Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M707637200 on January 3, 2008

J. Biol. Chem., Vol. 283, Issue 10, 6126-6135, March 7, 2008
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
283/10/6126    most recent
M707637200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Jeppesen, B.
Right arrow Articles by Tait, J. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Jeppesen, B.
Right arrow Articles by Tait, J. F.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Entropic and Enthalpic Contributions to Annexin V-Membrane Binding

A COMPREHENSIVE QUANTITATIVE MODEL*Formula

Brian Jeppesen{ddagger}, Christina Smith{ddagger}, Donald F. Gibson{ddagger}, and Jonathan F. Tait{ddagger}§1

From the Departments of {ddagger}Laboratory Medicine, §Medicine (Medical Genetics), and Pathology, University of Washington, Seattle, Washington 98195

Received for publication, September 11, 2007 , and in revised form, November 30, 2007.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Annexin V binds to membranes with very high affinity, but the factors responsible remain to be quantitatively elucidated. Analysis by isothermal microcalorimetry and calcium titration under conditions of low membrane occupancy showed that there was a strongly positive entropy change upon binding. For vesicles containing 25% phosphatidylserine at 0.15 M ionic strength, the free energy of binding was –53 kcal/mol protein, whereas the enthalpy of binding was –38 kcal/mol. Addition of 4 M urea decreased the free energy of binding by about 30% without denaturing the protein, suggesting that hydrophobic forces make a significant contribution to binding affinity. This was confirmed by mutagenesis studies that showed that binding affinity was modulated by the hydrophobicity of surface residues that are likely to enter the interfacial region upon protein-membrane binding. The change in free energy was quantitatively consistent with predictions from the Wimley-White scale of interfacial hydrophobicity. In contrast, binding affinity was not increased by making the protein surface more positively charged, nor decreased by making it more negatively charged, ruling out general ionic interactions as major contributors to binding affinity. The affinity of annexin V was the same regardless of the head group present on the anionic phospholipids tested (phosphatidylserine, phosphatidylglycerol, phosphatidylmethanol, and cardiolipin), ruling out specific interactions between the protein and non-phosphate moieties of the head group as a significant contributor to binding affinity. Analysis by fluorescence resonance energy transfer showed that multimers did not form on phosphatidylserine membranes at low occupancy, indicating that annexin-annexin interactions did not contribute to binding affinity. In summary, binding of annexin V to membranes is driven by both enthalpic and entropic forces. Dehydration of hydrophobic regions of the protein surface as they enter the interfacial region makes an important contribution to overall binding affinity, supplementing the role of protein-calcium-phosphate chelates.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The annexins are a family of calcium-dependent membrane-binding proteins with diverse functions (1–3), most of which involve interaction with membranes. Some members of the family have been implicated in disease states (4, 5), and some are being developed for diagnostic or therapeutic applications. Annexin V is a typical member of this protein family that has also received particular attention as a diagnostic agent to measure cell death in vivo in many different diseases (6, 7). Further work on understanding the functions and potential clinical applications of annexins will be greatly facilitated by a fuller understanding of the structural and mechanistic basis for their interaction with membranes.

Annexins bind to membranes with remarkably high affinity (8–10), and the reaction is strongly exothermic (11–13), but the mechanistic basis for this reaction is complex. There is broad agreement on the importance of calcium for the interaction of annexins with membranes at neutral pH. Calcium is thought to mediate binding via formation of calcium bridges between protein and membrane phospholipid head groups (14). Beyond this, there is little agreement on the key factors governing binding affinity, and many different theories have been suggested. General electrostatic interactions may occur between positively charged amino acids and negatively charged groups on the membrane. There may be insertion of specific hydrophobic residues, such as Trp-187, into the hydrophobic region of the membrane (15, 16). There may also be specific interactions between protein side chains and non-phosphate components of the phospholipid head groups (14, 17). Finally, formation of trimers and higher order multimers has been postulated to promote membrane binding via formation of favorable protein-protein interactions on the membrane surface (13, 18–21).

Recently, we developed methods for quantitative measurement of the binding of annexin V to cells and phospholipid vesicles, which allow one to determine the free energy of binding for individual protein molecules (22, 23). These methods rely on calcium titration performed at low levels of membrane occupancy, which overcomes the difficulties of analyzing a full binding isotherm previously identified by Bazzi and Nelsestuen (24, 25). This method allowed us to determine that under physiological conditions there are about eight high-affinity calcium binding sites, formed by the A and B helices in each of the four domains, that account for about 70% of the free energy of binding (23). In this study, we sought to investigate the source of the remaining ~30% of binding energy, and thereby gain insight into the structural and mechanistic basis for the binding reaction. We find that there is a strongly positive entropy change upon binding, which we attribute to hydrophobic interactions in the interfacial region of the membrane bilayer. We also found that nonspecific ionic interactions, protein-head group interactions, and protein-protein interactions did not contribute significantly to binding affinity. We present an overall quantitative model to explain these results under "Discussion."


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Materials—Recombinant proteins were prepared as described: human wild-type annexin V (26) and annexin V-128 (23). Both proteins were quantitated by absorbance at 280 nm using an extinction coefficient of 21,500 M–1 cm–1 (27). Calcium chloride stock solution (1 M) was prepared in water from ultrapure calcium chloride (99.99%; Aldrich) and the concentration determined by refractometry. Phospholipids were from Avanti%20Polar%20Lipids">Avanti Polar Lipids Inc. (Alabaster, AL) and unilamellar vesicles (diameter ~110 nm (28)) were prepared as previously described (8). Unless noted differently, the vesicles were composed of 25% PS (1-palmitoyl,2-oleoyl)2 (Avanti product number 840034), 2% R-PE (rhodamine-labeled phosphatidylethanolamine; number 810146), 20% C7-PC (1,2-diheptanoyl; number 850306), and 53% PC (1-palmitoyl,2-oleoyl; number 850457). Phospholipids with different anionic head groups tested were heart cardiolipin (number 840012), 1,2-dioleoyl-phosphatidylmethanol (number 840420), and 1-palmitoyl,2-oleoyl-phosphatidylglycerol (number 840457). Other reagents were from Sigma.

Construction, Expression, and Fluorescent Labeling of Annexin V Mutants—Expression vectors were constructed from the base plasmid pJ128 (23), which expresses a 325-amino acid protein (named annexin V-128) similar to wild-type human annexin V but with these changes: 1) an N-terminal extension of (Met)-Ala-Gly-Gly-Cys-Gly-His; 2) deletion of the initiator Met at position 1 of wild-type annexin V; 3) a point mutation C316S. Point mutations were made with the QuikChange Multisite-directed mutagenesis kit (Stratagene, La Jolla, CA). After construction and cloning, each mutant vector was sequenced through the entire protein-coding region to verify the presence of the desired mutation(s) and the absence of unintended mutations. Each annexin V mutant protein was named with an arbitrary numerical suffix. For protein expression, plasmids were transformed into Escherichia coli strain Tuner (DE3) pLacI (Novagen, Madison, WI) and expressed and purified as described (23). The yield and chromatographic behavior of all mutant proteins were comparable with that of annexin V-128, suggesting that all proteins were stable and did not have grossly altered conformations. Protein purity was ≥98% as judged by SDS-PAGE with Coomassie staining. Proteins were quantitated with an extinction coefficient at 280 nm of 0.6 ml/mg·cm (27) (with appropriate adjustment for the W187A mutant). Annexin V-128 and mutant proteins were labeled specifically on the N-terminal cysteine with IAF (Sigma) as described (22) and purified after labeling with the same MonoQ chromatography step used for unlabeled proteins. A few proteins used in FRET assays were also labeled with Alexa Fluor 546 C5 maleimide (Invitrogen), using the same procedure.

