The Mighty Arginine, the Stable Quaternary Amines, the Powerful

Amina S. Woods , Shelley N. Jackson , Ernest K. Lewis , Thomas Egan .... J. Wang, Andrew E. Taggi, Jerrold Meinwald, Roy A. Wise, and Amina S. Woods...
0 downloads 0 Views 382KB Size
Download PDF
The Mighty Arginine, the Stable Quaternary Amines, the Powerful Aromatics, and the Aggressive Phosphate: Their Role in the Noncovalent Minuet Amina S. Woods The National Institute on Drug Abuse, Intramural Research Program, National Institutes of Health, 5500 Nathan Shock Drive, Baltimore, Maryland 21224 Received October 23, 2003

In the age of proteomics, the role of certain amino acid residues and some post-translational modifications in noncovalent complex formation are gaining in importance, as the understanding of interactions between biological molecules, is at the heart of the structure function relationship puzzle. In this work, mass spectrometry is used to highlight ammonium- or guanidinium-aromatic interactions through Cation-π bonds and ammonium- or guanidinium-phosphate interactions through salt bridge formation. Such interactions are crucial factors in certain ligand-receptor interactions and receptorreceptor interactions. In addition, the ability of phosphorylated residues and phosphorylated lipids to form noncovalent complexes with guanidinium and quaternary ammonium (mostly through Coulombic interactions) is demonstrated, and could explain the stability of certain membrane embedded protein, or a possible role for phosphorylation in protein-protein interactions. Dougherty’s work demonstrates cation-π interactions in intra-protein interactions and folding, the present work explores inter-peptide interactions, i.e., the formation of noncovalent complexes between peptides’ epitopes containing adjacent aromatic residues and ones containing adjacent Arg as a model to better understand the role of cation-π complexes in protein-protein interaction. Complexes of peptides containing aromatic residues with quaternary amines as well as the interaction of aromatic compounds, with the guanidinium group of Arg are also investigated. Considering that an inordinate number of therapeutic compounds contain aromatic rings and quaternary amines, the above-described interactions could possibly be of great importance in better understanding their mechanism of action. Keywords: noncovalent interactions • quaternary amines • phosphate moieties

Introduction In the age of proteomics, the study of protein structure has been at the forefront of research, with the goal of better understanding the role of proteins in the biochemistry, physiology, and pathology of the cell, and in finding protein epitopes or whole proteins that could serve as disease markers.1 However, a physiological event often requires a group of proteins to interact as exemplified by G proteins,2 or a ligand to fit in a hydrophobic pocket3 or receptors to communicate directly such as the intramembrane Adenosine A2A receptor-Dopamine D2 receptor interaction resulting in heteromerization between A2AR and D2R. It is very probable that the loss of affinity to dopamine binding of the D2R when the A2AR is activated is due to conformational changes transmitted through the heteromeric interaction.4 To perform its function a protein has to fold properly. In the case of the prion protein, a misfolding results in copies of the protein aggregating and causing the deadly condition of “Mad Cow Disease”.5 Hence, the importance of the various types of interactions leading to intra-protein * To whom correspondence should be addressed. E-mail: awoods@ intra.nida.nih.gov.

478

Journal of Proteome Research 2004, 3, 478-484

Published on Web 03/12/2004

interactions resulting in the correct conformation, as well as inter-protein or protein-lipid interactions, which allow proteins to fit in their environment and perform their assigned function as some of the most important proteins are totally or partially embedded in the cell membrane lipid bilayer.6 The general consensus is that hydrophobic interactions are the main stay of protein-lipid interactions.7,8 However, when one looks at Dougherty’s recent work,9-14 a less acknowledged interaction, “cation-π”, seems to play a crucial role and is gaining in importance at a logarithmic rate. When dwelling on noncovalent interactions, most researchers think of hydrogen bonding, hydrophobic effects or electrostatic interaction. In the past few years, cationic-π bonds and ammonium- or aminesaromatic interactions are starting to command greater attention, as they can be important factors in ligand-receptor interaction or protein folding.9 Most of the drugs of abuse, such as cocaine, amphetamines, and morphine, as well as many therapeutic compounds contain aromatic rings. Tertiary amines such as nicotine and quaternary amines such as acetylcholine and many of its agonists and antagonists have to fit in the fairly hydrophobic pockets of the acetylcholine receptor. Such a fete 10.1021/pr034091l CCC: $27.50

