Environment-Controlled Interchromophore Charge Transfer

These spectra displayed three bands, e.g., band I between 44 and 50 kK (kK = 103 cm-1), band II at 53 kK, and band III above 55 kK, which were, respec...
1 downloads 0 Views 234KB Size
Download PDF
J. Phys. Chem. B 2006, 110, 13235-13241

13235

Environment-Controlled Interchromophore Charge Transfer Transitions in Dipeptides Probed by UV Absorption and Electronic Circular Dichroism Spectroscopy Isabelle C. Dragomir, Thomas J. Measey, Andrew M. Hagarman, and Reinhard Schweitzer-Stenner* Department of Chemistry, Drexel UniVersity, 3141 Chestnut Street, Philadelphia, PennsylVania 19104 ReceiVed: March 15, 2006; In Final Form: April 22, 2006

Charge transfer (CT) transitions between the C-terminal carboxylate and peptide group have been investigated for alanyl-X and X-alanine dipeptides by far-UV absorption and electronic circular dichroism (ECD) spectroscopy (where X represents different amino acid residues). The spectra used in the present study were obtained by subtracting the spectrum of the cationic species from that of the corresponding zwitterionic peptide spectrum. These spectra displayed three bands, e.g., band I between 44 and 50 kK (kK ) 103 cm-1), band II at 53 kK, and band III above 55 kK, which were, respectively, assigned to a nCOO- f π* CT transition, a πCOO- f π* CT transition, and a carboxylate π f π* (NV1) transition, respectively By comparison of the intensity, bandwidth, and wavenumber position of band I of some of the investigated dipeptides, we found that positive charges on the N-terminal side chain (for X ) K), and to a minor extent also the N-terminal proton, reduce its intensity. This can be understood in terms of attractive Coulomb interactions that stabilize the ground state over the charge transfer state. For alanylphenylalanine, we assigned band I to a nCOO- f π* CT transition into the aromatic side chain, indicating that aromatic side chains interact electronically with the backbone. We also performed ECD measurements at different pH values (pH 1-6) for a selected subset of XA and AX peptides. By subtraction of the pH 1 spectrum from that observed at pH 6, the ECD spectrum of the CT transition was obtained. A titration curve of their spectra reveals a substantial dependence on the protonation state of the aspartic acid side chain of AD, which is absent in DA and AE. This most likely reflects a conformational transition of the C-terminus into a less extended state, though the involvement of a side chain f peptide CT transition cannot be completely ruled out.

Introduction One of the most important aspects of peptide and protein dynamics involves the extremely rapid transfer of charges and excitation energy, which is often a prerequisite of energy pumping processes. In this regard, the understanding of interchromophore charge transfer (CT) processes is crucial. One type of CT transition, namely, the transition between electronic states of adjacent peptide linkages in a polypeptide chain has been thus far mostly addressed by theoretical studies. Serrano-Andre´s and Fu¨lscher1 and Hirst and associates2 employed excited-state calculations to predict no f π* and π f π* transitions between peptide groups of short peptides in vacuo. Their intensities and energies depend significantly on the dihedral angles between interacting peptide groups. Experimentally, Weinkauf and associates conducted detailed studies of the propagation of positive holes in polypeptides generated by the two-photon ionization of an aromatic side chain.3 Asher and co-workers combined UV absorption and resonance Raman spectroscopy to identify a band at 200 nm in the spectra of dipeptides with a deprotonated carboxylate group, which they assigned to an nCOO- f π* transition.4 In another study from that research group, the direction of the underlying electronic dipole moment was determined by UV resonance Raman single-crystal measurements on glycylglycine.5 Later, Sieler et al. proposed a concomitant πCOO- f π* transition on the basis of a detailed analysis of the visible and near resonance * Author to whom correspondence should be addressed. Phone: (215) 895-2268. Fax: (215) 895-1265. E-mail: [email protected].

