Haematoporphyrin and OO''-diacetylhaematoporphyrin binding by serum and cellular proteins. Implications for the clearance of these photochemotherapeutic agents by cells.

Haematoporphyrin derivative (HpD), a mixture of porphyrins, is currently used as a photochemotherapeutic agent in the treatment of neoplasias. The interaction of purified components of HpD with serum and cellular proteins was investigated using absorption and fluorescence spectroscopy. The interactions of haematoporphyrin and OO'-diacetylhaematoporphyrin with human albumin and with haemopexin, the two major serum porphyrin-binding proteins, show stoichiometries of 1 mol of porphyrin bound per mol of protein. The apparent dissociation constants, Kd, are in the range of 1-2 microM for albumin and 3-4 microM for haemopexin. These two major components of HpD would, after intravenous injection, bind to albumin and circulate in serum as albumin complexes. Free porphyrin rather than porphyrin bound to albumin interacts with Morris hepatoma tissue culture cells. A rapid high-affinity saturable transport system operates at free porphyrin concentrations of less than 2 microM. In addition, fluorescence spectra show that components in rat liver cytosol can bind haematoporphyrin and OO'-diacetylhaematoporphyrin and distinguish these binders from those present in rat serum.     

Biochemical properties of the heme oxygenase inhibitor, Sn-protoporphyrin. Interactions with apomyoglobin and human serum albumin

Sn-protoporphyrin is a strong competitive inhibitor of heme oxygenase and a potential pharmacological agent for the treatment of neonatal hyperbilirubinemia. Little is otherwise known about the biochemistry of tin porphyrins. We have investigated aspects of the chemistry of tin-protoporphyrin in aqueous solution and of its interactions with heme-binding proteins other than heme oxygenase, specifically apomyoglobin and human serum albumin. In the pH region 7-10, Soret region absorption studies of unbound Sn-protoporphyrin demonstrate a pH-dependent monomer-dimer equilibrium (KD congruent to 10(6) M-1 at pH 7) with little higher aggregation. Dissociation of the dimer is relatively slow at neutral pH, permitting interaction of protein ligands with monomeric and dimeric species to be distinguished and providing insights into kinetic mechanisms of porphyrin binding by heme-binding proteins. In the present study, the kinetics of interaction of Sn-protoporphyrin with apomyoglobin are presented as novel evidence that this binding proceeds by an induced fit mechanism. Binding of Sn-protoporphyrin to both apomyoglobin and serum albumin is unexpectedly weak. Between pH 7 and 9, the apparent affinity of Sn-protoporphyrin for apomyoglobin is less than 1/200 that of heme and, at pH 9, is also significantly less than that of protoporphyrin. The apparent affinity of Sn-protoporphyrin for human serum albumin is less than 1/1000 that of heme and 1/30 to 1/100 that of protoporphyrin. Competition studies between heme and Sn-protoporphyrin and between bilirubin and Sn-protoporphyrin indicate that Sn-protoporphyrin distributes differently among porphyrin-binding sites on serum albumin than does heme and that it is also not an effective competitor with bilirubin for bilirubin-binding sites. These results argue that Sn-protoporphyrin should not significantly alter normal mechanisms for the binding and transport of heme or of preformed bilirubin by serum albumin. From a more general perspective, the results indicate potentially unusual binding site selectivity by tin chelates; possible origins of this selectivity are discussed.     

Distribution of pyridoxal-5-phosphate between proteins and low molecular weight components of plasma: effect of acetaldehyde

It is found that approximately 65-70% of pyridoxal-P at physiological concentrations is bound to plasma proteins; 15% of its amount is bound to amino acids and peptides as a result of the Schiff base formation. Over 85% of pyridoxal-P associated with plasma proteins is bound to serum albumin. Inorganic phosphate and NaCl decrease the affinity of pyridoxal-P for albumin or other proteins. Acetaldehyde interacts with the alpha-amino group of the aspartic acid residue of the N-end of the polypeptide chain of the albumin molecule and with two epsilon-amino groups of the lysine residues having anomalously low value of pKa and deprotonated at physiological values of pH of the medium. Acetaldehyde competes with pyridoxal-P for the first (of the highest affinity) binding site of the coenzyme on serum albumin. Acetaldehyde is not bound at the second site of high affinity for pyridoxal-P on serum albumin.     