 2004 American Chemical Society

research articles

Arginine, Quaternary Amines, Aromatics, and Phosphate

is only possible through the interaction of the acidic π-cloud of the aromatic residues, lining the active site, with the basic tertiary and quaternary amines of the ligand. Dougherty’s laboratory has published an impressive array of papers that spans the fields of physical-, organic-, and bio-chemistry emphasizing the importance of ammonium and amines interactions with various aromatic organics such as benzenes, indols, pyridines, pyrimidines, and the aromatic amino acid residues phenylalanine, tyrosine, and tryptophan.10-16 Aromatic compounds have a π-cloud on their surface caused by sp2 hybridization. The positive charge on the nitrogen of a quaternary ammonium interacts with the negative delocalized charge on aromatic molecules forming complexes, with a binding enthalpy of 9 kcal/mol for choline and as much as 19 kcal for NH4.17 A study by Sussman et al.’s of ammonium interaction with aromatic molecules, indicates that the NH4+-π interaction has the largest nonelectrostatic interaction portion (∼47%) of the total binding energy, whereas the NH4+aromatic hydrogen interaction has 90% of the total electrostatic allocation. Sussman’s findings put in perspective the role of the nonelectrostatic component in cation-π bonding.17 The NH4+-cation-π portion of the interaction contains a large nonelectrostatic component. Schrader et al. have provided definite proof of cation-π interaction between aromatic amino acid residues or aromatic organic molecules and the guanidinium group of arginine.18 According to Gallivan and Dougherty9 Arg are much more likely than Lys to form cation-π interactions with aromatic residues suggesting that the nonelectrostatic effects predominate, as the side chain of Arg is larger and less solvated than that of Lys. They also deduced that Arg side chains form better van der Waals interactions with aromatic rings while at the same time donating hydrogen bonds and binding to the aromatic ring, whereas Lys usually loses hydrogen bonds to bind to aromatics. They also pointed out that the cation-π interaction is directional as the face not the edge of the aromatic residue or organic molecule must be in contact with the cation. Calculations for several ammonium formate-aromatic systems gave9,11 interaction energy in the range of 2.5-6.2 kcal/mol. The above cited literature shows that peptides containing Arg or biologically active amines such as tertiary and quaternary amines associate with aromatic residues in proteins via cation-π bonding. Most cation-π interactions, involving proteins that were previously studied, were between residues of the same protein. In the present work, peptide-peptide interaction and peptide-quaternary amines interaction are highlighted. Some of the peptides used in this study had phosphorylated Tyr (pY), it was noticed that such peptides formed better noncovalent complexes in both peptide-peptide and peptide-lipid interactions and added more quaternary amines than nonphosphorylated peptides, most likely due to the great negative electrostatic potential of the phosphate group. The phosphate groups on peptides or lipids interact with quaternary ammonium and guanidinium to form noncovalent complexes. In previous work, it was observed that peptide-peptide,19-21 protein-protein,22-24 peptide-excitatory amino acids,25 peptideoligonucleotides,26-29 peptide-small organic molecules31,32 and double stranded DNA32,33 could be detected by matrix assisted laser desorption/ionization (MALDI)19,22-33 or Ion MobilityMALDI,20,31 or AP-MALDI21 if a matrix solution with a higher pH range was used. Lecchi and Pannell32,33 discovered a less acidic matrix (6-aza-2-thiothymine [ATT]) while studying double stranded DNA. The above work demonstrates that ATT (work-

ing solutions pH 4.5-7.0) was used successfully to demonstrate various types of noncovalent complexes.19-21,25,28-33 These same studies illustrate that when R-cyano-4-hydroxy cinnamic acid (CHCA, pH 1-2) was used, noncovalent complexes were not seen. In the present work, cation-π interaction is detected by MALDI using ATT matrix, no complexes were seen with CHCA. However complexes of these same peptides’ mixtures and peptides-small molecules mixtures, were detected by ESI (data not shown). Only peptides containing two or more adjacent aromatic amino acid residues, or a phosphorylated Tyr and phosphorylated lipids formed complexes with peptides containing two or more adjacent Arg. Some of these same peptides were shown to interact with small organic compounds and form stable complexes. In addition the present work highlights the importance of linear epitope in protein-protein and peptide-protein interactions.