UV Raman spectra of the same peptide.6 Despite its potential relevance for the understanding of CT transitions between peptide backbone chromophores as well as between side chain and backbone, no further experimental studies have thus far been undertaken to explore physical and structural parameters of the obtained/proposed transitions. The present study employed UV/vis absorption spectroscopy to determine the structural and physical parameters governing the nCOO- f π* transition in a series of AX and XA dipeptides. We chose dipeptides as our model system because they only have one peptide bond into which the negative charge residing on the C-terminal carboxylate can be transferred. Moreover, a recent electronic circular dichroism (ECD) and 1H NMR study on the above peptides provided structural information,7 which allowed us to elucidate the relationship between the relative orientation of the C-terminal carboxylate and spectral parameters of the nCOO- f π* band. Additionally, we performed ECD measurements to disentangle the contribution of the nCOO- f π* transition to the ECD spectra of dipeptides at neutral pH. We also used this technique as a convenient tool to probe the influence of the negatively charged side chains on the pK value and the structure of the C-terminal carboxylate group. Materials and Methods Materials. L-Alanyl-L-alanine (AA), L-alanyl-L-glycine (AG), (GA), L-alanyl-L-leucine (AL), L-leucyl-Lalanine (LA), L-alanyl-L-serine (AS), L-seryl-L-alanine (SA), L-alanyl-L-valine (AV), L-valyl-L-alanine (VA), L-alanyl-Lisoleucine (AI), L-isoleucyl-L-alanine (IA), L-alanyl-L-threonine L-glycyl-L-alanine

10.1021/jp0616260 CCC: $33.50 © 2006 American Chemical Society Published on Web 06/15/2006

13236 J. Phys. Chem. B, Vol. 110, No. 26, 2006

Dragomir et al.

(AT), L-threonyl-L-alanine (TA), L-alanyl-L-proline (AP), Lprolyl-L-alanine (PA), L-alanyl-L-glutamate (AE), L-glutamateL-alanine (EA), L-alanyl-L-methionine (AM), L-methionyl-Lalanine (MA), L-alanyl-L-phenylalanine (AF), L-phenylalanylL-alanine (FA), L-alanyl-L-arginine (AR), L-arginyl-L-alanine (RA), L-acetyl-L-alanine (Ac-A), L-alanyl-L-aspartate (AD), and L-aspartate-L-alanine (DA) were purchased from Bachem Bioscience Inc. and used without any further purification. The purity of all the peptides used was greater than 98%, with the exception of AF (>96.3%). Methods. For the UV/vis absorption measurements, the peptides were dissolved in water to obtain concentrations of 0.01 or 0.02 M. Half of each solution was adjusted to a pH of about 1, the other to pH 7, by the addition of HCl and NaOH. A pH 12 sample was also made for AA. For the Ac-A, AA, AD, DA, and EA CD measurements, we prepared 0.01 M samples in water at ∼pH 1, 2, 2.5, 3, 3.5, 4, 4.5, 5, and 6, via the addition of small aliquots of HCl and NaOH. Both the UV absorption and the ECD spectra were measured at the Drexel University College of Medicine using a JASCO J-810 spectrapolarimeter purged with N2. Spectra were obtained in the 180-240 nm range at 0.5 nm resolution using 0.1 and 0.2 mm quartz cells, respectively, for the 0.02 and 0.01 M solutions. For all the spectra, 10 accumulations were averaged at 20°C ((1°C) using a scanning speed of 200 nm/min. Care has been taken to correct all peptide spectra for the background by subtracting the respective aqueous solution spectrum obtained for exactly the same pH and ionic strength values (i.e., identical Cl- and Na+ concentrations).8 Neglecting this yields nonreproducible results. Our UV absorption spectra were further analyzed by utilizing the program MULTIFIT,9 and the integrated absorbtivity was used to calculate the corresponding dipole strength of band I, via the following equation10

D≈

9.2 × 10 39 ν0

∫band (ν) dν

Figure 1. Absorption difference spectra of AX dipeptides obtained by subtracting the spectrum of the indicated dipeptide observed at pH 1 from that measured at pH 7.

