SEARCH

SEARCH BY CITATION

Keywords:

  • porphyrin;
  • DNA;
  • protein;
  • telomere;
  • bilirubin;
  • biliverdin;
  • polypeptides;
  • biogenic metal;
  • complex;
  • chiral recognition

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. INTERACTION OF DNA WITH BIOMOLECULES AND BIOGENIC METALS
  5. GUANINE TETRAMERS AND QUADRUPLEXES
  6. BIOINTERACTIONS OF BILE PIGMENTS
  7. POLYPEPTIDE AND PROTEIN INTERACTIONS WITH OTHER BIOMOLECULES
  8. LITERATURE CITED

Vibrational circular dichroism (VCD) spectra are reliable indicators of the spatial structure of chiral molecules. The specific and characteristic feature of vibrational spectroscopy, and therefore also of VCD, where the energy of some vibrational modes is predominantly focused to a specific part of the molecule, enables monitoring both the structure of the molecule dissolved in different solvents and under different physicochemical conditions and molecular interactions. This minireview deals with recent contributions covering structural information on the bioinspired interactions obtained by means of VCD, especially in the following areas: interaction of DNA with biomolecules and biogenic metals, guanine tetramers and quadruplexes, biointeractions of bile pigments, and polypeptide and protein interactions with other biomolecules. Chirality 21:E215–E230, 2009. © 2009 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. INTERACTION OF DNA WITH BIOMOLECULES AND BIOGENIC METALS
  5. GUANINE TETRAMERS AND QUADRUPLEXES
  6. BIOINTERACTIONS OF BILE PIGMENTS
  7. POLYPEPTIDE AND PROTEIN INTERACTIONS WITH OTHER BIOMOLECULES
  8. LITERATURE CITED

Since its discovery,1–5 vibrational circular dichroism (VCD) has had a wide range of applications. Because of the exact theoretical formulation of VCD,6 implementation of the Stephens' equations in the Gaussian7 and much experimental efforts of the leading group in the field, it became a very efficient method for the determination of the absolute configuration of chiral molecules.8–10 The pioneering experimental work helped VCD become an effective technique in the study of biologically important macromolecules, such as proteins and DNA.11–15 The specific and characteristic feature of vibrational spectroscopy, and therefore also of VCD, that the energy of some vibrational modes is restricted predominantly to specific parts of the molecule, enables studying not only the structure of molecules dissolved in different solvents and under different physicochemical conditions but also their intermolecular interactions. Among these, the most interesting interactions are those that are inspired by biosystems and the biofunctions. Many spectroscopic methods are involved in the characterization of these complex interactions. In this minireview, we will focus on the VCD contribution.

Circular dichroism (CD) methods are useful for studies of biomolecular interactions only if at least one partner of interaction is chiral, or if achiral units form a chiral structure. Although species resulting in an interaction should be exactly described as an integer entity, it is often dealt and described as a system of two or more components. Porphyrin–polypeptide and porphyrin–DNA adducts serve as examples of such description of interaction (for details see “Interaction of DNA with Biomolecules and Biogenic Metals” and “Polypeptide and Protein Interactions with other Biomolecules” sections). There are some possible effects accompanying the interactions that can be studied by the CD techniques. In the course of interaction, the spatial structure of the chiral component is changed. This structural variation can be sensed by CD. An example of this is DNA interacting with intercalating porphyrins and drugs (“Interaction of DNA with Biomolecules and Biogenic Metals” section). Induced CD (ICD) is another effect observable by CD. It originates in the previously achiral component while an interaction with a chiral component takes place. Chiral arrangement of achiral, e.g., planar, entities is another possibility which we meet in the case of biomolecular interactions. In this case, new CD signal originating in chiral structure composed of achiral components is detected. Chiral associates composed of planar achiral porphyrins on helical polypeptide matrices serve as an example. Bimolecular interactions quite often have the character of chiral recognition, when only specific diastereoisomer interacts with a chiral matrix. This is realized as an assembly of weak interactions: ionic interactions, hydrogen bonds, π–π interactions, or van der Waals interactions. Bilirubin (BR) interacting with serum albumin serves as an example (“Biointeractions of Bile Pigments” section).

In CD techniques, the left and right circularly polarized light possess chiral information, therefore more structural information can be extracted from CD spectra than in the case of parent absorption spectroscopies where unpolarized light is used for testing molecules. The structure of the chiral object is encoded in an observed position, and the intensity and sign of CD bands through rotational strength of transition are expressed by means of electric and magnetic transition moments that depend on the spatial structure. This structural information in the CD spectrum can be extracted at different levels. If the size of the studied system enables the complete conformation analysis and relevant calculations, the procedure of obtaining the structure of a product of biomolecular interaction based on density functional theory (DFT) is the same as that of an individual molecule. This can be obtained by following the protocol described.8–10 It includes a choice of tentative configuration, performing the conformational analysis, computation of CD spectrum, and comparison with experiment. If the computed pattern agrees in the position, sign, and relative band intensity with the experimental one, the tentative configuration is the right structure. In this procedure, VCD possesses some advantages when compared with widespread electronic CD (ECD): Instead of a few strongly overlapping electronic bands, better resolved vibrational bands (typically tens of them) are observed in the case of VCD. Instead of transitions involving excited electronic states (in the number needed to reproduce the experiment), the transitions between vibrational states within ground electronic state in harmonic approximation are taken in the case of VCD. However, typically the size and complexity of interacting biomolecules do not allow this procedure. Instead, the concept of characteristic vibrations typical to vibrational spectroscopy is used. Some of the vibration modes, known as the characteristic vibrations, focus the kinetic energy predominantly into a definite group of atoms, e.g., the C[DOUBLE BOND]O stretching, though a molecule, or an associate, vibrates as a whole with substantially less energy spread into the remaining part of the molecule. Experimentally, the transition involving the characteristic vibration is observed for different molecules in the same narrow spectral region, e.g., C[DOUBLE BOND]O stretching in the peptide bond, so-called amide I, is found at ∼1650 cm−1. It follows from the concept of characteristic vibrations that VCD, which is inherently sensitive to the slight structural variation, can be used to identify the part of the molecule whose structure is influenced by an interaction.