Material and Methods a. Peptides: YGGFLRR (Dynorphin 1-7), RRPYIL (Neurotensin 8-13), IRPKLKWNDQ and HCKFWW were purchased from Sigma (Saint Louis, MO). NpYISKGSTFL and pYVPML (p indicates a phosphorylated residue) from Anaspec (Palo Alto, CA), PHPFHFFVYK and YPW, YKW and cyclo-WW from Bachem (King of Prussia, PA), CAHPNDLpYVELPENIPFY, AAAYAAANH2, AAApYAAA-NH2 and AAAFAAA-NH2 were synthesized at the Johns Hopkins School of Medicine Sequencing and Synthesis Laboratory. All peptides were dissolved in milli-Q water to a concentration of 10 picomoles/µL. Peptide mixtures were made by adding equal amounts of the working solutions of YGGFLRR or RRPYIL or IRPKLKWNDQ to HCKFWW or NpYISKGSTFL or pYVPML or PHPFHFFVYK or YPW or AAAYAAA-NH2 or AAApYAAA-NH2 or AAAFAAA-NH2. b. Lipids: Sphingomyelin, cardiolipin and phosphatidylinositol phosphate were purchased from Avanti Polar Lipids (Alabaster, AL) and diluted in chloroform:methanol:water 8:4:3 at concentration of 50 pmoles/µL. c. Quaternary Ammonium Compounds: Acetylcholine chloride (Ach) and hexachlorobenzene (HXB) were purchased from Aldrich (Milwaukee, WI), and were diluted to 1 nanomole/µL. d. Sample solutions: Solutions containing an equal volume of HCKFWW or AAAFAAA-NH2 (10 pmole/µL) and Ach (1 nanomole/µL) were made. Solutions containing an equal volume of hexachlorobenzene (1 nanomole/µL) and YGGFLRR or RRPYIL (10 pmole/µL) were prepared, as well as solutions containing an equal volume of Sphingomyelin, cardiolipin or phosphatidylinositol phosphate (50 pmoles/µL) and YGGFLRR or RRPYIL (10 pmole/µL). e. Matrix: 6-aza-2-thiothymine (ATT) and R-cyano-4hydroxycinnamic acid (CHCA) were purchased from Aldrich (Milwaukee, WI). All matrixes were prepared fresh daily as a saturated solution in 50% ethanol. f. Sample Preparation: A 0.3 µL mixture + 0.3 µL matrix (ATT or CHCA). Complexes were only seen when ATT was used as matrix. ATT forms long spindly crystals that aggregate at the periphery of the dried sample-matrix mixture drop, which makes it more difficult to find a sweet spot. g. Instrument: Mass spectra were acquired in linear mode in both positive and negative ion mode on a DE-PRO MALDI from PE-Biosystems (Framingham, MA), equipped with a nitrogen laser (337 nm) and an extraction voltage of 20 kV. All spectra were the average of 50 shots. Journal of Proteome Research • Vol. 3, No. 3, 2004 479

research articles

Figure 1. Interaction between HCKFWW and YGGFLRR resulted in the formation of a noncovalent complex. In the right-hand corner is the electrostatic model of the peptides involved.

Results and Discussion a. Peptide Containing Adjacent Aromatic Residues: In the present study, the role of cation-π-guanidinium group interaction is highlighted. It plays an important role in peptide-peptide noncovalent complex formation. Only peptides containing a minimum of two adjacent aromatic residues, such as HCKFWW (Phe, Trp, Trp) and peptides containing a minimum of two adjacent Arg (basic residues) such as YGGFLRR and RRPYIL formed noncovalent complexes. No complexes were seen with IRPKLKWNDQ. MALDI Spectra of the peptide mixtures containing equal amounts of the working solutions of HCKFWW and YGGFLRR or RRPYIL respectively, showed the formation of noncovalent complexes between the guanidinium group of the adjacent Arg and the π-cloud of the adjacent aromatic residues Phe, Trp, and Trp at m/z 1775.1 (Figure 1) and 1724.1, respectively. Spectra of the peptide mixtures containing equal amounts of the working solutions of PHPFHFFVYK and YGGFLRR or RRPYIL respectively, showed the formation of noncovalent complexes between the guanidinium group of the adjacent Arg and the π-cloud of the adjacent Phe at m/z 2187.5 and 2136.5, respectively. Peptide mixtures containing equal amounts of the working solutions of cyclo-WW and YGGFLRR or RRPYIL respectively, showed the formation of noncovalent complexes between the guanidinium group of the adjacent Arg and the π-cloud of the adjacent aromatic residues Trp and Trp at m/z 1241.4 and 1190.4, even though the aromatic di-peptide was cyclic. However peptides containing nonadjacent aromatic residues such as YPW and YKW, or one aromatic residue, AAAFAAA-NH2 and AAAYAAA-NH2 did not form a noncovalent complex with either YGGFLRR or RRPYIL. Hence, our results suggest that for peptide-peptide interaction to take place, a minimum of two adjacent aromatic residues and two adjacent Arg are necessary for the complex of the two peptides to be stable enough to be detected by mass spectrometry. The data also implies that the position of single or adjacent aromatic residues in the peptide was irrelevant as to whether complex formation occurred. The relative intensity of the complexes varied from 10 to 40%. Complexes were only seen when ATT (pH 5.4) was used as a matrix, and not when CHCA (pH 1.5) was used. b. Addition of Hexachlorobenzine to Peptides Containing Arginine: To test whether other aromatic compounds would 480

Journal of Proteome Research • Vol. 3, No. 3, 2004

Woods

Figure 2. Interaction between RRPYIL and hexachlorobenzene (HXB) shows that one HXB adds to each Arg guanidinium group. In the right-hand corner the electrostatic model shows that as both HXB and guanidinium are flat their interaction involves stacking.