(1)

where ν0 is the wavenumber of band I and (ν) is the molar absorptivity. The absolute value of the transition dipole moment was calculated as the square root of the oscillator strength and obtained in units of esu cm. Results and Discussion Absorption Difference Spectra. The absorption difference spectra in Figures 1 and 2 were obtained by subtracting the absorption spectrum acquired at acidic pH from the corresponding spectrum measured at neutral pH, for a series of AX and XA dipeptides where A is alanine and X represents a subset of the 20 naturally occurring amino acids. These subtractions were carried out to isolate absorption bands resulting from CT transitions from the highest occupied n and π molecular orbitals (HOMOs) of the deprotonated carboxylate group into the π* lowest unoccupied molecular orbital (LUMO) of the peptide group. All difference spectra exhibit a weak band (band I) between 44 and 50 kK (kK ) 103 cm-1), which resembles the band Chen et al. observed for glycylglycine and assigned to a nCOO- f π* CT transition.4 A second band (band II) appears at around 53 kK, whose oscillator strength is generally comparable with that of band I. Suprisingly, this band was not obtained by Chen et al. for glycylglycine.4 We assign this band to the πCOOf π* transition proposed by Sieler et al.6 Alternatively, one might invoke a n f πCOO-* CT transition, but this would involve a charge separation and a 2-fold negative charge on the C-terminus, which makes this an unlikely transition. For

Figure 2. Absorption difference spectra of XA dipeptides obtained by subtracting the spectrum of the indicated dipeptide observed at pH 1 from that measured at pH 7.

some peptides, the band is overlapped with a more intense band with a peak position above 55 kK, which according to Chen et al.4 and Clark11 can be assigned to the carboxylate π f π* (NV1) transition. Its absence in some spectra is most likely due to an overlap with the corresponding band of the carboxyl group, which gives rise to a negative signal in the difference spectra. Interestingly, the position of band I correlates very well with the maxima of the couplet in the respective UV circular dichroism spectra for all peptides apart from FA.7 One might interpret this as suggesting that the band should be assigned to an intrapeptide no f π* rather than to a C-terminus f peptide CT transition. However, to explain the observed dipole strengths one would have to invoke substantial electronic mixing between the NV1 states of the peptide and carboxylate group, which should also, to some extent, be present when the C-terminus is protonated, in contrast to our observation. Our data therefore

