Opioids play a major role in pain management,1–3 and human serum albumin (HSA), a prominent protein in blood, is an important carrier for opioids and many other medications.4–7 When the concentration of proteins in the plasma changes, especially the concentration of HSA, the free concentration of opioids may also change. This indicates that HSA can bind opioids, making HSA an important carrier for opioids in the blood.8–10 However, the location of the binding sites and the significance of these bindings for opioids remain unclear.
HSA contains 3 helical domains (I–III), each further divided into 2 subdomains (A and B).5 The protein can bind to a large variety of lipophilic/amphipathic compounds and medications including fatty acid (FA), bilirubin, thyroxine, hemin, ibuprofen, aspirin, inhalation anesthetics, propofol, and opioids.4–9,11–13 Many compounds share and may compete at the same binding sites in HSA. This may be reflected in changes in their free concentrations with subsequent pharmacological consequences. General anesthetics and opioids are often administered together during anesthesia. It is unclear whether opioids share binding sites with general anesthetics in protein targets. Moreover, if opioids and general anesthetics share similar binding sites in HSA, it is reasonable that they could also share similar binding sites in the central nervous system.
Morphine (opioid agonist) and naloxone (opioid antagonist) possess very similar chemical structures. However, their pharmacologic effects are totally different. It is unclear whether this dramatic difference is ascribed to different atomic interactions in the same binding site or to interactions in distinct binding sites for each ligand. Because high-resolution structures of HSA and HSA with general anesthetics have been resolved,14 it is possible to reveal how opioids bind to HSA, including binding specificity and selectivity by structural analysis in combination with direct binding studies. Additionally, the results from the opioid binding analysis could be compared with the binding sites of general anesthetics.
In the present study, we characterized opioid–HSA interactions using affinity chromatography, fluorescence spectroscopy, isothermal titration calorimetry (ITC), protein geometrical analysis, and binding prediction to elucidate the following important clinically relevant issues: (a) the location of the binding site(s) for opioids in HSA; (b) whether naloxone (opioid antagonist) shares the binding site with morphine (opioid agonist); and (c) whether opioid agonists share their binding site(s) with general anesthetics.
HSA (essentially FA free), naloxone (chloride), fentanyl (citrate salt), and morphine (sulfate) were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO). Affi-Gel 10 was purchased from Bio-Rad Laboratories (Hercules, CA). All other chemicals were of reagent grade or better, purchased from Sigma-Aldrich Chemical Co. Fluorescence studies were performed with a spectrofluorophotometer (RF5301PC, Shimadzu, Japan). Chromatography studies were performed with Shimadzu pump (Shimadzu LC-600 Liquid Chromatograph pump, Columbia, MD) and detector (UV detector, Shimadzu SPD-6AV, Japan). ITC studies were performed with a VP-ITC (Microcal Inc., Northampton, MA).
Elution chromatography was used as described previously.15 Affi-Gel 10 3.0 mL was washed in cold distilled water and transferred to a graduated cylinder. Excess water was removed, and coupling was accomplished by adding 2.5 mL of the HSA solution (50 mg/mL) and gently mixing. After coupling for 2 hours at 25°C ± 1°C, the gel was allowed to settle, and the supernatant was removed. To block all remaining unreacted groups, we then incubated the gel with 0.3 M glycine, pH 7.0, for an additional 30 minutes. A control gel (without protein) was prepared by reacting 3.0 mL of 0.3 M glycine, pH 7.0, with 3.0 mL of gel for 2 hours, as described above.
The coupled gel was washed with degassed mobile phase (10 mM sodium phosphate, pH 7.0), and packed into Bio-Rad MT-2 columns (holding approximately 2.5 mL of gel). Columns were connected to a Shimadzu LC-600 Liquid Chromatograph pump (Columbia, MD) and then flushed with mobile phase at 0.3 mL/min until a steady baseline as detected by UV absorbance at selected wavelength (210 nm) (Shimadzu SPD-6AV).
After stable baseline is obtained at 0.3 mL/min, 50 μL 0.5 mM propofol, 20 μL 1 mg/mL naloxone, or morphine, or fentanyl in the mobile phase was loaded into the chromatography apparatus at time zero. Absorbance traces at 210 nm were monitored for subsequent 60 minutes and stored. Retention times for these compounds in the protein column were then compared with that of the control (glycine) column under the same conditions. The retention times compared with that of the control column were related with the degree of affinity of the ligands for HSA. Longer retention times indicated stronger affinity for HSA.
