is haemoglobin a buffer for blood plasma?

haemoglobin is an integral part of calculating base excess in the Siggaard-Andersen approach to acid-base analysis.
slide 9 in a presentation on this website lists the diverse components in this calculation.

but according to the physiocochemical or Stewart approach, the pH of body fluid depends on only 3 independent quantities:

1- the "strong ion difference" or SID
the difference between the completely dissociated or "strong" positive ions and their negative counterparts.
2- the total amount of weak acids in plasmaAtot
3- and the carbon dioxide partial pressure, PCO2
to quote Peter Stewart himself:
"Hemoglobin in RBCs is usually assigned an important role as a 'buffer' for plasma in whole blood. Because it is in the RBC-ICF, not in the plasma, there is no way that hemoglobin can affect the [SID] the Pc, or the [Atot] of the plasma surrounding the RBCs, and therefore no way it can affect plasma [H+]. In other words, it cannot possibly serve as a plasma 'buffer.'" How to Understand Acid-Base, page 176

there is a very Salomonic solution to these seemingly irreconciliable positions, though. and it dates back to the pioneering times of acid-base research, when Stewart still was in his diapers, and Ole Siggaard-Andersen had yet to be born:

Studies of Gas and Electrolyte Equilibria in the Blood
was published in The Journal of Biological Chemistry by Donald van Slyke et alii in july 1923 and is available on the net .
from page 780 they describe the theoretical basis of a phaenomenon observed even earlier, in 1920, by E.J.Warburg (reference in the van Slyke article): when whole blood is moved to higher pH values (in vitro) by reducing PCO2, the erythrocytes loose water and chloride to the serum. this is attributed to the Gibbs-Donnan-effect (very nicely described on an entomological website: this term means, in a nutshell, that a system with two solutions separated by a semipermeable membrane and different amounts of non-diffusing charged substances on the two sides of that membrane forces those ions capable of diffusing through it to assume different concentrations in the separated liquids.
the typical case is that of a protein acting as a multivalent anion, like albumin or haemoglobin, on one side of the membrane, repelling mobile anions like chloride to the other side, and attracting mobile kations like potassium to its own side.
in the case of blood the immobile actors are albumin in the serum and haemoglobin inside the cells, with albumin typically having around 10mEq/l negative charge and haemoglobin around 30, the Donnan effect of the latter is thus dominating, causing the intracellular chloride level to be lower than that of the serum.
there is a third actor, though: sodium. the action of the Na-K-pump causes this kation to essentially be restricted to the extracellular space, creating a Donnan effect of its own, further forcing chloride to assume a higher concentration outside of than inside of the cells.
now - as the pH moves to higher values, the proteins following their mass equilibrium liberate more H+, increasing their negative charge and displacing HCO3-. as this increase is much more pronounced inside the cells, the Donnan effect moves more chloride out to the serum, reducing SID. this happens every time blood moves through the lungs. both the reduced SID and the increase amount of ionised proteins make HCO3- convert to CO2, thus aiding in the removal of the latter.
the reverse is true for the transition from arterial to venous blood, assisting CO2 removal.

haemoglobin inside of red cells is thus capable of influencing serum acid-base balance - BUT: only by affecting two of the three independent variables of the Stewart approach: PCO2 and SID. as these are accounted for by the measurements at the base of the Stewart type acid-base analysis, the haemoglobin value is of no further relevance. if one does not measure serum chloride, the analysis is limited to the base excess method, and the haemoglobin plays its roll ....

hyperventilation accentuates the movement of chloride and water out of red blood cells.
this was experimentally corroborated in volunteers by Rapoport et alii in 1946, who also saw a reduction of serum Na+ levels in their hyperventilating subjects. source:
S. Rapoport, Charles D. Stevens, George L. Engel, Eugene B. Ferris, and Myrtle Logan THE EFFECT OF VOLUNTARY OVERBREATHING ON THE ELECTROLYTE EQUILIBRIUM OF ARTERIAL BLOOD IN MAN J. Biol. Chem., May 1946; 163: 411 - 427

later studies did not find any such effect, though:
Gaudenz et alii, Plasma Ionized Magnesium during Acute Hyperventilation in Humans; Clinical Science (1996) 91, 347
Steurer et alii, Elektrolytveränderungen während und nach willkürlicher Hyperventilation (Changes in Plasma Electrolyte Levels during and after Intentional Hyperventilation); Schweizerische Rundschau für Medizin (1995) 84, 328.

a recent pubmed search did not throw up more data. (i, RG, shall update this text, as soon i can get new data on this subject!) the old studies seem to have been meticulously done, and unlike the newer ones they provide a sound theoretical framework for their experimental observations. these theoretical frameworks with their emphasis on the changing degree of ionisation actually are quite close to the logic of the Stewart approach.
i did a n=4 informal in-vitro study of my own - it confirms the old data.

consult the glossary for other aspects of acid-base equilibria and the rules and mathematics behind our website:     Glossary