Biomolecular interactions are often mediated by specific physicochemical conditions in solutions, in aqueous solutions with specific pH, ionic strengths, concentration, and the temperature interval. Despite the X-ray refraction technique, which is the first choice among structural methods, VCD can be carried out depending on a quite broad range of such conditions, and the conditions can be tuned and modeled to reproduce bioinspired interactions in model systems. It should be noted that among the CD absorption techniques, ECD is usually the first choice. ECD is widespread, less time consuming, and demands 2–3 order less concentration when compared with VCD. The advantage of VCD is that it produce significantly better resolved bands when compared with only a few overlapping broad ECD bands. For example, VCD is able to distinguish aromatic vibrations from carbonyl stretching vibrations, whereas these electronic transitions are observed in ECD overlap. Also, ECD provides experimental data only if chromophores in an accessible UV/vis range are connected to chiral centers or a chiral axis, or are arranged in chiral structures. In contrast, each bond in a chiral molecule is a chromophore for VCD and each chiral molecule possesses VCD. From the biomolecular interaction point of view, VCD shows another advantage. The energy absorbed by studied molecules needed to obtain VCD is substantially less than is required to get ECD spectra (the IR photon is typically 10–30 times less energetic than a photon in the UV/vis region). The transition to higher electronic states requires higher energy to be absorbed, which can in some cases disturb weak interactions typical to interactions of biomolecules. Subtle variation of the spatial structure of interacting partners may also be out of experimental resolution of ECD. ECD and VCD are profitably used as complementary methods. Porphyrin–polypeptide and porphyrin-DNA interactions are very good examples: ECD is advantageously used to follow ICD in originally achiral planar porphyrin demonstrated in the Soret region caused by interaction with helical structures of polypeptides or DNA. VCD of this system characterizes structural variations of polypeptides and DNA in detail. Easily obtained ECD spectra are used first, followed by detailed VCD studies.

Generally, a protocol of how to use VCD for the studies of bimolecular interactions (similar to the determination of absolute configuration of small- or middle-sized molecules by VCD) has not been established. Rather, each problem needs specific procedure, technique, and solution as demonstrated in some examples in this minireview.

There are some common principles when using VCD on biomolecular interactions: It helps if the structure of interacting components, if chiral, are characterized by VCD. It is favorable if interacting molecules show characteristic signals in different spectral regions so that their spectral characteristic do not overlap. The isotope labeling can be used to shift characteristic vibration of one component and to open the spectral window for the other components. A study of ligand-induced secondary structure changes in calmodulin, in which the isotope labeling with C13 was used, serves as an example (“Protein–Protein Interactions” section). In the case of ICD, the spectral effects newly observed during interaction must be identified and characterized. Most VCD applications in the biomolecular interactions can be classified as empirical, although some attempts involving more advanced interpretation based on ab initio calculations appeared (“Guanine Tetramers and Quadruplexes” section). The specificity in using ab initio calculation on the study of biomolecular interactions can be summarized in the following points. Often, the size of at least one partner in biomolecular interaction is so big that it is not possible to optimize its structure so that the ab initio calculation could be performed. In this case, a simplification is applied. For example, helical structures of oligo or polynucleotides are replaced by very short segments. Special care should be paid in the order of such simplification so that it still includes all important properties for structural study. If the size of one of the components is suitable for the structure determination, the DFT calculation can be used to assign individual bands to specific parts of the molecule. Then, if some variations in spectra were observed during the interaction, these data could serve for localization of the part of the molecule involved in interaction (“Guanine Tetramers and Quadruplexes” section). In addition to the size of biomolecules, the narrow spectral region experimentally accessible, mostly given by strong absorption of used solvent, is unfavorable for using ab initio calculation. A few experimentally accessible VCD bands in the case of biomolecules, e.g., only carbonyl region, when compared with many tens of accessible bands in the case of rigid small- or medium-sized molecules, e.g., natural products, restrict reliability of a structural determination on the basis of correlation experimental and calculated spectra. Another weak point of using DFT calculation in structural studies of biomolecules is that aqueous solvents are used as the preferred environment. Sometimes, the explicit or implicit inclusion of solvents in calculation is needed. Summarizing, the use of DFT calculation on biomolecular interaction is restricted to some special cases. However, the experimental manifestation of biomolecular interaction in VCD spectra is very pronounced showing a number of new or changed spectral features, so it is hoped that the calculation of VCD spectra can also contribute to solve the structural aspects of biomolecular interactions.

INTERACTION OF DNA WITH BIOMOLECULES AND BIOGENIC METALS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. INTERACTION OF DNA WITH BIOMOLECULES AND BIOGENIC METALS
  5. GUANINE TETRAMERS AND QUADRUPLEXES
  6. BIOINTERACTIONS OF BILE PIGMENTS
  7. POLYPEPTIDE AND PROTEIN INTERACTIONS WITH OTHER BIOMOLECULES
  8. LITERATURE CITED

DNA alone attracted the VCD experts very soon16 when the VCD experiment was mature enough to study such a large system that requires an aqueous environment, buffered conditions, and salt. VCD was used by several groups in the study of natural DNA (often calf thymus, ctDNA) and the sequential oligonucleotides.15, 17, 18 The biomolecules studied frequently along with DNA are porphyrins and pharmaceuticals containing pyrrols and other heterocyclic compounds. Such studies are mostly motivated by the medical and pharmaceutical application of the DNA adducts: porphyrins are used in photochemical therapy, and many anticancer drugs serve on the basis of intercalation into DNA, thus causing selective inhibition of DNA replication. All the results discussed in this section use the concept of characteristic vibrations that localize the transition, observed as spectral bands, to special parts of DNA, especially to nucleic bases (even to their individual parts), sugars, or phosphates. A number of previously published VCD experimental materials is treated mostly empirically, without an attempt to model a structural consequence of interactions more exactly. Recently,19 the DFT calculation of IR absorption was performed on the guanosine monophosphate interaction with biogenetic metals by taking into account the implicit and explicit solvent effect of water. Such calculation can be taken as a first step, which can also expand VCD in the future.