Figure 3. Interaction between HCKFWW and acetylcholine (Ach) shows that one Ach adds to each aromatic amino acid residue, resulting in three additions. In the right-hand corner the model of the electrostatic interaction is seen.

interact with the guanidinium group of Arginine, hexachlorobenzene was added to the same peptides. The π-cloud on hexachlorobenzene (HXB) proved to be stable and strong enough to form a noncovalent complex with the guanidinium group of one Arg. Peptides YGGFLRR and RRPYIL each added two hexachlorobenzene at m/z 1152.7, 1438.4 and 1101.7, 1387.3 (Figure 2). The addition of hexachlorobenzene was better detected in negative than in positive mode. The relative intensity of the first addition was 60 to 75%, and the second 25-30%, again suggesting that aromatic rings form very stable complexes with the guanidinium group of Arg. However, each complexed hexachlorobenzene easily loses 4 chlorines (two at a time), giving molecular ions at m/z 959.7 and 1245.3 (Figure 2). Unlike the case of peptide-peptide interaction, where the minimum charge of two adjacent aromatic residues is required for complex formation with the two guanidiniums of the basic peptides (the interaction has to be strong enough for the two peptides each containing several noncharged residues to

Arginine, Quaternary Amines, Aromatics, and Phosphate

research articles

Figure 4. a. Interaction between NpYISKGSTFL and RRPYIL, resulted in complex formation. The relative intensity of the complex MH+ is over 60% suggesting a stable and strong interaction. b. Interaction between pYVPML and RRPYIL, resulted in complex formation. In the right-hand corner is the electrostatic model of the peptides involved. c. Interaction between AAApYAAA-NH2 and YGGFLRR, resulted in the formation of a noncovalent complex. In the right-hand corner is the electrostatic model of the peptides involved. d. Interaction of CAHPNDLpYVELPENIPFY with YGGFLRR, resulted in its addition to both the phosphorylated Tyr and the adjacent carboxyl terminal aromatic residues.

maintain their interaction), one guanidinium/hexachlorobenzene was sufficient, as the interaction has to only involve the aromatic ring and the guanidinium group. Gallivant and Dougherty pointed out that the interaction is directional,8 our results suggest that a small molecule like hexachlorobenzene can more easily position itself to interact with the peptides’ Arg side chains in a stacked geometry than two peptides can. c. Addition of Acetylcholine to Peptides Containing Aromatic Residues: The choline of acetylcholine (Ach) is made of an alkylated quaternary amine. A solution containing a 1:1 mixture of a peptide containing three aromatic residues HCKFWW added three Ach, resulting in three different noncovalent complexes at m/z 1052.3 (addition of one Ach), 1198.5 (addition of two Ach), and 1344.7amu (addition of three Ach) (Figure 3), while AAAFAAA-NH2, added just one Ach at amu 736.9. This result shows again that the size of Ach might have played a role in allowing it to position itself favorably to interact in a stacked geometry with the aromatic residues. It is also possible that small organic molecules such as Ach and HXB because of their size and geometry are more likely to maintain the noncovalent interaction with amino acid residues regardless of whether these molecules are positively charged quaternary

amines (Ach) interacting with aromatic residues or negatively charged aromatic molecules such as HXB interacting with the guanidinium group of Arg. d. Peptide Containing Phosphorylated Tyrosine Residues: Three phosphorylated peptides were used. NpYISKGSTFL contains a phosphorylated Tyr at position two and a Phe at position nine, formed noncovalent complexes with both YGGFLRR and RRPYIL at m/z 2078.3 and 2027.3, respectively (Figure 4 a). pYVPML which contains a phosphorylated Tyr at position one and no other aromatic residue interacted with the same two peptides forming noncovalent complexes at m/z 1570.8 and 1519.8, respectively (Figure 4b). AAApYAAANH2, which also contains a phosphorylated Tyr at position four and has an amidated carboxyl terminus, thus ensuring that no other acidic site is available for interaction, was mixed with both basic peptides and formed complexes at m/z 1555.7 (Figure 4c) and 1504.7, respectively. The fourth peptide CAHPNDLpYVELPENIPFY with a phosphorylated Tyr at position 8 and adjacent Phe and Tyr at 17 and 18, formed complexes with YGGFLRR at 3082.4 and 3950.4 (Figure 4d) and at 3031.4 and 3048.4 with RRPYIL. The results suggest, that in the presence of adjacent aromatic residues and a phosphate group, CAHPNDLpYVELPENIPFY added one or two molecules Journal of Proteome Research • Vol. 3, No. 3, 2004 481

