Synthesis and characterization of tubular amphiphilic networks with controlled pore dimensions for insulin delivery

Abstract

A convenient laboratory process for the preparation of thin-walled (∼0.02 cm) tubular amphiphilic membranes has been developed. The membranes are suitable for implantation and isolation of pancreatic islets from immune responses. The process involves the simultaneous free radical copolymerization/crosslinking of dimethyl acrylamide (DMAAm) and methacrylate ditelechelic polyisobutylene (MA-PIB-MA) in narrow-bore (∼4 mm inner diameter) glass tubes horizontally rotating in a thermostated oven. The pore sizes of the membranes can be controlled by the length, i.e. molecular weight, of the hydrophilic poly(dimethyl acrylamide) (PDMAAm) segment (M c,hydrophilic). Pore sizes, M c,hydrophilic's, and molecular weight cut-off (MWCO) ranges of designed amphiphilic membranes were characterized in terms of Stokes (or viscosity) radii (R s) and the relationships between these parameters were evaluated. Membranes were designed to allow the rapid diffusion of molecules such as insulin (M n = 5733 g/mol, R s = 1.34 nm) but to be opaque to serum albumin (M n > 66 000 g/mol, R s > 3.62 nm) and larger proteins such as immunoglobulins. The diffusion coefficients (D) and permeabilities (P) of tubular and flat-sheet amphiphilic membranes have been compared and were found to be similar. Membrane pore size dimensions of the tubular devices were determined by the out-diffusion of commercially available protein markers of known molecular weights (M n = 6500-66 000 g/mol) and dimensions (Rs = 1.50-3.62 nm). It was found that the minimum Mc,hydrophilic or Rs that still allows the diffusion of insulin is ∼ 800 g/mol or ∼ 1.34nm, respectively, and that the maximum Mc,hydrophilic or Rs that prevents the ingress of antibodies is ∼ 5000 g/mol or ∼ 3.62 nm, respectively. According to diffusion experiments, the presence or absence of lightly crosslinked 1% calcium alginate does not affect the rate of diffusion of glucose and insulin through our tubules. These membranes are being used in vivo for encapsulating islet cells for implantation to correct type 1 diabetes.