Interactions with Porphyrins and Cyclic Derivatives

The predominant experimental techniques that were used in the numerous studies of porphyrin–DNA interactions involve ECD, and absorption in the UV region where nitrogen bases serve as chromophores, and in the Soret visible region where the porphyrin signal originates. The ECD and other results were reviewed recently.20 The VCD can provide complementary information, especially on the DNA part of the porphyrin–DNA complexes. The C[DOUBLE BOND]O, C[DOUBLE BOND]N, and C[DOUBLE BOND]C characteristic vibrations of nitrogen bases are observed in the spectral region 1800–1600 cm−1, where the individual IR bands can be assigned to the groups of atoms.21 Therefore, the conformational changes of DNA caused by porphyrin or pyrrol binding, especially via the intercalation of pyrrol derivatives can be observed, and the sites of the DNA that are involved in interaction can be identified. Accordingly, the binding site can be identified when the characteristic vibrations perturbed by the interaction is assigned to the particular groups; for example, C6[DOUBLE BOND]O in guanine or C2[DOUBLE BOND]O in cytosine, which are localized in the major or minor groove regions, respectively (Scheme 1). The other IR spectral region suitable to follow spatial variations caused by the interaction, especially of the ionic type, is the phosphate–sugar region, 1300–1000 cm−1, where PO2− and sugar vibrations localized in the phosphate and sugar parts of the DNA backbone, respectively, are observed. The published studies involve natural ctDNA and targeted sequential oligonucleotides, e.g., (dG-dC)10 and (dA-dT)10 composed of one type of base pair or special sequences typically involved in drug–DNA interactions.

thumbnail image

Scheme 1. Nitrogen base pairs adenine-thymine (A-T) and guanine-cytosine (G-C) forming major (M) and minor (m) groove.

Download figure to PowerPoint

The use of the VCD technique in empirical studies of DNA interaction is demonstrated by the example of DNA–porphyrin interactions. Representative porphyrins (Scheme 2) with different tendencies of interaction were used in these studies.

thumbnail image

Scheme 2. Structure of porphyrins used in studies.

Download figure to PowerPoint

Distinct VCD spectral differences between complexes of individual porphyrins with natural DNA (Fig. 1) demonstrate different modes of interaction: VCD spectra for Fe(III)TMPyP complexes reflect minimal perturbation of the nitrogen bases by the interaction.22 The intercalative binding mode of the free base porphyrin, TMPyP, and metal porphyrin without an axial ligand, Cu(II)TMPyP, was demonstrated as a decrease in intensity and a shift to lower frequencies of the VCD couplet at 1694(−)/1674(+) cm−1, which corresponds to the C6[DOUBLE BOND]O stretching of guanine coupled with the C2[DOUBLE BOND]O stretching of cytosine. Similar spectral features were also observed for the system Cu(II)TMPyP-(dG-dC)10 (Fig. 2).23 Because the groups involved in interactions are located in the major and minor groove, one can anticipate that the intercalation of porphyrin between GC base pairs influences both grooves.

thumbnail image

Figure 1. IR (bottom) and VCD (top) spectra of DNA-Cu(II)TMPyP (A), DNA-Fe(III)TMPyP (B), and DNA-TMAP (C) complexes at different c(DNA)/c(porphyrin) concentration ratios. Reproduced with permission from Novy et al., Collect Czech Chem Commun, 2005, 70, 1799–1810, © Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech Republic.

Download figure to PowerPoint

thumbnail image

Figure 2. IR (bottom) and VCD (top) spectra of (dG-dC)10 (A) and (dA-dT)10 (B) with Cu(II)TMPyP at different c(DNA)/c(porphyrin) concentration ratios. c(DNA) ∼ 52 mmol/L per base pair. Reproduced with permission from Novy and Urbanova, Biopolymers, 2007, 85, 349–358, © Wiley Periodicals, Inc.

Download figure to PowerPoint

The specificity of the interaction was studied using the oligonucleotides (dG-dC)10 and (dA-dT)10 as matrices.23 For Cu(II)TMPyP-(dG-dC)10 at a high porphyrin load, the absorption shifted from 1652 to 1644 cm−1, and the IR absorption band corresponding to the C2[DOUBLE BOND]O cytosine stretching overlapping with the porphyrin pyridinium vibration significantly increased in intensity. These observations cannot be explained by superposition of unchanged C2[DOUBLE BOND]O vibration and an increased concentration of pyridinium C[DOUBLE BOND]N. Rather, it is reminiscent of single-strand characteristic vibrations and can be explained by the partial disruption of the CG Watson-Crick bond, following the intercalation of the free base porphyrin and the metalloporphyrin without an axial ligand into the GC region of DNA.

When Cu(II)TMPyP at low concentrations interacts with (dA-dT)10, the vibrations at 1691 and 1661 cm−1 assigned to the C2[DOUBLE BOND]O of thymine situated in the minor groove and the “in-plane” thymine, respectively, are shifted when compared with pure (dA-dT)10. Only a slight intensity decrease was observed for the VCD couplet at 1671(−)/1655(+) cm−1 assigned to the C4[DOUBLE BOND]O vibration situated in the major groove site. Thus, VCD and IR illustrate that the interaction takes place predominantly in the minor groove site in this case. At higher porphyrin load, the VCD couplet at 1671(−)/1655(+) cm−1 changed significantly. These signals are assigned to the thymine vibration C4[DOUBLE BOND]O localized in the major groove, indicating that at higher porphyrin concentrations, the major groove site is also involved in the interaction. These results are complementary to the ECD observations showing that the (dA-dT)10 conformation most likely changed from the B- to the Z-form.23 The minor groove widths of the Z- and B-forms are 0.2 and 0.6 nm, respectively; thus, the minor groove of the Z-form becomes too narrow to interact with porphyrin. The complementary VCD and ECD spectral results suggest that in natural ctDNA, where both GC and AT regions coexist, the interactions with TMPyP and CuTMPyP are described well by the superposition of interactions observed for (dG-dC)10 and (dA-dT)10.

