CLIN.CHEM.23/10, 1813-1818(1977)
MyoglobinBindingin HumanSerum
Lawrence J. Kagen and A. Butt
Seventy-five of 82 human blood sera demonstrated binding
for human myoglobin, the criterion for binding being the
failure of myoglobin to pass through CF 50A (Amicon
Corp.) ultrafilters. In most cases, when myoglobin was
present in a concentration of 2 mg/liter, 50-87.5% of it
was bound. Binding properties of sera were not related to
age or sex of the donors. Dilution of serum by more than
50-fold removed its binding property. The percent binding
could be increased by increasing the concentration of
myoglobin to a maximum of 5-15 mg/liter. The binding
factor of serum has a molecular weight exceeding 50 000
but less than 300 000, as determined by ultrafiltration
experiments. Serum fractions that were devoid of albumin
could bind myoglobin, as did two sera that lacked haptoglobin. Added hemoglobin did not interfere with myoglobin
binding. Ultracentrifugation studies suggested that bound
myoglobin exists in a complex that is slightly smaller in size
than immunoglobulin G (i.e., approximate mol. wt. 100 000
to 150 000). Serum that bound myoglobin interfered with
the complement fixation assay for the detection of myo-
globin; serum that did not contain the binding factor had
less such interference. We suggest that myoglobin binding
factor may be responsible for this interference. Interference was not noticed in double-diffusion precipitin
tests.
Additional Keyphrases:
proteins
cord-blood
assay
.
heart disease
infarction
serum
complement fixation
.
.
Myoglobin,
the oxygen-binding
heme protein of
striated muscle, may be released into the circulation
and, after renal clearance, appear in the urine in several
kinds of muscle disease or dysfunction
(1). Recent interest has also focused upon serum myoglobin assay in
the detection and assessment of myocardial infarction
(2, 3). Among the factors to be considered in interpretation of serum assays and analyses of urine is the possibility of protein binding of myoglobin in the circulation. Two earlier studies yielded conflicting data re-
The Phillip D. Wilson Research Foundation,
The Hospital for
Special Surgery affiliated with the New York Hospital-Department
of Medicine, Cornell University Medical College, 535 E. 70th St., New
York, N. Y. 10021.
Received June 17, 1977; accepted July 12, 1977.
garding binding of myoglobin by serum proteins (4,5).
In this report we present evidence for the presence of
binding of myoglobin by most human blood sera, along
with partial characterization
of this weak association
system.
Materials and Methods
Ultrafiltration
of myoglobin serum mixtures. Purified human myoglobin, prepared as described (7), was
added (final concn, 2 mg/liter) to human serum samples
and to 0.15 mol/liter saline, and the mixtures were allowed to incubate at 37 #{176}C
for 1 h. They were subsequently ultrafiltered
through cone-shaped
“Diaflo”
membranes (CF 50A; Amicon Corp., Lexington, Mass.
02173) at 1500 r.p.m. for 30 mm. Myoglobin was assayed
in the ultrafiltrates
so obtained by hemagglutinationinhibition
or complement
fixation
techniques,
or
both.
Ultracentrifugation
of myoglobin serum mixtures.
Myoglobin serum mixtures were subjected to ultracentrifugation
in linear sucrose gradients (50 to 250
g/liter) in a Model L-2 Ultracentrifuge
(Beckman Instruments Inc., Fullerton, Calif. 92634) at 35 000 r.p.m.
for 17 h at 5 #{176}C,
in the SW 41 rotor. The final concentration of myoglobin in these experiments
was 200
mg/liter. Thirty 0.3-ml fractions were collected after
centrifugation.
The volume of applied sample was 1.0
ml. Because some samples of frozen purified human
myoglobin contained aggregated material, these were
pre-centrifuged
in the gradients described,
and the
unaggregated myoglobin was recovered and used for the
preparation of mixtures for assay. Protein concentration
was determined in the fractions by the modified Folin
method (6).
Assays of myoglobin. Myoglobin and specific antiserum were prepared as described (7).
