HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 28, 20985-20995, July 14, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/28/20985    most recent M001038200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Petersen, C. E. Articles by Bhagavan, N. V. Search for Related Content PubMed PubMed Citation Articles by Petersen, C. E. Articles by Bhagavan, N. V. Social Bookmarking                      What's this?

A Dynamic Model for Bilirubin Binding to Human Serum Albumin^*

Charles E. Petersen, Chung-Eun Ha, Krishna Harohalli, Jimmy B. Feix^§, and Nadhipuram V. Bhagavan^¶

From the  Department of Biochemistry and Biophysics, John A. Burns School of Medicine, University of Hawaii, Honolulu, Hawaii 96822 and the ^§ Biophysics Research Institute, Medical College of Wisconsin, Milwaukee, Wisconsin 53226

Received for publication, February 7, 2000, and in revised form, March 31, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Site-directed mutagenesis of human serum albumin was used to study the role of various amino acid residues in bilirubin binding. A comparison of thermodynamic, proteolytic, and x-ray crystallographic data from previous studies allowed a small number of amino acid residues in subdomain 2A to be selected as targets for substitution. The following recombinant human serum albumin species were synthesized in the yeast species Pichia pastoris: K195M, K199M, F211V, W214L, R218M, R222M, H242V, R257M, and wild type human serum albumin. The affinity of bilirubin was measured by two independent methods and found to be similar for all human serum albumin species. Examination of the absorption and circular dichroism spectra of bilirubin bound to its high affinity site revealed dramatic differences between the conformations of bilirubin bound to the above human serum albumin species. The absorption and circular dichroism spectra of bilirubin bound to the above human serum albumin species in aqueous solutions saturated with chloroform were also examined. The effect of certain amino acid substitutions on the conformation of bound bilirubin was altered by the addition of chloroform. In total, the present study suggests a dynamic, unusually flexible high affinity binding site for bilirubin on human serum albumin.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The binding of bilirubin, a toxic metabolite of heme, to human serum albumin (HSA)¹ has been studied extensively for many years. Early medical interest in the bilirubin-HSA interaction arose when it became clear to physicians that prolonged high blood concentrations of bilirubin, which often occur in premature infants, could result in bilirubin encephalopathy (1-5). In this condition, significant amounts of bilirubin, which is toxic to all tissues, partitions from the blood to neuronal tissue causing irreversible brain damage. The prolonged high blood concentrations of bilirubin in premature infants results from underdevelopment of the liver, the organ responsible for conversion of bilirubin to a soluble form and its excretion into the bile. HSA binds bilirubin (K_d = 10⁷-10⁸ M) at a high affinity site and acts as a buffer preventing the transfer of bilirubin from blood to the tissues, thus playing a critical role in impairing the development of bilirubin encephalopathy. In total, other studies on the pathology of bilirubin encephalopathy in premature infants have highlighted the importance of HSA as a bilirubin transport molecule in normal neonates (who experience a transient hyperbilirubinemia after birth) and in normal adults. These findings provided the motivation for many years of study on the bilirubin-HSA binding mechanism, making it one of the most studied of the HSA-ligand interactions (6-11).

Although some studies have suggested that lower affinity binding components (K_d = 10⁶-10³ M) contribute to HSA-bilirubin binding, most studies have primarily attempted to locate the high affinity binding site and to identify amino acid residues involved in the high affinity binding process. A number of studies that measured the affinity of proteolytic fragments of HSA for bilirubin showed that the high affinity bilirubin binding site was located in subdomain 2A (12-15). Subdomain 2A, which corresponds approximately to amino acid positions 190-300, has recently been shown by x-ray crytallography to be one of two principal sites on HSA for small hydrophobic ligands (16, 17). One of the proteolytic studies described above identified a region containing amino acid positions 186-238 as common to all fragments that retained the ability to bind bilirubin (14). However, another study indicated that the region corresponding to amino acid positions 240-258 was critical for bilirubin binding (15). Based on the above proteolytic studies, the region corresponding to amino acid positions 186-258 was selected as a target for mutagenesis. Selection of specific residues within this region for mutagenesis were based on the following observations.

