Recombinant Human Albumin in Cell Culture

Human serum albumin is the most abundant protein in plasma, present at roughly 35-50 g/L. Structurally it is a single polypeptide of 585 amino acids with a molecular weight of about 66.5 kDa, folded into three homologous helical domains and cross-linked by 17 disulphide bonds, with one free cysteine at position 34. It is not glycosylated, which is a convenience for recombinant production because it removes the need for a host that performs human-type glycosylation.

In cell culture, albumin is doing several jobs at once, and confusion about which of them matters for a given application is the usual source of disappointment when substituting one albumin for another.

Carrier and solubiliser. This is the dominant function. Albumin is a general-purpose transport protein with multiple binding sites for fatty acids, and further sites for hormones, bilirubin, metal ions and a wide range of small hydrophobic molecules. In a defined medium it is how lipids and other poorly soluble nutrients are presented to cells in a usable, non-toxic form. Free fatty acids in aqueous solution are detergents and are cytotoxic; bound to albumin they are a nutrient.

Antioxidant and thiol buffer. The free thiol at cysteine-34 is a significant reducing species and scavenges reactive oxygen species and free radicals. In serum-free medium, where the antioxidant capacity normally supplied by serum is absent, this contribution is meaningful.

Scavenger of toxic species. Albumin binds heavy metal ions, excess free fatty acids and various lipid peroxidation products, sequestering compounds that would otherwise damage cells. This is part of why albumin often appears to "rescue" cultures that grow poorly in a minimal defined medium -- it is removing toxicity rather than adding nutrition.

Shear and interfacial protection. Albumin is surface-active and adsorbs at gas-liquid interfaces. In stirred, sparged or shaken culture it reduces damage from bubble rupture and hydrodynamic shear. It also coats plastic surfaces, reducing non-specific adsorption of medium components and of secreted product -- a practical concern at low protein concentrations and in small volumes.

Colloid osmotic contribution. At the concentrations used it makes a modest contribution to the osmotic properties of the medium, which is more relevant in perfusion and high-density processes than in flask culture.

Expression systems. rHSA is produced commercially in yeast -- principally Pichia pastoris, and also Saccharomyces cerevisiae -- and in transgenic rice (Oryza sativa). Yeast systems are well established: expression under the AOX1 methanol-inducible promoter in Pichia has been reported at 1.6 g/L in shake flask and 8.86 g/L in fermenter culture. Rice-expressed rHSA has been extensively characterised and compared against both plasma-derived albumin and yeast-expressed material.

Because albumin is non-glycosylated, all of these systems can in principle produce a protein with the correct primary sequence and no glycosylation mismatch. The practical differences between systems lie in host-derived impurities, purification process, and post-translational modifications that are not part of the native protein.

Why use recombinant rather than plasma-derived albumin:

• No human donor material. Plasma-derived albumin is a human blood product and carries the associated adventitious agent risk profile, donor screening requirements and supply constraints. Removing it from a manufacturing process removes an entire category of risk assessment.
• Animal-origin-free and xeno-free. For stem cell, cell therapy and regenerative medicine work, and increasingly for biologics manufacture generally, xeno-free status is a requirement rather than a preference.
• Supply and price stability. Recombinant production is a fermentation process that responds to demand. Plasma supply does not.
• Consistency. A defined fermentation and purification process is more controllable than pooled donor plasma.

An honest caveat on consistency. Recombinant does not automatically mean invariant. Published characterisation of rice-expressed rHSA from multiple suppliers has documented supplier-to-supplier and lot-to-lot variability in glycation -- non-enzymatic modification by sugars during expression and processing. Glycation alters binding properties and can affect performance. Recombinant albumin is considerably more consistent than serum, but it is still a protein produced by a living organism, and it should be lot-tested for demanding applications exactly as any other critical raw material would be.

Formats. rHSA is commonly supplied as a sterile-filtered concentrated solution, frequently at 200 mg/mL (20%), and also as a lyophilised powder. Solution format avoids the reconstitution step and the risk of foaming and denaturation that comes with dissolving a surface-active protein; powder gives longer ambient shelf life and cheaper shipping.

Typical range: 0.1-10 g/L, with most serum-free formulations using 0.5-5 g/L. The right figure depends entirely on which of albumin's functions you are relying on:

• As a lipid and fatty acid carrier, the concentration needs to be sufficient to bind the lipid load you are supplying. This usually places it in the low grams-per-litre region.
• As a shear protectant in stirred or sparged culture, 1-5 g/L is typical, though a synthetic surfactant such as poloxamer 188 is often used alongside or instead, and is considerably cheaper for that specific purpose.
• As a surface-blocking agent to prevent adsorption losses, much lower concentrations suffice -- fractions of a gram per litre.
• As an antioxidant and scavenger, the requirement scales with how stressed the culture is and how minimal the medium.

Optimise it rather than inheriting it. Albumin is frequently one of the more expensive components of a defined medium, and the concentration carried over from a published formulation is often higher than a given process needs. A simple titration -- three or four concentrations across the range, measuring growth, viability, product titre and product quality -- frequently identifies a substantial saving. Equally, if a defined medium is underperforming, raising albumin is a reasonable early thing to test.

Preparing solutions. Albumin is surface-active and foams readily; foaming denatures protein at the air-liquid interface. Dissolve powder by sprinkling onto the surface of the liquid and allowing it to hydrate without stirring, then mix gently. Do not vortex, do not shake vigorously, and do not sparge a solution being prepared.

Sterilisation. Filter through 0.22 um. Do not autoclave. Note that albumin solutions foul filters relatively quickly at high concentration; use an appropriately sized filter and consider a prefilter.

