The intrinsic tyrosine fluorescence of histone H1. Steady state and fluorescence decay studies reveal heterogeneous emission

In wavelength-resolved steady state spectra we observe three different kinds of emission from histone H1, a class A protein with only a single tyrosine residue. Unfolded H1 emissions that peak at approximately 300 and 340 nm can both be excited maximally at approximately 280 nm. Another, peaking much further to the red at approximately 400 nm, can be excited maximally at approximately 320 nm. The 300-nm fluorescence can be resolved by lifetime measurements into three components with decay times of approximately 1, 2, and 4 ns. On sodium-chloride-induced refolding of H1, simplification of the emission properties occurs. The 340 and 400-nm components disappear while the two shorter lifetime components of the 300-nm band diminish in amplitude and are replaced by the 4-ns decay. We believe that the 340-nm emission is tyrosinate fluorescence resulting from excited-state proton transfer. The origin of the 400-nm emission remains uncertain. We assign the 1 and 2-ns components of the 300-nm emission to two states of tyrosine in denatured H1 and the 4-ns decay to fluorescence of the single tyrosine residue in the globular region of refolded H1. Our results support the contention that salt induced folding of H1 is a cooperative two state process, and permit us to better understand the previously reported increases in fluorescence intensity and anisotropy on salt-induced folding.     

Reagentless optical sensing of glutamine using a dual-emitting glutamine-binding protein

Glutamine is a major source of nitrogen and carbon in cell culture media. Thus, glutamine monitoring is important in bioprocess control. Here we report a reagentless fluorescence sensing for glutamine based on the Escherichia coli glutamine-binding protein (GlnBP) that is sensitive in the submicromolar ranges. The S179C variant of GlnBP was labeled at the -SH and N-terminal positions with acrylodan and ruthenium bis-(2,2'-bipyridyl)-1,10-phenanthroline-9-isothiocyanate, respectively. The acrylodan emission is quenched in the presence of glutamine while the ruthenium acts as a nonresponsive long-lived reference. The apparent binding constant, K'(d), of 0.72 microM was calculated from the ratio of emission intensities of acrylodan and ruthenium (I(515)/I(610)). The presence of the long-lived ruthenium allowed for modulation sensing at lower frequencies (1-10 MHz) approaching an accuracy of +/-0.02 microM glutamine. Dual-frequency ratiometric sensing was also demonstrated. Finally, the extraordinary sensitivity of GlnBP allows for dilution of the sample, thereby eliminating the effects of background fluorescence from the culture media.     

A fluorescently labeled intestinal fatty acid binding protein. Interactions with fatty acids and its use in monitoring free fatty acids.

The fatty acid-binding protein from rat intestine (I-FABP) has been covalently modified with the fluorescent compound Acrylodan. Acrylodan was found to label Lys27, one of the few amino acid residues found by x-ray diffraction studies to change orientation upon fatty acid (FA) binding to I-FABP. Binding of FA to this Acrylodan-modified I-FABP (ADIFAB) induces a large shift in fluorescence emission wavelength from 432 to 505 nm. As a consequence, the ratio of emission intensities provides a direct measure of the concentration of FA bound to the protein. Binding of FA is well described by single site equilibrium for FA concentrations below the critical micelle concentration. ADIFAB dissociation constants (Kd) determined at 37 degrees C and at concentrations below the critical micelle concentration for oleate, palmitate, linoleate, arachidonate, and linolenate were, respectively, 0.28, 0.33, 0.97, 1.6, and 2.5 microM. The variation of these Kd values with FA molecular species is highly correlated with the solubility of the FA in water, suggesting that all these FA bind with a similar conformation in the I-FABP binding site. The ADIFAB response together with the measured equilibrium constants allows a direct determination of the concentration of long chain free fatty acid (FFA) in the concentration range, depending upon the FA molecular species, between 1 nM and > 20 microM. As an example of its use as a probe to measure FFA levels, ADIFAB is used here to monitor the time course for FFA release from IgE receptor- and ionomycin-activated rat basophilic leukemia (RBL) cells.     