In thermodynamic work on the bilirubin-HSA interaction, the H° and S° were found to be 13.5 kcal/mol and 0.0085 kcal/mol/degree, respectively (18). The observation that the large negative free energy for the interaction was mainly provided by a large enthalpy change suggested the importance of electrostatic rather than hydrophobic interactions. Because bilirubin contains two carboxyl groups, researchers proposed that key lysine or arginine residues in the binding site must form critical salt linkages with bilirubin carboxyl groups. One study claimed to verify the importance of lysine in the binding process by acetylating all the lysine residues in HSA and showing a dramatic reduction in bilirubin affinity (19). Although a further such study showed a reduction in bilirubin affinity only when buried lysine residues (20) were acetylated, these types of studies cannot identify the specific amino acid residues involved in the binding process and raise serious concerns about changes in the global structure of HSA. In another study, an analog of bilirubin was covalently linked to HSA, which was then digested with several proteases (15). The bilirubin analog was found to be covalently linked to a proteolytic fragment of HSA that contained only one positively charged residue, lysine 240. Lysine 240 was believed to be critical for bilirubin binding to HSA until a natural variant of HSA K240E was discovered (21) and found to have the same affinity as wild type HSA for bilirubin. In the present investigation, positively charged residues between amino acid positions 186 and 260 shown in the x-ray crystal structure of HSA to protrude into the subdomain 2A binding pocket were chosen as targets for mutagenesis. Because of the potential importance of interactions between aromatic amino acid residues and the pyrrole rings of bilirubin, three aromatic residues that protrude into the subdomain 2A binding site were also selected for mutagenesis. The following HSA mutants were synthesized in the yeast species Pichia pastoris: K195M, K199M, F211V, W214L, R218M, R222M, H242V, and R257M. Recombinant wild type HSA was also produced. The affinity of bilirubin for each HSA species was measured using fluorescence spectroscopy (22) and a well validated kinetic assay (23-24).

In the second phase of this study, circular dichroism and absorption spectroscopy were the methods chosen to investigate the conformation of bilirubin bound to the HSA mutants described above. The decision to use the above methods was based on previous work in which these methods were successfully used to elucidate the conformation of bilirubin bound to wild type HSA (25-27). In the present study, the absorption and CD spectra of bilirubin bound to each of the HSA mutants were intercompared, and the differences were used to draw conclusions about the contribution of various amino acid residues to the structure of the wild type HSA-bilirubin complex. The bisignate CD spectrum of bilirubin bound to wild type HSA can be inverted by lowering the pH from 7.4 to 4.8 (28) or by saturating an aqueous solution with chloroform (29-31). In previous work, these changes in the CD spectrum of bound bilirubin have been attributed to changes in the conformation of bound bilirubin resulting from structural changes induced in wild type HSA by either low pH or chloroform. The absorption and CD spectra of bilirubin bound to each of the HSA mutants in the presence of saturating chloroform were intercompared. The expectation of these experiments was that by monitoring how various mutations influence the chloroform-induced inversion of the CD spectrum that it might be possible to gain insights into how specific amino acid residues contribute to the dynamic flexibility of the bilirubin binding site.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Synthesis and Purification of Recombinant HSA

The techniques used in this study to mutate the HSA coding region and to synthesize and purify the mutated protein have been previously described (32-35).

Measurement of Binding by Fluorescence Quenching

Background-- The fluorescence emission spectrum of tryptophan at amino acid position 214, which is the only tryptophan in HSA and is located in the subdomain 2A binding pocket, overlaps significantly with the absorption spectrum of bilirubin. Hence, the net decrease in fluorescence of HSA upon serial additions of bilirubin is directly proportional to the fraction of HSA molecules with a bilirubin molecule bound. Extrapolation of a line through the initial portion of the stochiometric titration, where an insignificant amount of dissociation occurs to a bilirubin/HSA mole ratio of 1, can be used to determine the quenching efficiency, i.e. the fraction of the fluorescence remaining when every high affinity HSA bilirubin binding site is occupied.