Dipeptide CT Transitions Probed by UV/CD indicate that the n-orbitals of the peptide and carboxylate group have rather comparable energies. We have used our MULTIFIT program9 to obtain the peak wavenumber and the integrated absorptivity of band I for all peptides investigated. These and the respective electronic transition dipole moments are listed in Table S1 in the Supporting Information, where the dipole moments were obtained as described in the preceding section. The obtained parameter values suggest that the choice of the amino acid residue at the X-position has a limited influence on the electronic transition dipole moment, irrespective of whether X is N- or C-terminal. This is not surprising for XA because the C-terminus choice of X has a limited, though detectable impact on the φ-value of the alanine residue, which substantially populates the polyproline II (PPII) conformation (χPPII ) 0.58-0.7).7 We calculated average values of ν˜ 0 ) 47.916 ( 0.4 kK and µ ) 0.88 ( 0.15 D. An inspection of Table S1 reveals that only the parameter values of GA and KA deviate significantly from their respective averages. GA exhibits a larger wavenumber and a larger dipole moment, whereas a lower wavenumber and a significantly reduced dipole moment is characteristic for KA. This influence of the lysine residue is apparently positiondependent since AK does not show such a deviation from the respective average value of the AX series (Table S1; cf. also the discussion below). For the investigated AX peptides we calculated an average dipole moment of 0.96 D with a standard deviation of (0.17 D. Thus, only AG (1.25 D), AM (0.79 D), AS (0.70 D), and AF (0.75 D) deviate significantly from the average value. Hence, all these oscillator strengths are on the same order of magnitude, and some are even slightly larger (AG, AV, AT, AD, and AP) than the average value observed for the XA peptides. On a first view, this is a surprising result for the following reason. For the invoked nCOO- f π* transition, the oscillator strength would be maximal for an orthogonal and minimal for a coplanar conformation. A n f π* interaction was recently suggested as a decisive stabilizing force for the PPII conformation.12 Thus, our data would suggest that the carboxylate group of most of the XA and AX peptides exhibits very similar distributions of relative orientations, which are substantially tilted with respect to the peptide plane. Unfortunately, this is at variance with results earlier inferred from the 3JCRNH coupling constants of these peptides.7 They suggest that, for example, AV, AI, and particularly AT have a higher propensity for βs than for PPII (the obtained PPII fractions were 0.35, 0.39, and 0.19, respectively). This would imply a lower absorptivity for band I compared with AA, for which a PPII fraction of 0.64 was reported, since βs exhibits the carboxylate plane in a significantly less tilted orientation than the canonical PPII. This discrepancy, however, is somewhat reduced if one uses the residue specific 3J 13 CRNH constants, which Shi et al. recently inferred from the coil library of Avbelj and Baldwin.14 Their constants yield substantially larger PPII fractions for AV (0.45), AI (0.48), and particularly AT (0.32). These values seem to be more in agreement with the significant CD couplets observed for these peptides, and they make band I dipole strengths of the AX at least qualitatively understandable. However, one has to keep in mind that this dipole moment depends certainly on more than one parameter. Besides the orientation, one has to consider electrostatic interactions particularly with the charges of the C-terminal residues, since even aliphatic side chains exhibit partial charges. Positive partial charges certainly stabilize the ground state, whereas negative charges might stabilize the CT state. DFT calculations on alanine, valine, and isoleucine dipeptides revealed negative partial charges on the respective carbon atoms of the side chain with a tendency of higher charges

J. Phys. Chem. B, Vol. 110, No. 26, 2006 13237

Figure 3. PPII conformation of KA.

for the β-conformation compared with PPII (Cho, private communication and ref 15). These charges doubtlessly destabilize the ground state and thus increase the probability for a charge transfer transition, which could explain why the CT bands of AV and AI still exhibit substantial oscillator strength. A clarification of this issue requires extensive density functional theory/ab initio calculations, which are out of the scope of this paper. It is interesting to note that AG exhibits the highest peak wavenumber and the largest dipole strength of all investigated AX peptides, indicating that subconformations assignable to larger oscillator strengths contribute to the high-energy wing of band I. This notion is corroborated by the fact that all bands are slightly asymmetric in that they show a more pronounced wing on the low-energy side, so that the first moment value is lower than the peak wavenumber. Indeed, the average wavenumber of the βs preferring peptides AI, AV, AS, and AT (47.847 ( 0.139 kK) is significantly lower than the respective average value for the PPII-like peptides AA, AK, AD, and AG (48.339 + 0.224 kK). The very low band I oscillator strength observed for KA deserves a more detailed consideration. To rationalize the possible influence of the lysine proton on the CT transition, we analyzed the PPII conformation ((φ,ψ) ) (-70°,150°)) of KA.7 As can be inferred from Figure 3, the positive charge of the N-terminal lysine side chain is in closer proximity to the C-terminal carboxylate group than to the peptide carbonyl oxygen, which is the acceptor for the CT process. This can be expected to stabilize the ground state. The situation is different for RA. In a PPII conformation, the distance between the positive side chain charge and the carboxylate group is larger than its distance to the peptide carbonyl. This explains why RA, in contrast to KA, exhibits a rather normal transition dipole moment for nCOO- f π*. A conformational change, i.e., an equilibrium shift toward the more extended βs conformation, can be ruled out as a reason for the reduced oscillator strength because of the limited influence of N-terminal residues on the 3J CRNH coupling constants of the C-terminus residue in XA peptides.7 Whereas most CT transitions cause a charge separation, the nCOO- f π* transition produces a neutral radical at the C-terminus and reduces the distance between the negative charge (now located at the carbonyl oxygen) and the positive Nterminus charge. This, doubtless, lowers the Gibbs energy of