These immunological
reagents were used in three
assays: (a) Double diffusion in gels as described in 7 and
8. (b) Complement
fixation as described
in 2. (c)
Hemagglutination-inhibition
was performed
with
erythrocytes
coated with myoglobin and reagents supplied by Ortho Diagnostics, Inc., Raritan, N. J. 08869.
Reagents so prepared were mixed in microtiter plates
CLINICAL
CHEMISTRY,
Vol. 23, No. 10, 1977
1813
40-
60
#{149}
#{149}
Adult sera
Umbilical-cord
sera
30-
.
-
80
S..
S
S
..
a,
0
40
E
S
S
S
#{149}
S
#{149}
S.
0
20
40
60
% mgbbinding
80
Fig. 1. Binding of myoglobin (mgb) by 82 samples of human
serum
(Titertek; Linbro Scientific Inc., Hamden, Conn. 06517)
as follows: 25 &l of sample for assay (e.g., ultrafiltrate preparations),
75 jl of diluted antibody, and 50
il of a suspension
(2.5 ml/dl) of myoglobin-coated
erythrocytes. Settling patterns of the cells were assessed
after 2-3 h. Because undiluted serum interfered with
the settling patterns, only ultrafiltrates,
which lacked
interfering factors, were assayed by hemagglutination
inhibition. The difference in titer between the amount
of myoglobin present in ultrafiltrates
prepared from
myoglobin/saline
mixtures and that in myoglobin/
serum mixtures was used to calculate the percent
binding in serum.
With the hemagglutination
inhibition test we could
detect 50 g of myoglobin per liter. There was no reaction in the presence of normal serum or commercial
preparations
of human albumin, hemoglobin,
or
globulin in concentrations
ranging from 1 mg/liter to 30
g/liter.
Sources of proteins. Preparations
of albumin (fraction V), y-globulins
(fraction
II), and a-globulins
(fraction IV) were obtained from Pentex Research
Products Div., Kankakee, Ill.
Results
Binding of myoglobin by human blood serum. Myoglobin (2 mg/liter final concn) was added to myoglobin-free sera from 82 individuals (normal volimteers and
patients awaiting elective orthopedic surgery). These
sera were then centrifuged
through CF 50 A Amicon
filter cones, and the resulting ultrafiltrate
was assayed
for myoglobin. The difference in myoglobin content
between that of the filtrates from serum and of those
similarly prepared from 0.15 mol/liter saline was determined, to calculate the percent binding. As shown
(Figure 1), myoglobin was retained in the high-molecular-weight fraction after filtration in all but five sera
(6.1%). In most cases 50-87.5% of the added myoglobin
was so retained. There was no relation between the
subjects’ sex or age and myoglobin binding. Figure 2
includes data from 78 of the adults as well as from umCLINICAL CHEMISTRY, Vol. 23, No. 10, 1977
#{149}
.
S.
I
0
1(X)
.
.
V
20
‘:
#{149}
S
S
S
S
z
#{149}#{149}
S
#{149}
#{149}
S
0
2
1814
‘‘SS.
r
>
>‘
S
#{149}5 #{149}
#{149}#{149}
#{149}
S
#{149}
#{149}
#{149}
I
20
S
#{149}
S
I
I
40
60
Mg binding (%)
I
100
I
80
Fig. 2. Relationship of myoglobin(Mg)-binding properties of
human serum samples to age of donors
bilical cord blood samples from 14 newborns. Although
myoglobin was bound by 93.9% of the sera from adults
and 100% of those from newborns, it was never bound
completely. Free or low-molecular-weight
myoglobin
was detected in the ultrafiltrates
in all cases. Myoglobin
binding in eight serum samples studied did not change
after freezing at -20 #{176}C
and thawing or after storage for
four days at 4 #{176}C.
Variation in the binding of myoglobin
was studied in eight other individuals by repeated assay
of serum samples obtained at different times (Table 1).
In three cases there was no change after intervals of one
to 60 days, in three apparent binding increased, and in
two it diminished. The average change in binding was
14.1 ± 14.3%.
