| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Department of Biochemistry and Biophysics, University of Hawaii, 1960 East-West Rd., Honolulu, HI 96822.
2
Biophysics Research Institute, Medical College of
Wisconsin, 8701 Watertown Plank Rd., P.O. Box 26509, Milwaukee, WI
53226.
3
Department of Medicine, Institute of Clinical
Endocrinology, Tokyo Women's Medical College, 8-1 Kawada-cho,
Shinjuku-ku, Tokyo 162, Japan.
a Author for correspondence. Fax 808-956-9498; e-mail bhagavan{at}jabsom.biomed.hawaii.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods and Results: Recombinant R218H, R218P, and wild-type HSA were synthesized in yeast, and the affinities of these HSA species for l- and d-thyroxine were determined using fluorescence spectroscopy. The dissociation constants for the binding of wild-type, R218P, and R218H HSA to l-thyroxine were 1.44 x 10-6, 2.64 x 10-7, and 2.49 x 10-7 mol/L, respectively. The circular dichroism spectra of thyroxine bound to R218H and R218P HSA were markedly different, indicating that the structure of the thyroxine/HSA complex is different for either protein.
Conclusions: The Kd values for l-thyroxine bound to R218P and R218H HSA determined in this study were similar. The extremely high serum total-thyroxine concentrations reported previously for R218P FDH patients (10-fold higher than those reported for R218H FDH patients) are not consistent with the Kd values determined in this study. Possible explanations for these discrepancies are discussed.© 1999 American Association for Clinical Chemistry
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Because of our initial finding that the amino acid substitution R218H produces FDH, we thought that all cases of FDH resulted from this specific genotype. Recently, a new genotype that also causes FDH was identified in several Japanese patients (13). This newly identified genotype results from the substitution of cytidine for guanine at nucleotide 653, which gives rise to a substitution of proline for arginine at amino acid position 218 (R218P). We have identified two new Japanese FDH patients unrelated to those described previously who also have the R218P genotype.
Because both the R218H and R218P genotypes can be responsible for FDH, it becomes important to determine whether there are any clinically significant differences between the FDH HSA species resulting from these two distinct genotypes. Our new R218P cases findings were identified by analyzing archived DNA samples supplied by O. Isozaki. Unfortunately, the natural R218P HSA from these patients was not available for this study. R218H HSA from patients identified previously was also unavailable. To obtain R218P and R218H HSA for this study, these proteins were expressed in the yeast species Pichia pastoris. Recombinant wild-type HSA was also produced as a control. Because all FDH patients that have been identified are heterozygous, their serum contains a mixture of wild-type and either R218H or R218P HSA. A method has not been developed to separate wild-type from FDH HSA; therefore, all previous studies on natural FDH HSA had used a mixture of the two proteins. One advantage of expressing the recombinant proteins is that it allowed us to use homogeneous preparations of R218H and R218P HSA in these studies.
The affinity of l- and d-thyroxine for R218H and R218P HSA was measured using fluorescence spectroscopy. Circular dichroism (CD) was used to examine the structure of thyroxine bound to R218P and R218H HSA.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
synthesis and purification of recombinant hsa
Introduction of mutations into the HSA-coding region.
Specific
mutations were introduced into the HSA-coding region in a plasmid
vector containing the entire HSA-coding region (pHiL-D2 HSA), using
standard techniques as described previously (14).
Expression of recombinant HSA.
Each pHiL-D2 HSA expression
cassette coding for a particular HSA mutant was introduced into the
yeast species P. pastoris by electroporation. A yeast clone
that contained the expression cassette stably integrated into the
chromosomal DNA was isolated in each case.
Verification of the DNA sequence of HSA clones.
The DNA
sequence of the HSA-coding region was verified as described previously
(14).
Purification of recombinant HSA.