Tryptophan (Trp) Intrinsic Fluorescence Study
A Trp intrinsic fluorescence study was performed using the method described previously.12 Briefly, stock solutions of morphine, naloxone, and fentanyl were prepared in 130 mM NaCl and 20 mM sodium phosphate (pH 7) buffer at a concentration of 1 mM for morphine, 1 mM for naloxone, and 1 mM for fentanyl. Different concentrations of opioids were equilibrated with 0.075 mM HSA in a 10-mm path-length quartz cuvette (1.5 mL) with a polytetrafluorethylene stopper during fluorescence measurements. The inhibition of Trp fluorescence intensity was consistent with a binding site in proximity to the Trp residue. The point of 50% inhibition (EC50) of the fluorescence maxima in opioid-free species was calculated with Hill plots. Fluorescence measurements were performed with a Spectrofluorimeter RF-5301PC (Shimadzu, Columbia, MD). All fluorescence measurements were performed at room temperature (25°C ± 1°C). Excitation and emission slit widths were 3 and 5 nm, respectively. The background fluorescence was subtracted from each emission spectrum. The single Trp residue in HSA (Trp214) was excited at a wavelength of 295 nm. Emission spectra were recorded from 310 to 450 nm. Because opioids contain aromatic rings, we corrected the fluorescence intensity with the following equation to reduce the inner filter effect:
Fcorr is the corrected intensity and Fobs is the observed intensity. ODex and ODem are the optical densities at the excitation and emission wavelengths, respectively.
For the competition experiments, naloxone was titrated into the protein sample in the presence of 2 mM morphine or 2 mM fentanyl. If morphine or fentanyl shares the same binding site with naloxone near the Trp residue, the profile of fluorescence inhibition by naloxone should change in the presence of either morphine or fentanyl.
Isothermal Titration Calorimetry
ITC can measure the full thermodynamic profile of bimolecular binding. We and others have demonstrated that data obtained from ITC are in good agreement with data derived from other techniques.13,16 This technique can also be used for competition experiments using ligands with known binding site(s) to locate the binding site for tested compound as described previously.12 On the basis of the known HSA crystallographic binding sites for propofol and halothane under solution conditions,14 propofol and halothane were used as a probe of opioid binding sites using competition experiments. During an ITC experiment, heat release from the interaction between anesthetic and HSA can be measured. If a binding site is already occupied (e.g., by fentanyl, morphine, or naloxone), then there will be less measurable heat release during each aliquot of the titration.
Titrations of the ligands to HSA were performed and heat changes were measured at 20°C. The sample cell contained either 60 μM HSA solution in the presence or absence of morphine (2 mM), naloxone (2 mM), or fentanyl (2 mM), whereas the reference cell contained water. Ligands, 285 μL (stock concentrations of 10 mM for halothane and 0.5 mM for propofol), were titrated. The signals of ligand into buffer, buffer into protein, and buffer into buffer were subtracted after separate titrations. Origin 5.0 (Microcal Software, Inc., Northampton, MA) was used to fit thermodynamic parameters to the heat profiles.
Ligand protein docking is a computational method to predict potential binding sites in a specific protein target using available high-resolution protein structures. Docking calculations were performed using DockingServer (http://www.dockingserver.com).17,18 Semiempirical charges were added to the atoms in the ligands. The charges were calculated using MOPAC 2009 (http://www.cacheresearch.com/mopac.html).
Atomic coordinates of the halothane–HSA complex (1E7C.pdb)14 were obtained from the Protein Data Bank (www.rcsb.org/pdb). The structures of morphine, fentanyl, and naloxone were obtained from the Drug Bank.19 We used AutoDock tools and Autogrid program to add essential hydrogen atoms, solvation parameters, and Kollman united atom type charges and to generate the affinity (grid) maps (48 × 41 × 42 Å) with 0.375 Å spacing.20 The Solis & Wets local search method and Lamarckian genetic algorithm (LGA) are applied during docking simulations with initial random setup of the ligand molecule.21 The detailed method of the docking simulation can be reviewed on the docking server (http://www.dockingserver.com).17
Affinity Chromatography for Global Binding
In comparison with the control column, the prolongation of retention time of naloxone and morphine is similar (3.0 minutes vs. 2.9 minutes). This indicates that both have a similar affinity for HSA. In contrast, fentanyl displays a relatively longer retention time prolongation (3.7 minutes) (Fig. 1), indicating a slightly stronger affinity. Propofol has a much longer retention time prolongation (10.7 minutes) than all the opioids tested, indicating much stronger affinity. The data were repeated without significant change.
