10 and 80:20 DCM:EtOH solutions were 13.2 + 0 cps and 14.4 + 2.5 cps, respectively. Particle sizes of batches made 4 % w/v EC in 90:10 and 80:20 DCM:EtOH solutions showed a d50 of 6 – 7 microns and yields of 110.5% ± 16.0 and 123.5% ± 8.5, respectively. Cumulative protein leakage over 3 hours from batches manufactured using the 4 % w/v EC in 90:10 and 80:20 DCM:EtOH solutions were 76 ± 1.7 % and 80 ± 6.5 %. respectively.

Conclusion: A decrease in the molecular weight of EC did not affect yield or size of the microcapsules but increased EE. However, the increased EE of the microcapsules manufactured with the 7 cps EC did not reduce protein leakage and actually produced a higher burst release. Addition of ethanol to 4% w/v EC solutions in DCM reduced the viscosity of the polymer solutions. Microcapsules manufactured with 4% w/v EC in DCM:ethanol solutions produced modest reductions in protein leakage. These small reductions in leakage are unlikely to improve efficacy of our therapeutic approach. There is therefore a need to develop additional strategies to reduce the leakage of protein from EC microcapsules.


Viscosity profiles of various polymer solutions


Protein release from microcapsule batches


Encapsulation efficiency of various microcapsule batches