Fluorescence Quenching Vesicle Binding Assay—The affinity of IAF-annexin V-128 for phospholipid vesicles was determined by calcium titration as described by Tait et al. (22); binding is detected by quenching of fluorescein fluorescence due to resonance energy transfer upon binding to vesicles containing 2% rhodamine-labeled phospholipid. Unless noted otherwise, vesicle binding assays were performed at 25 °C with 1 nM IAF-annexin V-128 and 10 µM vesicles containing 25% PS and 2% R-PE in a buffer containing 50 mM HEPES-Na, pH 7.4, 100 mM NaCl, 3 mM NaN3, 0.1 mg/ml ovalbumin. A concentration of 10 µM vesicles corresponds to 100 nM of binding sites assuming about 100 phospholipid molecules per binding site (counting phospholipids on both faces of the bilayer) (22); the vesicle membrane was therefore ≤1% saturated with protein throughout the titration. Fluorescence measurements were made on a Victor3 Model 1420 fluorescence plate reader (PerkinElmer Life Sciences) with excitation at 485 ± 7 nm and emission at 535 ± 25 nm.

Calcium titrations were analyzed as described (22) based on the following model of the binding reaction between calcium ions, annexin V, and membrane binding sites in Equations 1, 2, 3, 4, 5.

Formula 1(Eq. 1)

Formula 2(Eq. 2)

Formula 3(Eq. 3)

Formula 4(Eq. 4)

Formula 5(Eq. 5)
The EC50 and calcium stoichiometry (n) values were determined by non-linear least-squares fitting of the calcium titration curves to the following function,

Formula 6(Eq. 6)
where Q is the observed fluorescence quenching at a given calcium concentration, and Qmax is the maximum quenching observed (about 0.8) when all fluorescent protein is bound to the vesicles.

Cell Binding Assay—Affinities of mutant proteins for RBC with exposed PS were measured by calcium titration under conditions of low membrane occupancy as described by Tait et al. (22). RBC binding assays were performed with 1 nM IAF-protein and 1 x 108 cells in 1 ml of a buffer containing 50 mM HEPES-Na, pH 7.4, 100 mM NaCl, 3 mM NaN3, 1 mg/ml bovine serum albumin. The concentration of binding sites was about 180 nM, assuming that each RBC contains 1.1 x 106 binding site (22); the membrane was therefore ≤0.6% saturated with protein throughout the titration.


Figure 1
View larger version (15K):
[in this window]
[in a new window]

  		FIGURE 1.
Isothermal titration calorimetry of annexin V as a function of annexin/phospholipid ratio. The calorimeter cell contained phospholipid vesicles (25% PS, 2% R-PE, 48% PC, 20% C7-PC) at an initial concentration of 2 mM and annexin V at initial concentrations from 0.2 to 20 µM, in 50 mM HEPES-Na, pH 7.4, 100 mM NaCl, 3 mM NaN3. Calcium chloride was added in small increments to give the indicated total concentrations, and the heat response integrated after each addition. The graph shows the cumulative heat response over the course of the titration. Titration of calcium into vesicles alone gave a small, linear, non-saturable exothermic response (224), which was subtracted from the data obtained in the presence of protein plus vesicles. Panel A, total heat evolved for protein concentrations of 20 µM ({square}), 14.6 µM ({circ}), 10 µM ({Delta}), 4 µM (+), 2 µM (x), and 0 µM (224). Panel B: total heat evolved for protein concentrations of 2 µM ({square}), 1.4 µM ({circ}), 0.8 µM ({Delta}), 0.4 µM (+), 0.2 µM (x), and 0 µM (224). Panel C, comparison of enthalpy and free energy of binding as a function of membrane occupancy for vesicles containing 25% PS. The molar enthalpy change ({circ}) for titrations given in A and B is plotted as a function of the annexin:phospholipid ratio in the titration. The upper x axis gives the corresponding value of membrane occupancy, assuming a ratio of 100 phospholipid molecules (on both faces of the bilayer) per annexin V when the vesicle surface is fully covered with annexin V (22). The dotted line is the average of all {Delta}H values obtained at all occupancy levels. The cumulative free energy of binding ({square}, solid line) was determined from the fluorescence quenching assay and data are taken from Fig. 2C.

 
FRET Assay for Formation of Annexin V Multimers—Assays were performed in a buffer consisting of 50 mM HEPES-Na, pH 7.4, 100 mM NaCl, 3 mM NaN3, 0.1 mg/ml ovalbumin. Phospholipid vesicles consisted of 25% PS, 20% C7PC, 55% PC, or 100% PS. A calcium concentration of 0.1 mM was used for most experiments with 100% PS, and a calcium concentration of 1 mM for most experiments with 25% PS. In most experiments ("vesicles last" protocol), buffer, calcium, IAF-labeled protein (0.5 nM), and Alexa Fluor 546-labeled protein (2.5 nM) were combined and mixed once; phospholipid vesicles were then added at various concentrations from 0 to 100 µM and the sample was mixed again and incubated 5 min at room temperature in the dark. In some experiments ("vesicles first" protocol), the vesicles were added in the first step, and the two fluorescent proteins in the second step. Fluorescence intensity (excitation at 485 nm, emission at 535 nm) was measured on triplicate 300-µl reaction aliquots on a Victor3 plate reader as described above. Fluorescent quenching was calculated relative to the fluorescence value of a sample without phospholipid vesicles.

Isothermal Titration Calorimetry—The isothermal titration calorimetry experiments were performed at 25 °C on a Micro-Cal VP-ITC (MicroCal LLC, Northampton, MA). Unless otherwise indicated, both the calcium titrant and the sample in the calorimetric cell were in a buffer of 50 mM HEPES-Na, pH 7.4, 100 mM NaCl, 3 mM NaN3 that was thoroughly degassed prior to the experiment. The syringe burette contained 300 µl of 20 mM CaCl2 in this buffer, and aliquots were injected at 300-s intervals. The first injection was 0.75 µl to expel the initial dead volume of the syringe and resulted in no heat release. The following injections were conducted in sets of five repetitions, and in each subsequent set, the injected volume was doubled, starting at 1.5 µl (5 x 1.5 µl, 5 x 3 µl, 5 x 6 µl, etc.). The titration was stopped when the heat of reaction was no longer detected and only the heat of dilution was observed. The calorimetric cell, stirred at 394 rpm, initially contained 1.37 ml of 2.0 mM vesicles and 0.2 to 20 µM annexin V. As CaCl2 was injected into the sample cell, the rate of heat release was measured against a reference cell containing water, whereas the system remained at 25 °C. The raw data were integrated with Origin 7.0 to obtain the total evolved heat per injection.