VCD is sensitive to the different modes of interaction between porphyrin TMAP with bulky mesosubstituents and natural ctDNA (Figs. 1 and 3),23 and especially pronounced changes are observed in VCD due to interaction with sequential oligonucleotides (dG-dC)10 and (dA-dT)10. The intercalation mode without substantial conformational change of the oligonucleotide matrix is indicated by a significant decrease in intensity without a change in the band positions in the case of (dG-dC)10. The only substantial change in the VCD pattern is observed at the high porphyrin load at 1633 cm−1, assigned to C2[DOUBLE BOND]O cytosine stretching localized in the minor groove. The original negative signal is changed to a positive one. This observation, together with the other spectral results, was interpreted as a manifestation of semi-intercalation,24 where the right-handed double helix is preserved and the wedged porphyrin causes an opening between two base pairs. The high porphyrin load in the TMAP-(dA-dT)10 complexes causes the shift of the absorption band assigned to the C2[DOUBLE BOND]O and in-plane vibration of thymine at 1690 and 1641 cm−1 to 1702 and 1634 cm−1. Such changes are characteristic of thymine being present in the single strand and can occur during semi-intercalation of TMAP, where the hydrogen bonds between thymine and adenine are broken. This suggestion is in agreement with molecular modeling.25 The observations in absorption are accompanied by a sign reversal of the originally positive couplet at 1699(−)/1684(+) cm−1 into a negative one. This result correlates with the UV-ECD and the Soret band ECD: both signals reverse the sign of their pattern, suggesting that the porphyrin stacking causes the B–Z transition of the DNA matrix. TMAP, in contrast to TMPyP, causes a significant perturbation of the standard B-form DNA.

thumbnail image

Figure 3. IR (bottom) and VCD (top) spectra of (dG-dC)10 (A) and (dA-dT)10 (B) with TMAP at different c(DNA)/c(porphyrin) concentration ratios. c(DNA) ∼ 52 mmol/L per base pair. Reproduced with permission from Novy and Urbanova, Biopolymers, 2007, 85, 349–358, © Wiley Periodicals, Inc.

Download figure to PowerPoint

The interaction of DNA with the heterocyclic drug daunomycin is surely among biologically important interactions. Daunomycin is a drug that has found wide application in cancer therapy. Its intercalation into DNA causes selective inhibition of DNA replication and RNA transcription. A detailed review of the VCD results with daunomycin complexes of six different octanucleotides was published recently,26 and it demonstrates the sensitivity of VCD in the nitrogen base region to intercalation of drugs into the different sequences of base pairs.

Interactions with Metal Ions

DNA–metal interactions are another group of biologically relevant interactions where VCD provided valuable information.21, 27–32 Among the metal–DNA adducts, cisplatin-oligonucleotide is interesting, because the compound cis-diamminedichlorplatinium(II) is a member of the widely used anticancer drug family, where the interaction with DNA plays a crucial role in inducing apoptosis. VCD was shown to sensitively monitor the structural changes of d(CCTG*G*TCC)·d(GGACCAGG) caused by cisplatin in the 1740–1500 cm−1 region and the phosphate spectral regions which are used for the oligonucleotide spectral studies (Fig. 4).32 Comparison of the vibrational spectra of the unmodified duplex and its cisplatin adduct reveals the conformational changes that take place upon binding. In connection with NMR and X-ray results, VCD in the nitrogen base region reveals the changes in conformation in the platinated complexes, especially in the frame of adjacent guanines during the intrastrand cisplatin interaction. Other spectral features are interpreted as an isomerization that leads to interstrand adduct. Minor changes of the main VCD couplet in the sugar–phosphate region (see Fig. 4) imply that the backbone has not been affected much by complexation with platinum.

thumbnail image

Figure 4. VCD spectra of the d(CCTG*G*TCC)·d(CCTGGTCC)-cisplatin complex in the base (A) and phosphate (B) regions as a function of time. Reproduced with permission from Tsankov et al., J Phys Chem B, 2003, 107, 6479–6485, © American Chemical Society.

Download figure to PowerPoint

Manganese(II) is an essential metal that mediates structural rather than enzymatic activity of DNA-binding proteins. Based on the combined use of VCD, IR absorption, ECD, and UV absorption, a model of Mn–DNA interactions was proposed.33 Characteristic features in VCD indicate that Mn2+ ions especially influence the vibration localized in guanine and cytosine, while there is no evidence of direct interaction of Mn2+ with N7 (Scheme 1). The increase of the intensity of the bands in the phosphate–sugar region upon interaction with metal ions demonstrates that phosphates are involved in the interaction. The observed results are interpreted as a reflection of the indirect interaction of Mn and DNA via water molecules.

VCD was used in the study of cfDNA condensation induced by Cr3+ ions.21 DNA condensation34 is a process of biological interest, e.g., in DNA packing in living cell. In vitro, this process was studied by spectroscopic and microscopic techniques. ψ-type spectrum (psi for “polymer and salt induced”) characteristic by an enhanced intensity and long “tail” in nonabsorbing regions was typically observed in ECD for condensed DNA. It was observed that the intensity enhancement in VCD also accompanied the DNA condensation. The spectral variations accompanied with the DNA condensation observed in VCD (Fig. 5) enable to identify that Cr3+ ions bind mainly to N7 in guanine and to phosphate groups, and DNA secondary structure remains in B-form.

thumbnail image

Figure 5. VCD spectra of DNA and DNA with Cr3+ ions at different [Cr]/[phosphate] ratios. All spectra are plotted in the same scale. Reproduced with permission from Andrushchenko et al., Biopolymers, 2002, 61, 243–260, © Wiley Periodicals, Inc.