Fatty acid and drug binding to a low-affinity component of human serum albumin, purified by affinity chromatography.

Binding equilibria for decanoate to a defatted, commercially available human serum albumin preparation were investigated by dialysis exchange rate determinations. The binding isotherm could not be fitted by the general binding equation. It was necessary to assume that the preparation was a mixture of two albumin components about 40% of the albumin having high affinity and about 60% having low affinity. By affinity chromatography we succeeded in purifying the low-affinity component from the mixture. The high-affinity component, however, could not be isolated. We further analyzed the fatty acid and drug binding abilities of the low-affinity component. The fatty acids decanoate, laurate, myristate and palmitate were bound with higher affinity to the mixture than to the low-affinity component. Diazepam was bound with nearly the same affinity to the low-affinity component as to the albumin mixture, whereas warfarin was not bound at all to the low-affinity component.     

Peptide-specific antibodies employed in determining the interspecies immunological cross-reactivity of haemopexin

1. Antibodies were raised in rabbits against nine peptides analogous to sequences of the human serum beta-glycoprotein haemopexin, and seven peptides were very antigenic. 2. One of these affinity-purified peptide-specific antibodies interacted with a highly conserved sequence of the haemopexin of five of the seven species tested. 3. Another antibody bound pig haemopexin even better than human haemopexin. 4. The overall, arbitrarily assessed, immunological cross-reactivity between the haemopexin of human and other species follows the order: rabbit greater than mouse greater than chicken greater than pig greater than rat greater than cow.     

Effects by heme, insulin, and serum albumin on heme and protein synthesis in chick embryo liver cells cultured in a chemically defined medium, and a spectrofluorometric assay for porphyrin composition.

Primary chick embryo liver cells, which had been previously cultured in Eagle's medium containing 10% fetal bovine serum, had the same characteristics (inducibility of delta-aminolevulinic acid synthetase and synthesis of plasma proteins) when cultured in a completely defined Ham F-12 medium containing insulin. Insulin was active in the physiological range; 2 to 3 nM were sufficient to increase the induced delta-aminolevulinic acid synthetase to 50% of the maximum effect obtained with a saturating amount of insulin (30 nM). Serum albumin added to the Ham-insulin medium caused protoporphyrin but not uroporphyrin, generated in the cultured liver cells, to be transferred to the medium. As little as 10 mug of human serum albumin per ml caused the transfer of one-half of the protoporphyrin. Bovine serum albumin was only about 1/30 as effective. A spectrofluorometric method and calculation procedure are described for quantitation, in the nanomolar range, of total porphyrin and the percentage of this that is protoporphyrin or uroporphyrin plus coproporphyrin. The method is satisfactory for the measurement of porphyrins generated by 1 mg wet weight of cells in culture in 20 hours. Heme (0.1 to 0.3 muM), when added to the medium as hemin, human hemoglobin, or chicken hemoglobin, specifically inhibited the induction of delta-aminolevulinic acid synthetase by one-half. This high sensitivity for heme was observed under conditions in which the defined medium was free of serum and where a chelator of iron was added to the medium to diminish the synthesis of endogenous heme. Heme endogenously generated from exogenous delta-aminolevulinic acid also inhibited the induction; chelators of iron prevented this inhibition. The migration of heme from the mitochondria to other portions of the cell is discussed in terms of the affinities of different proteins for heme. A hypothesis of a steady state of liver heme metabolism, controlled by the concentration of "free" heme, is presented. The different effects of heme on the synthesis of a number of proteins are summarized.     