research articles

Woods

of the positively charged peptide. The relative intensity of the first addition to peptide four is 100%, whereas that of second is 38-45%. The relative intensities of the complexes peaks, for the previous three peptides were 30 to 70%. The data infers that the increase in complex formation for the addition of one peptide is probably due to the summing of complexes formed with the phosphate only and the ones formed with the adjacent aromatic residues only, on peptide four, while the Relative intensity of the peak for the simultaneous addition of two peptides to peptide four (one at each site) is within the same range as the other phosphorylated peptides. The data indicates that the interaction with phosphorylated Tyr is rather stable and that one phosphate group’s charge is as powerful or surpasses that of two adjacent aromatic or acidic amino acid residues. The pKa of Asp and Glu side chains is 3.9 and 4.3, respectively, whereas that of Tyr is 10.1 and that of a phosphate group is 2.2. Hence, a post-translationally modified Tyr is far more acidic, as the phosphate group adds a negative electrostatic charge to the delocalized electrons of the aromatic ring. No complexes were seen with IRPKLKWNDQ or when CHCA was used as a matrix for all of the above interactions (data not shown). e. Lipids Containing Phosphates: Lipids such as sphingomyelin, phosphatidyl inositol phosphate and cardiolipin contain one, two, and two phosphate groups, respectively and did interact with peptides containing adjacent Arg, but not with peptide containing one Arg. Mixtures were made, by adding equal amounts of YGGFLRR or RRPYIL and sphingomyelin (Sph) or cardiolipin(Crd) or phosphatidylinositol phosphate (PIP). Spectra of all mixtures showed noncovalent complexes of peptide-lipid as seen in Figure 5a where complexes of Sph (containing fatty acid chains varying in length, the major speciesis seen at MH+ 732.1 amu, other species are seen at increments of 28 amu) and RRPYIL at amu 1521.9, 1549.1, and 1631.0. Sodiated peaks are seen at amu 1485.3, 1573.7, and 1653.6. Sph and YGGFLRR formed only one complex with the major speciesof Sph at amu 1600.1. YGGFLRR or RRPYIL when added to PIP show MH+ at amu 1834.5 and 1783.2, respectively. Crd and YGGFLRR or RRPYIL show MH+ at amu 2316.9 and 2265.9, respectively. The relative intensities of the complex peaks varied from 10 to 50%. f. Addition of Acetylcholine to Lipids Containing Phosphate: Mixtures of Sphingomyelin or cardiolipin and Ach were prepared. Sphingomyelin contains two fatty acids. The fatty acid side chains vary in length. Hence, the presence of multiple sphingomyelin species as reflected by the MH+ at amu 732.1, 760.1, 788.1, and 816.1. Each of these species added one Ach forming noncovalent complexes at amu 878.3, 906.3, 934.3, and 961.3, respectively (Figure 5b). Crd contains two phosphates and added as many as four Ach at MH+ 1594.1, 1740.3, 1886.5, and 2032.7. To find out if phosphorylated lipids can compete with phosphorylated peptides for Ach, a mixture of NpYISKGSTFL, crd and Ach was prepared. Three additions of Ach to the peptide were seen at amu 1355.5, 1501.7, and 1647.9. Three additions to crd were also seen at amu 1594.1, 1740.3, and 1886.5(Figure 5c). The relative intensity of the nonvovalent complexes varied from 10 to 25%. The relative intensity of complexes MH+ with two Ach molecules added was around 25%. This finding could imply that the addition of two Ach to the phosphate group enhances the stability of the complex, which might result in a decrease in in-source decay and metastable fragmentation. The result also suggests that lipids 482

Journal of Proteome Research • Vol. 3, No. 3, 2004

Figure 5. a. Interaction between sphingomyelin phosphate and RRPYIL, resulted in complex formation. As Sphingomyelin has fatty acid side chains that vary in length, complex formation with the various speciesare seen. In the right-hand corner is the electrostatic model of the RRPYIL and sphingomyelin. b. One Ach molecule added to each sphingomyelin as only one charge is available on the phosphate. However several peaks are seen because of the heterogeneity of the fatty acid side chains. c. A mixture of Ach, NpYISKGSTFL and cardiolipin, shows that both lipid and peptide can complex with Ach when present in the same solution. Each added as many as 3 Ach molecules.

under these conditions competed favorably with peptides in forming noncovalent complexes with quaternary amines. g. Ab Initio Calculations: Geometry optimization of Acetylcholine, hexachlorobenzene and peptides was carried out at the Hartree-Fock 6-31G** level of theory using Spartan ’02 (32).