Download figure to PowerPoint

DNA–Protein Complex

The spectroscopic study of complexes formed from DNA, the nonhistone chromatin protein HMGB1, and the linker histone H1 demonstrates the ability of VCD to provide structural information on large supramolecular complexes.35 This is a very challenging problem, because all the components overlap spectrally in the UV region (where amide groups and nitrogen base chromophores absorb), as well as in the 1700–1400 cm−1 region (where amide I and II of proteins and the nitrogen base vibrations are localized). The observed changes of C2[DOUBLE BOND]O vibrations of thymine and C2[DOUBLE BOND]O vibration of cytosine (Scheme 1) and weaker spectral variations of the vibrations localized in the major groove reveal that in the ternary complex, HMGB1 rather than H1 interacts predominantly with DNA bases. The role of histone H1 was identified as facilitating the interaction between HMGB1 and DNA, through binding of negatively charged groups of the DNA backbone and aspartic and glutamic acids of HMGB1. The presence of manganese(II) ions affects the structure of the supramolecular ternary complex DNA-HMGB1-H1 and was discussed in detail.35

GUANINE TETRAMERS AND QUADRUPLEXES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. INTERACTION OF DNA WITH BIOMOLECULES AND BIOGENIC METALS
  5. GUANINE TETRAMERS AND QUADRUPLEXES
  6. BIOINTERACTIONS OF BILE PIGMENTS
  7. POLYPEPTIDE AND PROTEIN INTERACTIONS WITH OTHER BIOMOLECULES
  8. LITERATURE CITED

The G-rich oligonucleotides may form G-tetramers in the presence of monovalent cations, in which each guanine base forms two hydrogen bonds with its neighbors.36 Depending on the sequence, the number of guanines, and the type of cations, G-tetramers can stack on top of each other and form a G-quadruplex structure.37, 38 G-quadruplexes occur in telomeres, specialized nucleotides located at the ends of chromosomes, playing a role in the aging of cells. VCD spectroscopy, in combination with ab initio calculations, can be used in the structural studies of these objects in solution, where the structure is not always the same as obtained from X-ray analysis of the crystal state. Indeed, there have been a few attempts to solve tetramer and quadruplex structures by VCD.39–41 Because of the complexity of molecular interaction, a simplification of the model used for DFT computation was adopted: instead of oligonucleotide, only tetramer and octamer were used.40 Another way to simplify the calculation is omitting the sugar part from the structure of tetramer and higher structures. However, involving the sugar part into the structure is important because sugar brings chirality into the planar nucleic base. The procedure of simplification is also described in this section.

G-Rich Sequential Oligonucleotides

VCD can clearly distinguish the molecular structure of (GGGA)5 in the presence of Li+ ions from that in the presence of K+ and Na+ ions (Fig. 6).39

thumbnail image

Figure 6. IR (bottom) and VCD (top) spectra of d(GGGA)5 in the presence of Li+ (solid lines), Na+ (dashed lines), and K+ (dotted lines) ions. c(ion) = 0.5 mol/L, ρ(d(GGGA)5 = 35 mg/mL (∼52 mmol/L per base pair). Reproduced with permission from Novy et al., Biopolymers, 2008, 89, 144–152, © Wiley Periodicals, Inc.

Download figure to PowerPoint

Apparently, the spectral pattern of G-rich oligonucleotides also differs from the spectra of natural DNA and of sequential oligonucleotides, for example of (dG-dC)10 and (dA-dT)10. The VCD obtained for the optimized structure of a nonplanar tetramer is in good agreement with the experimental spectra obtained for the Li+ adduct, while the experimental spectra of K+ and Na+ structures resemble optimized planar structure of G-tetramer. The simulation of VCD spectra obtained from the simplified model of the tetramer provides qualitative agreement with experimental data (Fig. 7).

thumbnail image

Figure 7. Comparison of experimental (bottom) and simulated (top) IR and VCD spectra of d(GGGA)5. Reproduced with permission from Novy et al., Biopolymers, 2008, 89, 144–152, © Wiley Periodicals, Inc.

Download figure to PowerPoint

VCD was also used for the investigation of quadruplex structures formed in the homopolynucleic acids polyriboguanylic (polyG) and polyriboinosinic acid (polyI).40, 42 Infrared absorption and VCD changes correlate with structural changes of both polymers. The computations reported here were performed in a rough manner and provide preliminary results.

A detailed structural study of the G-quartet was performed recently using guanosine-5′-hydrazide (G-hydrazide).41 This compound forms a stable hydrogel in the presence of alkali metal cations. It was proposed43 that gelation of the sample is caused by the formation of supramolecular aggregates consisting of G-quartets that may be stacked into columns by binding of monovalent metal cations between tetramers. VCD, having the advantage of exquisite sensitivity to the molecular structure of chiral molecular systems, clearly offers the possibility to characterize the structures of G-hydrazide, measured as a solution in DMSO-d6 and as a hydrogel in sodium phosphate/D2O buffer spectra, on the basis of their distinctively different VCD spectra (Fig. 8).

thumbnail image

Figure 8. IR and VCD spectra of guanosine-5′-hydrazide obtained from (a, exp) a solution in DMSO-d6 (dashed lines, 55 mmol/L), (a, calc) B3LYP/6-31G** VCD spectrum of monomeric compound, and (b) a gel state in deuterated sodium phosphate/D2O buffer (solid lines, 38 mmol/L). Adapted with permission from Setnicka et al., Langmuir, 2008, 24, 7520–7527, © American Chemical Society.