Hydrophobic surfaces of tubulin probed by time-resolved and steady-state fluorescence of nile red

Binding of Nile Red to tubulin enhances and blue-shifts fluorescence emission to about 623 nm with a "shoulder" around 665 nm. Binding is reversible and saturable with an apparent Kd of approximately 0.6 microM. Nile Red does not alter tubulin polymerization, and polymerization in 2-(N-morpholino)ethanesulfonic acid (Mes) buffer does not alter the spectrum of the Nile Red-tubulin complex. In contrast, polymerization in glutamate buffer results in a red shift, reduction of intensity, and a decrease in lifetime, suggesting an increase in "polarity" of the binding environment. Lifetimes of 4.5 and 0.6 ns fluorescence in Mes buffer are associated with the 623-nm peak and the 665-nm shoulder, respectively. Indirect excitation spectra for these components are distinct and the 4.5-ns component exhibits tryptophan to Nile Red energy transfer. Acrylamide quenching yields linear Stern-Volmer plots with unchanged lifetimes, indicating static quenching. Apparent quenching constants are wavelength-dependent; global analysis reveals a quenchable component corresponding to the 4.5 ns component and an "unquenchable" component superposing the 0.6-ns spectrum. Analysis of anisotropy decay required an "associative" model which yielded rotational correlation times of greater than 50 ns for the 4.5-ns lifetime and 0.3 ns for the 0.6-ns lifetime. Dilution of tubulin in Mes results in an apparent red shift of emission without lifetime changes, due only to loss of the 623-nm component. These data are reconciled in terms of a model with two binding sites on the tubulin dimer. The more "nonpolar" site is located in a region of subunit-subunit contact which accounts for the fluorescence changes upon dilution; this permits estimation of a subunit dissociation constant of 1 microM.     

Measuring the adsorption of Fatty acids to phospholipid vesicles by multiple fluorescence probes

Fatty acids (FA) are important nutrients that the body uses to regulate the storage and use of energy resources. The predominant mechanism by which long-chain fatty acids enter cells is still debated widely as it is unclear whether long-chain fatty acids require protein transporters to catalyze their transmembrane movement. We use stopped-flow fluorescence (millisecond time resolution) with three fluorescent probes to monitor different aspects of FA binding to phospholipid vesicles. In addition to acrylodan-labeled fatty acid binding protein, a probe that detects unbound FA in equilibrium with the lipid bilayer, and cis-parinaric acid, which detects the insertion of the FA acyl chain into the membrane, we introduce fluorescein-labeled phosphatidylethanolamine as a new probe to measure the binding of FA anions to the outer membrane leaflet. We combined these three approaches with measurement of intravesicular pH to show very fast FA binding and translocation in the same experiment. We validated quantitative predictions of our flip-flop model by measuring the number of H⁺ delivered across the membrane by a single dose of FA with the probe 6-methoxy-N-(3-sulfopropyl) quinolinium. These studies provide a framework and basis for evaluation of the potential roles of proteins in binding and transport of FA in biological membranes.     

Effect of omega-3 fatty acids on the modification of erythrocyte membrane fatty acid content including oleic acid in peritoneal dialysis patients

Erythrocyte membrane fatty acids (FA), such as oleic acid, are related to acute coronary syndrome. There is no report about the effect of omega-3 FA on oleic acid in peritoneal dialysis (PD) patients. We hypothesized that omega-3 FA can modify erythrocyte membrane FA, including oleic acid, in PD patients. In a double-blind, randomized, placebo-controlled study, 18 patients who were treated with PD for at least 6 months were randomized to treatment for 12 weeks with omega-3 FA or placebo. Erythrocyte membrane FA content was measured by gas chromatography at baseline and after 12 weeks. The erythrocyte membrane content of eicosapentaenoic acid and docosahexaenoic acid was significantly increased and saturated FA and oleic acid were significantly decreased in the omega-3 FA supplementation group after 12 weeks compared to baseline. In conclusion, erythrocyte membrane FA content, including oleic acid, was significantly modified by omega-3 FA supplementation for 12 weeks in PD patients.     

Fluorescence studies of the carboxyl-terminal domain of smooth muscle calponin effects of F-actin and salts.