Intrinsic Trp Fluorescence for Localized Binding Sites of Opioids
Naloxone inhibits the intrinsic Trp fluorescence of HSA in a dose-dependent manner with EC50 of 2.2 mM as shown in Figure 2, indicating that naloxone may bind to a site close to the single Trp residue (Trp214). Fentanyl does not change the Trp fluorescence signal significantly at the maximum concentration of 2.7 mM, and morphine inhibits Trp fluorescence slightly only at high concentration (2.4 mM).
The inhibition of Trp fluorescence in HSA by naloxone does not change in the presence of either morphine or fentanyl, indicating that the Trp binding site is a distinct binding site for naloxone, and it is not shared with opioid agonists such as morphine or fentanyl (Fig. 2).
Binding Site Overlapping with General Anesthetic Probed by ITC
Consistent with the affinity chromatography experiment, titration at a low concentration of morphine, naloxone, or fentanyl does not yield strong heat profiles for the opioid– HSA interaction, mainly because of their weak interaction (data not shown).
As shown in Figure 3, heat release for propofol–HSA interactions were significantly inhibited by morphine, fentanyl, and naloxone, suggesting that opioids share some sites in HSA with propofol. The heat release is almost completely inhibited by fentanyl, indicating that fentanyl may share the same 2 binding sites for propofol in HSA (domain IIIA and IIIB in reference).14 Morphine and naloxone showed similar profiles, indicating that they may share at least 1 of the propofol binding sites. In the case of the halothane–HSA interaction, the heat release was significantly inhibited by morphine, fentanyl, and naloxone, indicating that opioids share some sites in HSA with halothane. In the case of halothane competition, morphine and naloxone inhibit more heat release than fentanyl, clearly indicating that both morphine and naloxone have binding sites overlapping with halothane. The halothane– HSA interaction heat profile is not changed equally by morphine and naloxone; this is in agreement with the fluorescence experiment showing that both morphine and naloxone have distinct binding sites that are not shared with each other. The results of propofol and halothane titration to HSA are consistent with our previous data.12
As positive controls, automatic docking to HSA was performed for both halothane and propofol. The top ranked sites from docking experiments for halothane and propofol are consistent with crystal data published previously.14Figure 4 shows the top ranked sites for propofol and halothane. Some of the known binding sites for propofol and halothane, and the entire FA binding sites, are displayed for easy comparison.14
Morphine and naloxone share the binding site as indicated in Figure 5. This site is also one of the primary sites for the binding of halothane (HAL7) and FA2. It is located at the interface between the IA and IIA domains. In the case of naloxone, a binding site near the Trp214 was also identified by docking calculation, which is consistent with the Trp fluorescence result. This site is also known to bind FA (FA7), thyroxine 1, warfarin, azapropazone, and halothane (HAL6).7,14Figure 5 depicts this binding site and highlights the proximity of the Trp residue with naloxone. Consistent with the competition experiment in the fluorescence study, this site was not identified using morphine or fentanyl, indicating the selectiveness of the site for naloxone. This result is consistent with the results in the ITC competition experiment presented above.
The top-ranked site for fentanyl is located at the IA–IB domain (FA1/HAL8). One of the halothane and FA binding sites is shown in Figure 6. Because the ITC competition experiment indicates that fentanyl and propofol overlap the same binding site (heat release of propofol titration is almost completely inhibited by fentanyl), other possible binding sites were analyzed. One additional site was identified as a possible fentanyl binding site, and it corresponds to the site at the IIIA domain and is also the binding site of FA, thyroxine 4, diazepam, halothane (HAL3), propofol (PR1), and ibuprofen.7,14
The interaction of general anesthetics with HSA has been well characterized in our laboratory and others.12,14 However, little is known about the interactions between opioids and HSA. Using several direct binding techniques, we demonstrated that the interaction between opioids and HSA is relatively weaker than that between propofol and HSA. Some of the important binding sites for opioids are revealed. Interestingly, morphine and naloxone have similar affinity to HSA, share some sites in HSA, but naloxone has a distinctive binding site near Trp. Morphine and fentanyl share binding sites with general anesthetics (propofol and halothane) in HSA.