The effect of dilution of serum on binding is shown
in Figure 3. Binding was no longer detectable at dilutions of 50-fold or more. Reduction in binding appeared
with dilutions of fivefold or greater, in both sera tested.
Table 1. Repeated Assays of Myoglobin Binding in
Sera of Eight Individuals (Two Samples from
Each)
Myoglobin
First
Av.
sample
binding,
% a
Second sample
b
days
Change,
50.0
50.0
1
0
50.0
50.0
75.0
87.5
1
3
+25.0
+37.5
50.0
50.0
4
0
75.0
50.0
5
-25.0
93.8
87.5
75.0
87.5
30
60
-18.8
0
87.5
93.8
120
+6.3
68.0
71.0
%
14.1
Percent retention by CF 50A ultrafilter membranes of myoglobin, 2 mg/
liter.
#{176}lnterval
between obtaining the frst and second senan sample from the same
individual.
Table 2. Effect of Addition of Myogiobin (Mgb) to
Ultraf liter Pert usates of Serum before Testing for
Binding
Test sample
a,
%
Whole serum + Mgb
CF 50 perfusate + Mgb
XM 100 perfusate + Mgb
XM 300 perfusate + Mgb
0.15 mol/liter saline + Mgb
C
87.5
bo4
0
50
87.5
0
Percent retention of myoglobin, 2mg/liter, by CF 50A uftrafilter membrane.
Myoglobin was added to each sample. Serum, perfusates of serum, or saline
and the mixtures were then ultraflltered throu
1
2
5
Reciprocal
10
20
serum dilution
CF 50A membranes.
50
Fig. 3. Effect of dilution of serum in 0.15 mol/liter NaCI on
myoglobin(mgb)-binding ability
Table 3. Binding of Myoglobin (Mgb) by Two Sera
Containing No Detectable Haptoglobin
#{149}1ja
100
No. 1, undil.
dil. 5-fold
10-fold
80
No. 2, undil.
dii. 5-fold
10-fold
g60
75
75
50
75
75
50
C
Myogiobin, 2 mg/liter, was added to each of the above samples and the
percent retention by CF 50A ultraf liter membranes was measured.
20
40
60
MgbconcentrationigImI)
Fig. 4. Effect of concentration of myoglobin (Mgb) on its binding
by normal human serum
Binding of myoglobin by serum was related to the
amount of myoglobin present (Figure 4). At concentrations of 1 mg/liter or less, less than 50% of the added
myoglobin was present in the bound fraction. Myoglobin was not demonstrably
bound at 0.5 mg/liter. Binding was maximum between 5 and 15 mg/liter. At higher
concentrations,
binding
diminished
slightly.
Free
myoglobin was also present at all concentrations,
i.e.,
maximum binding never reached 100%.
Properties of the binding fraction of serum. Serum
alone was perfused through CF 50A cones and myoglobin was added to the resulting ultrafiltrates.
These
ultrafiltrates
were then refiltered through the same
cones and, as shown (Table 2), myoglobin was now
completely filterable, indicating the binding factor in
serum was not present in the original CF 50 ultrafiltrate.
In similar fashion, using XM 300 filtration of serum
before addition of myoglobin and CF 50A assay, 87.5%
of myoglobin was present in the bound fraction. With
XM 100 prefiltration
of serum, 50% of the added myoglobin could be bound. This indicated that the binding
factor in human serum was itself not filterable through
the CF 50A ultrafilters,
but was completely filterable
through
XM 300 ultrafilters
and partly filterable
through XM 100 ultrafilters.
Two sera were obtained from patients undergoing
severe hemolytic reactions at a time when neither had
any haptoglobin detectable by precipitin analysis with
specific antiserum. Each of these two sera bound myoglobin well (Table 3).
Partly purified commercial preparations
of human
serum proteins were tested for binding ability. At 30
g/liter, the most binding was found in the y-globulin
fraction II preparation,
while little or no binding was
evident in the albumin, or a-globulin fractions. The
‘y-globulin preparation,
upon analysis by immunoelectrophoresis,
was found to contain both immunoglobulins and i3-globulin components,
but no albumin.
