Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (3)
Right arrowRequest Permissions
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Guan, Y.
Right arrow Articles by Thomas, G. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Guan, Y.
Right arrow Articles by Thomas, G. J., Jr
Social Bookmarking
 Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Nucleic Acids Research Pages 3892-3999  


Structural polymorphism and Raman conformation markers of cyclic deoxytriadenylic acid
Introduction
Materials And Methods
   Sample preparations
   Raman spectroscopy
Results And Discussion
   R32 crystal structure of c(dAp)3
   Raman spectroscopy of the R32 crystal structure and solution structure of c(dAp)3
   Rationalizing the Raman and X-ray results on c(dAp)3 crystal structures
Summary And Conclusions
Acknowledgements
References


Structural polymorphism and Raman conformation markers of cyclic deoxytriadenylic acid

Structural polymorphism and Raman conformation markers of cyclic deoxytriadenylic acid

Yifu Guan, James M. Benevides, Y. G. Gao1, Andrew H.-J. Wang1 and George J.Thomas, Jr*

Division of Cell Biology and Biophysics, School of Biological Sciences, University of Missouri, Kansas City, MO 64110, USA and 1Biophysics Division and Department of Cell and Structural Biology, University of Illinois, Urbana-Champaign, Urbana, IL 61801, USA

Received June 22, 1998; Accepted July 22, 1998

ABSTRACT

X-ray analysis of two different trigonal crystal forms (space groups R32 and P3) of cyclic deoxytriadenylic acid [c(dAp)3] indicates for each an asymmetric unit consisting of two conformationally similar c(dAp)3 molecules. Raman spectroscopy supports the X-ray interpretation for the R32 crystal, but identifies another c(dAp)3 conformation not revealed in the P3 X-ray structure. The results for the P3 crystal can be explained if an additional c(dAp)3 conformer is present but not sufficiently ordered within the lattice to contribute to X-ray diffraction. The Raman signature of aqueous c(dAp)3, which differs from signatures of both the R32 and P3 crystals, exhibits backbone markers similar to those of thermally denatured DNA and indicates that c(dAp)3 molecules in solution populate a wider range of phosphoester ring conformations than in R32 and P3 crystals. Thus, polymorphism is observed for both crystal and solution structures of c(dAp)3. The results imply a highly flexible phosphoester ring that may be relevant to the function of cyclic oligonucleotides as biological effectors. A novel Raman marker at 821 cm-1 is demonstrated as diagnostic of phosphoester torsions [alpha] and [zeta] in the gauche+ range. Specific Raman markers are also identified for the S-type (C2[prime]-endo) deoxyadenosine conformations that occur in R32 and P3 crystal structures of c(dAp)3.

INTRODUCTION

Raman spectroscopy is a versatile probe of the conformations, interactions and dynamics of DNA and RNA. Theoretical and practical considerations in application of the method to nucleic acids and their complexes have been discussed in recent reviews (1,2). A useful approach is to combine Raman spectra of oligonucleotide single crystals with X-ray structure determinations to establish libraries of Raman bands diagnostic of specific nucleotide conformational parameters (3-5). The numerous Raman markers established in such investigations provide a definitive basis for identifying furanose conformation, glycosyl torsion and phosphodiester geometry in large DNA and RNA structures and their complexes (1,6-11), as well as in related nucleoside and nucleotide derivatives (12,13). A similar strategy has led to the identification of Raman bands diagnostic of quadruplex structures formed by guanine- and cytosine-rich DNA sequences (14,15). Raman markers of the nucleotides also provide a convenient means of monitoring the participation of specific base residues in nucleic acid dynamics, including structure transformations (16) and hydrogen-isotope exchange phenomena (17-19). It is appropriate to enlarge the existing database of Raman markers of nucleic acid structure by combined X-ray/Raman analysis of additional model nucleotides.

In this communication, we report a Raman spectroscopic analysis of the nucleic acid oligomer, cyclic deoxytriadenylic acid [c(dAp)3], in aqueous solution and in two different trigonal crystal forms (space groups R32 and P3). Raman spectra of the c(dAp)3 crystals are evaluated in conjunction with recent X-ray structure determinations (to 1.0 Å resolution) (20). In the R32 and P3 crystals, independent molecules of both crystal lattices possess perfect 3-fold symmetry, with the molecule shaped like a shallow bowl of diameter ~15 Å and thickness ~4 Å. Both crystal forms contain molecules in which the adenine base planes are roughly perpendicular to the 3-fold axis, as shown in Figure 1 (20).


Figure 1. Structure of the cyclic deoxytriadenylic acid molecule [c(dAp)3] crystallizing in space group R32, as viewed along the crystallographic c axis. The molecule has perfect 3-fold symmetry. The structure depicted (designated as conformer A1 in Table 1) exhibits the glycosyl torsion in the anti range([chi] = -143.9°) for each deoxyadenosine residue and is one of two conformationally similar molecules present in the crystallographic asymmetric unit. The second conformer of the asymmetric unit (A2, Table 1) differs appreciably only by virtue of having the glycosyl torsion in the high-anti range ([chi] = -106.9°). From data of Gao et al. (20).