Instrumentation and General Experimental Parameters-- For all measurements described in this report, all HSA samples were treated identically and all experiments were performed three times. Fluorescence intensity measurements were made on a QM-1 spectrafluorometer (Photon Technologies Int., South Brunswick, NJ). Samples were excited at 295 nm with a 2.0 nm bandpass, and emission intensity was collected through a monochrometer from 330 to 360 nm. The fluorescence emission intensity was determined to be the integrated area under the emission spectrum from 330 to 360 nm. All samples were suspended in 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na₂HPO₄, 1.4 mM KH₂PO₄, pH 7.4 (PBS). The fluorescence of a buffer blank was subtracted from all measurements. For all titrations, 800 µl of an HSA solution was placed in a dual-path length fluorescence cuvette (10 x 2 mm), with the short path length oriented toward the emission side, maintained at a temperature of 37 °C by a constant temperature circulator. All bilirubin stock solutions were prepared by dissolving bilirubin in 10 mM NaOH. Dilutions of bilirubin stock solutions were prepared by diluting the stock with distilled water. For stochiometric titrations, 10 µM HSA was titrated to a bilirubin/HSA mole ratio of 2.0. For K_d determinations, 0.4 µM HSA was titrated with bilirubin to a bilirubin/HSA mole ratio of 4.0. Because the HSA mutant W214L does not contain tryptophan, the K_d value for this mutant could not be measured by the above method.

Data Analysis-- For fluorescence spectroscopy, data bracketing the titration midpoint (0.4-0.6 bound) was fit to the single component simple binding equation shown below by the least squares method using the computer program Prism (Graphpad).

(Eq. 1)

Measurement of Binding by Rate of Peroxidase Oxidation of Bilirubin

Background-- The use of the peroxidase oxidation method to determine the K_d value for the bilirubin-HSA complex is based on several assumptions that have been shown to be valid for the wild type HSA-bilirubin complex in previous studies (23-24). A known amount of bilirubin is added to a certain amount of HSA so that less than 1% of the total bilirubin added is unbound. An optimal amount of the enzyme, horseradish peroxidase, and hydrogen peroxide are added to the bilirubin/HSA mixture, and the rate of oxidative destruction of bilirubin is determined by monitoring the reduction in absorbance of the mixture at 440 nm. It is assumed that only free bilirubin is available as a substrate for horseradish peroxidase so that there is a relationship between the rate of destruction of bilirubin and its free concentration. Because the concentration of free bilirubin is far below the K_m of the enzyme for bilirubin, the relationship between the free concentration of bilirubin and its oxidation rate is linear. It is assumed that the rate constant for the dissociation of bilirubin from HSA is much greater than the oxidation rate, so that equilibrium between free and bound bilirubin is instantaneously re-established as bilirubin is oxidized, i.e. oxidation of bilirubin by horseradish peroxidase is the rate determining step. From the start of the reaction until 5% or less of the total bilirubin is oxidized, it is assumed as a simplifying approximation that the free bilirubin concentration is constant. Linear regression through this initial rate data is used to calculate the oxidation rate. Experiments with lower concentrations of horseradish peroxidase and bilirubin in the absence of HSA are used to determine a rate constant for the oxidation process. This rate constant can be used to derive free bilirubin concentrations from the oxidation rates measured in the presence of HSA. The K_d values can then be calculated when the free and bound concentrations of bilirubin and the HSA concentration are known.