The effect of addition of hemoglobin
to serum is
shown in Table 4. Hemoglobin in concentrations
up to
500 mg/liter interfered very little, if at all, with myoglobin binding.
Ultracentrifugal
pattern of myoglobin in the presence of human blood sera. Figure 5 demonstrates
the
results obtained after subjecting myoglobin to ultracentrifugation
in linear gradients of sucrose-in-saline
solutions (50-150 g/liter). Myoglobin alone under these
conditions was found near the top of the gradient (max.
near 75 g/liter). When myoglobin was mixed with “high
binding serum,” it appeared deeper in the gradient, in
a wider zone of sucrose concentration,
from 80 to 130
CLINICAL CHEMISTRY, Vol. 23, No. 10, 1977
1815
Table 4. Effect of Addition of Hemoglobin to
Myoglobin (Mgb) on Myoglobin Binding by Human
Serum
% 14gb bour4a
0.15 mol/liter saline + Mgb
Serum + Mgb
+ Hemoglobin, mg/liter
Table 5. Double-diffusion Agar Gel Precipitin
(Ouchterlony) Reactions of Myoglobin in Saline
(0.15 mol/liter) or Human Serum with
Antimyoglobin Serum
0
87.5
Qualitative precipitin assay
Myoglobln,
mg/liter
In salIne
In serum a
100
4+
4+
50.0
50
3+
200
87.5
20
2+
3+
3+
50
20
10
5
50.0
87.5
50.0
50.0
10
5
1
0
500
a Percent retention of myoglobln, 2 mg/liter,
1+
0
a Av. results with six
1+
trace
0
human serum samples.
by CF 50A ultrafilter mem-
branes.
g/liter. Serum with no demonstrable
binding by the
ultrafiltration
technique did not bring about this increase in myoglobin density, although a slight increase
in depth of migration to 103 g/liter was seen. Six sera
obtained from patients with myoglobinemia, containing
myoglobin from 10 to 184 mg/liter, had a distribution
of myoglobin in the gradient from 61 to 105 g/liter. The
positions occupied by albumin (100 g/liter) and immunoglobulin G (130 g/liter) are indicated, and suggest
that the density of bound myoglobin was somewhat less
that that of immunoglobulin
G.
Interference
of serum with immunoassays
for myoglobin. Double-diffusion
precipitin analysis in agar gels
was not significantly altered by the presence of serum,
although there was a slight tendency for the precipitin
band to be more prominent
or easily seen in serum
compared to saline diluent (Table 5).
Certain sera, however, did interfere with complement
fixation assay. As shown in Figure 6, assay of 1 mg of
myoglobin per liter, in saline diluent, demonstrated
94%
complement
fixation. Serum from an individual with
no binding demonstrable
after ultrafiltration,
as described before, allowed less complement fixation (58%)
when tested undiluted,
but when diluted two- and
fivefold it allowed complement
fixation to increase to
near expected values. However, serum which demonstrated high binding (87.5%) interfered
to a greater
degree with complement
fixation, even when diluted.
Maximum complement
fixation was 43% at fivefold
dilution. Both sera were subjected to ultrafiltration
through XM-100 ultrafilters.
In both cases perfusates
LOW-BINDING
SERUM HIGH-BINDING
SERUM
Mm
j Myoglobinalone (8)
80
1 Mgb
low-binding
sera(3)
+
Mgb + high-binding
sera(7)
________
XM-100
retentate
C.,
serum
Clinical sera(6)
I
15%
IgG
I
10%
I
I
I
5%sucrose
Alb-5
bottom
top
Fig. 5. Ultracentrifugation of myoglobin added to human serum
samples in linear gradients (in 0.15 mol/liter NaCI)of sucrose
(50-150 g/Iiter)
Each fraction (0.3 ml) obtained after ultracentrifugation was tested for myoglobin
by dooble.dlffusion agar gel precipitin analysis. Ba’s indicate positions of positive
fractions as functions of sucrose concentrations. Myoglobin 200 mg/lIter In all
cases except those of clinical sara. “Low binding” serum caused no demonstrable retention by CF 50A fliters of myoglobin when tested at 2 mg/lIter. “High
binding” serum brought about 87.5% retention under these conditions. Dark
bars indicate position of fractions positive for myoglobin. Numbers indicate
number of determinations, with gray areas showing standard deviation of averages
1816 CLINICALCHEMISTRY,Vol. 23, No. 10, 1977
neat
2
5 neat
2
Reciprocal
dilution
5
Fig. 6. Complement fixation assay of myoglobin (1 mg/liter) in
saline and two human serum samples and their filtered components
In saline, myoglobin produced 94.5% complement fixation (broken line). ‘Low
binding’ serum caused no demonstrable retention of myogiobin by CF 50A ultraf lIter assay. ‘High binding” serum produced 87.5% myoglobin retention.