In control experiments, CaCl2 was titrated into a solution of vesicles and buffer only, which generated small, linear, nonsaturable exothermic heat responses. This control, accounting for heats of calcium dilution and interaction with vesicles, was subtracted from the experimental titrations that included both protein and vesicles in the calorimetric cell. The molar {Delta}H was calculated from the cumulative heat at the plateau divided by the amount of annexin V in the calorimeter cell.

To test for possible effects of denatured annexin V on the calculated molar enthalpy values, the relative amounts of active and inactive protein were determined by analyzing annexin V-vesicle mixtures by two methods. Gel filtration chromatography (Superose 12 resin) indicated that only 1% of the annexin V did not bind to vesicles. Centrifugal ultrafiltration (Centricon 100 concentrator) showed that only 5% of the annexin V passed through the filter.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Comparison of Enthalpy and Free Energy Values as a Function of Membrane Occupancy—To allow direct comparison between {Delta}H and {Delta}G values, isothermal calorimetry titrations were performed with the same reagents used to determine affinity in the fluorescence quenching vesicle titrations. Initial experiments performed at high ratios of protein to membrane were consistent with earlier studies of annexin V from other groups (11, 13). As shown in Fig. 1, calcium titration of 20 µM annexin V and 2 mM vesicles containing 25% PS in a buffer of 0.15 M ionic strength was strongly exothermic and reached a flat and stable end point. Calcium titration of 2 mM vesicles alone was slightly exothermic and showed a linear, non-saturable response, primarily reflecting weak interactions between calcium and membrane phospholipids. The heat released in the vesicle-only titration was taken as the most appropriate blank and was subtracted from the heat released in the vesicle + protein titrations to give the net heat release plotted in the figures.


Figure 2
View larger version (12K):
[in this window]
[in a new window]

  		FIGURE 2.
Calculation of free energy of binding as a function of membrane occupancy. Panel A, experimental data for titration to 100% occupancy. Calcium titrations were performed by fluorescence quenching assay with 1 nM IAF-annexin V-128 and 0.1 µM phospholipid vesicles (25% PS, 2% R-PE, 48% PC, 20% C7-PC). Results are expressed as a percentage of maximum quenching, and are the mean ± S.D. of three separate experiments. Panel B, theoretical curves for stepwise binding of individual annexin V molecules at different levels of membrane occupancy. Data of panel A were used to determine the calcium concentration required to induce binding of an additional annexin V molecule at occupancy levels of 10, 20, 30%, etc. These calcium concentrations were then used as EC50 values to calculate theoretical binding curves from Equation 6 assuming a constant calcium stoichiometry of 8 at all occupancy levels (see "Experimental Procedures" and Ref. 24). Curves from left to right represent individual binding events occurring at occupancy levels of 10, 20, 30, 40, 50, 60, 70, 80, and 90%. Panel C, stepwise and average {Delta}G values as function of membrane occupancy. Stepwise {Delta}G values ({Delta}) for individual binding events occurring at occupancy levels of 10, 20, 30% etc. were calculated with Equations 4 and 5 based on the EC50 values from panel B and a calcium stoichiometry of 8. The {Delta}G value at 1% occupancy is calculated directly from the experimentally determined pKd value given in Table 1. For comparison with {Delta}H values given in Fig. 1, which are cumulative, a cumulative {Delta}G ({square}) is calculated as the average of all preceding stepwise {Delta}G values up to that level of occupancy.

 
Calcium titrations were performed over a wide range of annexin V concentrations to measure the heat response as a function of membrane occupancy (Fig. 1). As expected from earlier work with binding assays, the titration curves became steeper and shifted to the left as the protein:phospholipid ratio was decreased. However, the enthalpy per mole annexin V remained fairly constant, staying in a range of about –35 to –41 kcal/mol over the range from 1 to 100% occupancy (Fig. 1C). The average {Delta}H value is –37.5 ± 0.6 kcal/mol (mean ± S.E., n = 16) over the whole range of membrane occupancy. Low protein concentrations generate small heat signals, which are more susceptible to experimental error and probably explain the higher variability in enthalpy at low membrane occupancy. At the higher end of the occupancy range, there is some uncertainty about the exact point of 100% occupancy under these assay conditions; if there is some free protein left in solution for the titration performed with 20 µM annexin V, this may account for the slightly lower molar enthalpy value obtained at nominal 100% occupancy.

To allow direct comparison between enthalpy and entropy values, the {Delta}G of binding was determined as a function of membrane occupancy. The experimental {Delta}G of binding at 1% occupancy is –52.8 ± l.5 kcal/mol (mean ± S.E., n = 25), based on a pKd value of 38.7 as determined by the calcium titration assay (see Table 1 below). The free energy of binding is about 15 kcal/mol more negative than the enthalpy of binding under the same conditions, indicating a large positive entropy change when annexin V binds to membranes (Fig. 1C).


View this table:
[in this window]
[in a new window]

  		TABLE 1
Binding affinity as a function of phospholipid head group

Calcium titrations were performed by fluorescence quenching assay with 1 nM IAF-annexin V-128 and 10 µM phospholipid (to 1% maximum occupancy) in a buffer of 0.15 M ionic strength. Vesicles contained 25% of the indicated anionic phospholipid, 1% R-PE, 25% C7-PC, and 48% PC. Results are also given for vesicles containing 98% PS and 2% R-PE. Results are given as mean ± S.D. for the indicated number of independent titrations. (The S.D. is larger for the 25% PS titrations because this average includes many different preparations of vesicles, and therefore includes variability introduced by batch to batch variation in exact phospholipid content.)

 
At higher occupancy, it is not possible to measure {Delta}G directly by fitting binding isotherms to a single binding constant due to the negative cooperativity of binding with respect to protein. The {Delta}G value was thus calculated from a model developed by Bazzi and Nelsestuen (24), in which one estimates a stepwise EC50 value at various levels of membrane occupancy for titrations carried out from 0 to 100% occupancy and then calculates the stepwise {Delta}G by assuming a constant calcium stoichiometry at all occupancies (Fig. 2, A and B). As occupancy increases, the calculated {Delta}G value progressively decreases (Fig. 2C), and the difference between enthalpy and free energy becomes smaller as the membrane approaches saturation. Control experiments (see "Experimental Procedures") showed only 1–5% inactive protein in the annexin V preparations used for calorimetry, indicating that the large difference between enthalpy and free energy was not simply due to an inaccurate protein denominator in the calculation of molar enthalpies.

These results were further confirmed by a limited number of experiments with vesicles containing 50% PS in a buffer with 0.05 M Na+. Over the range from 10 to 100% occupancy, the molar enthalpy was nearly constant, averaging –46 ± 2.0 (mean ± S.D., n = 3) kcal/mol annexin V. The corresponding experimental {Delta}G value was –63 kcal/mol (based on a pKd value of 46 at 1% occupancy) (23), again much larger than the enthalpy of binding under these conditions. The enthalpy and free energy of binding are both higher than the values obtained with 25% PS vesicles at 0.15 M ionic strength, consistent with higher affinity binding previously observed as sodium concentration decreases and membrane PS content increases (22). The calcium stoichiometry calculated from the linear portion of the isothermal calorimetry titration curve was about 12 mol Ca2+/ mol annexin V, consistent with earlier results showing calcium stoichiometries from 11 to 13 under conditions that maximally favor binding (13, 23, 29).