Isolation of the haemopexin-haem receptor from pig liver cells

Isolated pig liver plasma membranes interact specifically with the haemopexin-haem complex (Kd 4.4 X 10(-7) M). Affinity chromatography was used to isolate a membrane component which binds this complex with high affinity. Pig serum haemopexin was first isolated by affinity chromatography on haemin-Sepharose followed by HPLC gel filtration. Liver membranes solubilized with Triton X-100 were incubated with haemin-Sepharose saturated with haemopexin, and as a control, with affinity gel lacking haemopexin. SDS-poly-acrylamide gel electrophoresis of the eluted protein indicated that from the haemin-Sepharose emerglow-molecular-mass haemin-binding proteins whereas the eluate from haemopexin-haemin-Sepharose contained an additional 71 kDa protein, which did not bind free haemin. This protein appears to represent the haemopexin-haem receptor or a part of it. Haem from the haemopexin complex, as also free haemin, was accepted by a binder in the plasma membrane, which in gel filtration behaved like an 80 kDa molecule. This component probably represents a second functional subunit of the haemopexin-haem receptor.     

A dynamic model for bilirubin binding to human serum albumin

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.     

Binding of Streptococcal Lipoteichoic Acid to Fatty Acid-Binding Sites on Human Plasma Fibronectin

The ability of Streptococcus pyogenes lipoteichoic acid and palmitic acid to bind to purified human plasma fibronectin was investigated. Initial studies indicated that intact fibronectin formed soluble complexes with lipoteichoic acid, resulting in a change in the mobility of fibronectin in an electrical field. Fibronectin covalently linked to agarose beads bound radiolabeled lipoteichoic acid in the acylated form but not in the deacylated form. An 18-M excess of fibronectin inhibited binding of lipoteichoic acid to the immobilized protein by 92%. Fibronectin-bound [(3)H]lipoteichoic acid could be specifically eluted with unlabeled lipoteichoic acid, as well as by fatty acid-free serum albumin. Serum albumin, which is known to contain fatty acid-binding sites capable of binding to the lipid moieties of lipoteichoic acid, inhibited the binding of lipoteichoic acid to fibronectin in a competitive fashion. The fibronectin-bound lipoteichoic acid could be eluted by 50% ethanol and various detergents but not by 1.0 M NaCl, various amino acids, or sugars. Similarly, radiolabeled palmitic acid adsorbed to fibronectin could be eluted with 50% ethanol but not with 1.0 M NaCl. Fibronectin adsorbed to a column of palmityl-Sepharose was eluted with 50% ethanol in 0.5% sodium dodecyl sulfate but not with 1.0 M NaCl or 1% sodium dodecyl sulfate alone. The binding of lipoteichoic acid to fibronectin followed first-order kinetics and was saturable. A Scatchard plot analysis of the binding data indicated a heterogeneity of lipoteichoic acid-binding sites similar to that previously found for serum albumin. Nevertheless, fibronectin contains at least one population of high-affinity binding sites for lipoteichoic acid. The binding affinity (nKa approximately 250 muM(-1)) is 2 orders of magnitude greater than the binding affinity of serum albumin. These data suggest that human plasma fibronectin contains specific binding sites for fatty acids and that lipoteichoic acid binds to these sites by way of its glycolipid moiety.     

Binding of norgestrel to human plasma proteins.

Binding of [14, 15-3H](+/-)-norgestrel to human plasma proteins has been investigated. Norgestrel showed greater affinity to plasma than to human serum albumin indicating specific norgestrel binding protein(s) in the plasma. alpha1-acid glycoprotein showed high affinity for norgestrel when compared with human serum albumin. The binding protein was eluted at pH 5.8 by step by step elution on a DEAE-cellulose column. Norgestrel binding to plasma proteins was not affected at 60 degrees C. The optimal binding occurred between pH 7 and 8. Ligand specificity of the binding protein revealed that progesterone was able to compete for the norgestrel binding sites, whereas corticosterone, testosterone, oestradiol, and norethindrone acetate did not show much competition. The molecular weight of the binding protein was found to be approximately 43 000. Sucrose density gradient analysis indicated that norgestrel bound to a macromolecular component of sedimentation coefficient 2.9 S. The association constant (Kass) and dissociation constant (Kdiss) of norgestrel-binding plasma protein was found to be 1.4-10(6) M-1 and 0.7-10(-6) M respectively. The number of binding sites was 0.5-10(-9) mol/mg protein. Norgestrel-binding protein in the plasma appeared to be a protein different from human serum albumin, corticosteroid-binding globulin and sex-steroid-binding protein. This binding protein showed some similarities to alpha1-acid glycoprotein.     