Instrumentation and General Experimental Parameters-- All absorption measurements in this study were recorded with a Cary 1E spectrophotometer (Varian). Bilirubin was added, to a concentration of 20 µM, to 1 ml of a 40 µM HSA solution in PBS in a 1-cm pathlength cuvette equilibrated at 37 °C by a constant temperature circulator. Hydrogen peroxide was added to a final concentration of 100 µM, and the reaction was initiated by the addition of peroxidase to a final concentration of 10 nM. Reduction in the absorbance of the solution at 440 nm because of the oxidation of bilirubin was monitored for 5 min using kinetic data acquisition software provided with the instrument. Linear regression through this data using the computer program Prism (Graphpad) was used to calculate the oxidation rate. The rate constant for the bilirubin oxidation process was determined as follows. 0.1 nM peroxidase was added to a 1-ml solution of PBS without HSA, a peroxide concentration of 100 µM, and a bilirubin concentration of 1 µM, and the rate of bilirubin destruction was monitored by the reduction in absorbance at 440 nm as described above. Varying the bilirubin concentration from 1 to 10 µM did not significantly affect the value of the rate constant determined by the above method.

Absorption Spectrum of Bilirubin Bound to HSA

The absorption spectrum of bilirubin bound to each HSA species was obtained as follows. To 1 ml of a 40 µM HSA solution suspended in PBS in a 1-cm pathlength cuvette maintained at 37 °C, bilirubin was added to a concentration of 5 µM, and after allowing 10 min for equilibration, the spectrum was recorded from 300 to 550 nm. Under the above conditions, dissociation is insignificant, and therefore the spectrum corresponds approximately to that of HSA-bound bilirubin. The absorption spectrum of the appropriate HSA species in the absence of bilirubin was used as a blank in each case. The absorption spectrum of bilirubin bound to HSA in the presence of chloroform was obtained as follows. 1.2 ml of 40 µM HSA in PBS was mixed with 200 µl of chloroform previously equilibrated with PBS and allowed to stand at room temperature for 30 min. After the two phases were separated, 1 ml of the aqueous layer was removed, bilirubin was added to a concentration of 5 µM, and the spectrum was recorded as described above.

Circular Dichroism Spectrum of Bilirubin Bound to HSA

All experiments were performed at room temperature using a Jasco 700 spectrapolarimeter. 400 µl of a 40 µM HSA solution in PBS, either in the absence of chloroform or saturated with chloroform as described above, was placed in a 0.2-cm pathlength pockels cell and scanned from 200 to 550 nm. Bilirubin was added to a final concentration of 30 µM, and after mixing and allowing 10 min for equilibration, the sample was re-scanned from 200 to 550 nm. The scan for each HSA species in the absence of bilirubin was used as a blank and subtracted from the appropriate measurement in the presence of bilirubin to produce the final spectrum. The addition of bilirubin to PBS in the absence of HSA did not produce a measurable CD spectrum, thus confirming previous observations that free bilirubin in aqueous solutions does not exhibit opitical activity.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Dissociation Constants-- The initial hypothesis of this study, that mutation of certain positively charged and aromatic amino acid residues in subdomain 2A would result in HSA species with dramatically reduced affinity for bilirubin, was shown to be false by two independent methods (Fig. 1, Table I). The K_d values determined by either method varied slightly. This variation is expected as each method is subject to unique systematic errors. The K_d values reported for bilirubin binding to wild type HSA, using either method, vary widely in the literature. Most K_d values have fallen in a range from approximately 10⁷ to 10⁸ M (22-24, 17); for all HSA species examined in the present study, the K_d values were in this range.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Fluorescence binding isotherms. The bilirubin binding curves derived from fluorescence quenching experiments are shown for the mutants indicated. The y axis shows the log of the free bilirubin concentration (µM). The x axis shows the number of ligand molecules bound/HSA molecule. The data points shown for the binding of a particular HSA species with bilirubin represent an average data set obtained from the three values determined for free bilirubin concentration and the number bound for each point along the titration in three separate experiments. A smooth curve through the data is also shown.