13238 J. Phys. Chem. B, Vol. 110, No. 26, 2006

Figure 4. Absorption difference spectra of AA obtained (a) by subtracting the spectrum measured at pH 12 from that measured at pH 6 (solid line) and (b) by subtracting the spectrum measured at pH 1 from that measured at pH 6 (dashed line).

the CT state and contributes to the electron-transfer rate. This implies that a deprotonation of the N-terminus should reduce the transition dipole moment and thus the absorptivity of the CT band. We therefore measured the absorption spectrum of anionic alanylalanine at pH 12 and subtracted it from the spectrum of the zwitterionic species. Figure 4 compares the difference spectra “zwitterionic - cationic” and “zwitterionic - anionic”. The latter exhibits a small band at 46 kK and surprisingly a rather broad band at 51 kK. A comparison of the two difference spectra suggests that the deprotonation eliminates the low-energy wing of the CT band. This is consistent with the observation of Chen et al.,4 whose spectra suggest that the CT band of acetylglycine is significantly blue-shifted compared with that of glycylglycine. On the contrary, the protonated lysine in KA bleaches mostly the high-energy side of the CT band. These results resemble observations obtained by kinetic holeburning experiments16 on heme proteins and corroborates the above-mentioned notion that band I reflects a conformational distribution of the carboxylate group, which gives rise to inhomogeneous broadening. Apparently, these conformations differ with respect to the influence of the side chain and terminal charges on the transition dipole moments. Band I in the spectra of AF and AP differ from that of the remaining AX peptides in that they are significantly red-shifted from the AA band position (by approximately 2.3 and 2.2 kK, respectively). Moreover, the transition dipole moment of AF (0.75 D) is significantly lower than the average value, and the band itself is narrower and more symmetric. When compared to AA, AF’s bandwidth (∼1706.1 cm-1) is less than half of that of AA (∼3765.8 cm-1). These spectral changes were not observed for FA. All these observations suggest that for AF band I is assignable to a different mechanism. Since its position is close to the lower-energy absorption band of phenylalanine (at 48 kK),17 it is reasonable to assign it to a nCOO- f πF* transition into the LUMO of the aromatic side chain. This transition wins the competition with the no f π* into the peptide orbital owing to the much closer proximity of the phenylalanine with the carboxylate group. ECD Spectra and pH Dependence. We measured the ECD spectra of some XA and AX peptides as a function of pH between pH 1 and 6. The spectra are shown in the left column of Figure 5. These experiments were aimed at (1) identifying the ECD spectrum associated with the nCOO- f π* transition, (2) investigating the influence of the side chains on the ECD signal and the C-terminal pK value, and (3) exploring the possibility that CT transitions are also possible from side chains with carboxylate groups such as aspartic acid and glutamic acid. The ECD spectra in Figure 5 reflect a pronounced pH dependence for AA, DA, AE, and AD. In each case, the maximum between 205 and 215 nm disappears and converts to