Binding of estrogen-3-sulfates to stallion plasma and equine serum albumin.

The binding of estrone-3-sulfate (E1-3-S) and estradiol-3-sulfate (E2-3-S) to adult stallion plasma was determined and compared with the binding to equine serum albumin (ESA). On the ESA molecule, two binding sites for E1-3-S with an association constant of 1.3 x 10(5) M-1 and several sites of weaker affinity were found; the data for E2-3-S showed the existence of four binding sites of moderate affinity (1 x 10(5) M-1) and several sites of weaker affinity. The removal of albumin from the stallion plasma resulted in the absence of binding of E1-3-S or E2-3-S, whereas the removal of glycoproteins resulted in binding parameters similar to those obtained with whole plasma. These results indicate that ESA is the only estrogen sulfate binder in horse plasma. Under physiological conditions, 95% of E1-3-S was bound to ESA.     

Multiple fatty acid binding to albumin in human blood plasma

Binding equilibria of long-chain fatty acids to human serum albumin, in serum or plasma, were studied by a dialysis exchange rate technique. Palmitate was added to citrated plasma in vitro and it was observed that between six and ten palmitate molecules were bound to albumin with nearly equal affinity. Observations in vivo gave similar results in the following series: (a) in two volunteers with increased fatty acid concentrations after fasting, exercise, and a cold shower: (b) in three male volunteers in whom high concentrations of non-esterified fatty acids, up to 4.6 mM, were induced by intravenous administration of a preparation of lecithin/glycocholate mixed micelles, and (c) in 81 patients with diabetes mellitus, type I. The binding pattern of palmitate in serum or plasma is essentially different from that observed with palmitate added to buffered solutions of pure albumin when two molecules are tightly bound and about four additional molecules with lower affinity. The differences may partly be explained by the presence of chloride ions in blood plasma, reducing the affinity for binding of the first two fatty acid molecules, and partly by facilitated binding of several molecules of mixed fatty acids, as found in plasma.     

Identification of alpha-1 acid glycoprotein as a lysophospholipid binding protein: a complementary role to albumin in the scavenging of lysophosphatidylcholine

Alpha-1 acid glycoprotein (AGP, orosomucoid), a major acute phase protein in plasma, displays potent cytoprotective and anti-inflammatory activities whose molecular mechanisms are largely unknown. Because AGP binds various exogenous drugs, we have searched for endogenous ligands for AGP. We found that AGP binds lysophospholipids in a manner discernible from albumin in several ways. First, mass spectrometric analyses showed that AGP isolated from plasma and serum contained lysophosphatidylcholine (LPC) enriched in mono and polysaturated acyl chains, whereas albumin contained mostly saturated LPC. Second, AGP bound LPC in a 1:1 molar ratio and with a higher affinity than free fatty acids, whereas albumin bound LPC in a 3:1 ratio but with a lower affinity than that of free fatty acids. Consequently, free fatty acids displaced LPC more avidly from albumin than from AGP. Competitive ligand displacement indicated the highest affinity for AGP to LPC20:4, 18:3, 18:1, and 16:0 (150-180 nM), lysophosphatidylserine (Kd 190 nM), and platelet activating factor (PAF) (Kd 235 nM). The high affinity of AGP to LPC in equilibrium was verified by stopped-flow kinetics, which implicated slow dissociation after fast initial binding, being consistent with an induced-fit mechanism. AGP also bound pyrene-labeled phospholipids directly from vesicles and more efficiently than albumin. AGP prevented LPC-induced priming and PAF-induced activation of human granulocytes, thus indicating scavenging of the cellular effects of the lipid ligands. The results suggest that AGP complements albumin as a lysophospholipid scavenging protein, particularly in inflammatory conditions when the capacity of albumin to sequester LPC becomes impaired.     