Figure 3
View larger version (15K):
[in this window]
[in a new window]

  		FIGURE 3.
Summary of enthalpy and free energy of binding as a function of membrane PS content. Enthalpy, •; free energy, {blacksquare}. Enthalpy value for 18% PS is from Ref. 11. Free energy value for 15% PS is from Ref. 22 and for 50% PS from Ref. 23; other enthalpy and free energy values were measured in this study.

 


Figure 4
View larger version (14K):
[in this window]
[in a new window]

  		FIGURE 4.
Effect of urea on free energy of binding. Calcium titrations were performed at 25 °C with 1 nM IAF-annexin V-128 and 10 µM vesicles containing 25% PS in a buffer of 50 mM HEPES-Na, pH 7.4, 100 mM NaCl, 3 mM NaN3 plus the indicated concentration of urea. Plotted values are pKd ({square}) and {Delta}{Delta}G({Delta}), calculated as the difference in free energy of binding compared with the value observed in the absence of urea. Results are mean ± S.D. of five or more separate assays at each concentration of urea.

 
Fig. 3 summarizes the enthalpy and free energy values determined in this study and previous studies. It is evident that free energy consistently exceeds enthalpy when one takes into account the differences in vesicle PS content and buffer salt concentrations. The data suggest that the change in the free energy of binding as a function of membrane PS content is due primarily to changes in enthalpy; the entropic contribution to binding appears roughly constant throughout the range of PS values. The increase in overall binding affinity as PS content increases can be explained by an increase in the number of high affinity calcium binding sites that come into play; this number is about eight for membranes containing 25% PS, and reaches a maximum of about 12 when the membrane consists of essentially pure PS (see value for 98% PS in Table 1 below).

Effect of Urea on Binding Affinity—To evaluate the possible role of hydrophobic effects in causing entropy changes during the annexin-membrane binding reaction, calcium titrations were performed in the presence of urea. Increasing urea caused a progressive decline in binding affinity that then plateaued in the range from 4 to 5 M urea (Fig. 4). The binding reaction was abruptly abolished above 5 M urea, consistent with denaturation of annexin V previously shown to occur above 5 M urea (30). The average pKd value at 4 and 5 M urea corresponds to a reduction in the overall {Delta}G of binding of about 17 kcal/mol compared with the {Delta}G value obtained in the absence of urea. Thus, urea decreased binding free energy by an amount approximately equal to the calculated entropy change (15 kcal/mol) in the overall binding reaction.


Figure 5
View larger version (12K):
[in this window]
[in a new window]

  		FIGURE 5.
Effect of protein surface charge on binding affinity. Single amino acid residues on the membrane binding face of the protein were mutated to give different changes in net charge: Lys or Arg to Glu (–2), Lys or Arg to Gln (–1), Asp to Asn (+1), or Asp to Lys (+2). Binding affinity was measured with the RBC-calcium titration assay. The line connects the mean pKd value for each group of mutants. The R2 value is the correlation coefficient of a linear regression fit of pKd to charge change. The mutants used are listed in Table S1.

 
Effect of Varying Protein Surface Charge and Interfacial Hydrophobicity—To study the role of surface charge and hydrophobicity in regulating membrane binding affinity, 12 charged residues were selected that are spread across all four domains on the membrane binding face of the molecule (mutated residues are listed in supplemental Table S1). Using the crystal structures of human and rat annexin V (14, 31), we chose Lys and Arg residues that were located on the same face of the protein as calcium ions, and Asp residues that were similarly situated, provided they did not participate directly in known calcium binding sites. Each residue was mutated either to a neutral residue (Gln or Asn) or to a residue of the opposite charge; this allowed a change in charge of +2 to –2 relative to unmodified protein.

As shown in Fig. 5, changes in binding affinity were small to zero for most of these mutants, and there was no overall correlation (r2 = 0.06) between the net charge change and the change in binding affinity. Notably, the two mutants in which a negatively charged Asp was changed to a Lys both had slightly decreased binding affinity relative to wild-type protein. The opposite result would be expected if positively charged residues on the protein surface were attracted to negatively charged phosphate groups on the membrane. Thus, there was no evidence that general charge-charge interactions were important in regulating binding affinity.

However, a different trend began to emerge when we correlated the change in binding affinity with the change in interfacial hydrophobicity for pairs of mutants in which either Glu or Gln was present at a given amino acid position. Because a carboxyl to amide mutation is isosteric, these mutants should only reflect changes due to charge and/or hydrophobicity. As shown in Fig. 6, in 7 of 10 cases, the mutant with the neutral glutamine residue had a higher binding affinity than the mutant with the negatively charged glutamate residue. For the whole set of 10 mutant pairs, the average {Delta}{Delta}G value was 1.60 kcal/mol. This agrees closely with the predicted value of 1.44 kcal/mol for a Glu to Gln mutation based on the Wimley-White scale of interfacial hydrophobicity (32). Thus, the presence of neutral residues on the binding surface increases binding affinity relative to results obtained with charged residues in the same position, and the mean change observed was consistent with expectations from the Wimley-White scale.

We also evaluated the role of Trp-187, which has been previously postulated to participate in hydrophobic interactions during membrane binding (15, 16). Results showed a small but consistent decrease in binding affinity for the W187A mutant (Table S1), consistent with earlier data (33). The {Delta}{Delta}G was –2.18 kcal/mol, versus a predicted Wimley-White value of –2.02 for a Trp to Ala mutation (32). Thus, the data for the W187A mutant support the role of interfacial hydrophobicity in binding, but also indicate that the contribution of any single residue to overall binding affinity via this mechanism is quite small.

Constant Free Energy of Binding to Membranes with Different Phospholipid Head Groups—To assess whether structural features of the phospholipid head group could contribute to binding affinity, calcium titration assays were performed with membranes containing 25% of phospholipids with different anionic head groups (Table 1). The free energy of binding was essentially the same for all phospholipids tested, with pKd values around 40. Consistent with earlier work, the EC50 values were substantially lower for cardiolipin compared with PS or phosphatidylglycerol. However, the lower EC50 value was offset by a compensating decrease in slope, leading to a calculated pKd value (cf. Equation 4) that was the same as for PS. The simplest phospholipid tested, phosphatidylmethanol, which contains a single methyl group esterified to the phosphate, was essentially the same as all the others. This indicated that structural features of the phospholipid head group, aside from the phosphate group itself, were not involved in determining binding affinity under low-occupancy conditions.


Figure 6
View larger version (12K):
[in this window]
[in a new window]

  		FIGURE 6.
Affinity difference between pairs of glutamate and glutamine mutants across the membrane binding face of annexin V. At 10 different amino acid positions, mutants were prepared that contained either a glutamine or a glutamate residue. Binding affinities were measured with the RBC-calcium titration assay. The {Delta}pK value plotted in the figure is the difference between the pKd of the glutamine and the glutamate mutants at each amino acid position. The mutants used are listed in supplemental Table S1.

