228
AUSTEN IUGGS
some of the 0,-binding sites (hemes); if n = 1, the sites are functionally
independent of one another. If two or more components are present
which differ substantially in oxygen affinity an undulating curve may
be produced such that n, at y = #, may be appreciably less than 1.0.
Although values of n < 1 have frequently been interpreted as implying
"negative heme-heme interactions" ( Manwell, 1964), such values can
also reflect the presence of multiple components which differ sufficiently
in oxygen affinity. This difference may arise experimentally in a number
of ways other than the existence of different components with intrinsically different oxygen affinities. So, for example, two hemoglobin components may be present with identical oxygen equilibria but different
sensitivities to oxidation of the iron. This situation sometimes results in
the formation of significant quantities of methemoglobin in one of the
components but not the other during the preparation or during the
equilibrium measurements. Since the presence of ferric heme can
increase the oxygen affinity of the unoxidized hemes of the same molecule, oxygen equilibrium measurements of a mixture will show a lower
apparent value of n than would be found in the absence of oxidation.
The oxygen equilibrium curves then appear to have higher than expected
oxygenation levels at low pressures. These considerations are mentioned
here because most fish hemolyzates have multiple components, and
at least some of these differ considerably in sensitivity to oxidation.
The value of n should not be interpreted as reflecting the energy of
interaction. Wyman (1964) has shown that high values of n (up to 5)
in invertebrate hemoglobins may in fact be associated with a lower overall free energy of interaction than the lower value of n ( 22.8) characteristic of mammalian hemoglobins. He has devised a graphical method for
obtaining the overall free energy of interaction (see Fig. 4). Unfortunately, this method requires getting data at both very low and at very
high oxygenation levels ( <3% and >97%)-values that are often difficult
to obtain in practice.
The cooperativity between 0,-binding sites in mammalian hemoglobins is largely independent of pH, at least within the normal physiological range, and is uniform between 10 and 901% saturation ( n z 2.83.0). At sufficiently low or high oxygenation levels, as shown in Fig. 4,
the value of n approaches unity. For some fish hemoglobins, however,
n is a function of pH between 10 and 90% oxygenation and often changes
greatly with the degree of oxygenation. This pH dependence differs in
sign in different hemoglobins. Thus, for example, in the hemoglobin of
the barndoor skate, n increases from 1.2 to 1.8 between pH 6.5 and 7.5
( Manwell, 1958b), whereas in the coelacanth, Latimeria, n decreases
from 1.6 to 1.1 between pH 7 and 8 (Bonaventura and Riggs, 1969).
AUSTEN IUGGS
some of the 0,-binding sites (hemes); if n = 1, the sites are functionally
independent of one another. If two or more components are present
which differ substantially in oxygen affinity an undulating curve may
be produced such that n, at y = #, may be appreciably less than 1.0.
Although values of n < 1 have frequently been interpreted as implying
"negative heme-heme interactions" ( Manwell, 1964), such values can
also reflect the presence of multiple components which differ sufficiently
in oxygen affinity. This difference may arise experimentally in a number
of ways other than the existence of different components with intrinsically different oxygen affinities. So, for example, two hemoglobin components may be present with identical oxygen equilibria but different
sensitivities to oxidation of the iron. This situation sometimes results in
the formation of significant quantities of methemoglobin in one of the
components but not the other during the preparation or during the
equilibrium measurements. Since the presence of ferric heme can
increase the oxygen affinity of the unoxidized hemes of the same molecule, oxygen equilibrium measurements of a mixture will show a lower
apparent value of n than would be found in the absence of oxidation.
The oxygen equilibrium curves then appear to have higher than expected
oxygenation levels at low pressures. These considerations are mentioned
here because most fish hemolyzates have multiple components, and
at least some of these differ considerably in sensitivity to oxidation.
The value of n should not be interpreted as reflecting the energy of
interaction. Wyman (1964) has shown that high values of n (up to 5)
in invertebrate hemoglobins may in fact be associated with a lower overall free energy of interaction than the lower value of n ( 22.8) characteristic of mammalian hemoglobins. He has devised a graphical method for
obtaining the overall free energy of interaction (see Fig. 4). Unfortunately, this method requires getting data at both very low and at very
high oxygenation levels ( <3% and >97%)-values that are often difficult
to obtain in practice.
The cooperativity between 0,-binding sites in mammalian hemoglobins is largely independent of pH, at least within the normal physiological range, and is uniform between 10 and 901% saturation ( n z 2.83.0). At sufficiently low or high oxygenation levels, as shown in Fig. 4,
the value of n approaches unity. For some fish hemoglobins, however,
n is a function of pH between 10 and 90% oxygenation and often changes
greatly with the degree of oxygenation. This pH dependence differs in
sign in different hemoglobins. Thus, for example, in the hemoglobin of
the barndoor skate, n increases from 1.2 to 1.8 between pH 6.5 and 7.5
( Manwell, 1958b), whereas in the coelacanth, Latimeria, n decreases
from 1.6 to 1.1 between pH 7 and 8 (Bonaventura and Riggs, 1969).