The secreted HSA was isolated
from growth medium as follows. The medium was brought to 50%
saturation with ammonium sulfate at room temperature. The temperature
was then lowered to 4 °C, and the pH was adjusted to 4.4, the
isoelectric point of HSA in a solution 50% saturated with ammonium
sulfate. The precipitated protein was collected by centrifugation and
resuspended in distilled water. Dialysis was carried out for 72 h
against 100 volumes of phosphate-buffered saline (PBS; 137 mmol/L NaCl,
2.7 mmol/L KCl, 4.3 mmol/L
Na2HPO4, 1.4 mmol/L
KH2PO4, pH 7.4) with one
change of buffer. The solution was loaded onto a column of cibacron
blue immobilized on Sepharose 6B (Sigma). After the column was washed
with 10 bed volumes of PBS, HSA was eluted with 3 mol/L NaCl. The
eluent was dialyzed against PBS and passed over a column of
Lipidx-1000 (Packard Instruments) to remove hydrophobic ligands
possibly bound to the HSA (15). The resulting protein
exhibited only one band on sodium dodecyl sulfate-polyacrylamide gel
electrophoresis.
thyroxine binding studies
Background.
As shown previously, the fluorescence emission
spectrum of HSA overlaps substantially with the absorption
spectrum of thyroxine (16). It has been shown that the
quenching of the fluorescent emission of tryptophan 214 (located in the
subdomain 2A binding site) on serial additions of thyroxine is
primarily a result of the binding of thyroxine to this site
(16). To determine the dissociation constant for the
thyroxine/HSA equilibrium, two separate experiments are required. A
high concentration of HSA is first titrated with ligand to approximate,
as near as possible, stoichiometric binding. In this case, a plot of
fluorescence vs the ligand/HSA molar ratio shows an initial monotonic
decrease in fluorescence, which then plateaus at a minimum value that
reflects the fraction of fluorescence not quenched by bound ligand. A
lower concentration of HSA is then titrated with ligand, and the
fraction of ligand-bound HSA can be calculated on the basis of the
quenching efficiency determined from the stoichiometric titration
described above. The free concentration of thyroxine can then be
calculated on the basis of the number of ligand-bound HSA molecules and
the amount of thyroxine added to the cuvette at each point along the
titration.
Experimental conditions.
Fluorescence intensity
measurements were made on a QM-1 spectrafluorometer (Photon
Technologies). Samples were excited at 295 nm with a 2.0 nm
bandpass, and the emission intensity was collected through a
monochromator from 330 to 360 nm. The fluorescence emission intensity
was the integrated area under the emission spectrum from 330 to 360 nm.
All samples were suspended in PBS. The fluorescence of a buffer blank
was subtracted from all measurements. For all titrations, 800 µL of a
HSA solution was placed in a dual-pathlength fluorescence cuvette
(10 x 2 mm) with the short pathlength oriented toward the
emission side, and the temperature of the cuvette was maintained at
37 °C by a constant temperature circulator. All thyroxine stock
solutions were prepared by dissolving the ligand at a concentration of
1 mmol/L in 10 mmol/L sodium hydroxide. Dilutions of the stock were
prepared by diluting the stock with distilled water. For stoichiometric
titrations 10 µmol/L HSA was titrated to a thyroxine/HSA molar ratio
of 4. For Kd determinations,
0.4 µmol/L HSA was titrated with thyroxine.
Analysis of data.
The data for each
Kd determination were fit to
the equation shown below by nonlinear regression (least-squares
method), using the computer program Prism (Graphpad):
 |
The variable H is the Hill coefficient and is a measure of the degree to which the relationship between the number of molecules bound and the log of the free ligand concentration deviates from simple binding. For simple binding with no positive or negative cooperativity, the Hill coefficient is 1. In this case, the Hill coefficient is a measure of the degree to which the curve that best fits the data deviates from an ideal shape.
cd measurements on bound thyroxine
Background.
It had been shown that although pure solutions of
either D- or L-thyroxine exhibit no CD at
255–400 nm, each of these enantiomers exhibits a distinct induced CD
spectrum in the presence of HSA (17)(18).
Previous work with commercial HSA has attributed this induced CD
spectrum to an asymmetric orientation of the outer and inner rings of
thyroxine when thyroxine is bound to its high affinity binding site on
HSA. In this study, we compared the induced CD spectra of the
D and L enantiomers of
thyroxine when bound to wild-type, R218H, and R218P HSA to determine
whether the spectra differed significantly among these HSA species.
Experimental conditions.
