Background: Human serum albumin (HSA) is an abundant plasma protein that binds a wide variety of hydrophobic ligands including fatty acids, bilirubin, thyroxine and hemin. Although HSAheme complexes do not bind oxygen reversibly, it may be possible to develop modified HSA proteins or heme groups that will confer this ability on the complex.

We describe molecular capturing properties of protein nanotubes with a controllable ligand binding affinity and size selectivity. These practical biocylinders were prepared using an alternating layer-by-layer (LbL) assembly of protein and oppositely charged poly(amino acid) into the nanoporous polycarbonate (PC) membrane (pore diameter, 400 nm), with subsequent dissolution of the template. The tube wall typically comprises six layers of poly-L-arginine (PLA) and human serum albumin (HSA) [(PLA/HSA)(3)]. Use of high molecular weight PLA (M(w) = ca. 70 000) yielded robust nanotubes, which are available as lyophilized powder. The (PLA/HSA)(3) nanotubes swelled considerably in water, although the outer diameter was almost unaltered. Uranyl ion, 3,3'-diethylthiacarbocyanine iodide, and zinc(II) protoporphyrin IX (ZnPP) were bound to the HSA component in the cylinder wall. Similar nanotubes comprising recombinant HSA mutant [rHSA(His)], which has a strong binding affinity for ZnPP, captured this ligand more tightly. Furthermore, addition of excess myristic acid released ZnPP from the tubes through a ligand replacement reaction. The hybrid nanotubes bearing a single avidin layer as an internal surface captured FITC-biotin efficiently. Biotin-labeled nanoparticles are also incorporated into the tubes when their particle size is sufficiently small to enter the pores. Subsequent TEM observation revealed a line of loaded nanoparticles (100 nm) in the one-dimensional space interior.

Hemoglobin (Hb, Mw: 64 500) and albumin (Mw: 66 500) are major protein components in our circulatory system. On the basis of bioconjugate chemistry of these proteins, we have synthesized artificial O(2) carriers of two types, which will be useful as transfusion alternatives in clinical situations. Along with sufficient O(2) transporting capability, they show no pathogen, no blood type antigen, biocompatibility, stability, capability for long-term storage, and prompt degradation in vivo. Herein, we present the latest results from our research on these artificial O(2) carriers, Hb-vesicles (HbV) and albumin-hemes. (i) HbV is a cellular type Hb-based O(2) carrier. Phospholipid vesicles (liposomes, 250 nm diameter) encapsulate highly purified and concentrated human Hb (35 g/dL) to mimic the red blood cell (RBC) structure and eliminate side effects of molecular Hb such as vasoconstriction. The particle surface is modified with PEG-conjugated phospholipids, thereby improving blood compatibility and dispersion stability. Manipulation of physicochemical parameters of HbV, such as O(2) binding affinity and suspension rheology, supports the use of HbV for versatile medical applications. (ii) Human serum albumin (HSA) incorporates synthetic Fe(2+)porphyrin (FeP) to yield unique albumin-based O(2) carriers. Changing the chemical structure of incorporated FeP controls O(2) binding parameters. In fact, PEG-modified HSA-FeP showed good blood compatibility and O(2) transport in vivo. Furthermore, the genetically engineered heme pocket in HSA can confer O(2) binding ability to the incorporated natural Fe(2+)protoporphyrin IX (heme). The O(2) binding affinity of the recombinant HSA (rHSA)-heme is adjusted to a similar value to that of RBC through optimization of the amino acid residues around the coordinated O(2).