Download figure to PowerPoint

It was suggested that the strong enhancement of the VCD signal observed in a gel state originates from the formation of a highly ordered supramolecular structure, the G-quartets, and the very weak VCD of G-hydrazide in DMSO-d6 solution indicates the presence of monomeric molecules. The nonplanarity of the guanine skeleton is generally small, and thus guanine alone is considered to be optically inactive,44 especially in the UV region where the skeleton is the principal chromophore. In G-hydrazide, however, the guanine moiety is attached to the chiral ribose; therefore, a VCD signal can be induced and observed. On the basis of ab initio DFT at the B3LYP/6-31G** level, the structure of the monomer was optimized from several starting geometries, and the VCD spectrum was calculated for the lowest-energy conformers. In spite of a weak and noisy VCD spectrum caused by the weak inherent chirality of G-hydrazide, a very good agreement of the simulated and experimental spectra is evident (see Fig. 8). In addition, the combination of experimental and theoretical approaches allowed an assignment of the fundamental modes in the spectral range 1800–1600 cm−1. According to mass spectroscopy,43 the preferred species in the gel state is the G-tetramer. It must be hypothesized, however, that the aggregation in a gel state may involve dimers and other intermediates. The analysis of structures with different orientations of sugar-hydrazide moieties revealed that the spatial orientation of the ribose moiety has a negligible effect on the spectra in the region 1800–1600 cm−1, and was therefore omitted. The theoretical conformational analysis has shown that, among the studied structures, only cyclic tetramers lead to absorption and VCD spectra that correspond to the experiments. The optimized structure of the cyclic tetramer was significantly altered by the sodium cation positioned in the center of the G-quartet (Fig. 9, structure B). A near-planar arrangement of guanine bases gave rise to a slightly twisted structure, with close C4h symmetry. Very good agreement was observed between spectrum b and the experimental spectrum in Figure 9. The only feature not fully estimated by DFT simulation was band 6 at 1568 cm−1 found in experimental spectra. This band was assigned to the C[DOUBLE BOND]C and C[DOUBLE BOND]N vibrations of the guanine moiety, and the corresponding absorption peak was previously observed in highly ordered structures.45 Such disagreement can be understood, because staking between tetramers that influences the guanine skeleton vibrations was not considered in the optimized structure.

thumbnail image

Figure 9. B3LYP/6-31G** structures and spectra of G-quartets in the absence (a) and in the presence (b) of sodium cation, and experimental spectrum (exp) of gel. Top view (upper structures) and side view (bottom structures). Adapted with permission from Setnicka et al., Langmuir, 2008, 24, 7520–7527, © American Chemical Society.

Download figure to PowerPoint

BIOINTERACTIONS OF BILE PIGMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. INTERACTION OF DNA WITH BIOMOLECULES AND BIOGENIC METALS
  5. GUANINE TETRAMERS AND QUADRUPLEXES
  6. BIOINTERACTIONS OF BILE PIGMENTS
  7. POLYPEPTIDE AND PROTEIN INTERACTIONS WITH OTHER BIOMOLECULES
  8. LITERATURE CITED

Biliverdin (BV) is the first product of heme oxidation, and in birds, reptiles, and amphibians, it leaves the body as the predominant end product of heme degradation. In mammals, BV is reduced to BR, an orange cytotoxic pigment (Scheme 3). High levels of unconjugated BR in serum may cause kernicterus, damage to the brain centers of infants. Both bile pigments are poorly soluble in water. They form noncovalent complexes with serum albumin, and BR in this form is transferred to the liver, where it is further conjugated. Because of the crucial role of pigment complexation in neutralization of their toxicity, the interaction of BR with serum albumin and its model systems is one of the most studied albumin–ligand interactions.46–48

thumbnail image

Scheme 3. The structures of bilirubin-IX α (BR) and biliverdin-IX α (BV) showing the interconversion of the intramolecular hydrogen-bonded ridge/tile conformer of BR and the helical conformer of BV.

Download figure to PowerPoint

VCD was used49, 50 in studies of the structural aspects of the interactions of bile pigments in model systems. Because of fast racemization of pure pigments in solution, an appropriate chiral recognition must be involved prior to a spectral study by chiroptical methods. Polypeptides [poly(L-Lys), poly(D-Lys), poly(L-Arg)] and β-cyclodextrin recognize single enantiomers and at the same time serve as a model of pigment interactions with serum albumin and enzymes, respectively. Using β-cyclodextrin that recognizes the M-helical enantiomer of BR is advantageous for VCD, because there is minimal overlap between the pigment and matrix bands in the middle IR region. This enables identification of the negative band at ∼1560 cm−1, assigned to asymmetric COO, as a marker of M-helicity of the pigment part of the complex. Polypeptides are less advantageous systems from the spectral point of view, because amide vibrations overlap characteristic signals of BR, except for the COO marker. The sign of the band at ∼1560 cm−1 (Fig. 10) is opposite for poly(L-Lys) and poly(L-Lys) with dodecanoate ions and indicates that poly(L-Lys) recognizes the P-helical conformer of BR, while poly(L-Lys) with dodecanoate ions forms complexes with the M-helical conformer of BR.

thumbnail image

Figure 10. VCD spectra of complex bilirubin-poly(L-lysine) (BR+PLL), pH 10.8, and bilirubin-poly(L-lysine)-dodecanoate (BR+PLL+C12), pH 10.5. Reproduced with permission from Goncharova and Urbanova, Anal Bioanal Chem, 2008;392:1355–1365, © Springer-Verlag.

Download figure to PowerPoint

The same VCD marker can be used for the determination of the chiral structure of BV complexes with biogenic metals Cu, Cd, Mn, Zn, Fe, Ag (some of the spectra shown in Fig. 11).51 The sign of the main ECD and VCD patterns did not change for any of the studied metals other than Zn, for which an inversion of the signal sign was observed. For BV, the inversion of the ECD and VCD signals can be attributed to the opposite helicity of this pigment in the Zn-chelate. Spectral data show that complexation of BV with Zn takes place in two steps: initial formation of an olive-green complex with similar spectral properties to those of the Cd-chelate, followed by rapid inversion of the helicity and formation of the brown-green chelate. In the case of BR, the Zn-chelate did show an inversion signal in ECD, but did not show an inversion sign in VCD signal. This observation may be rationalized in terms of a flattening of the molecule, i.e., an increase in the angle between the pyrrinone chromophores without an inversion of helicity.

thumbnail image

Figure 11. VCD (A) and IR absorption spectra (B) of biliverdin bound to β-cyclodextrin and its chelates with Mn(III), Zn(II), and Fe(III).