Measurements were made on a J-700
spectrapolarimeter (Jasco) at room temperature. All samples were
suspended in PBS. A pockel cell with a pathlength of 1 mm
containing 330 µL of sample was scanned from 190 to 255 nm and from
255 to 400 nm at a scan speed of 0.5 nm/s at sensitivities of 1000 and
2 millidegrees, respectively. Each wavelength range was scanned
three times, and the scans were averaged automatically by the
instrument software.
For all HSA species, a 40 µmol/L solution of HSA in PBS was scanned before the addition of thyroxine. The sample was titrated with either L- or D-thyroxine to the following thyroxine/HSA molar ratios, 0.25, 0.50, 0.75, 1.00, 1.50, 2.00, and 2.50. After each addition, the sample was mixed and scanned. The expectation of the experiment was that the CD spectral changes resulting from the binding of thyroxine to its high affinity binding site would reach a plateau at a molar ratio near 1.00 and that the spectrum would change only very slightly as the molar ratio was increased from 1.00 to 2.50 because binding to the high affinity site should be nearly saturated in the region of the titration in which the molar ratio is adjusted from 1.00 to 2.50.
The induced CD spectrum for each enantiomer of thyroxine was calculated as the difference spectrum equal to the CD spectrum of a thyroxine/HSA solution at a molar ratio of 2 (corrected for dilution) minus the CD spectrum of HSA without thyroxine. Thyroxine in the absence of HSA does not exhibit a measurable CD spectrum in the wavelength range examined. The CD difference spectra were smoothed using a noise reduction feature of the Standard Analysis program (Jasco), which is included in the software package provided with the instrument.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
R218H and R218P HSA have similar and significantly higher affinities
for L-thyroxine than does wild-type HSA. The
Kd values for
L-thyroxine binding to R218P and R218H were 2.64 x
10-7 and 2.93 x
10-7 mol/L, respectively (Table 1
and Fig. 1
). The Kd values
determined for D-thyroxine binding to R218P and R218H HSA
were 2.49 x 10-7 and 4.12 x
10-7 mol/L, respectively, similar to the values
for L-thyroxine binding. For binding to wild-type HSA, the
values for L- and D-thyroxine (1.44 x
10-6 and 1.28 x
10-6 mol/L, respectively) were nearly identical.
|
|
Other researchers have shown that although L-thyroxine
exhibits no measurable CD spectrum from 255 to 400 nm, when
L-thyroxine binds to wild-type HSA (commercial), negative
peaks in an induced CD spectrum occur at positions near peaks observed
in the thyroxine absorption spectrum. These peaks in the absorption
spectrum occur at 292 and 325 nm (18)(19). The
peaks in the CD spectrum near 325 and 292 nm may be attributable to the
outer and the inner rings of thyroxine, respectively
(17)(18). For D-thyroxine,
previous researchers found that the magnitude of the peak in the CD
spectrum associated with aromatic ring A was reduced in magnitude but
was still negative, whereas the peak associated with aromatic ring B
was reduced in magnitude and became positive, indicating an altered
orientation of aromatic ring B for D-thyroxine
bound to wild-type HSA (commercial). We found spectra similar to those
described above for D- and
L-thyroxine bound to commercial and recombinant
wild-type HSA (Fig. 2
). At 190–255 nm, the CD spectrum arises from the secondary
structure of the protein. In this region, thyroxine binding did not
cause any measurable changes in the CD spectra for all HSA
species.
|
For all HSA samples at a concentration of 40 µmol/L, the magnitude of the induced CD spectra at 255–400 nm changed most dramatically as the concentration of thyroxine was increased from a thyroxine/HSA molar ratio of 0.00 to 1.00. For all HSA species, the magnitude of the induced CD spectra increased only slightly as the thyroxine/HSA molar ratio was increased from 1.00 to 1.50. When the ratio was increased from 1.5 to 2.5, there was no detectable change. There were no detectable changes in the shapes of the spectra for any HSA species as the molar ratio was increased from 1.00 to 2.5. The changes in the CD spectra as a function of the thyroxine/HSA molar ratio described above suggest that the induced CD spectra for the range of thyroxine concentrations studied arise mainly from thyroxine binding to its high affinity binding site on HSA. Based on the above observations, we determined that the induced CD spectrum of thyroxine bound to its high affinity site was the difference spectrum equal to the spectrum of a thyroxine/HSA mixture at a molar ratio of 2.00 minus the spectrum of HSA alone. The approximate molar ellipticity of bound thyroxine was calculated by assuming that under these conditions one-half of the thyroxine added was bound to HSA.