HSA has an extraordinary binding capacity for various ligands.4,5,7,11 HSA affinity column has been used to detect the binding with various compounds, including both IV and inhaled general anesthetics.12,22 As indicated by the zonal elution chromatograph, the retention time for naloxone, morphine, and fentanyl is prolonged significantly when compared with control column. This indicates that the tested opioids interact with HSA, which is consistent with the findings reported previously using other techniques.10
The changes in retention time are similar for naloxone and morphine, indicating that morphine and naloxone have similar affinity for HSA. The change in retention time for propofol is almost 3 times longer than that of morphine and naloxone, indicating that propofol has much stronger affinity for HSA. The results of fluorescence experiments and ITC are consistent with the affinity chromatography data, indicating a weak interaction between HSA and naloxone with an EC50 only around 1 mM.
The primary structure of HSA has 585 amino acid residues, and the molecular weight is 66 kD.23,24 HSA has predominantly an α helical secondary structure (approximately 67%),4,23,25 which mimics the base structure for membrane protein receptors. On the basis of this fact, HSA may be able to serve as a reasonable surrogate for the study of opioid–protein receptor interactions. However, as found in this study, the interaction of opioids with HSA is rather weak in comparison with the reported affinity with their native receptors. Thus, the interaction between opioids and HSA may not mimic the interaction between opioids and the opioid receptors well.
Distinct and Shared Sites in HSA for Opioids
On the basis of the structural similarity of morphine and naloxone (Fig. 5), it is predictable that they may share sites in HSA. This is demonstrated by both experimental data and docking calculations. Although the structure of fentanyl shows significant structural differences with that of halothane and propofol, fentanyl shares a site located in domain III of HSA with these two anesthetics. The current hypothesis for the antagonist effect of naloxone is that naloxone replaces opioid agonists (morphine or fentanyl) because of its stronger affinity to the opioid receptors. However, this feature is not observed in HSA. Interestingly, a distinct binding site (near Trp214) for naloxone is shared with halothane but not with morphine despite the structural similarities between morphine and naloxone. HSA contains only 1 Trp residue (Trp214) that can be used to identify and determine the affinity of compounds that bind near the site.24 Naloxone induces inhibition of intrinsic Trp fluorescence that is dependent on the concentration, indicating that naloxone binds to a site close to the Trp residue. This is consistent with the docking predictions of naloxone in HSA (Fig. 5). In contrast to naloxone, morphine inhibits the Trp fluorescence signal slightly only at high concentrations, and fentanyl has no effect. Consistent with these findings, the docking prediction failed to identify this site for either morphine or fentanyl. If naloxone has such selectivity in this protein with multiple binding sites, it is highly possible that naloxone may have distinct binding site(s) among the proteins in the central nervous system. Its antagonist effect might result from binding to distinct sites in receptor proteins rather than merely the displacement of opioid agonists from their sites.
Pharmacological Significance of the Shared Sites with General Anesthetics and with Other Medications
The results of the competition study using ITC and computational predictions (docking) are consistent in indicating that opioid agonists share some binding sites with halothane and propofol in HSA. A site in domain III has been reported as an important site shared by both halothane and propofol in a crystal study and was confirmed with direct binding techniques using engineered domain III in solution conditions.12 This site is also an opioid binding site as demonstrated in this study, indicating that general anesthetics may share binding sites with opioids in the central nervous system, a potential mechanism for minimal alveolar concentration decrease.
