Albumin is not a buffer in plasma.

Abstract

Dear Sir, While not disagreeing with general conclusions of Caironi and Gattinoni in their recent paper1 and actually very much appreciating their detailed review of albumin physiology, we should like to state one basic point of disagreement. Referring to albumin's imidazole residues authors write In fact, having a pH of about 6.75, residues may both give up or accept H+ from environment depending on surrounding pH, thereby acting as a buffer molecule. Apart from fact that 6.75 represents pK and not pH, in our opinion albumin is not a buffer under physiological conditions. Buffers are defined, according to Donald van Slyke, as substances which by their presence in solution increase amount of acid or alkali that must be added to cause unit change in pH2. Buffering capacity is measured in mEq/L (per pH unit), commonly symbolised as value. Even though almost 90 years old, van Slyke's conclusion still holds true: ability of blood in vivo to neutralize non-volatile acids without change in pH is not due to buffer action ... . It is due to volatility of one of [the] acids, CO2, and ability of respiratory apparatus to remove... excess CO2 .... . If albumin is to be considered a buffer, its concentration ought to be positively correlated with b value of whole blood. Its acid-base properties have been well characterised in human plasma by Figge3 and Constable and Stampfli4. The formulae for relation of albumin dissociation with pH can easily be incorporated into well-known set of six basic equations proposed by Peter Stewart5, i.e. mass conservation, electroneutrality, Henderson-Hasselbalch formula, and three dissociation mass equilibria for water, non-volatile weak acids and bicarbonate/carbonate. We integrated these equations into our acid-base computer programme for pH calculation and simulated titration by acid to plasma, varying buffer base (Strong Ion Difference, SID)6. In complete accordance with Stewart's own results5 the properties of complex systems ... are often quite contrary to qualitative predictions and adding a weak acid to a CO2 solution actually decreases buffering ability of that solution rather than increases it (chapter 7.3). The buffer strengths of solutions with albumin values of 20, 40 and 60 g/L and normal electrolyte composition and pCO2 are calculated to be 71, 62 and 54 mEq/L per unit change in pH; graphs can be created on our website6.