A comparison of the induced CD spectra resulting from the binding of
L-thyroxine to the various HSA species in this study shows
several important trends. For wild-type, R218H, and R218P HSA bound to
L-thyroxine, the peak CD signal near 325 nm was negative,
with peak molar ellipticities of -12 340, -28 520, and -16 800,
respectively (Fig. 2
). The peak in the CD spectrum near 292 nm was
negative for L-thyroxine bound to wild-type and R218H HSA,
with peak molar ellipticities of -6260 and -12 640, respectively.
For L-thyroxine bound to R218P HSA, the peak in the CD
spectrum near 292 nm was positive and substantially blue shifted, with
a peak molar ellipticity of 12 580. A comparison of the induced CD
spectra resulting from the binding of D-thyroxine to R218H
HSA and R218P HSA showed them to be similar to those obtained for
L-thyroxine (Fig. 2
).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). The
Kd value determined in this
study for L-thyroxine binding to R218P was 2.64 x
10-7 mol/L, which is similar to the
Kd of 1.11 x
10-7 mol/L determined in a previous study using
the natural protein (13). The
Kd values determined for R218H
binding to L-thyroxine in this study and two
previous studies were 2.93 x 10-7,
2.22 x 10-7 (9), and 1.70
x 10-7 mol/L (19), respectively.
Despite these similarities, a case was made in the only study other
than this one in which the binding of R218P to
L-thyroxine was measured (13) that
R218P HSA has a much greater affinity for
L-thyroxine than does R218H HSA. The argument in
that study was as follows. The
Kd values for the binding of
L-thyroxine to wild-type and R218P HSA were
9.09 x 10-6 and 1.11 x
10-7 mol/L (13), respectively, an
80-fold difference (Table 1
). The magnitude of this difference between
the Kd values was compared with
a previous study in which Kd
values of 2.22 x 10-6 and 2.22 x
10-7 mol/L were reported for wild-type and R218H
HSA, respectively (9). The 10-fold difference between these
previously reported values was then compared to the 80-fold difference
described above. The conclusion of authors of the study was that R218P
HSA had an affinity eightfold greater than R218H HSA for
L-thyroxine, although the absolute
Kd values of R218P and R218H
HSA were similar. The analysis described above is seriously flawed. The
major flaw is that the conclusion critically depends on the
Kd value determined for the
binding of L-thyroxine to wild-type HSA. This
Kd value for wild-type HSA
determined for comparison to R218P HSA in the study described above was
much greater than that determined in our study and many other studies
of R218H HSA (Table 1
). The binding of
L-thyroxine to wild-type HSA has been well
studied by a variety of methods, with values ranging from 6.25 x
10-7 to 3.64 x 10-6
mol/L (20)(21)(22)(23)(24). The large value for the wild-type
Kd used in the comparative
analysis described above, 9.09 x 10-6
mol/L, would magnify the difference between the affinities of R218P and
wild-type HSA, creating a systematic error in the comparative analysis.
If we assume that the Kd values
for L-thyroxine binding to purified R218H and R218P HSA are
similar, as this study indicates, we are left with a major paradox. In
two previous studies of FDH patients with the R218H genotype, the mean
serum total-thyroxine concentrations were 151.6 (11) and 240
(10), and the reference values were 92.8 and 140 nmol/L,
respectively (Table 1
). In the study of six FDH patients with the R218P
genotype described above, the mean serum total-thyroxine concentration
was 2293.7 nmol/L (13) compared with 111.3 nmol/L for
healthy controls. On the basis of these serum total-thyroxine
concentrations, we would expect the affinity of R218P HSA for
L-thyroxine to be increased at least 20-fold over
that of R218H HSA.