These sara were uftrafiltered through XM-100 membranes and myoglobin was
then added to serum, the retentate, and the perfusate for testing by complement
fixation assay
so prepared did not interfere with complement fixation
at dilutions of two- and fivefold, although the perfusate
from the low binding serum did display some interference when tested undiluted.
Interference
was found
primarily in the retained fraction and, as shown, accounted for the pattern of interference
of the high
binding serum.
Interference
by whole serum in the hemagglutination-inhibition
assay could not be determined because
whole serum, as well as that retained by XM-100 ultrafilters, produced artifacts of erythrocyte settling in
the hemagglutination-inhibition
assay. Therefore this
technique
was only applied to CF 50A ultrafiltrate
fractions.
Discussion
The apparent molecular size of myoglobin increased
after it was added to most (77 of 82) human sera, so that
it no longer passed through CF 50A membrane ultrafilters. This phenomenon
could conceivably have resuited from myoglobin binding to serum proteins or
other components, from myoglobin aggregation, or from
interference
with the function of the filters by other
serum-related
factors. We suggest that nonspecific
aggregation or membrane artifacts were not responsible
because the percent of binding described by this technique was related to the concentration
of myoglobin,
being lowest at low concentrations of myoglobin, 0.5-1.0
mg/liter and greatest at 5-15 mg/liter. Above this concentration, binding was again decreased. These observations are compatible with a binding system of low
affinity and capacity rather than with nonspecific occlusion of membrane filtration surfaces. In addition,
ultracentrifugation
experiments
also indicated
that
myoglobin is present in serum, in part, in a heavy form,
slightly less dense than immunoglobulin
G (mol. wt.
approx. 150 000). For these reasons it seems likely that
myoglobin was bound by a component of most human
blood sera. This component was completely filterable
through Amicon XM 300 filters, not filterable through
CF 50A filters, and filterable through XM 100 filters.
This suggests, from the manufacturer’s
assessment of
the molecular permeability
ranges of the ultrafilters,
that the binding factor had a molecular weight near
100 000. If this were the case and the binding factor,
perhaps protein in nature, associated with one or two
myoglobin molecules, the observed behavior in the ultracentrifuge
would be expected (e.g., the size of the
complex would be in the 120 000 to 145 000 range).
The binding factor in human serum was not haptoglobin, because two sera devoid of detectable haptoglobin bound myoglobin as well as other sera, and
moreover, addition of hemoglobin to serum did not inhibit its binding of myoglobin. Nor was albumin involved in this myoglobin binding, because partly purified serum fraction II, which contained no detectable
albumin, bound myoglobin to the greatest extent of
those commercial
fractions tested. The myoglobin
binding factor was present in most human sera tested
(93.9%) and could be demonstrated
in all of 14 samples
of umbilical cord blood at birth. Its binding capacity was
low, because binding could be prevented by diluting
serum 50-fold, and weak, because free myoglobin was
found to coexist with the bound form at all concentrations. Myoglobin was most efficiently bound in a concentration range of 5-15 mg/liter, where over 90% was
present in the high-molecular-weight
form. Above this
concentration,
although
more myoglobin
could be
bound, the proportions of free myoglobin increased to
a greater extent.
Previously,
two reports on myoglobin binding in
human serum came to differing conclusions. Javid et al.