Absorption and Circular Dichroism Spectra-- Quite dramatic changes were seen in the absorption and CD spectra of bilirubin bound to some of the HSA mutants (Fig. 2). The CD spectra in the wavelength range from 200 to 300 nm in the presence or absence of bilirubin were identical to those of wild type HSA for all mutants, indicating that the secondary structure of HSA was not altered by any of the mutations or by the addition of bilirubin. In the CD spectrum of HSA-bound bilirubin corresponding to the wavelength range from 300 to 550 nm, striking deviations from the CD spectrum of bilirubin bound to wild type HSA were observed for some mutants. Surprisingly, these spectral changes were most dramatic for two HSA mutants containing substitutions of aromatic residues, F211V and H242V. R222M also exhibited moderate changes in the bilirubin CD spectrum. The three HSA mutants described above also had bilirubin absorption spectra significantly different from that of wild type HSA. K195M, K199M, and R218M had normal bilirubin absorption spectra but exhibited slight changes in their CD spectra. W214L and R257M had bilirubin absorption spectra and CD spectra similar to that of wild type HSA. Saturation of an aqueous wild type HSA/bilirubin mixture with chloroform resulted in an inversion of the bilirubin CD spectrum as reported previously for commercial HSA (Fig. 3). All of the HSA mutants except F211V exhibited chloroform-induced bilirubin CD spectra similar to that of bilirubin bound to wild type HSA in the presence of chloroform (Fig. 3). The absorption spectrum of bilirubin bound to H242V and R222M HSA in the presence of chloroform was also similar to that observed for bilirubin bound to wild type HSA in the presence of chloroform. The absorption spectrum of bilirubin bound to F211V HSA was not significantly altered by the addition of chloroform (Fig. 3).

View larger version (62K): [in this window] [in a new window]   Fig. 2.   Circular dichroism and absorption spectra of bilirubin bound to HSA mutants. Panels A-H correspond to the HSA mutants K195M, K199M, F211V, W214L, R218M, R222M, H242V, and R257M, respectively. Each panel shows an absorption and circular dichroism spectrum for bilirubin bound to wild type HSA and a particular HSA mutant. The bilirubin absorption spectrum, which is at the top of each panel, is aligned with the bilirubin circular dichroism spectrum, shown at the bottom of each panel. For all panels, the absorption spectrum and circular dichroism spectrum for bilirubin bound to wild type HSA are shown as dashed lines. The corresponding spectra of bilirubin bound to a particular HSA mutant are shown as solid lines. The y-axis shows the molar ellipticity (Mol. Ellip.) equal to 3298 (El  Er), where El and Er refer to the extinction coefficient for left and right circularly polarized light, respectively.

View larger version (53K): [in this window] [in a new window]   Fig. 3.   Circular dichroism and absorption spectra of bilirubin bound to HSA mutants in the presence of chloroform. Panels A-I correspond to K195M, K199M, F211V, W214L, R218M, R222M, H242V, R257, and wild type HSA, respectively. Each panel shows an absorption and circular dichroism spectrum for bilirubin bound to a particular HSA species in the presence and absence of chloroform. The bilirubin absorption spectrum, which is at the top of each panel, is aligned with the bilirubin circular dichroism spectrum, shown at the bottom of each panel. For all panels, the absorption and circular dichroism spectra of bilirubin bound to a particular mutant in the absence of chloroform are shown as dashed lines. The corresponding spectra of bilirubin bound to a particular HSA mutant in the presence of chloroform are shown as solid lines. The y-axis shows the molar ellipticity (Mol. Ellip.) equal to 3298 (El  Er), where El and Er refer to the extinction coefficient for left and right circularly polarized light, respectively.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The initial goal of this study, to determine which amino acid residues are critical for high affinity bilirubin binding to HSA, was unsuccessful, because none of the HSA mutants had a significantly altered affinity for bilirubin. There are several potential explanations for this finding. First, positively charged and/or aromatic residues may not be critical in providing the free energy for the HSA-bilirubin interaction. Second, critical, positively charged, or aromatic residues may not have been mutated in this study. One of the inherent limitations of site-directed mutagenesis is that it requires the selection of a reasonably small number of residues as targets for mutagenesis. In this report, information from proteolytic, x-ray crystallographic, and thermodynamic studies were combined to select targets for mutagenesis. It is possible that the solution structure of the subdomain 2A bilirubin binding site is significantly different from that observed in crystals, so that positively charged or aromatic residues in subdomain 2A that appear to be protruding into the binding pocket in the crystal structure are unimportant in the solution structure of subdomain 2A. Third, it could be possible that, despite the evidence from proteolytic studies, that bilirubin does not bind to subdomain 2A. Finally, the large number of amino acid residues in subdomain 2A with the potential to interact favorably with bilirubin, coupled with the unusually high degree of flexibility of HSA, could result in a dynamic situation in which no single amino acid residue is critical in providing the free energy of binding, i.e. low energy adjustments in the structure of the bilirubin-HSA complex allow favorable amino acid contacts to be replaced easily by similar neighboring residues,