The fluorescence parameters of the environment-sensitive acrylodan, selectively attached to Cys273 in the C-terminal domain of smooth muscle calponin, were studied in the presence of F-actin and using varying salt concentrations. The formation of the F-actin acrylodan labeled calponin complex at 75 mm NaCl resulted in a 21-nm blue shift of the maximum emission wavelength from 496 nm to 474 nm and a twofold increase of the fluorescent quantum yield at 460 nm. These spectral changes were observed at the low ionic strengths (< 110 mm) where the calponin : F-actin stoichiometry is 1 : 1 as well as at the high ionic strengths (> 110 mm) where the binding stoichiometry is a 1 : 2 ratio of calponin : actin monomers. On the basis of previous three-dimensional reconstruction and chemical crosslinking of the F-actin-calponin complex, the actin effect is shown to derive from the low ionic strength interaction of calponin with the bottom of subdomain-1 of an upper actin monomer in F-actin and not from its further association with the subdomain-1 of the adjacent lower monomer which occurs at the high ionic strength. Remarkably, the F-actin-dependent fluorescence change of acrylodan is qualitatively but not quantitatively similar to that earlier reported for the complexes of calponin and Ca2+-calmodulin or Ca2+-caltropin. As the three calponin ligands bind to the same segment of the protein, encompassing residues 145-182, the acrylodan can be considered as a sensitive probe of the functioning of this critical region. A distance of 29 A was measured by fluorescence resonance energy transfer between Cys273 of calponin and Cys374 of actin in the 1 : 1 F-actin-calponin complex suggesting that the F-actin effect was allosteric reflecting a global conformational change in the C-terminal domain of calponin.     

Interactions of fusidic acid and elongation factor G with lipid membranes

Fusidic acid (FA) is a potent antibiotic and blocks the protein synthesis by binding to elongation factor G (EF-G) directly. Here we hypothesized that the antibiotic activity of FA would be potentiated by several orders of magnitude if both FA and EF-G would be residing in the lipid membranes and, hence, the probability of interaction would transform from three-dimensional to two-dimensional. Such detailed information could lead to more effective therapeutic interventions if they are understood on a molecular level. Interactions between FA and various lipid membranes composed of 1-palmitoyl-2-oleyl-sn-glycero-3-phosphocholine (POPC) and cholesterol (Chol) were studied by capillary electrochromatography (CEC). The influence of the lipid vesicle size--sonicated liposomes and liposomes extruded through 30-, 50-, and 100-nm filters--on the packing of vesicles on the silica capillary surface was investigated by CEC and dissipative quartz crystal microbalance. The CEC results evidenced that FA interacts with and resides in phospholipid membranes. Likewise, monolayer, asymmetrical flow field flow fractionation, and CEC studies confirmed that EF-G is hydrophobic and incorporated into POPC and POPC/Chol membranes. Including EF-G in phospholipid vesicles did not improve the binding of FA to the membranes.     

Role of caveolin-1 and cholesterol in transmembrane fatty acid movement

We have created by transfection a series of HEK 293 cell lines that express varying amounts of caveolin-1 to test the possible effect of this protein on the transport and metabolism of long chain fatty acids (FA) in cells with this gain of function. We used an extracellular fluorescent probe (ADIFAB) to monitor binding of exogenous FA to the plasma membrane and an intracellular pH probe to monitor FA equilibration across the plasma membrane. Real-time fluorescence measurements showed rapid binding of oleic acid to the extracellular side of the plasma membrane and a rapid translocation across the lipid bilayer by the flip-flop mechanism (<5 s). Two cell lines expressing levels of caveolin-1 roughly comparable to that of adipocytes, which have a very high level of endogenous expression of caveolin-1, showed a relatively slow change in intracellular pH (t(1/2) < 100 s) in addition to the fast changes in fluorescence. We interpret this additional second phase to represent translocation of additional FA from the outer to inner leaflet of the plasma membrane. The slower kinetics could represent either slower flip-flop of FA across highly organized, rigid regions of the plasma membrane or binding of FA to caveolin-1 in the intracellular leaflet of the plasma membrane. The kinetics of palmitate and elaidate (a trans FA) transmembrane movement were identical to that for oleate. These results were observed in the absence of the putative FA transport protein, CD36, and in the absence of any changes in expression of fatty acid transport proteins (FATP) 2 and 4, and are in direct correlation with increased cellular free cholesterol content. FA metabolism was slow in all cell lines and was not enhanced by caveolin-1 expression. We conclude that transport of FA across the plasma membrane is modulated by caveolin-1 and cholesterol and is not dependent on the putative FA transport proteins CD36 and FATP.     