Download figure to PowerPoint

POLYPEPTIDE AND PROTEIN INTERACTIONS WITH OTHER BIOMOLECULES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. INTERACTION OF DNA WITH BIOMOLECULES AND BIOGENIC METALS
  5. GUANINE TETRAMERS AND QUADRUPLEXES
  6. BIOINTERACTIONS OF BILE PIGMENTS
  7. POLYPEPTIDE AND PROTEIN INTERACTIONS WITH OTHER BIOMOLECULES
  8. LITERATURE CITED

The use of VCD to study the interactions of peptides and proteins has benefited from the pioneering studies of Keiderling's group and others on peptides alone,52 polypeptides,11, 53–55 and proteins56–61 under different conditions and in different solvents. The main conformations of the polypeptide chain, as α-helix, β-sheet, and random coil, possess characteristic VCD shapes (for a review see Refs.11 and61). VCD has also played a crucial role in identifying the random coil structure as local left-handed helixes related to the polyproline II structure, whose loose helix geometry is established by interaction with solvents rather than by intrastrand hydrogen bonding.11, 53, 54 Although the X-ray structure of the protein is probably the most detailed source of structural information, the solution structure may differ because of the flexibility of the protein that cannot be followed by X-ray, but can be established by VCD, as was demonstrated for the structure of α-lactalbumin and its homolog lysozyme.62

Most authors use the concept of characteristic vibrations and empirical rules when interpreting the observation accompanied by polypeptide and protein interactions. VCD yields more detailed information when compared with ECD due to more pronounced spectral manifestation of different peptide conformations in the amide I and amide II regions than is seen in UV/vis. In addition, isotopic substitution which can be localized in the molecule shifts the relevant characteristic vibration. Hence, the possible changes in conformation due to interactions can be localized in the molecule.

Interactions with Porphyrins

Studies of the interaction of peptides and proteins with porphyrins have been inspired by chlorophyll–protein complexes taking part in the first steps of photosynthesis, by hemoglobin, abundant porphyrin–protein complexes, and also by molecular systems used in photochemical therapy. The porphyrin–polypeptide complexes were studied using many methods, including VCD, which we discuss here. Different complementary peptide matrices are used for anionic and cationic porphyrins. Interactions with anionic meso-tetrakis(4-sulfonatophenyl)porphine (TPPS) was studied using (L-lysyl-L-alanyl-L-alanine)n, n = 1, 2 ((KAA)n), and the oligopeptide poly(L-lysine).63, 64 Interactions with the cationic porphyrins meso-tetrakis(1-methyl-4-pyridyl)porphin tetra-p-tosylate (TMPyP) and meso-tetrakis(a-trimethyl-ammonio-p-tolyl)porphine (TATP) was studied using poly(L-glutamic acid).64, 65 VCD can provide quite detailed information on the peptidic part of the complexes, while the porphyrin segment is more profitably observed using spectroscopies in the visible region. The interaction with porphyrins changed the spatial structure of matrices depending on the length of the polypeptide chain, as is evident from Figs. 12 and 13. For the tripeptide KAA, the amide I′ shape, typical for PPII segment (+/−), changes to a shape which is consistent with the signal of the oligopeptide in the β-sheet conformation. With the intrinsic sensitivity of VCD to peptide and protein structures having been shown in many previous studies, it is evident that the interaction caused a new type of well-organized structure, which is similar to β-sheet. Keeping in mind that the matrix is composed of three amino acid residues, the similarity is only local. The patterns of decreased intensity characteristic for both PPII structures and β-sheet were observed in the case of poly(L-lysine) consisting of about 10 aminoacid residues, while only marginal changes of the PPII pattern were observed for poly(L-lysine) consisting of hundreds of amino acid residues (see Fig. 13).

thumbnail image

Figure 12. IR and VCD spectra of KAA alone (black, pH 7.2) and in the presence of TPPS (red, pH 7.4) at c(KAA)/c(TPPS) = 1. c(TPPS) = 0.14 mol/L was kept constant. Reproduced with permission from Setnicka et al., J Porphyrins Phthalocyanines, 2008, 12, 1270–1278, © Society of Porphyrins and Phtalocyanines.

Download figure to PowerPoint

thumbnail image

Figure 13. VCD spectra of poly(L-lysine), MW ∼ 1250 (A) and 34,300, with TPPS in D2O, c(poly(L-lysine)) = 0.39 mol/L per aminoacid residuum. (A) c(poly(L-lysine))/c(TPPS), pH: (a) pure poly(L-lysine), MW ∼ 1250, 1.9; (b) 20, 6.8; (c) 5, 8.2. (B) c(poly(L-lysine))/c(TPPS), pH: (a) pure poly(L-lysine), MW ∼ 34,300, 6.2; (b) 60, 7.7; (c) 20, 8.2; (d) pure TPPS, 9.8. Reproduced with permission from Urbanova et al., Biopolymers, 2001, 60, 307–316. © John Wiley and Sons Inc.