Dragomir et al. a minimum between 200 and 210 nm at neutral pH. AD is an exception from the rule in that another maximum appears at 226 nm. For each of these dipeptides, as well as Ac-A, the ECD spectrum at pH 1 was subtracted from that at pH 6 to obtain the difference spectra shown in Figure 6, namely, ∆(pH ) 6) - ∆(pH ) 1). There are two possible explanations for our observation. One might suppose that the deprotonation of the C-terminal carboxylate causes a structural change due to the electrostatic Coulomb interaction between the terminal charges. However, this option can be ruled out in view of the fact that Ac-A shows more or less the same behavior as AA. We therefore interpret the results, instead, as indicating that a substantial rotational strength is associated with the no f π* transition, which gives rise to the ECD spectra shown in Figure 6. The titration curves in Figure 5 (right column) depict the pH dependence of ∆∆ ) ∆(pH) - ∆(pH ) 1) at λ ) 206.5, 209, 207, 210, and 209 nm and 210.5 nm, respectively, for AcA, AA, AD, DA, and AE. With the exception of AD, all peptides exhibit a single phase, indicating that only one protonable group is involved, i.e., the C-terminal carboxylate. AD, however, clearly shows a second phase at higher pH (titration curve I, between pH 4.0 and 6.0. As mentioned above, the ECD spectrum measured at neutral pH depicts a maximum at 226 nm. The titration curve (titration curve II) for this wavelength is also shown in the right-hand column of Figure 5. Apparently it is different from that obtained at the lower wavelength. Altogether, the data reveal a more complex pH dependence of this dipeptide. We invoked a previously utilized protonation model to determine the pK values reflected by the titration curves.18 For our current purpose we assumed that no more than two protonation processes are involved. The experimentally observed ∆∆(pH) can be written as 1

∆∆(pH) )

∑ χij∆∆ij

(2)

i,j)0

where i,j ) 0,1 label the proton occupation of the protonable groups, χij denotes the mole fraction of the titration states (i,j) ) (0,0), (1,0), (0,1), and (1,1), and ∆∆ij the respective ∆∆ values. The mole fractions χij are written as

[

χ00 ) 1 +

[ [

χ10 ) 1 +

χ01 ) 1 +

[

χ11 ) 1 +

]

[H3O+] [H3O+] [H3O+]2 + + K2 K1 K1K2

] ]

[H3O+] K1 + K2 K2

-1

[H3O+] K2 K2 + + + K1 [H O ] K1

-1

K1 [H3O+]

+

3

K1 [H3O+]

+

K2 [H3O+]

+

-1

(3a)

(3b)

K1K2 [H3O+]2

(3c)

]

-1

(3d)

where [H3O+] is the hydronium ion concentration and K1 and K2 are the equilibrium constants of the two considered protonation processes. We inserted eqs 3a-3d into eq 2 and fit the thus derived equation to the titration curves in Figure 5. To this end we used ∆∆ij, K1, and K2 as free parameters. For obvious reasons, ∆∆11 ≈ 0 was fixed. In a first step, we assumed only one protonation process to contribute to ∆∆ (K2 . K1 and K2 . [H3O+]), and

Dipeptide CT Transitions Probed by UV/CD

J. Phys. Chem. B, Vol. 110, No. 26, 2006 13239

Figure 5. Left column: Electronic CD spectra of Ac-A, AA, AD, DA, and AE taken at pH 1, 2, 2.5, 3, 3.5, 4, 4.5, 5, and 6. The arrows point from pH 1 to 6. Right column: ∆∆ )∆(pH) - ∆(pH ) 1) as a function of pH obtained from the ∆ values at λ ) 206.5, 209, 207, 210, and 209 nm, respectively, for Ac-A, AA, AD, DA, and AE. For AD, an additional titration curve (titration curve 2, pink boxes) was obtained from the ∆ values at λ ) 226 nm.