Storage. Follow the supplier's specification. Concentrated solutions are typically stored refrigerated or frozen; repeated freeze-thaw should be avoided, and aliquoting on first thaw is worthwhile. Lyophilised material stores at 2-8 C or below and should be protected from moisture.

Endotoxin. Because albumin is used at grams per litre, its endotoxin contribution to the medium is proportionally large. A material at even a modest endotoxin specification can dominate the endotoxin load of a defined medium. Check the specification on the certificate of analysis and select a low-endotoxin grade for sensitive cells -- primary immune cells and stem cells in particular.

This is the point that causes the most confusion when people substitute one albumin for another and get a different result, and it deserves its own treatment.

Albumin is almost never delivered empty. Native albumin purified from plasma or serum arrives with a substantial cargo of bound fatty acids and other small molecules. Recombinant albumin from a fermentation process carries a different and generally smaller cargo. Suppliers also sell explicitly fatty-acid-free grades, prepared by charcoal treatment or similar processes to strip bound lipids.

These behave differently, and which one you want depends on the job:

• If albumin is your lipid delivery vehicle, you may want it loaded -- either supplied lipid-rich or complexed with a defined lipid mixture before addition. Many serum-free formulations deliberately pair albumin with a chemically defined lipid concentrate for exactly this reason.
• If you are studying fatty acid metabolism, lipid signalling, or anything where the lipid environment is the independent variable, you need fatty-acid-free albumin. Using a lipid-loaded preparation means adding an uncharacterised and variable dose of the thing you are trying to control. This is a recurring source of irreproducibility in metabolism research.
• If albumin is present purely for shear protection or surface blocking, the fatty acid content is largely irrelevant.

The practical rule: when a medium behaves differently after an albumin change, check the fatty acid specification before suspecting anything else. Two albumins with identical protein purity and identical concentration can support quite different growth if one is lipid-loaded and the other stripped. Record the grade and the supplier in your methods, because "1 g/L albumin" does not adequately describe what was added.

The same logic applies to other bound species -- metals, and residual process components from the purification. Charcoal-stripped and low-endotoxin grades exist because these differences are real and measurable.

Bovine serum albumin is the traditional and much cheaper choice, and for a great deal of routine work it remains perfectly adequate. The decision is not about which protein is better but about what the application requires.

Choose rHSA when:

• The process must be animal-origin-free or xeno-free -- cell therapy, stem cell culture, regenerative medicine, and increasingly biologics manufacture generally.
• The cells are human and you want a homologous protein. Species differences in albumin sequence and binding properties are real, and human cells cultured with human albumin avoid introducing a xenogeneic protein whose binding profile differs.
• You need to eliminate a bovine-derived adventitious agent pathway from a process, along with its documentation burden.
• Batch consistency matters enough to justify the cost. Recombinant material from a controlled fermentation is generally more consistent than BSA fractionated from pooled bovine serum, though as noted it is not immune to lot variation.

BSA remains reasonable when:

• The work is routine research on established lines, cost matters, and animal origin is not a constraint.
• You need a specific characterised grade -- low-endotoxin, low-fatty-acid, protease-free -- that is well established in BSA form and validated for your assay.
• Historical continuity matters, for instance in a long-running study where changing the albumin would introduce a step change in the data.

Do not substitute casually in either direction. rHSA and BSA differ in amino acid sequence, in binding affinities for fatty acids and other ligands, in typical fatty acid loading, and in impurity profile. A medium optimised on one will not necessarily perform identically on the other, and the concentration may need re-optimising. Test in parallel across several passages, measuring growth, viability and the product or functional endpoint you care about, before converting a process.

On cost. rHSA is substantially more expensive than BSA per gram, and at grams per litre this is a material line item. The honest framing is that the cost is justified by regulatory position, xeno-free status and consistency, not by superior performance as such -- for many cell lines in routine culture, well-chosen BSA performs equivalently.

Albumin is one component of the standard toolkit used to rebuild serum's functions from defined parts. A typical serum-free supplement package includes:

• Insulin, usually recombinant human insulin, driving glucose and amino acid uptake and providing a survival and growth signal.
• Transferrin, increasingly recombinant human transferrin, delivering iron in a bound, non-toxic form. Iron is essential and free iron is a potent generator of reactive oxygen species, so the delivery mechanism matters as much as the amount.
• Selenium, as sodium selenite, a cofactor for glutathione peroxidase and other selenoproteins, functioning as part of the antioxidant system.
• Albumin, providing the carrier, antioxidant, scavenger and shear-protection functions described above.
• Defined lipids, typically supplied as a chemically defined lipid concentrate, often paired with albumin as the carrier.
• Ethanolamine, a phospholipid precursor, included in some formulations.
• A shear-protective surfactant such as poloxamer 188 for suspension and bioreactor culture.

The first three of these are supplied together as the widely used ITS supplement, and albumin is frequently added alongside. Where a formulation is intended to be animal-origin-free, all components need to be recombinant or synthetic -- an animal-origin-free medium containing bovine albumin or porcine-derived insulin is not animal-origin-free, and this is a common oversight when assembling a formulation from individually sourced components.

A practical note on troubleshooting defined media. When cells grow poorly in a defined medium, the temptation is to add more of everything. A more efficient approach is to test the categories separately: is the problem a missing growth signal (test insulin and growth factors), a missing carrier or protective function (test albumin), an iron delivery problem (test transferrin), an oxidative stress problem (test selenium and antioxidants), or a lipid deficiency (test a defined lipid supplement)? Albumin in particular tends to mask several different underlying problems at once, because it addresses toxicity, lipid delivery and oxidative stress simultaneously -- which makes it a useful diagnostic when added and removed deliberately.