 


Figure 7
View larger version (15K):
[in this window]
[in a new window]

  		FIGURE 7.
Measurement of inter-annexin FRET as a function of membrane occupancy. A, predicted effects of random versus clustered annexin binding at low and high membrane occupancy. Protein labeled with donor, {blacktriangleup}, or acceptor, {circ}, fluorophores are shown. B, assays with wild-type protein. IAF-annexin V-128 (0.5 nM) and Alexa Fluor 546-annexin V-128 (2.5 nM) were combined with various concentrations of phospholipid vesicles, and the fluorescence intensity at 535 nm was measured 5 min later. Results are expressed as quenching of fluorescence relative to the level obtained in the absence of phospholipid vesicles. Dotted lines, vesicles containing 25% PS, with vesicles added first ({square}) or vesicles added last ({circ}). Solid lines, vesicles containing 100% PS, with vesicles added first ({square}) or vesicles added last ({circ}). C, assays with mutant proteins. Assays were performed with vesicles containing 100% PS as in A with donor-acceptor pairs of mutant proteins annexin V-129 (W187A mutant; {square}) and annexin V-247 (R25Q, K29Q mutant; {circ}). Vesicles were added last.

 
Lack of Multimer Formation at Low Occupancy—To assess whether multimer formation contributes to binding affinity at low occupancy, we used FRET to detect the formation of annexin multimers (13, 34). If annexin V always binds as part of a trimer or higher-order structure, FRET should be high at low occupancy and should stay high as occupancy increases (Fig. 7A). Initial results seemed consistent with this model for experiments performed on vesicles containing 25% PS (Fig. 7B, dashed lines). However, we began to notice inconsistencies in results depending on the order of addition of reagents (Fig. 7B, squares versus circles) as well as the incubation time and the concentration of calcium used. We therefore considered an alternative explanation: annexin V might bind to pre-existing clusters of PS in mixed PS/PC vesicles. Calcium is known to promote the segregation and clustering of PS in mixed vesicles (35, 36). This would make the clustering of annexin V seem more pronounced when vesicles were added to the reaction mixture first, or when there was insufficient time for PS clusters, formed in the absence of annexin V, to relax to a new equilibrium distribution following the addition of annexin V.

We therefore repeated the FRET assays with vesicles consisting of 100% PS (Fig. 7B, solid lines), in which segregation and clustering of PS and PC could not occur. With these vesicles, there was no FRET observed at low occupancy, and no effect of the order of reagent addition. However, the level of FRET at high occupancy was the same as for the 25% PS vesicles, demonstrating that FRET could still occur to the same degree when annexin molecules are close together due to crowding. Control experiments showed that the total amount of protein bound to the vesicles was constant at all concentrations of phospholipid added. We concluded that annexin V clusters do not form as a result of favorable protein-protein interactions at low occupancy.


Figure 8
View larger version (11K):
[in this window]
[in a new window]

  		FIGURE 8.
Schematic view of the role of the hydrophobic effect in annexin-membrane binding. The A and B helices of a single domain are shown schematically either free in solution (left) or bound to the membrane (right). In solution, there is ordered water clustered around the more hydrophobic residues on the protein surface. When free calcium concentration reaches a sufficient level, calcium ions are taken up to link the AB and B helix calcium binding sites to membrane phosphate groups, and ordered water is released as portions of the annexin molecule enter the more hydrophobic interfacial region. Dimensions of the interfacial region are from Ref. 52. The diagram is schematic; molecular components are not drawn to a uniform scale.

 
We also performed assays with proteins with mutations at sites previously identified as being located at trimer interfaces in crystal structures. Arg-25 and Lys-29 are part of a basic cluster at a key trimer interface, and mutation of these residues disrupts salt bridges between monomers in the trimer unit (37). However, mutation of these residues to neutral glutamine had no effect on the curve of FRET as a function of occupancy, which was identical to that of the wild-type protein (Fig. 7C). Similarly, Trp-187 is at another protein-protein interface in the crystal structure (16). The W187A mutant also showed no change in its curve of FRET as a function of occupancy (Fig. 7C). Thus, the data were consistent with lack of multimer formation on 100% PS vesicles at low occupancy for both wild-type and mutant proteins.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
This study provides a comprehensive quantitative picture of the forces promoting binding of annexin V to membranes. Measurements of the enthalpy and free energy of binding have allowed us to assess for the first time the quantitative contribution of all factors previously postulated to determine the overall affinity of annexin-membrane binding. These studies have led to several novel observations that alter our picture of the forces promoting binding of annexin V to membranes. We summarize these observations and their implications as follows.

Large Positive Entropy Change upon Membrane Binding—One surprising conclusion is the strongly positive entropy change that occurs upon binding, particularly at lower levels of membrane occupancy (Figs. 1 and 3). Although the net entropy increase becomes smaller as membrane occupancy increases, the difference persists up to nearly complete saturation of the membrane. This was unexpected based on earlier analyses, which have generally implied that the net entropy change should be negative for several reasons. Multiple studies have shown that membrane phospholipids are clustered (38, 39) and/or immobilized (40, 41) upon binding of annexins, and the membrane becomes more rigid (42), all factors that should cause a reduction in entropy. Likewise, the formation of ordered multimeric complexes of annexin V on the vesicle surface (13, 18–21) should also reduce entropy. The protein also loses translational and rotational freedom upon binding to the membrane, and calcium binding loops become much less mobile (43, 44). However, all these factors together are clearly not enough to counteract a large positive source of entropy in the binding reaction.

Role of Surface Hydrophobicity—What is the main source of the large increase in net entropy when annexin V binds to relatively empty membranes? One likely candidate is the dehydration of substantial areas of the protein surface upon transfer from the bulk aqueous environment to the relatively non-polar interfacial region. This would release ordered water molecules clustered around these residues in solution, resulting in an increase of entropy according to the usual concept of the hydrophobic effect. This concept is shown schematically in Fig. 8, which shows the changes postulated to occur when a portion of a single annexin domain binds to the membrane. Interestingly, one of the very first papers in the annexin field (45) presaged the role of calcium-dependent hydrophobic binding by using calcium-dependent hydrophobic binding to phenothiazine-Sepharose as a key step in the purification of annexins.

There is experimental evidence that annexin V causes dehydration of the vesicle surface upon binding (46), and spectroscopic data indicate that the membrane binding face of the protein enters a more hydrophobic environment upon binding (15). The total membrane surface area covered by a single annexin V molecule is roughly 3200 Å2, based on a coverage of about 50 phospholipid head groups on the external face of the bilayer (47) and an area per phospholipid head group of about 65 Å2 (10). Considering that exposure of hydrophobic regions of a protein to water is unfavorable by roughly 25 cal/Å2 (48), one can see that dehydration of only about 600 Å2 of hydrophobic protein surface area would be needed to explain a free energy difference of 15 kcal/mol. Dehydration of the membrane surface probably also contributes to the effect, which would further reduce the amount of hydrophobic protein surface area required to explain the overall effect.

The experimental results with urea (Fig. 4) also provide support for the hypothesis that hydrophobic effects are important in promoting annexin-membrane binding. Although studies continue to probe the detailed mechanisms by which urea promotes the solvation of hydrophobic molecules (e.g. Refs. 49–51), it is reasonable to use the observed effect of urea as one indicator of the importance of hydrophobic interactions in promoting annexin-membrane binding.