CD and absorption spectroscopy were useful in distinguishing among the above possibilities. The prevailing theory on the structure of bilirubin bound to wild type HSA (derived from absorption and CD spectra) and its application to the determination of the structure of mutant HSA-bilirubin complexes follows (Fig. 4). In aqueous solutions, bilirubin exists as an equimolar mixture of two intramolecularly hydrogen-bonded conformational isomers of P- and M-helicity referred to as "ridge tile" structures (36, 37) (Fig. 4). In the ridge tile conformation, the two dipyrrole chromophores of bilirubin are approximately perpendicular to each other. The appearance of a bisignate CD spectrum when bilirubin binds to wild type HSA has been interpreted as HSA-bound bilirubin adopting a ridge tile conformation of P-helicity (26, 27). In two extreme conformations of bilirubin, with  = 180 and 0° (Fig. 4), bilirubin would exist in the extended planar and compact planar conformations, respectively. In these two conformations, the electronic transition dipole moments of both dipyrrole chromophores are in the same plane, and therefore no exciton splitting of the absorption band is expected. For both the compact planar and extended planar conformations, previously described spectroscopic theories (26, 27) would predict a monosignate CD spectrum. These theories (26, 27) would also predict a red shift in the absorption spectrum in changing from the ridge tile to the compact planar conformation (26, 27). In changing from the ridge tile to the extended planar conformation, theory would predict a blue shift in the absorption spectrum (26, 27). When F211V HSA binds to bilirubin, the absorption spectrum of bilirubin red shifts relative to that of bilirubin bound to wild type HSA, and absorption peak splitting is greatly reduced. Additionally, the CD spectrum of bilirubin bound to F211V becomes monosignate (Fig. 2). These findings would be consistent with bilirubin bound to F211V adopting a conformation with a reduced value for  relative to the ridge tile conformation of bilirubin bound to wild type HSA, i.e. the conformation of bilirubin bound to F211V approaches that of the compact planar conformation (Fig. 4). Bilirubin bound to H242V HSA exhibits a similar red shift and a reduction in splitting of the absorption band, although to a lesser degree than observed for F211V HSA. Bilirubin bound to H242V exhibits a CD spectrum intermediate between wild type and F211V HSA. Although the CD spectrum is still bisignate, the intensity of the positive peak is greatly reduced so that there is a dramatic difference in the magnitude of the positive and negative peaks. Bilirubin bound to R222M exhibits changes in its absorption and CD spectrum similar to those of H242V but to a lesser degree. In total, the two aromatic residues phenylalanine and histidine at amino acid positions 211 and 242, respectively, appear to play a major role in maintaining the ridge tile conformation of bilirubin bound to wild type HSA. Mutation of these residues, and to some extent arginine at amino acid position 222, appears to result in structures for HSA-bound bilirubin with a decreased angle between the two dipyrrole chromophores. The magnitude of decrease in the angle between the two chromophores depends on the specific mutation, being greatest for bilirubin bound to F211V. Bilirubin bound to K195M and K199M HSA had an absorption spectrum similar to that of wild type HSA, but the CD spectrum of bilirubin bound to the above HSA mutants was increased in intensity. In the case of K199M, the CD spectrum was red-shifted relative to that of wild type HSA. For bilirubin bound to R218M, the absorption spectrum was normal, but the CD spectrum was decreased in intensity. It is noteworthy that for bilirubin bound to W214L significant changes in the CD and absorption spectra were not seen. This supports the findings of an earlier study in which chemical modification of tryptophan 214 did not affect bilirubin binding (38). Despite the fact that significant differences in bilirubin affinity were not seen for any HSA mutants studied, the observation that a variety of mutations in subdomain 2A can affect the structure of the high affinity bilirubin-HSA complex suggests that the high affinity bilirubin binding site is in fact located in subdomain 2A.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Conformations of bilirubin. Four possible conformations of bilirubin are shown. The two ridge tile enantiomers of M- and P-helicity originally proposed by Lightner (26, 27) are shown. Also shown are the extended and compact planar conformations of bilirubin with angles between the two dipyrrole chromophores of 180 and 00, respectively.