research articles

Arginine, Quaternary Amines, Aromatics, and Phosphate

The plots of the electrostatic potential surface of peptides, HCKFWW is seen in Figures 1a, that of RRPYIL in Figures 2, 4b and 5a, that of pYVPML in Figure 4b and that of AAApYAAANH2 in Figure 4c. The electrostatic potential surface of hexachlorobenzene is seen in Figure 2, and that of Acetylcholine in Figure 3. The electrostatic potential surfaces were generated by mapping the 6-31G** electrostatic potentials. Briefly, red colors represent regions of negative potential while blue colors represent areas of positive potential. Orange, yellow, green, and light green represent regions of intermediate potential. Quaternary ammonium and guanidinium groups are blue, aromatic rings can vary in color according to their electronegativity, and carboxyl groups are red. Calculations for each molecule were done separately. Several researchers have used gas phase techniques to better understand solution phase phenomena. Mainly, mass spectrometry was widely utilized in the study of noncovalent interactions involving biological molecules. Recently, the following researchers have implemented mass spectrometry in the study of cation-π interactions. Dunbar et al. performed extensive investigations on the gas-phase interactions of metal ions with aromatic molecules.35-39 Rodriguez-Cruz and coworkers studied gas-phase dissociation reactions of hydrated alkaline earth metal ions with benzene.40 In addition, the binding energies of benzene to Na+ and K+ ions have been determined by CID experiments. Ryzhov and co-workers provided an elegant example of the gas phase determination of metal-ion affinities of the aromatic amino acids,41,42 providing insight into how these constituents might interact in biological systems, as exemplified by peptide-peptide interactions, and peptide small molecules interactions. Pletneva et al.43 used NMR successfully to explore cation-π interactions in model systems between poly-Lys peptides and acidic peptides containing various aromatic residues and Waters et al.44 used circular dichroism to prove that cation-π interactions between aromatic residues and protonated amines stabilized R-helices in water. Their data as well as our previous results, demonstrating that noncovalent complexes were stable enough to withstand enzymatic digests,19,30 confirmed that peptides are well suited for the study of amino acid side chain interactions. In conclusion, the present experimental results using mass spectrometry as an analytical tool and peptide mixtures emphasized and confirmed that the negative charge on the aromatic amino acid side chains and the positive charge on the Arg side chain result in cation-π interactions and that peptide model systems seem to be better suited than large proteins for studying intra- and inter-protein interactions. Peptides as small as trimers and as large as 51 amino acid residues were used to model the interactions. In addition we demonstrated that although adjacent aromatic residues are required to complex with peptides containing adjacent Arg, just one phosphorylated Tyr was needed to form a stable noncovalent complex between two peptides. There may be a positional dependence at work, the phosphate group may be able to project from the R-helical peptide structure to interact, whereas the aromatic residues may be shielded by the secondary structure. However, the short peptides used in the above experiments were random coils, hence we could not necessarily make that claim. The fact that phosphate groups can form noncovalent complexes, suggests that phosphorylation or dephosphorylation could play a role in intra and inter-protein interactions, thus leading to conformational changes that could be relevant to protein function. The formation of noncovalent

complexes between phosphorylated lipids and peptides, suggests that hydrophobic interaction is not the only interaction involved in stabilizing proteins embedded in a lipid bi-layer. Small organic molecules such as hexachlorobenzene and acetylcholine interact with just one aromatic and one Arg residue respectively, implying that in addition to charge the size of the interacting molecules does matter, for complex formation, as peptide interactions need adjacent aromatic residues on one peptide and adjacent Arg on the other. The ab initio calculations gave the electrostatic picture of the peptides involved and supported the conclusions reached through the analytical data. Such experimental approach could be helpful in determining the site of action, as well as the mechanism involved in certain protein-protein interaction as well as therapeutic compounds and protein interactions.