Fluorescence lifetime and spectral study of the acid expansion of bovine serum albumin

The fluorescence lifetimes of the tryptophan residues of bovine serum albumin were measured in the native and acid-expanded conformation. A three-exponential process is required to fit the fluorescence decay data. The results are interpreted empirically in terms of two emitting species. The emission at longer wavelength (360 nm) has slower rates of decay than that at shorter wavelength (325 nm). For both emitting species the average lifetime decreases when the N-F transition occurs and shortens further when the protein expands. Rotational correlation times, derived from the decay of the fluorescence anisotropy of the tryptophan residues, suggest that longer emission wavelengths are associated with somewhat shorter correlation times. There is no certain indication of any independent motion of the tryptophans in any conformation, although some very fast process, perhaps Raman scattering, appears to occur. On acid expansion the long correlation times decrease to around 10 ns in the fully expanded form. Static quenching experiments using I- or acrylamide suggest a greater average exposure of the tryptophans when the protein is most greatly expanded. This is despite the fact that the fluorescence emission maximum shifts to shorter wavelength under these conditions. Also, there is no difference in accessibility to quenching between the longer and shorter wavelength emissions.     

Fatty acid binding proteins from different tissues show distinct patterns of fatty acid interactions

Fatty acid binding proteins (FABP) form a family of proteins displaying tissue-specific expression. These proteins are involved in fatty acid (FA) transport and metabolism by mechanisms that also appear to be tissue-specific. Cellular retinoid binding proteins are related proteins with unknown roles in FA transport and metabolism. To better understand the origin of these tissue-specific differences we report new measurements, using the acrylodated intestinal fatty acid binding protein (ADIFAB) method, of the binding of fatty acids (FA) to human fatty acid binding proteins (FABP) from brain, heart, intestine, liver, and myelin. We also measured binding of FA to a retinoic acid (CRABP-I) and a retinol (CRBP-II) binding protein and we have extended to 19 different FA our characterization of the FA-ADIFAB and FA-rat intestinal FABP interactions. These studies extend our previous analyses of human FABP from adipocyte and rat FABPs from heart, intestine, and liver. Binding affinities varied according to the order brain approximately myelin approximately heart > liver > intestine > CRABP > CRBP. In contrast to previous studies, no protein revealed a high degree of selectivity for particular FA. The results indicate that FA solubility (hydrophobicity) plays a major role in governing binding affinities; affinities tend to increase with increasing hydrophobicity (decreasing solubility) of the FA. However, our results also reveal that, with the exception of the intestinal protein, FABPs exhibit an additional attractive interaction for unsaturated FA that partially compensates for their trend toward lower affinities due to their higher aqueous solubilities. Thermodynamic potentials were determined for oleate and arachidonate binding to a subset of the FABP and retinoid binding proteins. FA binding to all FABPs was enthalpically driven. The DeltaH degrees values for paralogous FABPs, proteins from the same species but different tissues, reveal an exceptionally wide range of values, from -22 kcal/mol (myelin) to -7 kcal/mol (adipocyte). For orthologous FABPs from the same tissue but different species, DeltaH degrees values were similar. In contrast to the enthalpic dominance of FA binding to FABP, binding of FA to CRABP-I was entropically driven. This is consistent with the notion that FA specificity for FABP is determined by the enthalpy of binding. Proteins from different tissues also revealed considerable heterogeneity in heat capacity changes upon FA binding, DeltaC(p) values ranged between 0 and -1.3 kcal mol(-1) K(-1). The results demonstrate that thermodynamic parameters are quite different for paralogous but are quite similar for orthologous FABP, suggesting tissue-specific differences in FABP function that may be conserved across species.     

Interactions of oleic acid with liver fatty acid binding protein: a carbon-13 NMR study

13C NMR spectroscopy was used to probe the structural interactions between carboxyl-13C-enriched oleic acid (18:1) and rat liver fatty acid binding protein (FABP) and the partitioning of 18:1 between FABP and unilamellar phosphatidylcholine (PC) vesicles. Spectra of systems containing 2-8 mol of 18:1/mol of FABP (but no PC) exhibited one carboxyl resonance (182.2 ppm) corresponding to FABP-bound 18:1. At pH values less than 8.0, an additional carboxyl resonance, corresponding to unbound 18:1 in a lamellar phase, was observed. Both resonances exhibited ionization shifts with estimated apparent pKa values of less than 5 (bound 18:1) and greater than 7 (unbound 18:1). The intensity of the resonance corresponding to FABP-bound 18:1 increased with increasing 18:1/FABP mole ratio and at 8/1 mole ratio indicated that at least 2 and 6 mol of 18:1/mol of FABP were FABP-bound at pH 7.4 and 8.6, respectively. NMR spectra of systems containing equal concentrations (w/v) of FABP and PC and from 1 to 4 mol of total fatty acid (FA)/mol of FABP exhibited two 18:1 carboxyl resonances (182.2 and 178.5 ppm, pH 7.4). The downfield resonance corresponded to FABP-bound 18:1 and the upfield resonance to PC vesicle bound 18:1. At 1/1 mole ratio (FA/FABP), the intensities of both resonances were approximately equal, but at 4/1 mole ratio the resonance for PC vesicle bound 18:1 was 3-fold more intense than that for FABP-bound 18:1. The following conclusions are reached: (i) The carboxyl groups of 18:1 bound to liver FABP experience only one type of binding environment (the aqueous milieu adjacent to the protein surface).(ABSTRACT TRUNCATED AT 250 WORDS)     