Binding of physostigmine to rat and human plasma and crystalline serum albumins

The binding of 3H-physostigmine (3H-Ph) to human and rat plasma proteins and crystalline serum albumin was studied by ultrafiltration technique. This study showed that the percentage of 3H-Ph bound to rat plasma slightly decreased from 49% to 41% whereas human plasma showed an increase in binding from 29% to 43% over a 50-fold increase in drug concentration. Human plasma samples which were collected in a bag coated with citrate phosphate dextrose adenine-1 solution bound 50% less 3H-Ph than samples collected with EDTA indicating a drug-drug interaction between 3H-Ph and anticoagulants. No significant change in binding was observed if the samples were frozen prior to use. Scatchard plots for binding of 3H-Ph resulted in a positive slope for human plasma and a negative slope for rat plasma; whereas curvilinear Scatchard plots with negative slopes were obtained for binding to human and rat crystalline serum albumin.     

Development of mammalian serum albumin affinity purification media by peptide phage display

Several phage isolates that bind specifically to human serum albumin (HSA) were isolated from disulfide-constrained cyclic peptide phage-display libraries. The majority of corresponding synthetic peptides bind with micromolar affinity to HSA in low salt at pH 6.2, as determined by fluorescence anisotropy. One of the highest affinity peptides, DX-236, also bound well to several mammalian serum albumins (SA). Immobilized DX-236 quantitatively captures HSA from human serum; mild conditions (100 mM Tris, pH 9.1) allow release of HSA. The DX-236 affinity column bound HSA from human serum with a greater specificity than does Cibacron Blue agarose beads. In addition to its likely utility in HSA and other mammalian SA purifications, this peptide media may be useful in the proteomics and medical research markets for selective removal of mammalian albumin from serum prior to mass spectrometric and other analyses.     

Cicletanine binding to human plasma proteins and erythrocytes, a particular HSA-drug interaction

The binding of cicletanine to human serum, isolated proteins and red blood cells was studied in vitro by equilibrium dialysis. Our results show this drug is highly bound to serum (97.3%) at therapeutic levels. No saturation to the binding sites was seen. Human serum albumin was shown to mainly responsible for this binding (93.5%) with a saturable process characterized by one binding site with a moderate affinity (K = 75800 M-1) and a non saturable process with a low total affinity (nK = 6400 M-1). Like many basic lipophilic drugs, cicletanine showed a saturable binding to alpha-1-acid glycoprotein with one site and a moderate affinity (K = 38,800 M-1). Its binding to lipoproteins and red blood cells was weak and non saturable. Over the range of therapeutic concentrations, the unbound fraction in blood remains constant (3.6%). Moreover, interactions were studied using bilirubin and non esterified fatty acids at pathological concentrations and these endogenous compounds did not alter cicletanine binding human serum or to human serum albumin likewise cicletanine shared the diazepam-site on HSA but no inhibition could take place between cicletanine and the drugs sharing the same binding site in serum at therapeutic levels.     

Role of apolipoprotein D in the transport of bilirubin in plasma

Apolipoprotein D (apo D) is a 30-kDa glycoprotein of unknown function that is associated with high-density lipoproteins (HDL). Because unconjugated bilirubin has been shown to bind apo D with a 0. 8:1 stoichiometry, we examined the contribution of this protein to transport of bilirubin in human plasma. Density gradient centrifugation analysis using physiological concentrations of [(14)C]bilirubin reveals that 9% of unconjugated bilirubin is associated with HDL, with the remaining pigment bound primarily to serum proteins (i.e., albumin). The percentage of total plasma bilirubin bound to HDL was found to increase proportionally with bilirubin concentration. Affinity of human apo D for bilirubin was determined by steady-state fluorescence quenching, with Scatchard analysis demonstrating a single binding site for unconjugated bilirubin with an affinity constant (K(a)) of approximately 3 x 10(7) M(-1). Incorporation of apo D into phosphatidylcholine vesicles had no effect on K(a), suggesting that a lipid environment does not alter the affinity of the protein for bilirubin. Using stopped-flow techniques, the first-order rate constant for bilirubin dissociation from apo D was measured at 5.4 s(-1) (half-time = 129 ms). Our findings indicate that HDL is the principal nonalbumin carrier of bilirubin in human plasma and further support the proposition that the affinity of HDL for bilirubin is primarily the result of binding to apo D.     