Download figure to PowerPoint

The spatial structure of poly(L-glutamic acid), which serves as a model of the protein binding site for cationic ligands, can be easily tuned by pH at a concentration <0.025 mol/L per amino acid residue. In acidic pH, the PPII structure is converted into an α-helix. At a concentration about 10 times higher needed for VCD measurements, the addition of methanol or dioxan is necessary to prevent precipitation at acidic pH. It was shown65 that in the ternary system consisting of poly(L-glutamic acid), TPPS, and TATP, the interaction prevents the precipitation of poly(L-glutamic acid) at acidic pH, and the VCD spectrum in the amide I region can be simulated by the mixture of α-helical and PPII structures of polypeptide part of complex (Fig. 14).

thumbnail image

Figure 14. VCD and IR absorption spectra of poly(L-glutamic acid)-TATP-TPPS in D2O at pH ∼ 5 (a), simulated spectra (b). VCD and IR absorption spectra of poly(L-glutamic acid)-TATP in 40% methanol-d6/D2O at pH ∼ 5 (c) and pH ∼ 8 (d). c(polypeptide per aminoacid residuum)/c(porphyrin) = 100. Reproduced with permission from Palivec et al., J Pept Sci, 2005, 11, 536–545, © European Peptide Society and John Wiley and Sons, Ltd.

Download figure to PowerPoint

Biotin (hexahydro-2-oxo-1H-thieno[3,4-d]imidazole-4-pentanoic acid) is a water-soluble vitamin belonging to the B-complex. Its interaction with avidin, a basic tetrameric glycoprotein, belongs to the strongest protein–ligand interactions and was studied by IR, X-ray, and also VCD.66 The temperature-dependent VCD spectra and 2D-VCD correlation spectroscopy reveal the results on structure of complex, its temperature stability, and also on sequences involved in the unfolding of alone protein avidin. It was shown that although the structure of avidin–biotin complex is not significantly different from that of avidin, the binding of biotin increases the number of nonequivalent amide groups. The temperature dependence of VCD for avidin and avidin–biotin complex is different (Fig. 15) and reveals that the complex does undergo some reversible unfolding with increasing temperature, but unlike avidin does not undergo cooperative structural transition. The unfolding of avidin with temperature involves the disruption of β-sheet structure which is followed by the formation of antiparallel β-strand, leading to aggregation of protein.

thumbnail image

Figure 15. IR absorption (a) and VCD (b) spectra of avidin (A) and avidin–biotin complex (B) from 35 to 95°C, increasing temperature in 10°C interval, protein concentration ∼10% (w/v). Reproduced with permission from Wang and Polavarapu, J Phys Chem B, 2001, 105, 7857–7864. © American Chemical Society.

Download figure to PowerPoint

Protein–Protein Interactions

The difficulties in using VCD in structural studies of proteins, where different structural elements contribute to the entire molecule, are caused by the overlap of the characteristic spectral features typical to α-helix, β-sheet, PPII, etc., in the amide I and amide II spectral range. The set of chosen globular proteins with the known X-ray structures, so-called training set of proteins, were characterized by principal component analysis of their VCD, ECD, and IR spectra by Pancoska et al.,56, 67–70 and it has been shown that such analysis enables the estimation of the relative content of typical structures in the molecule.71 The difficulties caused by spectral overlap are even exaggerated in the case of protein–protein interactions.

VCD was effectively used for the characterization of intrapeptide interaction of avian pancreatic polypeptide (aPP).72 This interaction was realized in an equimolar mixture of two polypeptide segments in D2O and in 30% TFE in buffered D2O. The first segment was composed of the first 11 amino acid residues, aPP (1-11), showing spectral signature of non-α-helical structures. The second one was the acetylated segment of the 25 amino acid residues, Ac-aPP (12-36), showing substantial α-helical content. The VCD and ECD studies demonstrated that both fragments refold into the conformation of the full aPP when combined in an equimolar mixture. The refolded structure shows a different spectrum than that of the sum of the individual fragments. VCD reflects that the folding process changes the structure of the individual components.

Isotope substitution was successfully applied to structural studies of peptides whose residues were selectively labeled by isotopes,73–75 and detailed information about the peptides was obtained. Isotope labeling with 13C was also used to study the induction of conformational changes in the peptide. Calmodulin (CaM), a small calcium regulatory protein of 17 kDa, in the presence of Ca2+ ions, a synthetic peptide derived from the CaM-binding domains of smooth muscle myosin light chain kinase (smMLCKp), and intact protein apo-lactoferrin (apo-Lac) were studied.76 The study of the described interactions also uses the characteristic couplets observed in VCD for α-helix and PPII structures. It was first shown that the 13C labeling shifts the amide I′ couplet by about 43 cm−1 and opens the spectral window for other unlabeled ligands with signals in the same region. The interaction of Ca2+-13C-CaM with smMLCKp further increases the α-helical couplet of CaM. A clear separation of signals originating from two different proteins, one labeled with 13C, was achieved for apo-13C-CaM and unlabeled apo-Lac.

VCD was used in the spectroscopic investigation of the interaction of four terpene trilactones from Ginkgo biloba with the peptide Aβ(25-35) that serves as a model of amyloid beta peptide (Aβ).77 Such interaction is studied because of the possible inhibition of the pathological aggregation of Aβ by G. biloba, which is used in the “treatment” of dementia. The trilactones were followed by the band at ∼1790 cm−1, which is assigned to carbonyl groups and their couplings. The peptide was monitored by IR and VCD using predominantly the band at ∼1618 cm−1, which is assigned to β-sheet structure for the peptide alone in methanol. The signals of both studied components are well resolved. Under the conditions used for this study, 50/50 v/v mixture of EtOH/D2O, a systematic change of the trilactone terpene IR and VCD at ∼1790 cm−1 was not observed. A slight increase with time in both IR absorption and VCD at ∼1618 cm−1 was observed without any variation in the band shape or position. Considering the high sensitivity of VCD to conformational changes, this observation shows that the mutual influence of the interaction on both components is very small. Therefore, the authors rule out a direct therapeutic effect of the studied terpenes on the fragments of Aβ protein.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. INTERACTION OF DNA WITH BIOMOLECULES AND BIOGENIC METALS
  5. GUANINE TETRAMERS AND QUADRUPLEXES
  6. BIOINTERACTIONS OF BILE PIGMENTS
  7. POLYPEPTIDE AND PROTEIN INTERACTIONS WITH OTHER BIOMOLECULES
  8. LITERATURE CITED