Contribution of Individual Amino Acid Residues to Interfacial Hydrophobicity—The mutagenesis data confirm that interfacial hydrophobicity modulates binding affinity in a manner generally consistent with the predictions of the Wimley-White scale. This effect is general rather than residue-specific. The Wimley-White scale (32) provides some means to estimate how many amino acid residues might be needed to explain a free energy difference of 15 kcal/mol. In this scale, Trp is the most hydrophobic, at 1.8 kcal/mol, followed by Phe (1.1), Tyr (0.9), Leu (0.6), and Ile (0.3). These values reflect transfer to the interfacial region rather than the hydrophobic core of the membrane. Thus, contributions from many hydrophobic residues would be required to explain the observed effect, no single amino acid could explain such a large difference. For example, Trp-187 has been implicated in binding to membranes (15, 16), but this residue alone could account for only about 10% of the observed effect based on a predicted value of 1.8 kcal/mol for interfacial hydrophobicity. Our experimental data confirm the modest contribution of Trp-187 (see "Results"). Considering that the depth of the interfacial region is about 15 Å (52), whereas the dimension of annexin V normal to the plane of the bilayer is about 30–35 Å (31), there is considerable scope for a sizable fraction of the protein surface area to interact with components of the interfacial region (see Fig. 8). In conclusion, although the free energy change for any one residue is small (<2 kcal/mol), the overall effect of transferring tens of residues to the interfacial region would be substantial and could account for the 10–20 kcal/mol difference between enthalpy and free energy under lower occupancy conditions.

The mutagenesis results also show that nonspecific electrostatic attraction between positively charged protein residues and negatively charged membrane phospholipids makes no contribution to the overall affinity of binding. We included in our study the four highly conserved Lys residues 29, 101, 186, and 260, which are all located near the tip of the AB loop in their respective domains. These residues are highly conserved between different annexins and across species (2). Nevertheless, our results show that they make almost no contribution to binding affinity. Thus, it appears that they have some other function besides regulating binding affinity.

Enthalpy and Entropy as a Function of Membrane Occupancy—The calculated difference between free energy and enthalpy decreases at higher membrane occupancy (Fig. 1C). This indicates that the large positive entropy change seen at low occupancy is increasingly counteracted by negative entropy changes, most likely due to clustering and immobilization of phospholipids, and increased ordering of protein as the membrane becomes more crowded. These effects are likely to increase nonlinearly with occupancy. For example, experimental studies with fluorescence photobleaching show that membrane phospholipids are very mobile in the presence of low levels of annexin IV, but phospholipid mobility drops precipitously once a threshold concentration of protein is reached (40). Thus, whereas the hydrophobic effect is probably constant on a molar basis at different levels of occupancy, it is gradually offset by other factors, leading to a smaller net entropy increase as occupancy increases. However, at present we are unable to determine the entropic change due to immobilization of protein or lipid per se, so we cannot yet provide a quantitative model of this process.

Given the near constancy of {Delta}H with occupancy, it is reasonable to speculate that the enthalpy change largely reflects the formation of protein-calcium-phospholipid chelates, which is expected to occur with roughly constant stoichiometry throughout the occupancy range. The observed {Delta}H value of –37.6 kcal/mol obtained with 25% PS vesicles would correspond to an average enthalpy value of –9.4 kcal/mol for each of the four AB helix + B helix sites (each with two calcium ions) formed upon protein-membrane binding under these conditions. Our previous work with mutational analysis shows that the corresponding average free energy contribution for each AB + B site is –10.1 kcal/mol measured under the same conditions (23), nearly the same as the enthalpy value. Thus, the free energy change due to formation of calcium chelates during the binding reaction can be largely explained by the heat released.

Lack of Phospholipid Head Group Specificity—Another surprising observation was the complete lack of head group specificity in the overall free energy change of the binding reaction. Our results clearly show that the free energy of binding is unaffected by the chemical structure of the phospholipid head group, suggesting that only the phosphate group is critical for high affinity binding. At first glance this may appear inconsistent with a large body of data that shows differences in apparent binding affinity of annexin V for different phospholipids (e.g. Refs. 10, 53, and 54). However, closer inspection of these studies shows that comparisons were almost always made on the basis of EC50 values in calcium titration assays. As previously discussed (22, 23), in a highly cooperative system, the EC50 value is insufficient as a sole measure of affinity, one must also take into account the calcium stoichiometry of the reaction. For example, our results show that the lower EC50 value for cardiolipin is compensated by a decrease in the slope of the titration curve, resulting in the same pKd value for the overall reaction. This may be rationalized by considering that cardiolipin vesicles, which have a higher concentration of fixed negative charges due to the two negatively charged phosphate groups on each molecule, will have a higher concentration of reaction intermediates with calcium pre-bound to phosphate groups at any given calcium concentration; thus, the experimentally observable calcium titration curve will start to rise at a lower calcium concentration for cardiolipin vesicles compared with PS vesicles. However, the final complex formed is the same, i.e. a complex of annexin V coordinating about 8 calcium and 8 phosphate groups (for vesicles containing 25% anionic phospholipid). Thus, the net free energy change is the same for the total reaction leading from unbound, calcium-free protein and unbound, calcium-free phospholipid to the fully formed protein-calcium-phospholipid complex. This can be seen by breaking down the overall binding reaction into several component steps, as shown in Scheme 1.

Formula 6(Step 1)

Formula 6(Step 2)

Formula 6(Step 3)

Formula 10Scheme 1

Formula 10

Most methods used to measure binding of annexins to membranes directly measure only step 3 above, and thus will not directly reveal differences in the concentrations of intermediate complexes in steps 1 and 2.

Lack of Multimer Formation at Low Occupancy—The FRET experiments show that annexin V multimers do not form under conditions of low membrane occupancy on vesicles containing 100% PS, despite the extremely high affinity of binding observed with these vesicles. At low membrane occupancy, there is essentially no FRET from one protein molecule to another (Fig. 7), indicating that individual protein molecules are far apart on the surface of the vesicles (with a diameter of about 110 nm). If protein-protein interactions were a driving force in promoting the binding of annexin V to membranes, one would expect to see substantial levels of multimer formation and high levels of FRET even when there are relatively few protein molecules on the vesicle surface. As the membrane becomes more crowded, FRET does occur to an increasing degree, but this can be simply explained by the reduction of average intermolecular distance as the membrane becomes more crowded.

We propose an alternative explanation for the highly ordered patterns observed at high levels of membrane occupancy: formation of regular arrays of annexin V molecules on the membrane surface is best viewed as a process that minimizes unfavorable interactions, rather than maximizing favorable interactions. Formation of annexin clusters at lower occupancy levels on mixed PS/PC vesicles most likely reflects preferential binding to PS-rich clusters that provide higher affinity binding sites. Calcium promotes the clustering of PS in PS/PC mixed membranes (35, 36), and this phenomenon occurs in the calcium concentration range that is typically used to assess vesicle binding and multimer formation. Thus, under these conditions the observed FRET is a reflection of calcium-induced PS clustering rather than clustering induced by protein-protein interaction. A very similar model has been recently proposed to explain the high affinity binding of annexin I to PS/PC membranes (55).