TABLE 1: ∆Eij Values for the Protonation States of Dipeptides Inferred from an Analysis of the Titration Curves in Figure 5

∆∆00 ∆∆01 ∆∆10 ∆∆11

Ac-A (206.5 nm)

AA (209 nm)

AD (207 nm)

AD (226 nm)

DA (210.5 nm)

AE (209 nm)

LA (210.5 nm)

-3.57 0.12

-3.73 0.00

-2.60 0.00 -1.90 0.17

0.97 0.20 0.50 0.10

-1.73 0.07

-3.29 0.01

-2.29 0.03

as expected, all titration curves besides that of AD could be reproduced. The respective pK values are all indicated in Figure 5. Most of them are very similar, though the slightly higher pK value of Ac-A (compared with AA) indicates an influence of the N-terminal charge, as observed earlier for diglycine.6 For AE, however, the pK value is substantially lower, indicating a side chain influence. A similar tendency, though less pronounced, exists for the respective pK values of the free amino acids.19 The ∆∆00 values are very similar for Ac-A, AA, and AE, as shown in Table 1. The respective values for DA are significantly smaller. A similar result observed for LA (data

not shown) indicates that bulky N-terminal residues affect the orientation of the C-terminal carboxylate, as already suggested by Hagarman et al.7 Since their φ-value distributions are nearly identical, these structural differences must be assigned to the pseudo-ψ angle. As shown in Figure 5, a single titration (dotted line) cannot explain titration curve I of AD. We therefore assumed two protonable sides for a fit to the data. Subsequently, we used the same pK values to fit titration curve II. Thus, we obtained pK1 ) 2.82 and pK2 ) 4.55, which are assignable to the C-terminal and aspartic acid carboxylate, respectively. The

13240 J. Phys. Chem. B, Vol. 110, No. 26, 2006

Dragomir et al.

Figure 6. Electronic CD difference spectra of Ac-A, AA, AD, DA, and AE obtained by subtracting the pH 1 spectrum from the respective pH 6 spectrum.

former value is close to that observed for AE. The second value is slightly higher than the respective pK value of the free amino acid (3.8).19 The ∆∆ values in Table 1 indicate that titration curve II is dominated by the contribution assignable to the side chain protonation. There are two possible explanations for the influence of aspartic acid deprotonation on the ECD spectrum of AD. First, the most obvious interpretation would invoke a nCOO- f π* transition from the side chain carboxylate to the peptide LUMO. Doubtless, the distance between both groups and the relative orientation, particularly in the PPII conformation, do not rule out this possibility. The absence of any comparable contribution in the ECD spectrum of DA could be easily understood in terms of a much smaller orientation angle between the side chain carboxylate and the peptide plane, which would be consistent with the very similar ψ-values of PPII and βS. In AE, the carboxylate-peptide distance is too large to allow a CT transition. Second, it is possible that aspartic acid deprotonation causes a structural change at the C-terminus owing to the repulsion between the two negatively charged carboxylate groups. This notion is strongly supported by the peculiar changes in the ECD spectrum, namely, the appearance of the ∆ maximum at λ ) 226 nm. The ECD signal of AD measured at acidic pH is weak compared with those of the XA and AX peptides investigated in this and our previous study,7 indicating a more extended, βs-like conformation. Apparently, the protonation of both carboxylates switches the C-terminus into a more bent PPII or turnlike structure. Summary Taken together, our results show that the CT band assignable to a nCOO- f π* transition reported by Chen et al.4 can be found in spectra of other dipeptides. Most of the N- and C-terminal residues investigated were found to have a limited influence on the dipole moment as well as on the wavenumber position of the band. Exceptions from this role, however, are noteworthy. The positive side chain charge of lysine, for instance, significantly reduces the band’s oscillator strength, most likely because it stabilizes the ground state by an electrostatic interaction with the negatively charged C-terminal carboxylate. For AF, we obtained a qualitatively and quantitatively different band, which we assigned to a CT transition into a π* state of the phenyl