There are four plausible explanations for the paradox created by the
findings of this study. Natural R218P HSA could be substantially
different from recombinant R218P HSA such that the binding values
presented in this report for recombinant R218P HSA do not reflect the
true values that would be obtained with the natural protein. This seems
unlikely because the Kd values
obtained in this study and previous studies with recombinant R218H HSA
are similar to those obtained for natural R218H HSA. Second, the
binding methodology used in this study to determine the
Kd value for
L-thyroxine binding to R218P could be inaccurate. The
technique we used in this study to measure the binding of
L-thyroxine to R218P HSA is a well-validated
(19)(20)(25) and generally accepted
technique; therefore, it is hard to imagine a scenario in which this
technique could yield a Kd
value that is inaccurate by a factor of 10. Third, the presence of
R218P HSA in FDH patient serum samples could interfere with the method
used in the previous study to measure total thyroxine in such a way as
to give spuriously high values. It is noteworthy in this regard that
the initial discovery of the FDH phenotype was because FDH HSA
interfered with a standard clinical assay of free thyroxine in patient
serum samples, producing spuriously high values for free thyroxine
(1)(2)(3)(4)(5)(6)(7). Finally, there may be some serum component that
greatly increases the affinity of R218P HSA, but not R218H HSA, for
L-thyroxine. Comparison of the CD spectra of
L-thyroxine bound to R218P and R218H HSA revealed
quite dramatic differences in the structure of the protein/thyroxine
complex for either protein. The portion of the CD spectrum shown in
previous studies of wild-type HSA to be attributable to the inner
aromatic ring of thyroxine (17)(18)(Fig. 2
) was
the region that was primarily different when the spectra of
L-thyroxine/R218H HSA and
L-thyroxine/R218P HSA were compared. Our previous
mutagenesis/thyroxine analog studies on R218H HSA indicated that amino
acid position 218 is next to the inner aromatic ring of thyroxine when
thyroxine is bound to wild-type and R218H HSA (25). The
finding that the structure of the thyroxine/HSA complex for either
R218P or R218H HSA may be different suggests that these two
protein/ligand complexes may respond differently to allosteric
effectors.
Free-thyroxine concentrations are within the reference interval in R218P FDH patients; thus, they are clinically euthyroid (13). However, release of a substantial portion of the huge reservoir of HSA-bound thyroxine proposed for R218P patients, [estimated to be ~2200 nmol/L) (13)] could cause a sudden and severe thyrotoxicosis. The concentration of free thyroxine-binding globulin sites (~300 nmol/L) could serve as a buffer against a small release of thyroxine, but these sites could easily be saturated by a large release of the huge R218P HSA-bound reservoir. If an as yet unidentified allosteric effector for R218P HSA exists, fluctuations in the concentration of this effector could profoundly influence the thyroid status of R218P FDH patients. In addition, if the previously reported total-thyroxine concentrations for R218P FDH patients are correct, these patients may be especially susceptible to a drug-induced thyrotoxicosis caused by administration of many therapeutics, such as warfarin, aspirin, and furosemide, that bind to subdomain 2A of HSA, the location of the thyroxine binding site. A compendium of subdomain 2A ligands and an explanation of HSA binding sites the reader is presented in Ref. (26).
Our preliminary study of R218P HSA leaves many unanswered questions that will need to be resolved by future studies. Preferably, future studies would be conducted using natural R218P HSA obtained from FDH patients and the serum total-thyroxine concentrations would be measured by a variety of independent methods to eliminate the possibility of erroneous readings.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  I. Petitpas, C. E. Petersen, C.-E. Ha, A. A. Bhattacharya, P. A. Zunszain, J. Ghuman, N. V. Bhagavan, and S. Curry Structural basis of albumin-thyroxine interactions and familial dysalbuminemic hyperthyroxinemia PNAS, May 27, 2003; 100(11): 6440 - 6445. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Pannain, M. Feldman, U. Eiholzer, R. E. Weiss, N. H. Scherberg, and S. Refetoff Familial Dysalbuminemic Hyperthyroxinemia in a Swiss Family Caused by a Mutant Albumin (R218P) Shows an Apparent Discrepancy between Serum Concentration and Affinity for Thyroxine J. Clin. Endocrinol. Metab., August 1, 2000; 85(8): 2786 - 2792. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, R218H; 
, R218P.