Trafficking of exogenous fatty acids within Caco-2 cells.

Dietary fatty acids (FAs) crossing the apical plasma membrane of small intestinal enterocytes are targeted to different metabolic pathways than serum FAs crossing the basolateral membrane. This apparent compartmentalization of FA metabolism in enterocytes was further investigated using a model human enterocyte-like intestinal cell line. [3H]Oleic acid bound to bovine serum albumin (BSA) was added to the apical or basolateral surfaces of confluent monolayers of Caco-2 cells growing on uncoated polycarbonate filters. In other experiments, [3H]oleic acid incorporated into micelles with taurocholate (+/- 2-monoacylglycerol) was added apically. Caco-2 cells absorbed oleic acid bound to BSA from both the apical and basolateral surfaces at the same rate. Oleic acid in micellar solution was absorbed more efficiently than oleic acid bound to BSA. Regardless of its site or mode of presentation, the majority of the incorporated oleic acid was found in triglycerides. Only a small fraction was subjected to beta-oxidation or esterification into phospholipids. Most of the incorporated oleic acid was still retained intracellularly at 24 h. The polarity of triglyceride secretion was influenced by the experimental conditions. Triglyceride secretion was not significantly polarized when oleic acid-BSA was presented apically. However, the ratio of basolateral to apical secretion at 24 h was 9:1 for oleic acid-BSA presented basolaterally. For oleic acid in taurocholate micelles there was a trend toward polarity of secretion to the apical media (apical to basolateral ratio = 2:1). The inclusion of 2-monoacylglycerol in oleic acid-taurocholate micelles did not augment triglyceride synthesis or secretion. These differences indicate that compartmentation of FA metabolism in Caco-2 cells is influenced by the site of FA presentation. Northern and Western blot hybridization studies indicated that the liver fatty acid-binding protein but not the intestinal fatty acid-binding protein gene is expressed in these cells. The absence of this latter 15 kDa protein indicates that it is not required by Caco-2 cells for the synthesis of triglycerides or for the polarized export of triglyceride. These studies indicate that the Caco-2 cell line will be a useful model system for studying the polarization of FA trafficking/metabolism in enterocytes and defining the role of intracellular fatty acid binding proteins in these processes.     

Differences between apo and three holo forms of the intestinal fatty acid binding protein seen by molecular dynamics computer calculations

It is commonly believed that binding affinity can be estimated by consideration of local changes of ligand and protein. This paper discusses a set of molecular dynamics simulations of intestinal fatty acid binding protein addressing the protein's response to presence or absence of different ligands. A 5-ns simulation was performed of the protein without a ligand, and three simulations (one 5-ns and two 2-ns) were performed with different fatty acids bound. The results indicate that, although the basic protein structure is unchanged by the presence of the ligand, other properties are significantly affected by ligand binding. For example, zero-time covariance patterns between protein, bound waters, and ligand vary between the different simulations. Moreover, the interaction energies between ligand and specific residues indicate that different ligands are stabilized in different ways. In sum, the results suggest that binding thermodynamics within this system will need to be calculated not from a subset of nearby protein:ligand interactions, but will depend on a knowledge of the motions coupling together water, protein, and ligand.     