Acknowledgment. The author would like to thank Dr. Susanne Moyer for her invaluable intellectual contribution in shaping this paper, and Dr. Hay-Yan Wang for his help with molecular modeling. References (1) McDonald; W. H.; Yates, J. R. 3rd. Shotgun proteomics and biomarker discovery. Dis. Markers 2002, 18, 99-105. (2) Befort, K.; Tabbara, L.; Kling, D.; Maigret, B.; Kieffer, B. L. Role of aromatic transmembrane residues of the delta-opioid receptor in ligand recognition. J. Biol. Chem. 1996, 271, 10161-10168. (3) Arias, H. R. Topology of ligand binding sites on the nicotinic acetylcholine receptor. Brain Res. Rev. 1997, 25, 133-191. (4) Canals, M.; Marcellino, D.; Fanelli, F.; Ciruela, F.; de Benedetti, P.; Goldberg, S. R.; Fuxe, K.; Agnati, L. F.; Woods, A. S.; Ferre´, S.; Lluis, C.; Bouvier, M.; Franco, R. Adenosine A2A-dopamine D2 receptor-receptor heteromerization. Qualitative and quantitative assessment by fluorescence and bioluminiscence energy transfer. J. Biol. Chem. 2003, 278, 46741-46749. (5) Soto; C.; Saborio, G. P.; Anderes, L. Cyclic amplification of protein misfolding: application to prion-related disorders and beyond. Trends Neurosci. 2002, 25, 390-394. (6) Silverman, B. D. Hydrophobicity of transmembrane proteins: Spatially profiling the distribution. Protein Sci. 2003, 12, 586599. (7) May, S. Protein-induced bilayer deformations: the lipid tilt degree of freedom Eur. Biophys. J. 2000, 29, 17-28. (8) Anderson, R. G.; Jacobson, K. A role for lipid shells in targeting proteins to caveolae, rafts, and other lipid domains. Science 2002, 296, 1821-1825. (9) Gallivan, J. P.; Dougherty, D. A. Cation-π interactions in structural biology. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 9459-9464. (10) Dougherty, D. A. Cation-π interactions in chemistry and biology: a new view of benzene, Phe, Tyr, and Trp. Science 1996, 271, 163-168. (11) Ma, J. C.; Dougherty, D. A. The cation-π interaction. Chem. Rev. 1997, 97, 1303-1324. (12) Zacharias N.; Dougherty D. A. Cation-π interactions in ligand recognition and catalysis. Trends Pharmacol. Sci. 2002, 3, 281287. (13) Zhong, W.; Gallivan, J. P.; Zhang, Y.; Li, L.; Lester, H. A.; Dougherty, D. A. From ab initio quantum mechanics to molecular neurobiology: a cation-π binding site in the nicotinic receptor. Proc. Natl. Acad. Sci. U S A. 1998, 95, 12088-12093. (14) Dougherty, D. A.; Lester, H. A. Neurobiology. Snails, synapses and smokers. Nature 2001, 411, 252-253. (15) Smit, A. B.; Syed, N. I.; Schaap, D.; van Minnen, J.; Klumperman, J.; Kits, K. S.; Lodder, H.; van der Schors, R. C.; van Elk, R.; Sorgedrager, B.; Brejc, K.; Sixma, T. K.; Geraerts, W. P. A gliaderived acetylcholine-binding protein that modulates synaptic transmission. Nature 2001, 411, 261-268. (16) Petersson; E. J.; Choi, A; Dahan, D. S.; Lester, H. A.; Dougherty, D. A. A perturbed pK(a) at the binding site of the nicotinic acetylcholine receptor: implications for nicotine binding. J. Am. Chem. Soc. 2002, 124, 12662-12663. (17) Zhu, W.-L.; Tan, X.-J.; Puah, C. M.; Gu, J.-D.; Jiang, H.-L.; Chen, K.-X.; Felder, C. E.; Silman, I.; Sussman, J. L. How does ammonium interact with aromatic groups? A density functional theory (DFT/B3LYP) investigation. J. Phys. Chem. A 2001, 104, 9573-9580.

Journal of Proteome Research • Vol. 3, No. 3, 2004 483

research articles (18) Rensing, S.; Arendt, M.; Springer, A.; Grawe, T.; Schrader, T. Optimization of a synthetic arginine receptor. Systematic tuning of noncovalent interactions J. Org. Chem. 2001, 66, 5814-5821. (19) Woods, A. S.; Huestis, M. A. A study of peptide-peptide interaction by matrix-assisted laser desorption/ionization, J. Am. Soc. Mass. Spectrom. 2001, 12, 88-96. (20) Woods, A. S.; Koomen, J. M.; Ruotolo, B. T.; Gillig, K. J.; Russel, D. H.; Fuhrer, K.; Gonin, M.; Egan, T. F.; Schultz, J. A. A study of peptide-peptide interactions using MALDI ion mobility o-TOF and ESI mass spectrometry, J. Am. Soc. Mass. Spectrom. 2002, 13, 166-169. (21) Moyer, S. C.; Marzilli, L. A.; Laiko, V. V.; Doroshenko, V. M.; Woods, A. S.; Cotter, R. J. Atmospheric pressure matrix assisted laser desorption/ionization mass spectrometry (AP MALDI) on a quadrupole ion trap mass spectrometer Int. J. Mass. Spec. 2003, 226, 133-150. (22) Woods, A. S.; Buchsbaum, J. C.; Worrall, T. A.; Cotter, R. J.; Berg, J. M. Matrix assisted laser desorption ionization of non-covalently bonded compounds, Anal. Chem. 1995, 67, 4462-4465. (23) Zehl, M.; Allmaier, G. Investigation of sample preparation and instrumental parameters in the matrix-assisted laser desorption/ ionization time-of-flight mass spectrometry of noncovalent peptide/peptide complexes. Rapid Commun. Mass Spectrom. 2003, 17, 1931-1940. (24) Farmer, T. B.; Caprioli, R. M. Determination of protein-protein interactions by matrix-assisted laser desorption/ionization mass spectrometry. J. Mass Spectrom. 1998, 33, 697-704. (25) Woods, A. S.; Zangen, A. A Direct chemical interaction between dynorphin and excitatory amino acids. Neurochem. Res. 2001, 26, 395-400. (26) Juhasz, P.; Biemann, K. Mass spectrometric molecular-weight determination of highly acidic compounds of biological significance via their complexes with basic polypeptides. PNAS 1994, 91, 4333-4337. (27) Tang, X.; Callahan, J. H.; Zhou, P.; Vertes, A. Noncovalent proteinoligonucleotide interactions monitored by matrix-assisted laser desorption/ionization mass spectrometry. Anal. Chem. 1995, 67, 4542-4548. (28) Lin, S.; Long, S.; Ramirez, S. M.; Cotter, R. J.; Woods, A. S. Characterization of the “helix clamp” motif of HIV-1 reverse transcriptase using MALDI-TOF MS and surface plasmon resonance. Anal. Chem. 2000, 72, 2635-2640. (29) Lin, S.; Cotter, R. J.; Woods, A. S. Detection of non-covalent interaction of single and double stranded DNA with peptides by MALDI-TOF. Proteins Struct. Func. Genet. 1998, 2, 12-21. (30) Woods, A. S.; Moyer, S. C.; Wang, H. Y.; Wise, R. Y. Interaction of chlorisondamine with the neuronal nicotinic acetylcholine receptor. J. Proteome Res. 2003, 2, 207-212.