The results for binding of opioids to HSA have relevance to practical pharmacotherapy and potential drug–drug interactions. High percentages of opioids (e.g., up to 40% for morphine, 85% for fentanyl) in the blood are bound to plasma proteins. Although it is unclear what percentage of opioids bind to HSA, it has been demonstrated that the free concentrations of opioid agonists (i.e., morphine and sufentanil) change upon variation in the concentration of HSA, indicating the important role of HSA.26,27 However, the free concentration of sufentanil (varies from 19% in neonates to 8% in adults) correlates better with α1-acid glycoprotein plasma concentration than that with HSA, indicating that HSA may play less of a role in affecting the free concentration of this compound.27 The implication of opioids sharing binding sites with other drugs is that the free concentration of each medication may vary in relation to administration of the pure compound owing to competition for the same binding site(s). It has been demonstrated that the free fraction of thiopental significantly increases with increasing fentanyl concentration.28 The variation in free concentrations of opioids might be affected by high concentrations of commonly used general anesthetics (including inhaled agents and propofol), the binding of HSA to other drugs and compounds (e.g., FA and porphyrin compounds), or both.7,11,12 It is worth noting also that some of the halothane binding sites may not be the binding site for other inhalation anesthetics. Future binding studies involving mixtures of multiple drugs and known HSA-binding compounds will be needed to resolve some of these more subtle pharmacologically relevant issues and the effective free concentrations of each molecular species. Plasma protein binding for naloxone is about 50%, but it seems that HSA is less important in changing its free concentration than glycoprotein.29
HSA has a strong buffering capacity. It has been demonstrated that the total amount of FA bound to HSA did not change significantly in the presence of various compounds including propofol and diazepam.30 Opioids have weaker interactions with HSA, and thus the total amount of FA bound to HSA should not change in the presence of opioids.
Cocrystallization and other structural studies of opioids with HSA would confirm and assess the findings in this study. However, it is possible that the opioids may not be seen in the crystal structures because of the weak interaction with HSA.
In conclusion, the interaction between opioids and HSA is relatively weak. Morphine and naloxone have similar affinity to HSA. There are multiple binding sites for opioids in HSA. Fentanyl sites overlap with propofol binding sites. Morphine and naloxone share sites in HSA; however, naloxone has its own distinct site near the sole Trp in HSA, which is not shared with morphine. Opioids morphine and fentanyl share some binding sites with the general anesthetics halothane and propofol in HSA; one of the sites would be in domain IA and another in domain IIIA (both also sites for FA).
Name: Renlong Zhou, MD, PhD.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Renlong Zhou has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Jose Manuel Perez-Aguilar, PhD candidate.
Contribution: This author helped conduct the study, analyze the data, and write the manuscript.
Attestation: Jose Manuel Perez-Aguilar has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Qingcheng Meng, PhD.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Qingcheng Meng has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Jeffery G. Saven, PhD.
Contribution: This author helped analyze the data and write the manuscript.
Attestation: Jeffery G. Saven has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Renyu Liu, MD, PhD.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Renyu Liu has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
This manuscript was handled by: Marcel E. Durieux, MD, PhD.
Dr. Renyu Liu thanks Professor Roderic G. Eckenhoff for his great mentorship and supervision. The authors thank the technical support of Jin Xi, MS, and Jingyuan Ma, candidate for BA in Chemistry at the University of Pennsylvania. Funding support from the Department of Anesthesiology and Critical Care at the University of Pennsylvania and the Foundation for Anesthesia Education and Research is appreciated.