Ultraviolet and violet light: attractive orientation cues for the Indian meal moth, Plodia interpunctella

The Indian meal moth (IMM), Plodia interpunctella (Hübner) (Lepidoptera: Pyralidae), engages in long-distance or foraging flights in the twilight hours of the scotophase when blue light dominates the irradiance spectrum of the sky. We tested the hypothesis that IMM uses wavelengths of visible blue/violet light as orientation cues that trigger phototactic responses. In four-choice laboratory experiments, blue light (400–475 nm) was significantly more effective than green (475–600 nm), orange (575–700 nm), or red (590–800 nm) light in attracting males and mated females. In subsequent experiments that tested light emitting diodes (LEDs) emitting peak wavelengths in the blue/violet-light range, the 405-nm 'violet' LED was significantly more effective than the 435-, 450-, or 470-nm 'blue' LED in attracting males as well as virgin and mated females. In electroretinogram recordings, the 405-nm wavelength elicited significantly stronger receptor potentials from female and male eyes than the 350-nm (UV) wavelength, and in a behavioral experiment it significantly enhanced the known attractiveness of UV light. Equal attraction of IMMs to 405-nm LEDs at 600–700 µW/cm² with or without UV light, and significantly stronger attraction to a 405-nm LED than to a 350-nm LED at maximum light intensities, suggest that the deployment of violet instead of UV light could become one of several management tactics for control of IMMs.     

Determining molecular binding sites on human serum albumin by displacement of oleic acid

An NMR method was developed for determining binding sites of small molecules on human serum albumin (HSA) by competitive displacement of (13)C-labeled oleic acid. This method is based on the observation that in the crystal structure of HSA complexed with oleic acid, two principal drug-binding sites, Sudlow's sites I (warfarin) and II (ibuprofen), are also occupied by fatty acids. In two-dimensional [(1)H,(13)C]heteronuclear single quantum coherence NMR spectra, seven distinct resonances were observed for the (13)C-methyl-labeled oleic acid as a result of its binding to HSA. Resonances corresponding to the major drug-binding sites were identified through competitive displacement of molecules that bind specifically to each site. Thus, binding of molecules to these sites can be followed by their displacement of oleic acids. Furthermore, the amount of bound ligand at each site can be determined from changes in resonance intensities. For molecules containing fluorine, binding results were further validated by direct observations of the bound ligands using (19)F NMR. Identifying the binding sites for drug molecules on HSA can aid in determining the structure-activity relationship of albumin binding and assist in the design of molecules with altered albumin binding.     

Fluorescence and bioluminescence of bacterial luciferase intermediates.

An intermediate in the luciferase-catalyzed bioluminescent oxidation of FMNH2, isolated and purified by chromatography at -20degrees, was postulated to be an oxygenated reduced flavin-luciferase. Maintained and studied at -20 to -30degrees, this material exhibits a relatively weak fluorescence emission peaking about 505 nm when excited at 370 nm. It may comprise more than one species. Upon continued exposure to light at 370 nm, the intensity of this fluorescence increases, often by a factor of 5 or more, and its emission spectrum is blue shifted to a maximum at about 485 nm. Upon warming its fluorescence is lost and the fluorescence of flaving mononucleotide appears. If warming is carried out in the presence of a long chain aldehyde, bioluminescence occurs, with the appearance of a similar amount of flavine fluorescence. The bioluminescence yield is about the same with irradiated and nonirradiated samples. The bioluminescence emission spectrum corresponds exactly to the fluorescence emission spectrum of the intermediate formed by irradiation, implicating the latter as being structurally close to the emitting species in bioluminescence.     

Human serum albumin binding assay based on displacement of a non selective fluorescent inhibitor

In this paper, we describe a fluorescent antibacterial analog, 6, with utility as a competition probe to determine affinities of other antibacterial analogs for human serum albumin (HSA). Analog 6 bound to HSA with an affinity of 400+/-100 nM and the fluorescence was environmentally sensitive. With 370 nm excitation, environmental sensitivity was indicated by a quenching of the 530 nm emission when the probe bound to HSA. Displacement of dansylsarcosine from HSA by 6 indicated it competed with compounds that bound at site II (ibuprofen binding site) on HSA. Analog 6 also shifted the NMR peaks of an HSA bound oleic acid molecule that itself was affected by compounds that bound at site II. In addition to binding at site II, 6 interacted at site I (warfarin binding site) as indicated by displacement of dansylamide and the shifting of NMR peaks of an HSA bound oleic acid molecule affected by warfarin site binding. Additional evidence for multiple site interaction was discovered when a percentage of 6 could be displaced by either ibuprofen or phenylbutazone. A competition assay was established using 6 to determine relative affinities of other antibacterial inhibitors for HSA.