The above results are consistent with earlier studies described below, which suggest that positively charged and aromatic amino acid residues play an important role in determining the conformation of bilirubin bound to wild type HSA. In one study, analogs of bilirubin with either one or both of the carboxyl groups substituted with another functional group, were bound to HSA, and the CD spectra of these analogs were recorded. It was shown that the intense bisignate CD spectrum characteristic of the bilirubin ridge tile (Fig. 4) conformation was maintained by those analogs containing one carboxyl group but not by those in which both carboxyl groups were removed (27). Moving the carboxyl groups to other positions within the bilirubin molecule also affected the CD spectrum. It was concluded from the above study that the interaction of only one of the two carboxyl groups of bilirubin with a single positively charged amino acid residue, such as arginine or lysine, was required to maintain the ridge tile bilirubin conformation observed for wild type HSA. Other studies found that a bilirubin CD spectrum similar to that observed for bilirubin bound to wild type HSA could be induced by forming a complex between bilirubin and various chiral amines (39, 40). The intensity of the bisignate CD signal induced in bilirubin by several chiral amines was found to be greatly increased when a phenyl group was attached to each of the chiral amines. Furthermore, the magnitude of the bisignate CD spectrum observed for bilirubin complexed with various chiral amines with phenyl groups strongly depended on the number of carbon atoms separating the phenyl group from the chiral carbon. Based on these results, it was hypothesized that the ridge tile conformation of bilirubin bound to wild type HSA results from a combination of interactions with specific aromatic and positively charged amino acid residues in the subdomain 2A binding pocket. Our finding that both positively charged and aromatic residues affect the absorption and CD spectra of bilirubin would be consistent with the above hypothesis.

The inversion of the CD spectrum of bilirubin bound to wild type HSA upon saturation of an aqueous bilirubin/wild type HSA mixture with chloroform and other volatile anesthetics has been well studied (29-31). This inversion of the bilirubin CD spectrum has generally been interpreted as the interaction of chloroform with wild type HSA, causing a conformational change in the subdomain 2A bilirubin binding site, which results in a change in the conformation of bound bilirubin from a ridge tile structure with P-helicity to the mirror image structure with M-helicity (Fig. 4) (29-31). An examination of the chloroform-induced absorption and CD spectra of bilirubin bound to the HSA mutants produced in this study provided valuable insights into the structure of the high affinity subdomain 2A binding site. In the absence of chloroform, the absorption and CD spectra of bilirubin bound to H242V and R222M are quite different from those observed for bilirubin bound to wild type HSA. However, in the presence of chloroform, the CD and absorption spectra of bilirubin bound to H242V, R222M, and wild type HSA are similar. This chloroform-induced normalization of the CD and absorption spectra of bilirubin bound to H242V and R222M suggests that histidine and arginine at amino acid positions 242 and 222, respectively are unimportant in determining the structure of the wild type HSA-bilirubin complex in the presence of chloroform. A similar normalization is seen for bilirubin bound to K199M, because the CD spectrum of bound bilirubin is not red-shifted in the presence of chloroform. In contrast, the addition of chloroform does not significantly affect the shape of the absorption or CD spectrum of bilirubin bound to F211V, suggesting that phenylalanine at amino acid position 211 plays an important role in determining the structure of the HSA-bilirubin complex in both the presence and absence of chloroform.