1. Christo PJ, Mazloomdoost D. Cancer pain and analgesia. Recent Adv Clin Oncol 2008;1138:278–98
2. Przewlocki R, Przewlocka B. Opioids in chronic pain. Eur J Pharm 2001;429:79–91
3. Ripamonti C, Dickerson ED. Strategies for the treatment of cancer pain in the new millennium. Drugs 2001;61:955–77
4. Ascenzi P, Bocedi A, Notari S, Fanali G, Fesce R, Fasano M. Allosteric modulation of drug binding to human serum albumin. Mini Rev Med Chem 2006;6:483–9
5. Carter DC, Ho JX. Structure of serum albumin. Adv Protein Chem 1994;45:153–203
6. Zunszain PA, Ghuman J, McDonagh AF, Curry S. Crystallographic analysis of human serum albumin complexed with 4Z,15E-bilirubin-IXalpha. J Mol Biol 2008;381:394–406
7. Ghuman J, Zunszain PA, Petitpas I, Bhattacharya AA, Otagiri M, Curry S. Structural basis of the drug-binding specificity of human serum albumin. J Mol Biol 2005;353:38–52
8. Chauvin M, Lebrault C, Levron JC, Duvaldestin P. Pharmacokinetics of alfentanil in chronic renal failure. Anesth Analg 1987;66:53–6
9. Wilson AS, Stiller RL, Davis PJ, Fedel G, Chakravorti S, Israel BA, McGowan FX Jr. Fentanyl and alfentanil plasma protein binding in preterm and term neonates. Anesth Analg 1997;84:315–8
10. Wood M. Plasma drug binding: implications for anesthesiologists. Anesth Analg 1986;65:786–804
11. Zunszain PA, Ghuman J, Komatsu T, Tsuchida E, Curry S. Crystal structural analysis of human serum albumin complexed with hemin and fatty acid. BMC Struct Biol 2003;3:6
12. Liu RY, Meng QC, Xi J, Yang JS, Ha CE, Bhagavan NV, Eckenhoff RG. Comparative binding character of two general anaesthetics for sites on human serum albumin. Biochem J 2004;380:147–52
13. Liu R, Eckenhoff RG. Weak polar interactions confer albumin binding site selectivity for haloether anesthetics. Anesthesiology 2005;102:799–805
14. Bhattacharya AA, Curry S, Franks NP. Binding of the general anesthetics propofol and halothane to human serum albumin—High resolution crystal structures. J Biol Chem 2000;275:38731–8
15. Chan K, Meng QC, Johansson JS, Eckenhoff RG. Low-affinity analytical chromatography for measuring inhaled anesthetic binding to isolated proteins. Anal Biochem 2002;301:308–13
16. Zhang T, Johansson JS. An isothermal titration calorimetry study on the binding of four volatile general anesthetics to the hydrophobic core of a four-alpha-helix bundle protein. Biophys J 2003;85:3279–85
17. Hazai E, Kovacs S, Demko L, Bikadi Z. DockingServer: molecular docking calculations online. Acta Pharm Hung 2009;79:17–21
18. Mayo SL, Olafson BD, Goddard WA. Dreiding—A generic force-filed for molecular simulations. J Physical Chem 1990;94:8897–909
19. Wishart DS, Knox C, Guo AC, Cheng D, Shrivastava S, Tzur D, Gautam B, Hassanali M. DrugBank: a knowledge base for drugs, drug actions and drug targets. Nucleic Acids Res 2008;36:D901–6
20. Morris GM, Goodsell DS, Halliday RS, Huey R, Hart WE, Belew RK, Olson AJ. Automated docking using a Lamarckian genetic algorithm and an empirical binding free energy function. J Comput Chem 1998;19:1639–62
21. Solis FJ, Wets RJB. Minimization by random search techniques. Math Oper Res 1981;6:19–30
22. Liu R, Pidikiti R, Ha CE, Petersen CE, Bhagavan NV, Eckenhoff RG. The role of electrostatic interactions in human serum albumin binding and stabilization by halothane. J Biol Chem 2002;277:36373–9
23. Dockal M, Carter DC, Ruker F. The three recombinant domains of human serum albumin. Structural characterization and ligand binding properties. J Biol Chem 1999;274:29303–10
24. He XM, Carter DC. Atomic structure and chemistry of human serum albumin. Nature 1992;358:209–15
25. Peters T Jr. Serum albumin. Adv Protein Chem 1985;37:161–245
26. Mashayekhi SO, Hain RD, Buss DC, Routledge PA. Morphine in children with cancer: impact of age, chemotherapy and other factors on protein binding. J Pain Palliat Care Pharmacother 2007;21:5–12
27. Meistelman C, Benhamou D, Barre J, Levron JC, Mahe V, Mazoit X, Ecoffey C. Effects of age on plasma protein binding of sufentanil. Anesthesiology 1990;72:470–3
28. Russo H, Audran M, Bressolle F, Brès J, Maillols H. Displacement of thiopental from human serum albumin by associated drugs. J Pharm Sci 1993;82:493–7
29. Asali LA, Brown KF. Naloxone protein binding in adult and foetal plasma. Eur J Clin Pharmacol 1984;27:459–63
30. Simard JR, Zunszain PA, Hamilton JA, Curry SJ. Location of high and low affinity fatty acid binding sites on human serum albumin revealed by NMR drug-competition analysis. Mol Biol 2006;361:336–51