Plasma protein binding study of oxybutynin by high-performance frontal analysis

Plasma protein binding of oxybutynin (OXY) was investigated quantitatively and enantioselectively using high-performance frontal analysis (HPFA). An on-line HPLC system which consists of HPFA column, extraction column and analytical column was developed to determine the unbound concentrations of OXY enantiomers in human plasma, in human serum albumin (HSA) solutions, and in human alpha1-acid glycoprotein (AGP) solutions. OXY is bound in human plasma strongly and enantioselectively. The bound drug fraction in human plasma containing 2-10 microM (R)- or (S)-OXY was higher than 99%, and the unbound fraction of (R)-OXY was 1.56 times higher than that of (S)-isomer. AGP plays the dominant role in this strong and enantioselective plasma protein binding. The total binding affinities (nK) of (R)- and (S)-OXY to AGP were 6.86 x 10(6) and 1.53 x 10(7) M(-1), respectively, while the nK values of (R)- and (S)-OXY to HSA were 2.64 x 10(4) and 2.19 x 10(-4) M(-1), respectively. The binding affinity of OXY to AGP is much higher than that to HSA, and shows high enantioselectivity (SIR ratio of nK values is 2.2). It was found that both enantiomers are bound competitively at the same binding site on an AGP molecule. The binding property between OXY and low density lipoprotein (LDL) was investigated by using the frontal analysis method incorporated in high-performance capillary electrophoresis (HPCE/FA). It was found the binding is non-saturable and non-enantioselective.     

Flavonoid–serum albumin complexation: determination of binding constants and binding sites by fluorescence spectroscopy

After a meal rich in plant products, dietary flavonols can be detected in plasma as serum albumin-bound conjugates. Flavonol–albumin binding is expected to modulate the bioavailability of flavonols. In this work, the binding of structurally different flavonoids to human and bovine serum albumins is investigated by fluorescence spectroscopy using three methods: the quenching of the albumin fluorescence, the enhancement of the flavonoid fluorescence, the quenching of the fluorescence of the quercetin–albumin complex by a second flavonoid. The latter method is extended to probes whose high-affinity binding sites are known to be located in one of the two major subdomains (warfarin and dansyl-l-asparagine for subdomain IIA, ibuprofen and diazepam for subdomain IIIA). Overall, flavonoids display moderate affinities for albumins (binding constants in the range 1–15×10⁴ M⁻¹), flavones and flavonols being most tightly bound. Glycosidation and sulfation could lower the affinity to albumin by one order of magnitude depending on the conjugation site. Despite multiple binding of both quercetin and site probes, it can be proposed that the binding of flavonols primarily takes place in subdomain IIA. Significant differences in affinity and binding location are observed for the highly homologous HSA and BSA.     

Fluorescence and analytical ultracentrifugation analyses of the interaction of the tyrosine kinase inhibitor, tyrphostin AG 1478-mesylate, with albumin

Quantifying the interaction of drugs with carrier proteins in plasma is of importance for understanding effective drug delivery to disease-affected tissues. In this study, we employed analytical ultracentrifugation and steady-state fluorescence spectroscopy to characterize the interaction of a potential new anticancer drug, AG 1478-mesylate, with plasma proteins in a suspension of normal serum albumin (NSA). We found that mesylate salt of AG 1478, an epidermal growth factor receptor kinase inhibitor, sediments in 0.1%(w/v) NSA as a complex with a sedimentation coefficient of 3.8 S. This is consistent with the size of human serum albumin. This interaction was quantitated by meniscus depletion sedimentation and fluorescence titration analyses. AG 1478-mesylate binds to albumin with an apparent single-site affinity (K(d)) of 120 microM. In this article, we show that the cyclodextrin carrier molecule, Captisol, increases the apparent affinity of the hydrophobic AG 1478-mesylate for albumin (K(d)=4-6 microM), and we propose that the AG 1478-mesylate-Captisol (1:1) complex binds to albumin with at least 10-fold higher affinity than does AG 1478-mesylate ligand alone. A fluorenylmethoxycarbonyl-sulfonic acid (FMS) derivative of the 6-aminoquinazoline analog of AG 1478, which was designed to have improved serum-binding properties, was shown by fluorescence analysis to bind with approximately 100-fold greater affinity than the parent compound. This has significant implications in the effective delivery of therapeutic agents in vivo.