The overall results of this study can be summarized and interpreted as follows. Spectroscopic studies of the mutant HSA-bilirubin complexes revealed that despite the similarity of the K_d values, quite dramatic changes had occurred in the structure of mutant HSA-bilirubin complexes. Studies on the structure of bilirubin-mutant HSA complexes in the presence of chloroform showed that some of the amino acid substitutions that had altered the structure of the bilirubin-HSA complex in the absence of chloroform had no significant effect on the structure of the complex in the presence of chloroform, i.e. the structure of the bilirubin-wild type HSA complex in the presence chloroform results from a different combination of amino acid-bilirubin contacts than the structure of the complex in the absence of chloroform. It seems likely that the extreme flexibility of the high affinity HSA-bilirubin binding site allows for facile lower energy adjustments of the structure of the protein-ligand complex in response to the substitution of amino acid residues involved in binding. Because of the large pool of amino acid residues in subdomain 2A that can form potentially favorable interactions with bilirubin and the flexibility of the binding site, one could speculate that there may be very few single amino acid substitutions of residues in subdomain 2A that will dramatically alter the affinity of the HSA for bilirubin. In the first phase of the present study, we attempted to apply a static model, i.e. classic lock-and-key thinking, to the study of HSA-bilirubin interactions, and we were unsuccessful in identifying key amino acid residues. Although the dynamic nature of proteins complicates the identification of key amino acid residues in all studies of protein-ligand binding, our results suggest that these complications may be especially significant for the binding of bilirubin to HSA. The present study and our previous studies on the binding of warfarin and thyroxine to subdomain 2A, represent the first examples in the literature of the application of site-directed mutagenesis to the study of HSA-ligand interactions. Our results with bilirubin highlight the importance of monitoring conformational changes in the ligand and/or HSA in future mutagenesis studies of ligand binding to HSA.

The major treatment for jaundice in newborns, phototherapy, has sparked a large number of studies focused on the photophysics of bilirubin bound to HSA (41-47). In phototherapy, when infants are illuminated with light, bilirubin primarily bound to its high affinity HSA binding site is converted by light to more soluble nontoxic structural isomers that are easily eliminated from the circulatory system. The rate of formation of isomers and the isomer composition of the reaction products have been shown to be highly dependent on the conformation and electronic environment of bilirubin bound to HSA. Our next phase of experiments will be the attempt to determine whether bilirubin bound to any of the HSA mutants synthesized in this study produces a composition of photoproducts on illumination different from that observed for wild type HSA. One goal of these future experiments would be to identify an HSA mutant that produces soluble bilirubin photoproducts at a greater rate than wild type HSA upon illumination. Administration of such an HSA mutant to jaundiced newborns could potentially improve the efficacy of phototherapy in certain patients, especially those premature infants with low levels of HSA.

	FOOTNOTES

^* This work was supported in part by a grant-in-aid from the American Heart Association, Hawaii affiliate.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed: Dept. of Biochemistry and Biophysics, John A. Burns School of Medicine, University of Hawaii, 1960 East-West Rd., Honolulu, HI 96822. Tel.: 808-956-8130; Fax: 808-956-9498; E-mail: bhagavan@jabsom.biomed.hawaii.edu.

Published, JBC Papers in Press, April 11, 2000, DOI 10.1074/jbc.M001038200

	ABBREVIATIONS

The abbreviations used are: HSA, human serum albumin; PBS, phosphate-buffered saline (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na₂HPO₄, 1.4 mM KH₂PO₄).

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				I. Londono and M. Bendayan Glomerular handling of native albumin in the presence of circulating modified albumins by the normal rat kidney Am J Physiol Renal Physiol, December 1, 2005; 289(6): F1201 - F1209. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 275/28/20985    most recent M001038200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Petersen, C. E. Articles by Bhagavan, N. V. Search for Related Content PubMed PubMed Citation Articles by Petersen, C. E. Articles by Bhagavan, N. V. Social Bookmarking                      What's this?