(4) failed to find any evidence of myoglobin binding to
serum proteins after electrophoresis
on starch gel and
staining with benzidine or naphthalene
black, using
myoglobin at a concentration
of 1.25 g/liter. Wheby et
al. (5), however, using similar staining techniques
on
both paper and starch gel electrophoresis,
were able to
find evidence of myoglobin binding by a non-haptoglobin globulin. The binding capacity of this protein was
thought to be low and ranged from 0-230 mg/liter in the
sera tested. Horse myoglobin was also thought to be
bound to the /31-globulin fraction of human serum after
immunoelectrophoresis
(11).
Studies in the dog (9) indicate that myoglobin is
bound by an a- or f.-globulin to a maximum of 210
mg/liter and that binding is inhibited by hemoglobin.
Protein-bound
myoglobin was not subject to renal excretion.
The importance of myoglobin binding may be of two
sorts: (a) its effect on renal clearance and (b) its effect
on serum assay methods. Little is known about the renal
clearance of myoglobin in relation to protein binding,
but the weak binding observed would probably not
impair renal excretion. One study determined the renal
clearance of myoglobin to be 25-fold more rapid than
that of hemoglobin (10).
The present studies propose a weak-affinity
myoglobin binding factor in human serum capable of association with approximately
two molecules of myoglobin
and suggest that myoglobin binding by serum may be
the mechanism of its interference with the complement
fixation immunoassay.
This may lead to underestimation of the myoglobin content of sara, particularly at low
concentration
where extensive serum dilution is not
possible (2). Interference
with less sensitive precipitation methods was not demonstrated.
The assistance of Dr. David Bishop, Ortho Diagnostics Inc., Raritan, N. J., in making available reagents for use in the hemagglutination-inhibition
assay is gratefully acknowledged.
This study was supported in part by Grant 1 ROl AM 14650 from
the USPHS and a grant from the Muscular Dystrophy Association,
Inc.
References
1. Kagen, L. J., Myoglobin.
Biochemical,
Physiological
and Clinical
Columbia Univ. Press, New York and London, 1973.
2. Kagen, L. J., Scheidt, S., Roberts, L., et al., Myoglobinemia following acute myocardial infarction. Am. J. Med. 58, 177 (1975).
Aspects.
CLINICAL CHEMISTRY, Vol. 23, No. 10, 1977
1817
3. Stone, M. J., Willerson, J. T., Gomez-Sanchez, C. E., and Waterman, M. R., Radioimmunoassay
of myoglobin in human serum. Results in patients with acute myocardial infarction. J. Clin. Invest. 56,
1334 (1975).
4. Javid, L., Fischer, D. S., and Spaet, T. H., Inability of haptoglobin
to bind myoglobin. Blood 14,683 (1959).
5. Wheby, M. S., Barrett, O, Jr., and Crosby, W. H., Serum binding
of myoglobin, hemoglobin and hematin. Blood 16, 1579 (1960).
6. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.,
Protein measurements with the Folin phenol reagent. J. Biol. Chem.
193, 265 (1951).
7. Kagen, L. J., and Christian,
1818
C. L., Immunological
CLINICAL CHEMISTRY,Vol. 23, No. 10, 1977
measurement
of myoglobm in human adult and fetal skeletal muscle. Am. J. Physiol.
211,656(1966).
8. Kagen, L. J., and Gurevich, R., Precipitin reactions of antihuman
myoglobin serum with several human and animal muscle extracts.
Immunology
12,667 (1967).
9. Lathem, W., The binding of myoglobin by plasma protein. J. Exp.
Med. 111,65(1960).
10. Yuile, C. L., and Clark, W. F., Myohemoglobinuria. A study of the
renal clearance of myohemoglobin
in dogs. J. Exp. Med. 74, 187
(1941).
11. Biserte, G., Havez, R., Haym, A., and Laturaze, J., Isolement du
sbromucoide flu (fluA-globuline). Etude de son pouvoir de liason avec
lea chromoprot#{233}ides.
C. R. Acad. Sci. 250,418 (1960).