Overall Model of Binding—We summarize our current overall model of the binding reaction as follows (Table 2). The formation of protein-calcium-phosphate chelates is the single most important factor, with a free energy contribution of about –10 kcal/mol per combined AB helix + B helix site (with two calcium ions) in a single domain, or a total contribution of about –38 to –40 kcal/mol for all four domains (with eight calcium ions) for binding to membranes containing about 25% PS. Under conditions very favorable to binding, such as membranes containing nearly 100% PS, up to four additional weaker calcium binding sites come into play (presumably in the four DE helices), affinity increases correspondingly, and total calcium uptake increases to about 12.


View this table:
[in this window]
[in a new window]

  		TABLE 2
Summary of factors contributing to membrane binding affinity

 
The contribution of interfacial hydrophobicity can be estimated as about –15 kcal/mol. It is likely that the contributions of calcium binding sites and interfacial hydrophobicity are essentially unaltered as a function of membrane occupancy (the experimental data with calorimetry supports this view). However, as membrane occupancy increases, the overall reaction becomes less favorable due to the other unfavorable entropic factors discussed above, which increase nonlinearly as the membrane becomes more crowded. Our results effectively rule out a role for the other listed factors in determining affinity, at least under low occupancy conditions.

Practical Implications and Conclusion—Recently, we showed that chemical modification of lysine residues on annexin V with a variety of charged agents used to produce probes for diagnostic imaging decreased binding affinity to a degree that in vivo imaging sensitivity was compromised (56). This study may provide an explanation for these findings: if the bulky charged groups become attached to lysine residues on the membrane binding face of the protein, this may decrease the overall hydrophobicity of the membrane-binding face, thereby decreasing the large favorable entropic contribution to binding seen for the unmodified protein.

In conclusion, binding of annexin V to membranes is dominated by two critical factors: formation of protein-calcium-phosphate chelates, supplemented by dehydration of the protein and membrane surfaces as they combine in the interfacial region of the bilayer. This study provides a comprehensive framework for analyzing the interactions of annexins with membranes, and will facilitate the design of improved annexin V derivatives for diagnostic and therapeutic applications.


   FOOTNOTES
 
* This work was supported by United States Public Health Service Grant CA-102348. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

Formula The on-line version of this article (available at http://www.jbc.org) contains supplemental Table S1. Back

1 To whom correspondence should be addressed: Dept. of Laboratory Medicine, University of Washington, Box 357110, Seattle, WA 98195-7110. Tel.: 206-598-6131; Fax: 206-598-6189; E-mail: tait{at}u.washington.edu.

2 The abbreviations used are: PS, phosphatidylserine; C7-PC, diheptanoyl-phosphatidylcholine; EC50, calcium concentration causing 50% of maximum protein-membrane binding; FRET, fluorescence resonance energy transfer; IAF, iodoacetamidofluorescein; PC, phosphatidylcholine; R-PE, rhodamine-labeled phosphatidylethanolamine; pKd, negative logarithm of equilibrium dissociation constant Kd; RBC, red blood cells. Back