The crystal and solution structures of cyclic oligonucleotides are of interest for several reasons. First, although cyclic mononucleotides and dinucleotides can participate in a number of important biological functions, such as transmission of cellular information, inhibition of polymerase activity and activation of biosynthesis, cyclic trinucleotides appear to be of markedly lower biological activity. The structural basis for this difference in activity is not known. Second, we expect cyclic oligonucleotides, including c(dAp)3, to be useful for assessing the extent to which the vibrational Raman signature is sensitive to well-defined changes in phosphoester geometry and to eversion of the attached base residues. The covalent closure of the phosphoester ring may introduce unusual intramolecular constraints and provide an opportunity for novel intermolecular interactions involving adenines in the crystal lattice. The Raman spectrum should be highly sensitive to such phenomena. Finally, small oligonucleotides serve as models for larger nucleic acids. The structures of cyclic oligonucleotides, in particular, may be relevant to folding intermediates involved in the condensation and packaging of genomic DNA in chromatin and viral capsids.

High resolution structures of cyclic dinucleotides of adenine and guanine have been reported previously (21-24). These studies have provided evidence of considerable conformational diversity in the cyclic phosphoester backbone and have revealed novel base-stacking mechanisms. Here, we demonstrate that the crystal and solution structures of c(dAp)3 are different from one another and manifest additional novel conformations for the nucleic acid phosphoester moiety. The c(dAp)3 molecule is also found to accommodate unusual glycosyl torsions, while its everted base residues apparently do not participate in significant base stacking interactions.

MATERIALS AND METHODS

Sample preparations

Cyclic trideoxyadenylic acid, c(dAp)3, was synthesized according to published procedures (25). The molecule crystallized in space group P3 from a solution containing 2.5 mg trinucleotide, 12.5 mM MgCl2, 6 mM glycine (pH 4.5) and 7% 2-methyl-2,4-pentanediol (2-MPD). The R32 form was crystallized similarly, except that 20 mM CoCl2 replaced 12.5 mM MgCl2. Crystal data are as follows. P3 form: chemical formula C30H33N15O15P3; a = b = 23.328 Å (2) and c = 9.680 Å (3); 3059 observed reflections with final R-factor 0.109. R32 form: chemical formula C30H33N15O15P3; a = b = 22.638 Å (2) and c = 44.58 Å (1); 2159 observed reflections with final R-factor 0.137. Further details of the crystallographic structure determinations are given by Gao et al. (11).

Raman spectroscopy

Single crystals of c(dAp)3, authenticated by X-ray diffraction to be either in the R32 or P3 space group, were sealed with mother liquor in quartz capillary cells for Raman spectroscopic analysis. Capillaries were mounted on a thermoelectrically cooled (6°C) stage of an Olympus (Lake Success, NY) model BHSM microscope, which was optically coupled to an ISA/Jobin-Yvon (Edison, NJ) model 3000 Raman triple spectrograph and charge-coupled-device detector. Raman spectra were excited with the 514.5 nm line of a Coherent (Santa Clara, CA) Innova 70-2 argon laser. An 80X Olympus ULWD-MS-plan objective served to focus both the incident laser beam on the sample and the Raman scattered radiation on the entrance slit of the spectrograph. Radiant power at the crystal surface was maintained below 15 mW. Raman spectra presented below are averages of 20 or more exposures, each obtained with an integration time of 120 s. Data collection and processing were performed using ISA/Jobin-Yvon software. Other details of sample handling and data collection for Raman microspectroscopy of crystalline samples have been described (26).

Solutions of c(dAp)3, poly(dA) and poly(rA) were prepared at concentrations of ~30 mg/ml at neutral pH. Spectra were collected on a Spex (Metuchen, NJ) model 1877 Triplemate Raman spectrometer, equipped with a Princeton Instruments (Princeton, NJ) charge-coupled-device detector. Typically, 10 exposures, each of 120 s duration, were accumulated and averaged to generate the Raman spectra of solutions presented below. Further details of the instrumentation and data processing have been described (27).

Table 1. X-ray structural parameters of cyclic deoxytriadenylic acid in R32 and P3 space groups and corresponding parameters for double-stranded A DNA and B DNAa
Torsion R32 Crystal P3 Crystal A DNA B DNA
A1 A2 A1 A2
[alpha] 55.0 55.8 67.1 65.5 -70 -60
[beta] 160.3 163.9 167.0 167.1 180 180
[gamma] 44.3 46.0 49.2 54.1 60 50
[delta] 140.7 140.0 144.0 143.5 80 140
[epsis] -151.9 -155.2 -165.8 -164.8 160 180
z 60.3 62.9 56.4 53.2 -80 -90
[chi] -143.9 -106.8 -143.8 -141.0 -159 -119
aData on R32 and P3 crystal forms of c(dAp)3 are from reference 20, and data on A DNA and B DNA are from reference 28 and citations therein. A1 and A2 refer to the two c(dAp)3 molecules per asymmetric unit. Angles are in deg units.

RESULTS AND DISCUSSION

R32 crystal structure of c(dAp)3

The X-ray structure of c(dAp)3, as viewed along the c axis of the R32 crystal, is represented in Figure 1. Each nucleotide of the cyclic trimer is characterized by virtually the same set of torsion angles, [alpha], [beta], [gamma], [delta], [epsis], [zeta] and [chi], which are compared in Table 1 with corresponding torsions observed in A DNA and B DNA oligonucleotide crystal structures (28). Included in Table 1 are torsions for the P3 crystal structure (20), which is further discussed, below. The data of Table 1 indicate that in the R32 structure the deoxyadenosine (dA) furanose pucker ([delta]) corresponds to the S type (or C2[prime]-endo family) and the glycosyl torsion ([chi]) is in the anti range, similar to deoxynucleoside conformations associated with B DNA. Torsions [beta], [gamma], [delta] and [epsis] in the R32 structure also assume values that are similar to those of B DNA. On the other hand, torsions [alpha] and [zeta] are both in the gauche+ range (g+/g+ conformation), rather than in the g-/g- conformation typical of DNA duplexes (28).


Figure 2. Polarized Raman spectra of an oriented single crystal of c(dAp)3 (R32 space group). The I^ spectrum (upper trace) was obtained with electric vectors of incident and scattered radiation perpendicular to the crystallographic c axis and the I|| spectrum (lower trace) with the corresponding electric vectors parallel to c. Spectra were excited at 514.5 nm and corrected for contributions from the mother liquor.

On the basis of the X-ray determined structure (Fig. 1), the Raman spectrum of the R32 crystal is expected to exhibit marker bands diagnostic of C2[prime]-endo/anti deoxyadenosine and g+/g+ phosphodiester conformation. Further, the equivalency of backbone torsions in all nucleotides of the cyclic trimer is expected to produce Raman bands of relatively narrow width in comparison to the larger Raman bandwidths observed for other B DNA oligonucleotide crystals (7).

Raman spectroscopy of the R32 crystal structure and solution structure of c(dAp)3

General characteristics of the spectra. Figure 2 shows the polarized Raman spectrum of an oriented single-crystal of c(dAp)3 of space group R32. The I^ spectrum (upper trace) was obtained with electric vectors of incident and scattered radiation perpendicular to the crystallographic c axis, and the I||spectrum (lower trace) was obtained with the corresponding electric vectors parallel to c. In Figure 3, the spectrum of c(dAp)3 in H2O solution is compared with solution spectra of poly(rA) and poly(dA), as well as with the R32 crystal spectrum. For comparison with the solution spectrum, Figure 3 includes the sum of traces I^ and I|| (from Fig. 2), which more appropriately simulates the Raman signature of unoriented molecules. Table 2 lists the Raman frequencies and assignments for all forms of c(dAp)3 investigated here, including R32 and P3 crystals, amorphous powder, H2O solution and D2O solution. Depolarization ratios ([rho]) are also given for prominent Raman bands in the solution spectra. (Note that solution depolarization rates are not to be confused with polarized Raman intensities of oriented single crystals.)


Figure 3. Comparison of Raman spectra of the c(dAp)3 crystal (R32) and solution with Raman spectra of model polynucleotides. (A) Orientationally averaged spectrum (I^+ I||) of the R32 crystal from Figure 2. (B) Spectrum of c(dAp)3 in H2O solution. Perpendicular and parallel components of the solution spectrum of c(dAp)3 were also obtained (data not shown), and band depolarization ratios ([rho]) were calculated, as listed in Table 2. (C) Spectrum of poly(rA) in H2O solution. (D) Spectrum of poly(dA) in H2O solution. Solution spectra were excited at 514.5 nm from solutions containing c(dAp)3 at 40 mg/ml and pH 7.


The data of Figure 2 show that most of the prominent Raman bands of the R32 crystal are sharp, as expected for a homogeneous crystal structure. Figure 2 shows further that the majority of prominent Raman bands are considerably more intense in the I^ spectrum than in the I|| spectrum (i.e. I^ >> I||). Notable examples are the bands at 661, 729, 1204, 1252, 1309, 1342, 1378, 1483, 1509 and 1580 cm-1, all of which originate from in-plane vibrations of the adenine bases (29-33). These bands typically exhibit Raman tensors with the largest component in the plane of the base (26,34). The observed polarizations are consistent with the R32 structure of Figure 1, in which the planes of adenine residues of c(dAp)3 are close to perpendicular to the crystallographic c axis.

Table 2. Raman frequencies and assignments for cyclic deoxytriadenylic acid in crystalline (P3, R32), powder, and solution (H2O and D2O) states and solution depolarization ratios ([rho])a >
Crystal Powder Solution Assignmentb
R32 P3 H2O [rho] D2O [rho]
625             dAc,d
  640           dAd
661 665 652 656 0.50 654 0.36 dAd
729 732 727 727 0.11 720 0.10 [nu](ring)
793 790 789 790 0.07 790 0.06 [nu](OPO)
  813           [nu](OPO)
821             [nu](OPO)
838 838           [nu](OPO)
  870       854 0.11 d
892 894 882 883 0.20 884 0.46 d
927 921 907 911 0.63 907 0.90 d
1005 1013 1008 1008 0.22 1008 0.78 [delta](N6H2)
1068 1062 1059 1059 0.48 1045 0.57 d
1100 1097 1097 1092 0.06 1092 0.06 [nu]s(PO2-)
1112              
1204 1201   1204        
1215     1222        
1252 1252 1249 1250 0.34 1256 0.32 [nu](N1-C2)+[delta](C2H)
1309 1308 1303 1306 0.35 1305 0.18 [nu](C8-N9)+[nu](C2-N3)
1342 1340 1335 1337 0.25 1341 0.30 [delta](C2H)+[nu]((C2-N3)
1378 1382 1374 1376 0.26 1380 0.24 [nu](C1[prime]-N9)
1421   1417 1420 0.39 1420 0.39 [delta](C2[prime]H2)
1442 1441   1458 0.40 1458 0.36 [delta](C5[prime]H2)
1483 1485 1478 1481 0.15 1482 0.17 [delta](C8H)+[nu](N7-C8)
1509 1510 1506 1508 0.29 1518 0.34 [delta](C2H)+[nu](N2-C3)
1580 1582 1576 1580 0.63 1576 0.63 [nu](C4-C5)+[nu](C5-C6)+[nu](C5-C7)
          1623 0.67 [nu](ring)
1650 1661           [nu](ring)+[delta](N6H2)
aFrequencies are in cm-1 units. The depolarization ratio ([rho]) is listed for prominent bands in the solution spectra.
bFrom references 30 and 33.
cd, deoxyribose; s, symmetric; [nu], stretching mode; [delta], in-plane bending mode.
dCoupled vibration involving both adenine and sugar moieties.

Raman bands of Figure 2 that are not highly polarized in favor of the c direction (i.e. I^ ~ I||) can be attributed to vibrational modes either not localized in the plane of the adenine ring or characterized by intrinsically isotropic Raman tensors. The Raman band due to the phosphodioxy stretching mode near 1090-1100 cm-1, for example, has been shown to have an isotropic tensor in both A DNA and B DNA (34), and is therefore expected to exhibit I^ ~ I||. Figure 2 confirms that this is the case. Conversely, the polarization of the companion band at 1112 cm-1 suggests an altogether different Raman tensor, with its largest component perpendicular to c. This band, which is observed only in the R32 crystal, is tentatively assigned also to c(dAp)3 phosphates, possibly reflecting a perturbation associated with the presence of CoCl2 in the crystallization liquor.

Raman markers of deoxyadenosine conformation. The Raman spectrum of the R32 crystal is expected to be informative of the deoxyadenosine sugar ring pucker and glycosyl torsion in c(dAp)3. In Figure 3, the spectrum of the R32 crystal (Fig. 3A) is compared with spectra of the polynucleotide models, poly(rA) (Fig. 3C) and poly(dA) (Fig. 3D), which contain dA conformations in the C3[prime]-endo/anti (640 cm-1 Raman marker) and C2[prime]-endo/anti (661 cm-1 Raman marker) families, respectively (32). Thus, the 661 cm-1 band in the R32 spectrum is consistent with the C2[prime]-endo/anti dA conformation, in accord with the X-ray structure. The polarization of the 661 cm-1 band (I^ >> I||) is as expected for an in-plane adenine ring vibration in the R32 structure of Figure 1. The solution spectrum of c(dAp)3 (Fig. 3B) indicates that the C2[prime]-endo/anti dA conformer (~660 cm-1 Raman marker) also occurs in the solution structure.

Table 3. Raman markers of backbone and nucleoside conformations in c(dAp)3 and DNA
Marker type c(dAp)3 B DNAa A DNAa Z DNAb
R32 P3 Solution
Backbone
O-P-O 793 790 790 790 ± 5    
  813       807 ± 3  
  821   838     835 ± 7
PO2- 1100 1097 1092 1092 ± 1 1099 ± 1 1095 ± 2
2[prime]-CH2 1421 1413 1420 1422 ± 2 1418 ± 2 1425 ± 2
5[prime]-CH2 1441 1441 1458 1465 1463 (1425)
Deoxyadenosine
C2[prime]-endo/anti 661 665 656 663 ± 2    
C3[prime]-endo/anti   640     644 ± 4  
C2[prime]-endo/high-anti 625          
C3[prime]-endo/syn           624 ± 3
aFrom reference 37.
bFrom reference 46.

The X-ray structure shows further that a second c(dAp)3 conformer with high-anti glycosyl torsion exists in the R32 lattice (Table 1). The 625 cm-1 Raman band, with I^ ~ I|| (Figs 2 and 3A), exhibits the frequency, intensity and polarization properties compatible with assignment to a dA marker for which the plane of the adenine ring is not perpendicular to the crystallographic c axis. The 625 cm-1 band is therefore assigned to the population of c(dAp)3 molecules with high-anti ([chi] = -106.8°) glycosyl torsion. It is of interest to note that a similarly weak 625 cm-1 band is present in H2O and D2O solution spectra of c(dAp)3 (Fig. 3B and Table 2). No similar band occurs in spectra of poly(rA) (Fig. 3C) or poly(dA) (Fig. 3D). All of these findings are consistent with a high-anti dA conformation populating the R32 crystal and solution structures of c(dAp)3, but not the model polynucleotides. Further support for this assignment scheme is given below. Interestingly, an adenine ring mode near 625 cm-1 has also been reported in an oligonucleotide crystal containing a syn conformer of dA (35).

Raman bands of the 750-1100 cm-1 interval include putative markers of the phosphodiester g+/g+ conformation. By analogy with previous studies of DNA crystals and fibers (1,34), the well resolved Raman bands of the R32 crystal at 793, 821, 838, 892, 927, 1005, 1068 and 1100 cm-1 can be assigned to vibrational modes localized in the sugar-phosphate moiety of c(dAp)3. Table 3 shows that these c(dAp)3 backbone markers differ from those observed for any canonical DNA structure, although the differences are least for B DNA. This is consistent with the X-ray results, which show that all but two of the backbone torsion angles of c(dAp)3 fall within ranges characteristic of B DNA. Yet, it is clear that the 821 cm-1 band has no close counterpart in B DNA. Therefore, we propose this band as diagnostic of the unusual g+/g+ conformation found in the R32 crystal structure. Insensitivity of the 821 cm-1 band to deuteration of the crystal (data not shown) is in accord with its assignment to a vibrational mode localized in the cyclic phosphoester group.

The solution spectrum of c(dAp)3 (Fig. 3B) shows considerable divergence from that of the R32 crystal throughout the region 750-1100 cm-1. The partially resolved band of the crystal at 821 cm-1 (with a possible weak shoulder at 838 cm-1) is not apparent in the solution spectrum; additionally the sharp and intense peak observed at 793 cm-1 in the R32 crystal is replaced in the solution spectrum by a broad band centered near 790 cm-1. The latter is reminiscent of the broad Raman band near 790 cm-1 in thermally denatured DNA (16). Raman bands of the deoxyribose group at 892 and 927 cm-1 in the crystal are also weaker and broader in the solution spectrum, again reminiscent of heat-denatured DNA. These findings are consistent with a large distribution of sugar conformations in the solution structure, as opposed to the highly regular furanose conformation of the crystal structure. Specific assignments for the deoxyribose Raman bands of the 850-950 cm-1 interval have been given previously (34,36,37). The peak at 927 cm-1 and its accompanying high-frequency shoulder (~940 cm-1) are displaced from the corresponding band of B DNA (Table 3) and, like the 821 cm-1 band, are considered diagnostic of the phosphodiester g+/g+ conformation found in the R32 crystal structure.

Raman bands of the 1100-1700 cm-1 interval are informative of adenine interactions. Raman frequencies above 1100 cm-1 are due to in-plane ring vibrations of the adenine residue and have been discussed in detail elsewhere (29,38). Key assignments are listed in Table 1. Many Raman bands assigned to ring vibrations of DNA bases are hypochromic, i.e. the Raman intensities decrease when the bases are stacked in the native DNA secondary structure. Raman intensity profiles for selected bands of the c(dAp)3 crystal (R32) and solution (H2O) are compared with one another and with those of aqueous poly(dA) in Figure 4. Each of these structures contains the C2[prime]-endo/anti dA conformation. With few exceptions, the c(dAp)3 solution exhibits the highest Raman intensities, indicating the least base stacking. Base stacking is intermediate in the R32 crystal structure of c(dAp)3 and greatest in aqueous poly(dA). [In the few instances where Raman intensities are higher for the crystal than for the solution of c(dAp)3, the bands in question are due to the cyclic phosphoester ring rather than to the adenine residue, as noted above.]


Figure 4. Representation of relative Raman intensities of the R32 crystal of c(dAp)3 (solid black), H2O solution of c(dAp)3 (light stippling), and H2O solution of poly(dA) (dark stippling). Intensities are normalized to the band near 1100 cm-1, which is assigned to the symmetric stretching vibration of the PO2- group.

In DNA of mixed base composition, Raman intensity at 1421 cm-1 is due primarily to a deformation vibration of the deoxyribose C2[prime] methylene group, with a relatively small contribution from an overlapping adenine ring vibration (34). In c(dAp)3, however, the two overlapping contributors are comparably intense. This explains the surprisingly high intensity of the 1421 cm-1 band as well as its polarization (Fig. 2). On the other hand, the band is not particularly sensitive to amino group deuteration (Table 2), indicating that the involved adenine mode is localized within the purine heterocyclic ring.

Several bands in this region, including those at 1483, 1509, 1605 (weak shoulder) and 1650 cm-1, may serve as markers of hydrogen bonding interactions involving donor (N6H2) and acceptor (N1, N3, N7) sites of adenine (39). The small differences in frequencies of all of these bands in the crystal (Fig. 2) vis-à-vis the solution (Fig. 3B) are consistent with slightly different hydrogen-bonding states for adenine donor and acceptor groups in crystal and solution structures.

A putative adenine hydrogen bonding marker near 1005 cm-1. The moderately intense band near 1005 cm-1 in both crystal and solution spectra of c(dAp)3 is virtually eliminated by deuteration (Table 2). Like other deuteration-sensitive bands of c(dAp)3, it is likely due to a vibration involving the adenine N6 amino group. Assignment of the 1005 cm-1 band to the adenine residue is consistent with its polarization properties (Fig. 2) and with its appearance in 257-nm excited UVRR spectra of c(dAp)3 (Z. Q. Wen and G. J. Thomas, Jr, unpublished results). The slightly different frequencies observed for the band in crystal and solution structures may reflect somewhat different amino-group hydrogen-bonding states in these structures, as suggested above.

Comparison of Raman spectra of the P3 and R32 crystals of c(dAp)3. Figure 5 compares Raman spectra of P3 (top) and R32 (bottom) crystals of c(dAp)3, and clearly demonstrates different molecular conformations in the two crystalline forms. Prominent backbone markers occur at 790, 813 and 838 cm-1 for P3, rather than at 793 and 821 cm-1, as observed for R32 (Table 1). The band at 813 cm-1 in the P3 crystal is similar to backbone markers observed for RNA (41) and A DNA (37,42). It suggests a phosphoester geometry similar to that occurring in nucleic acid structures of the A form. The P3 crystal also exhibits an adenosine marker at 640 cm-1, diagnostic of the C3[prime]-endo/anti dA conformation (32), again consistent with A-form phosphoester geometry. At the same time, the P3 crystal contains markers at 838 and 661 cm-1, indicating a second molecular conformation with B-form phosphoester geometry and C2[prime]-endo/anti dA. The large number of Raman bands in the 850-1100 cm-1 interval and the apparent doubling of many bands in the region above 1100 cm-1 are also consistent with more than one c(dAp)3 conformation in the P3 crystal.


Figure 5. Polarized (I||) Raman spectra of P3 (top) and R32 (bottom) crystals of c(dAp)3. Each spectrum was obtained from a single crystal oriented with the crystallographic c axis parallel to the electric vector of the incident and scattered laser beams.

In summary, the Raman signature of the P3 crystal indicates the coexistence of two c(dAp)3 conformations. These are distinguished by different sugar pucker and phosphoester geometry: the A-like phosphoester geometry is associated with C3[prime]-endo dA conformers (640 cm-1), while the B-like phosphoester geometry is associated with C2[prime]-endo dA conformers (665 cm-1). Conversely, in the R32 crystal, c(dAp)3 molecules exhibit predominantly the C2[prime]-endo dA conformation. The intensity ratio (I665/I640) of the 665 and 640 cm-1 markers in the P3 crystal is consistent with a distribution of approximately two C2[prime]-endo conformers for each C3[prime]-endo conformer.

Rationalizing the Raman and X-ray results on c(dAp)3 crystal structures

The X-ray crystal structure indicates one c(dAp)3 conformation in the P3 lattice. Yet, the Raman spectrum indicates a second c(dAp)3 conformer in the same crystal. This can be explained by considering the arrangement of c(dAp)3 molecules in the P3 lattice, which is depicted in Figure 6A. The apparently very low molecular packing density leads to very large solvent channels in the P3 lattice. These channels are of sufficient size to allow rotation and/or translation of `disordered' c(dAp)3 molecules, a phenomenon observed previously in oligonucleotide single crystals (42,43). We propose that the two conformers revealed in the Raman spectrum of the P3 crystal include one that is well ordered and gives rise to the observed X-diffraction and one that is rotationally and translationally disordered and not detected by X-ray scattering. Because the X-ray results indicate that the ordered c(dAp)3 conformer in the P3 crystal exhibits C2[prime]-endo sugar pucker, we conclude that the disordered conformer exhibits the C3[prime]-endo Raman marker and corresponding A-like phosphoester geometry.


Figure 6. Molecular packing in the P3 (A) and R32 (B) crystal structures of c(dAp)3, each determined to 1.0 Å resolution by Gao et al. (20). In each case the view is approximately along the c axis. For both the R32 and P3 lattices, there are two very similar c(dAp)3 molecules per asymmetric unit (Table 1). The packing arrangement in the P3 lattice provides solvent channels sufficiently large to accommodate rotationally and translationally disordered c(dAp)3 molecules not observed in the crystal structure but detected by Raman spectroscopy.

The Raman spectrum of aqueous c(dAp)3 shows that in the solution environment C2[prime]-endo sugar pucker is preferred by the deoxyadenosine residues. It is therefore surprising that the population of mobile c(dAp)3 molecules in the P3 crystal exhibits the C3[prime]-endo conformation. This may reflect a decrease in water activity in the crystal environment. In oligonucleotide crystals shown previously to incorporate disordered molecules in a solvent channel (42,43), the ordered molecules exhibited the A DNA conformation and the disordered molecules occupying the solvent channels were of the B form. In the case of the P3 crystal of c(dAp)3, it is the more B-like conformation that is ordered in the lattice.

SUMMARY AND CONCLUSIONS

Cyclic trideoxyadenylic acid, c(dAp)3, crystallizes in R32 and P3 space groups. Both crystals diffract to near atomic resolution, exhibit similar sets of torsion angles in the phosphoester ring, and incorporate similar C2[prime]-endo (S-type) deoxyribose pucker and anti glycosyl torsions for the three deoxyadenosine residues per molecule. However, molecular packing is significantly different in the two crystal structures and leads to much larger solvent channels in the P3 lattice.

Polarized Raman microspectroscopy of an oriented R32 crystal provides confirmation of the molecular model proposed from X-ray analysis (Figs 1 and 6). The Raman signature of the R32 structure also confirms previously proposed Raman markers for the C2[prime]-endo/anti deoxyadenosine conformation and indicates novel Raman markers for the unusual [alpha] and [zeta] torsions (gauche+/gauche+) of the phosphoester ring.

Conversely, the Raman signature of the P3 crystal differs fundamentally from that of the R32 crystal and appears to be in conflict with the simple structure proposed from X-ray analysis (Fig. 6). The apparent conflict is resolved by recognizing that the P3 crystal comprises two populations of c(dAp)3 molecules, only one of which is ordered (S type) in the lattice and contributes to the observed X-ray diffraction map. The second conformer, presumed to exhibit translational and rotational disorder through the large solvent channels of the P3 lattice, exhibits C3[prime]-endo (N type) sugar pucker.

In aqueous solution, the c(dAp)3 molecule exhibits a Raman signature that is different from that of either the R32 or P3 crystal. The solution spectrum is characterized by relatively broad bands, reminiscent of heat-denatured nucleic acids and suggests a wide distribution of torsion angles for the cyclic phosphoester and deoxyribose rings. The crystal and solution results together suggest a highly flexible backbone (phosphoester ring) for c(dAp)3. This high degree of flexibility contrasts sharply with the rather rigid structures revealed for cyclic dinucleotides by X-ray analyses (22-25), and may be a factor that limits the biological activity of cyclic trinucleotides vis-à-vis the cyclic dinucleotides and cyclic mononucleotides.

Effects of crystal packing forces on DNA conformation have been well documented (44,45). The c(dAp)3 molecule, an inherently pleiotropic structure, appears to be one more example of an oligonucleotide in which crystal packing forces play a major role in determining the conformation of the sugar-phosphate backbone.

ACKNOWLEDGEMENTS

This paper is Part LXX in the series Raman Spectral Studies of Nucleic Acids. Support of this research by grants GM54378 (G.J.T.) and GM41612 (A.H.J.W.) from the National Institutes of Health is gratefully acknowledged.

REFERENCES

1. Thomas,G.J., Jr and Tsuboi,M. (1993) Adv. Biophys. Chem., 3, 1-70.

2. Peticolas,W.L. (1995) Methods Enzymol., 246, 389-416. MEDLINE Abstract

3. Benevides,J.M., Wang,A.H.-J., van der Marel,G.A., van Boom,J.H., Rich,A. and Thomas,G.J., Jr (1984) Nucleic Acids Res., 12, 5913-5925. MEDLINE Abstract

4. Benevides,J.M., Wang,A.H.-J., Rich,A., Kyogoku,Y., van der Marel,G.A., van Boom,J.H. and Thomas,G.J., Jr (1986) Biopolymers, 25, 41-50.

5. Benevides,J.M., Wang,A.H.-J., van der Marel,G.A., van Boom,J.H. and Thomas,G.J., Jr (1988) Biochemistry, 27, 931-938. MEDLINE Abstract

6. Peticolas,W.L., Kubasek,W.L., Thomas,G.A. and Tsuboi,M. (1987) In Spiro,T.G. (ed.), Biological Applications of Raman Spectroscopy: Raman Spectra and the Conformations of Biological Molecules. Wiley-InterScience, New York, Vol., 1, pp. 81-133.

7. Thomas,G.J., Jr and Wang,A.H.-J. (1988) In Eckstein,F. and Lilley,D.M.J. (eds), Nucleic Acids and Molecular Biology. Springer-Verlag, Berlin, Vol., 2, pp. 1-30.

8. Benevides,J.M., Weiss,M.A. and Thomas,G.J., Jr (1994) J. Biol. Chem., 269, 10869-10878. MEDLINE Abstract

9. Duguid,J., Bloomfield,V.A., Benevides,J.M. and Thomas,G.J., Jr (1993) Biophys. J., 65, 1916-1928. MEDLINE Abstract

10. Duguid,J., Bloomfield,V.A., Benevides,J.M. and Thomas,G.J., Jr (1995) Biophys. J., 69, 2623-2641. MEDLINE Abstract

11. Wen,Z.Q., Overman,S.A. and Thomas,G.J., Jr (1997) Biochemistry, 36, 7810-7820. MEDLINE Abstract

12. Dijkstra,S., Benevides,J.M. and Thomas,G.J., Jr (1991) J. Mol. Struct., 242, 283-301.

13. Tsuboi,M., Komatsu,K., Hoshi,J., Kawashima,E., Sekine,T., Ishido,Y., Russell,M.P., Benevides,J.M. and Thomas,G.J., Jr (1997) J. Am. Chem. Soc., 119, 2025-2032.

14. Miura,T. and Thomas,G.J., Jr (1994) Biochemistry, 33, 7848-7856. MEDLINE Abstract

15. Benevides,J.M., Kang,C.H. and Thomas,G.J., Jr (1996) . Biochemistry, 35, 5747-5755. MEDLINE Abstract

16. Duguid,J., Bloomfield,V., Benevides,J.M. and Thomas,G.J., Jr (1996) Biophys. J., 71, 3350-3360. MEDLINE Abstract

17. Reilly,K.E. and Thomas,G.J., Jr (1994) J. Mol. Biol., 241, 68-82. MEDLINE Abstract

18. Miura,T., Benevides,J.M. and Thomas,G.J., Jr (1995) J. Mol. Biol., 248, 233-238. MEDLINE Abstract

19. Tuma,R., Bamford,J.H.K., Bamford,D.H., Russell,M.P. and Thomas,G.J., Jr (1996) J. Mol. Biol., 257, 87-101. MEDLINE Abstract

20. Gao,Y.-G., Robinson,H., Guan,Y., Liaw,Y.-C., van Boom,J.H.,van der Marel,G.A. and Wang,A.H.-J. (1998) J. Biomol. Struct. Dynam., in press.

21. Frederick,C.A., Coll,M., van der Marel,G.A., van Boom,J.H. and Wang,A.H.-J. (1988) Biochemistry, 27, 8350-8361. MEDLINE Abstract

22. Egli,M., Gessner,R.V., Williams,L.D., Quigley,G.J., van der Marel,G.A., van Boom,J.H., Rich,A. and Frederick,C.A. (1990) Proc. Natl Acad. Sci. USA, 87, 3235-3239. MEDLINE Abstract

23. Liaw,Y.-C., Gao,Y.-G., Robinson,H., Sheldrick,G.M., Sliedregt,L.A.J.M., van der Marel,G.A., van Boom,J.H. and Wang,A.H.-J. (1990) FEBS Lett., 264, 223-227. MEDLINE Abstract

24. Guan,Y., Gao,Y.-G., Liaw,Y.-C., Robinson,H. and Wang,A.H.-J. (1993)J. Biomol. Struct. Dyn., 11, 253-276. MEDLINE Abstract

25. de Vroom,E., Broxterman,H.J.G., Sliedregt,A.J.M., van der Marel,G.A. and van Boom,J.A. (1988) Nucleic Acids Res., 16, 4607-4620. MEDLINE Abstract

26. Benevides,J.M., Tsuboi,M., Wang,A.H.-J. and Thomas,G.J., Jr (1993)J. Am. Chem. Soc., 115, 5351-5359.

27. Lamba,O.P., Becka,R. and Thomas,G.J., Jr (1990) Biopolymers, 29, 1465-1477. MEDLINE Abstract

28. Schneider,B., Neidle,S. and Berman,H.M. (1997) Biopolymers, 44, 113-124.

29. Lord,R.C. and Thomas,G.J., Jr (1967) Spectrochim. Acta, 23A, 2551-2591.

30. Tsuboi,M., Takahashi,S. and Harada,I. (1973) In Duchesne,J. (ed.), Physico-Chemical Properties of Nucleic Acids. Academic Press, New York, Vol., 2, pp. 92-143.

31. Hirakawa,A.Y., Ohada,H., Sasagawa,O. and Tsuboi,M. (1985) Spectrochim. Acta, 41A, 209-216.

32. Thomas,G.J., Jr and Benevides,J.M. (1985) Biopolymers, 24, 1101-1105. MEDLINE Abstract

33. Letellier,R., Ghomi,M. and Taillandier,E. (1987) J. Biomol. Struct. Dyn., 4, 663-683. MEDLINE Abstract

34. Thomas,G.J., Jr, Benevides,J.M., Overman,S.A., Ueda,T., Ushizawa,K., Saitoh,M. and Tsuboi,M. (1995) Biophys. J., 68, 1073-1088. MEDLINE Abstract

35. Thomas,G.J., Jr, Benevides,J.M. and Prescott,B. (1986) In Sarma,R.H. and Sarma,M.H. (eds), Biomolecular Stereodynamics. Adenine Press, Guilderland, New York, Vol., 4, pp. 227-254.

36. Brown,E.B. and Peticolas,W.L. (1975) Biopolymers, 14, 1259-1271. MEDLINE Abstract

37. Prescott,B., Steinmetz,W. and Thomas,G.J., Jr (1984) Biopolymers, 23, 235-256. MEDLINE Abstract

38. Thomas,G.J., Jr and Livramento,J. (1975) Biochemistry, 14, 5210-5218. MEDLINE Abstract

39. Toyama,A., Takeuchi,H. and Harada,I. (1991) J. Mol. Struct., 242, 87-98.

40. Thomas,G.J., Jr (1970) Biochim. Biophys. Acta, 213, 417-423. MEDLINE Abstract

41. Erfurth,S.C., Kiser,E.R., Kiser,J. and Peticolas,W.L. (1972)Proc. Natl Acad. Sci. USA, 69, 938-941. MEDLINE Abstract

42. Doucet,J., Benoit,J.-P., Cruse,W.B.T., Prange,T. and Kennard,O. (1989) Nature, 337, 190-192.43. Fabian,H., Hölzer,W., Heinemann,U., Sklenar,H. and Welfle,H. (1993) Nucleic Acids Res., 21, 569-576.

44. Saenger,W. (1983) Principles of Nucleic Acid Structure. Springer-Verlag, New York.

45. DiGabriele,A.D. and Steitz,T.A. (1993) J. Mol. Biol., 231, 1024-1039. MEDLINE Abstract

46. Benevides,J.M. and Thomas,G.J., Jr (1983) Nucleic Acids Res., 11, 5747-5761. MEDLINE Abstract


*To whom correspondence should be addressed. Tel: +1 816 235 5247; Fax: +1 816 235 1503; Email: thomasgj@cctr.umkc.edu


This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 14 Aug 1998
Copyright©Oxford University Press, 1998.

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (3)
Right arrowRequest Permissions
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Guan, Y.
Right arrow Articles by Thomas, G. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Guan, Y.
Right arrow Articles by Thomas, G. J., Jr
Social Bookmarking
 Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?