484

Journal of Proteome Research • Vol. 3, No. 3, 2004

Woods (31) Woods, A. S.; Fuhrer, K.; Gonin, M.; Egan, T.; Ugarov, M.; Gillig, K. J.; Schultz, J. A. AngiotensinII/acetylcholine non-covalent complexes analyzed with MALDI-ion mobility-TOFMS. J. Biomol. Tech. 2003, 14, 1-8. (32) Lecchi, P.; Pannell, L. K. The detection of intact double-stranded DNA by MALDI. J. Am. Soc. Mass Spectrom. 1995, 6, 972-975. (33) Lecchi, P.; Le, M. M.; Pannell, L. K. 6-Aza-2-thiothymine: a matrix for MALDI spectra of oligonucleotides. Nucleic Acids Res. 1995, 11, 1276-1277. (34) Spartan ’02, Wavefunction, Inc., Irvine, CA. (35) Yang, Y.-C.; Klippenstein, S. J.; Dunbar, R. C. Binding Energies of Ag+ and Cd+ Complexes from Analysis of Radiative Association Kinetics, J. Phys. Chem. A. 1997, 101, 3338-3347. (36) Lin, Y.-C.; Dunbar, R. C. Radiative Association Kinetics and Binding Energies of Chromium Ions with Benzene and Benzene Derivatives, Organometallics 1997, 16, 2691-2697. (37) Gapeev, A.; Dunbar, R. C. Binding of alkaline earth halide ions MX+ to benzene and mesitylene. J. Phys. Chem. A. 2000, 104, 4084-4088. (38) Gapeev, A.; Dunbar, R. C. Cation-π interactions and the gas-phase thermochemistry of the Na+/phenylalanine complex. J. Am. Chem. Soc. 2001, 123, 8360-8365. (39) Gapeev, A.; Dunbar, R. C. Reactivity and binding energies of transition metal halide ions with benzene, J. Am. Soc. Mass Spectrom. 2002, 13, 477-484. (40) Rodriguez-Cruz, S. E.; Williams, E. R. Gas-phase reactions of hydrated alkaline earth metal ions, M2+(H2O)n (M ) Mg, Ca, Sr, Ba and n ) 4-7), with benzene J. Am. Soc. Mass Spectrom. 2001, 12, 250-257. (41) Ryzhov, V.; Dunbar, R. C. Interactions of phenol and indole with metal ions in the gas phase: Models for Tyr and Trp side-chain binding. J. Am. Chem. Soc. 1999, 121, 2259-2268. (42) Ryzhov, V.; Dunbar, R. C.; Cerda, B.; Wesdemiotis, C. Cation-π effects in the complexation of Na+ and K+ with Phe, Tyr, and Trp in the gas phase. J. Am. Soc. Mass Spectrom. 2002, 11, 10371046. (43) Pletneva, E. V.; Laederach, A. T.; Fulton, D. B.; Kostic, N. M. The role of cation-π interactions in biomolecular association. Design of peptides favoring interactions between cationic and aromatic amino acid side chains. J. Am. Chem. Soc. 2001, 123, 6232-6245. (44) Tsou, L. K.; Tatko, C. D.; Waters, M. L. Simple cation-π interaction between a phenyl ring and a protonated amine stabilizes an R-helix in water. J. Am. Chem. Soc. 2002, 124, 14917-14921.

PR034091L