Human Serum Albumin (HSA)

HSA is usually isolated from blood plasma with bound fatty acids and bilirubin (see later). Its drug binding capacity means it plays an important role in pharmacodynamics. It displays many activities including antioxidant activity and lipophilic hormone binding. Low albumin levels are associated with poor disease prognosis. Glycated serum albumin is also a marker of glucose intolerance. Diseases associated with HSA are shown in the section HSA in medicine.

There is a considerable literature on HSA: Over 5,000 articles with HSA in the title reported in Pub Med between 1939 and 2024. In recent times papers have focussed upon clinical medical / pharmaceutical applications and developments. For example: Human Serum Albumin: From Molecular Aspects to Biotechnological Applications ( Alessandra di Masi 2023 [open access] ). Human Serum Albumin Misfolding in Aging and Disease (Tsao and Meyer 2022[open access] . Human Serum Albumin: from bench to bedside ( Fanali et al. 2012 ).

It was clear from earlier work on the analogous bovine serum albumin that the protein's single chain was folded into at least 3 domains. This is evident from the loop structure shown in the schematic image of the main chain. Using limited proteolysis I was able to seperate and purify individual HSA domains and investigate their binding properties. It was clear from circular dichroism (CD) measurements that the protein chain was largely alpha-helical and the isolated domains retained this secondary structure and bining properties. But it was also clear that pH changes caused significant relative movement between domains. Bound bilirubin changed its conformation when HSA solution was acidified and tightly bound fatty acids could be recovered by lowering the pH in the presence of activated charcoal. Experience with isolated domains representing two-thirds and one third of the molecule was to prove useful later in working with recombinant HSA at Delta Biotechnology Ltd. (see the section on recombinant HSA)

Anyway, I had plenty of HSA and using the so called "Dansyl-Edman" sequence method together with carboxypeptidase was able to characterise my large HSA fragments as residues 1-386 (P44), 49-307 (P29) and 308 - 584 (P31). To cut a long story short, P44 represented the first two domains and P31 encompassed the third domain with a bit of N-terminal sequence overlap of the middle domain. The fragments remained folded (HSA has 17 disulphide bridges) and I was able to allocate the binding site for N-acetyl tryptophan, bilirubin (haem breakdown product) to the first two domains.

Taken together with parallel studies of ligand binding to BSA, we concluded that HSA binding sites for many hydrophobic ligands clustered in the N-terminal region of the protein. Binding sites for long chain fatty acids were later reported to be present in the C-terminal half of the molecule. A hydrogen bond formed between stearic acid and tyr 401. Competition between natural ligands and drugs was known to occur. At that time we suggested that the common location for binding would probably have important consequences in pharmacodynamics (Open access: Geisow and Beavan 1977)

Many years later at Delta Biotechnology (Nottingham) Ltd we were able to express recombinant HSA in yeast (S. Cerevisiea). Initially in inclusion bodies which were insoluble and required extensive disulphide bridge reduction and refolding to gain an N-terminally blocked product. Largely through the initiative of Andrew Goodey and his team at Delta, we were eventually able to obtain a correctly folded product secreted into the fermentation medium. The signal sequence was correctly processed and high secretion levels were obtained.(Production of Recombinant Human Serum Albumin from Saccharomyces cerevisiae).

I was able to carry out extensive sequencing and mapping on Delta HSA to validate the product, but the most impressive demonstration of identity with the natural product (from donated blood plasma) was by mass spectrometry.

I took purified product to FISONS instruments where the firm was developing electrospray mass spectrometry instrumentation. Despite the instrument used being an early prototype I was "blown away " when the measured mass of Delta rHSA came within 2 to 3 proton masses of that calculated from the amino acid sequence. Had even one of the 17 disulphide bridges not been formed, we would have been able to detect the difference! For me, as a protein chemist used to obtaining rough protein molecular weights by gel electrophoresis, this was like "Saul's revelation on the road to Damascus" and I later becane a consultant to FISONS Instruments as my first consultancy contract.

Recalling the first structural studies I carred out on HSA at the MRC labs at Mill Hill, I suggested that Delta express recombinant forms of 1- and 2- domain HSA. A number of recombinant proteins were designed and successfuly secreted: rHSA 1-585 (full length); 1-387; 1-407 and 1-194. The rationale behind expressing shorted versions of HSA was that (in principle) with 2/3 smaller molecular weight less protein would have to be infused for a given colloid osmotic pressure (oncotic) effect. Patents were applied for but in fact this idea was not pursued in practice

Recombinant HSA was taken through to production in a purpose build pilot plant in Nottingham and a full quality assurance programme initiated in preparation for Phase I clinical trials. After various sales and acquistion, Delta IP and production of rHSA was taken up by ALBUMEDIX (Nottingham). Phase I comparability of recombinant human albumin and human serum albumin has been published: Bosse D1, Praus M, Kiessling P, Nyman L, Andresen C, Waters J, Schindel F. J Clin Pharmacol. 2005 Jan;45(1):57-67.

Albumedix/Sartorius' recombinant albumin (recombumin) can be followed up on their website: Sartorius Recombumin are now marketing rHSA and partnering other businesses to use rHSA for applications such as gene and cell therapies, vaccine stabilisation, protein and peptide formulations and medical device coating. For much background since Delta, interested readers should visit Sartorius web.

Research around albumin and its biology has increased dramatically during the past decade. Several albumin binding proteins and receptors have been identified, and it has been demonstrated that these play an important role in the transport of albumin between different compartments, as well as its internalization, d egradation, salvage and recycling. Most well understood is the interaction of albumin with the neonatal F_c receptor (F_cR_n) and the impact that this receptor has on the long serum half-life of albumin (approximately three weeks).