   ACKNOWLEDGMENTS
 
We thank Dr. Barry Stoddard for generously allowing use of the ITC apparatus and Dr. Andy Hoofnagle for helpful comments on the manuscript.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Gerke, V., and Moss, S. E. (2002) Physiol. Rev. 82, 331–371[Abstract/Free Full Text]
  2. Moss, S. E., and Morgan, R. O. (2004) Genome Biol. 5, 219[CrossRef][Medline] [Order article via Infotrieve]
  3. Gerke, V., Creutz, C. E., and Moss, S. E. (2005) Nat. Rev. Mol. Cell Biol. 6, 449–461[CrossRef][Medline] [Order article via Infotrieve]
  4. Rand, J. H., and Wu, X. X. (2004) Thromb. Res. 114, 383–389[CrossRef][Medline] [Order article via Infotrieve]
  5. Cesarman-Maus, G., Rios-Luna, N. P., Deora, A. B., Huang, B., Villa, R., Cravioto Mdel, C., Alarcon-Segovia, D., Sanchez-Guerrero, J., and Hajjar, K. A. (2006) Blood 107, 4375–4382[Abstract/Free Full Text]
  6. Blankenberg, F. G., Katsikis, P. D., Tait, J. F., Davis, R. E., Naumovski, L., Ohtsuki, K., Kopiwoda, S., Abrams, M. J., Darkes, M., Robbins, R. C., Maecker, H. T., and Strauss, H. W. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 6349–6354[Abstract/Free Full Text]
  7. Lahorte, C. M., Vanderheyden, J. L., Steinmetz, N., Van de Wiele, C., Dierckx, R. A., and Slegers, G. (2004) Eur. J. Nucl. Med. Mol. Imaging 31, 887–919[CrossRef][Medline] [Order article via Infotrieve]
  8. Tait, J. F., Gibson, D., and Fujikawa, K. (1989) J. Biol. Chem. 264, 7944–7949[Abstract/Free Full Text]
  9. Thiagarajan, P., and Tait, J. F. (1990) J. Biol. Chem. 265, 17420–17423[Abstract/Free Full Text]
  10. Andree, H. A., Reutelingsperger, C. P., Hauptmann, R., Hemker, H. C., Hermens, W. T., and Willems, G. M. (1990) J. Biol. Chem. 265, 4923–4928[Abstract/Free Full Text]
  11. Plager, D. A., and Nelsestuen, G. L. (1994) Biochemistry 33, 13239–13249[CrossRef][Medline] [Order article via Infotrieve]
  12. Patel, D. R., Jao, C. C., Mailliard, W. S., Isas, J. M., Langen, R., and Haigler, H. T. (2001) Biochemistry 40, 7054–7060[CrossRef][Medline] [Order article via Infotrieve]
  13. Patel, D. R., Isas, J. M., Ladokhin, A. S., Jao, C. C., Kim, Y. E., Kirsch, T., Langen, R., and Haigler, H. T. (2005) Biochemistry 44, 2833–2844[CrossRef][Medline] [Order article via Infotrieve]
  14. Swairjo, M. A., Concha, N. O., Kaetzel, M. A., Dedman, J. R., and Seaton, B. A. (1995) Nat. Struct. Biol. 2, 968–974[CrossRef][Medline] [Order article via Infotrieve]
  15. Meers, P., and Mealy, T. (1993) Biochemistry 32, 5411–5418[CrossRef][Medline] [Order article via Infotrieve]
  16. Concha, N. O., Head, J. F., Kaetzel, M. A., Dedman, J. R., and Seaton, B. A. (1993) Science 261, 1321–1324[Abstract/Free Full Text]
  17. Meers, P., and Mealy, T. (1994) Biochemistry 33, 5829–5837[CrossRef][Medline] [Order article via Infotrieve]
  18. Concha, N. O., Head, J. F., Kaetzel, M. A., Dedman, J. R., and Seaton, B. A. (1992) FEBS Lett. 314, 159–162[CrossRef][Medline] [Order article via Infotrieve]
  19. Voges, D., Berendes, R., Burger, A., Demange, P., Baumeister, W., and Huber, R. (1994) J. Mol. Biol. 238, 199–213[CrossRef][Medline] [Order article via Infotrieve]
  20. Langen, R., Isas, J. M., Luecke, H., Haigler, H. T., and Hubbell, W. L. (1998) J. Biol. Chem. 273, 22453–22457[Abstract/Free Full Text]
  21. Oling, F., Santos, J. S., Govorukhina, N., Mazeres-Dubut, C., Bergsma-Schutter, W., Oostergetel, G., Keegstra, W., Lambert, O., Lewit-Bentley, A., and Brisson, A. (2000) J. Mol. Biol. 304, 561–573[CrossRef][Medline] [Order article via Infotrieve]
  22. Tait, J. F., Gibson, D. F., and Smith, C. (2004) Anal. Biochem. 329, 112–119[CrossRef][Medline] [Order article via Infotrieve]
  23. Jin, M., Smith, C., Hsieh, H. Y., Gibson, D. F., and Tait, J. F. (2004) J. Biol. Chem. 279, 40351–40357[Abstract/Free Full Text]
  24. Bazzi, M. D., and Nelsestuen, G. L. (1991) Biochemistry 30, 7970–7977[CrossRef][Medline] [Order article via Infotrieve]
  25. Bazzi, M. D., and Nelsestuen, G. L. (1993) Cell Signal. 5, 357–365[CrossRef][Medline] [Order article via Infotrieve]
  26. Wood, B. L., Gibson, D. F., and Tait, J. F. (1996) Blood 88, 1873–1880[Abstract/Free Full Text]
  27. Funakoshi, T., Heimark, R. L., Hendrickson, L. E., McMullen, B. A., and Fujikawa, K. (1987) Biochemistry 26, 5572–5578[CrossRef][Medline] [Order article via Infotrieve]
  28. Tait, J. F., and Smith, C. (1999) J. Biol. Chem. 274, 3048–3054[Abstract/Free Full Text]
  29. Evans, T. C., Jr., and Nelsestuen, G. L. (1994) Biochemistry 33, 13231–13238[CrossRef][Medline] [Order article via Infotrieve]
  30. Budisa, N., Huber, R., Golbik, R., Minks, C., Weyher, E., and Moroder, L. (1998) Eur. J. Biochem. 253, 1–9[Medline] [Order article via Infotrieve]
  31. Huber, R., Berendes, R., Burger, A., Schneider, M., Karshikov, A., Luecke, H., Romisch, J., and Paques, E. (1992) J. Mol. Biol. 223, 683–704[CrossRef][Medline] [Order article via Infotrieve]
  32. Wimley, W. C., and White, S. H. (1996) Nat. Struct. Biol. 3, 842–848[CrossRef][Medline] [Order article via Infotrieve]
  33. Campos, B., Mo, Y. D., Mealy, T. R., Li, C. W., Swairjo, M. A., Balch, C., Head, J. F., Retzinger, G., Dedman, J. R., and Seaton, B. A. (1998) Biochemistry 37, 8004–8010[CrossRef][Medline] [Order article via Infotrieve]
  34. Ladokhin, A. S., and Haigler, H. T. (2005) Biochemistry 44, 3402–3409[CrossRef][Medline] [Order article via Infotrieve]
  35. Haverstick, D. M., and Glaser, M. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 4475–4479[Abstract/Free Full Text]
  36. Menke, M., Gerke, V., and Steinem, C. (2005) Biochemistry 44, 15296–15303[CrossRef][Medline] [Order article via Infotrieve]
  37. Mo, Y., Campos, B., Mealy, T. R., Commodore, L., Head, J. F., Dedman, J. R., and Seaton, B. A. (2003) J. Biol. Chem. 278, 2437–2443[Abstract/Free Full Text]
  38. Bazzi, M. D., and Nelsestuen, G. L. (1991) Biochemistry 30, 7961–7969[CrossRef][Medline] [Order article via Infotrieve]
  39. Junker, M., and Creutz, C. E. (1993) Biochemistry 32, 9968–9974[CrossRef][Medline] [Order article via Infotrieve]
  40. Gilmanshin, R., Creutz, C. E., and Tamm, L. K. (1994) Biochemistry 33, 8225–8232[CrossRef][Medline] [Order article via Infotrieve]
  41. Cézanne, L., Lopez, A., Loste, F., Parnaud, G., Saurel, O., Demange, P., and Tocanne, J.-F. (1999) Biochemistry 38, 2779–2786[CrossRef][Medline] [Order article via Infotrieve]
  42. Megli, F. M., Selvaggi, M., Liemann, S., Quagliariello, E., and Huber, R. (1998) Biochemistry 37, 10540–10546[CrossRef][Medline] [Order article via Infotrieve]
  43. Isas, J. M., Langen, R., Hubbell, W. L., and Haigler, H. T. (2004) J. Biol. Chem. 279, 32492–32498[Abstract/Free Full Text]
  44. Isas, J. M., Kim, Y. E., Jao, C. C., Hegde, P. B., Haigler, H. T., and Langen, R. (2005) Biochemistry 44, 16435–16444[CrossRef][Medline] [Order article via Infotrieve]
  45. Moore, P. B., and Dedman, J. R. (1982) J. Biol. Chem. 257, 9663–9667[Abstract/Free Full Text]
  46. Kohler, G., Hering, U., Zschörnig, O., and Arnold, K. (1997) Biochemistry 36, 8189–8194[CrossRef][Medline] [Order article via Infotrieve]
  47. Tait, J. F., and Gibson, D. (1992) Arch. Biochem. Biophys. 298, 187–191[CrossRef][Medline] [Order article via Infotrieve]
  48. Richards, F. M. (1977) Annu. Rev. Biophys. Bioeng. 6, 151–176[CrossRef][Medline] [Order article via Infotrieve]
  49. Lee, M. E., and van der Vegt, N. F. (2006) J. Am. Chem. Soc. 128, 4948–4949[CrossRef][Medline] [Order article via Infotrieve]
  50. Rezus, Y. L., and Bakker, H. J. (2006) Proc. Natl. Acad. Sci. U. S. A. 103, 18417–18420[Abstract/Free Full Text]
  51. Rezus, Y. L., and Bakker, H. J. (2007) Phys. Rev. Lett. 99, 148301[CrossRef][Medline] [Order article via Infotrieve]
  52. White, S. H., Ladokhin, A. S., Jayasinghe, S., and Hristova, K. (2001) J. Biol. Chem. 276, 32395–32398[Free Full Text]
  53. Raynal, P., and Pollard, H. B. (1994) Biochim. Biophys. Acta 1197, 63–93[Medline] [Order article via Infotrieve]
  54. Blackwood, R. A., and Ernst, J. D. (1990) Biochem. J. 266, 195–200[Medline] [Order article via Infotrieve]
  55. Kastl, K., Menke, M., Luthgens, E., Faiss, S., Gerke, V., Janshoff, A., and Steinem, C. (2006) Chembiochem 7, 106–115[CrossRef][Medline] [Order article via Infotrieve]
  56. Tait, J. F., Smith, C., Levashova, Z., Patel, B., Blankenberg, F. G., and Vanderheyden, J. L. (2006) J. Nucl. Med. 47, 1546–1553[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 JNMHome page
J. F. Tait
Imaging of Apoptosis
J. Nucl. Med., October 1, 2008; 49(10): 1573 - 1576.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
283/10/6126    most recent
M707637200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Jeppesen, B.
Right arrow Articles by Tait, J. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Jeppesen, B.
Right arrow Articles by Tait, J. F.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2008 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement