peptide conformational analysis using the tripos force field

6

Click here to load reader

Upload: craig-g-wall

Post on 06-Jul-2016

219 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: Peptide conformational analysis using the TRIPOS force field

INTERNATIONAL JOURNAL OF QUANTUM CHEMISTRY. VOL. 44.543-548 (1YY2)

Peptide Conformational Analysis Using the TRIPOS Force Field

CRAIG G. WALL Department of Chemistry, University of Texas at Austin, Austin, Texas 78712

EAMONN F. HEALY Department of Chemistry, University of Texas at Austin, Austin, Texas 78712 and

Department of Chemistry, St. Edwards University, Austin, Texas 78704

MARYE ANNE FOX Department of Chemistry, University of Texas at Austin, Austin, Texas 78704

Abstract

To explain observed electron transfer rates between the terminal aromatic moieties in the aryl- substituted alanine dimer 1, a conformational analysis of the dipeptide was performed using a standard molecular mechanics force field. The resulting low-energy conformers could conve- niently be grouped into two families with an average energy difference of ca. 2 kcal mpl-' and populations of 64% and 30%, respectively. These results correlate well with the 78:22 ratio of the two experimentally distinguishable decay processes for the radical anion of 1.

Introduction

Attaining an understanding of the factors that control rates of electron trans- fer between appropriate donors and acceptors has recently been an actively pur- sued goal of physical, organic, and inorganic chemists [l]. Among those factors that exert a decisive influence over such rates are the distance separating a donor-acceptor pair, the orientation of the two chromophores, the energetics for the forward and reverse electron transfers, and the polarity of the surrounding medium. In intramolecular electron transfer, therefore, the conformation and chemical composition of the bridging atoms intervening between the donor and acceptor are expected to alter significantly the magnitude of the electron- exchange interact ion.

Because of the ubiquity of proteins as matrices for biologically significant elec- tron transfers [2], a complete orbital analysis of short-chain peptides [3] as link- ing agents between model donor-acceptor pairs would be quite valuable. Although conformational analysis of short-chain peptides has been sometimes undertaken experimentally [4], convincing calculational descriptions of con- tributing conformational populations are lacking. Even very short chain peptides are much too complex, for example, for a rigorous ab initio treatment, and when the molecular structure is further complicated by the attachment of large aro- matic moieties as spectroscopic probes for electron location, only semiempirical

0 1992 John Wiley & Sons, Inc. CCC 0020-7608/92/040543-06

Page 2: Peptide conformational analysis using the TRIPOS force field

544 WALL, HEALY, AND FOX

methods offer any realistic hope for a reasonable chemical or physical descrip- tion within realistic computational times.

Several experimental methods are available to test for conformational popula- tions of frameworks connecting electron transfer donors and acceptors. For ex- ample, the use of fluorescent probes to provide information on peptide structure is well established [5] , as is the use of kinetic profiles to monitor the distance de- pendence of electron transfer from a radical anion formed by pulse radiolysis to either a covalently linked acceptor [6] or to a dispersed acceptor held at a fixed distance in a rigid glass [7]. Since recent work in our laboratories on characteriz- ing electron transfer between the donor naphthoyl radical anion and an acceptor biphenylamino group in a small conformationally flexible oligopeptide 1 yielded two kinetically distinct electron transfer rates [8], we were intrigued by the possi- bility that the biexponentiality might reasonably be assigned to equilibrating conformational families:

If intramolecular electron transfer were faster than conformational equilibra- tion, then the relative contributions of these rates might be expected to correlate with conformational populations, the different rates reflecting differences in the average donor-acceptor distances for each conformational family. Aecordingly, we report here a conformational study of dipeptide 1, in which we attempt to correlate the calculated conformational distribution with the experimentally ob- served populations (78: 22 associated with electron transfer rates of 5.6 x lo9 s-' and 5.2 x lo8 s-l, respectively [S]).

In this study, we have employed semiempirical methods as implemented in the standard AM-1 version [9] in the AMPAC package of computer programs [lo] to obtain optimized geometries and orbital characteristics of the aromatic donor and acceptor groups appended to the termini of a dialanine peptide. We have then performed a full conformational analysis through a molecular mechanics program [11] to address the physical separation and the magnitude of the elec- tronic interaction between these two groups.

Method

All calculations were performed on an Evans & Sutherland PS39O/VAX 3600 system using the Sybyl modeling software from TFUPOS Associates. Initial en- ergetically optimized input geometries for the aromatic caps in 1 were obtained from the AMPAC semiempirical package. The conformational analyses were

Page 3: Peptide conformational analysis using the TRIPOS force field

PEPTIDE CONFORMATIONAL ANALYSIS 545

performed via the Sybyl CSearch alogorithm using the TRIPOS 5.2 force field [ll]. The default conjugate gradient procedure was used for all energy minimiza- tions. Though containing no explicit hydrogen-bond terms, the TRIPOS 5.2 force field does allow for hydrogen bonding through its 6-12 potential function. For the conformational search, all nine rotatable bonds of the peptide backbone, in addition to the central bond in the biphenyl moiety, were allowed to vary in increments of 60". This gave a possible 6'' conformations. The resulting low- energy conformers were grouped by donor-acceptor distance, as given by the ipso-C14-ipso-Cz9 distance, and by the average energy of each conformer fam- ily. The relative population, in mole fraction xi, for each conformer was calcu- lated as the Boltzmann population for that conformer:

The energy of each conformer was weighted by the Boltzmann population for that conformer, and these normalized energies were then used to yield an aver- age energy for each of the conformer families.

Results

Of the 3300 conformers found within 10 kcal mol-' of tke minimum energy structure, the first 207 together compose 95% of the total composition. Group- ing of these 207 conformers by donor-acceptor distance and normalized energies initially yields three different conformational families, summarized in Table I. Representative structures from each of these families, A-C, are shown, respec- tively, in Figures 1-3, with that shown in Figure 1 depicting the lowest-energy conformer found. While the conformers in family C exhibit a different backbone configuration from those in families A and B, the conformers in A and B differ only in orientation about the central bond of the biphenyl moiety. Combining the conformers from A and B gives a set of 16 conformers with an average weighted energy of 20.2 kcal mol-' and a mole fraction of 0.64, with all structures within the set exhibiting the same folded backbone configuration and a C14-Cz1 dis- tance of 5.45 A. The remaining 191 conformers have an average energy of 22.25 kcal mol-' and a mole fraction of 0.31, the average C14-cZ1 distance be-

TABLE I . Characteristics of major conformational families of peptide 1.

Avg. energy Conformational family No. configurations ,yL d(Cl4-C2g) (A) (kcal mol-')

A 8 0.4688 5.45 20.075 B 8 0.1749 5.45 20.659 C 191 0.3061 7.60 22.251

Page 4: Peptide conformational analysis using the TRIPOS force field

Figure 1. A representative conformation of folded conformational family A.

Figure 2. A representative conformation of folded conformational family B.

Page 5: Peptide conformational analysis using the TRIPOS force field

PEPTIDE CONFORMATIONAL ANALYSIS

Figure 3. A representative conformation of extended conformational family C.

547

ing 7.60 A. Conformers in this second set typically have a more extended back- bone configuration.

That the electron transfer decay involving the donor and acceptor end groups of the dipeptide 1 in N-methylpyrollidinone (NMP) was best fit to two exponen-

Page 6: Peptide conformational analysis using the TRIPOS force field

548 WALL, HEALY, AND FOX

tials (with rate constants of 5.6 x lo's-' and 5.2 x lo8 s-' and relative contri- butions of 78% and 22%, respectively) [8] indicates that 1 must exist in at least two major conformational populations in NMP, correlating well with the calcu- iational grouping of energetically accessible conformations of 1 into two con- formational families with populations of 64% and 31%, respectively. These conformational families represent a set of folded conformers (as in Figs. 1 and 2) and a set of more extended conformers (as in Fig. 3). The calculated difference in donor-acceptor separation (from the nearest ips0 carbons) is fully consistent with the observed tenfold difference in decay rate constants when a standard value for the distance dependence is employed [6]. The above correlation there- fore implies that the backbone folding required to interconvert between these two conformational families is much slower than the time scale of the electron transfer measurement. This agreement between the experimentally observed conformer populations and the results of the conformational search thus sug- gests that in cases where electron transfer is faster than the conformational equi- libriation electron transfer kinetics can be used to probe conformational distribution in solution.

Acknowledgments

This work was supported by the Robert A. Welch Foundation through grants to MAF and to the Chemistry Department of St. Edward's University. We are grateful to Dr. Mark Meier for valuable suggestions.

Bibliography

[ l ] M. A. Fox and M. Chanon, Eds., Photoinduced Electron Transfer (Elsevier, Amsterdam,

[2] R.A. Marcus and N. Sutin, Biochim. Biophys. Acta 811, 265 (1985) and references cited therein.

[3] See, e.g., (a) K. S. Schanze and L. A. Cabana, J. Phys. Chem. 94, 2740 (1990); (b) M. Sisido, Y. Inai, and Y. Imanishsi, Macromol. 23, 1655 (1990); (c) K. S. Schanze and K. Sauer, J. Am. Chem. SOC. 110, 1180 (1988); (d) S. S. Isied, A. Vassilian, R. H. Magnuson, and H. A. Schwarz, J. Am. Chem. SOC. 107, 7432 (1985); (e) M. R. DeFelippis, M. Faraggi, and M. H. Klapper, J. Am. Chem. Soc. 112, 5640 (1990) and references cited therein.

[4] (a) J. A. Tanaka, H. Masuhara, N. Mataga, R. Goedeweeck, and F. C. DeSchryver, Polym. J. (Tokyo) 18, 331 (1986); (b) J. Vandendriessche, R. Goedeweeck, P. Collart, and F.C. DeSchryver, NATO Adv. Sci. Ser. C 225 (1986); (c) F. Ruttens, R. Goedeweeck, A. F. Lopez, and F. C. DeSchryver, Photochem. Photobiol. 42, 341 (1985); (d) R. Goedeweeck and F. C. DeSchryver, Photochern. Photobiol. 39, 515 (1984).

1988), VOIS. A-D.

[5] L. Stryer, Science 162, 526 (1968). [6] (a) L.T. Calcaterra, G. L. Closs, and J. R. Miller, J. Am. Chem. SOC. 105,670 (1983); (b) J. R.

[7] J. R. Miller, Science 189, 221 (1975). [8] M. S. Meier, M. A. Fox, and J. R. Miller, J. Org. Chem. 56, 5380 (1991). [9] M. J. S. Dewar, E. G. Zoebisch, E . F. Healy, and J. J. P. Stewart, J. Am. Chem. Soc. 107,3902

[lo] QCPE Publication 506 (Department of Chemistry, Indiana University, Bloomington, IN

[ll] M. Clark, R. D. Crarner, and N. van Opdenbosch, J. Cornp. Chem. 10, 982 (1989).

Received October 14, 1991 Accepted for publication November 12, 1991

Miller, L.T. Calcaterra, and G. L. Closs, Zbid. 106, 3047 (1984).

(1985).

47405).