ring. AD was found to exhibit a peculiar behavior in that its ECD spectrum is affected by the protonation of the carboxylate side group. It is likely that this causes the C-terminus of the molecule to switch into a less extended, more PPII-like conformation. The oscillator strengths and the earlier reported ECD spectra of AX peptides7 suggest a C-terminal conformation with a significant tilt angle between the carboxylate and the peptide group. Acknowledgment. We thank Dr. Patrick Loll for his help regarding the use of the JASCO spectrapolarimeter. We thank Dr. Minhaeng Cho for sending us his tables listing the partial charges of alanine, valine, and isoleucine dipeptides. I.D. performed her work under support from the Maryanoff summer research program. Financial support to R.S.S. was provided from the National Science Foundation (Grant No. MCB 0318749) and from the NIH-COBRE II grant for the Center for Research in Protein Structure, Function, and Dynamics (P20 RR1643901). Supporting Information Available: A table indicating the wavenumber position, molar absorptivity, and dipole moments of an optical band assignable to a nCOO- f π*(peptide) charge transfer transition of AX and XA dipeptides. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Serrano-Andre´s, L.; Fu¨lscher, M. P. J. Phys. Chem. B 2001, 105, 9323-9330. (2) Gilbert, A. T. B.; Hirst, J. D. J. Mol. Struct. 2004, 675, 53-60. (3) Weinkauf, R.; Schlag, E. W.; Martinez, T. J.; Levine, R. D. J. Phys. Chem. A 1997, 101, 7702-7710. (4) Chen, X. G.; Li, P.; Holtz, J. S. W.; Chi, Z.; Pajcini, V.; Asher S. A.; Kelly, A. L. J. Am. Chem. Soc. 1996, 118, 9705-9715. (5) Pajcini, V.; Chen, X. G.; Bormett, R. W.; Geib, S. J.; Li, P.; Asher, S. A.; Lidiak, E. G. J. Am. Chem. Soc. 1996, 118, 9716-9726. (6) Sieler, G.; Schweitzer-Stenner, R.; Holtz, J. S. W.; Pajcini, V.; Asher, S. A. J. Phys. Chem. B 1999, 103, 372-384. (7) Hagarman, A.; Doddasomayajula, R. S.; Measey, T.; Dragomir, I.; Eker, F.; Griebenow, K.; Schweitzer-Stenner, R. J. Phys. Chem. B. 2006, 110, 6979-6986. (8) Zimmerman, G.; Strong, F. C. J. Am. Chem. Soc. 1957, 79, 20632066. (9) Jentzen, W.; Unger, E.; Karvounis, G.; Shelnutt, J. A.; Dreybrodt, W.; Schweitzer-Stenner, R. J. Phys. Chem. 1996, 100, 14184-14191.

Dipeptide CT Transitions Probed by UV/CD (10) Nafie, L. A.; Dukor, R. K.; Freedman, T. B. Vibrational Circular Dichroism. In Handbook of Vibrational Spectroscopy; Chalmers, J. M., Griffiths, P. R., Eds.; John Wiley and Sons: Chichester, U. K., 2002. (11) Clark, L. B. J. Am. Chem. Soc. 1995, 117, 7974-7986. (12) Hinderacker, M. P.; Raines, R. T. Protein Sci. 2003, 12, 11881194. (13) Shi, Z.; Chen, K.; Liu, Z.; Ng, A.; Bracken, W. C.; Kallenbach, N. R. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 17964-17968. (14) Avbelj, F.; Baldwin, R. L. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 57421.

J. Phys. Chem. B, Vol. 110, No. 26, 2006 13241 (15) Lee, H.; Kim, S.-S.; Choi, J. H.; Cho, M. J. Phys. Chem. B 2005, 109, 5331-5340. (16) Campbell, B. F.; Chance, M. R.; Friedman, J. M. Science 1987, 238, 373-376. (17) Gratzer, W. B. In Poly-R-Amino Acids; Fasman, G. D., Ed.; Edward Arnold: London, 1967; pp 177-238. (18) Brunzel, U.; Dreybrodt, W.; Schweitzer-Stenner, R. Biophys. J. 1986, 49, 1069-1076. (19) Biophysics; Hoppe, W., Lohmann, W., Markl, H., Ziegler, H., Eds.; Springer: New York, 1983.