J Biol Chem, Vol. 274, Issue 35, 24749-24752, August 27, 1999

Structure of d(CGCGAATTCGCG) in the Presence of Ca²⁺ Ions^*

Jie Liu^§ and Juan Antonio Subirana^¶

From the  Departament d'Enginyeria Química, Universitat Politècnica de Catalunya, Diagonal 647, E-08028, Barcelona, Spain

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The dodecamer d(CGCGAATTCGCG) was the first oligonucleotide to be crystallized as a B-DNA duplex. Its structure was analyzed in detail in the early 1980s. Here we show that, in the presence of Ca²⁺, it crystallizes in a different way (R3 space group). The dodecamers form parallel columns of straight duplexes with ten base pairs in the B form. The terminal cytosines in each molecule are disordered, whereas the terminal guanines are placed in the minor groove of neighbor duplexes. The central GAATTC region is practically identical to that found in the classic structure of the same dodecamer crystallized in the P2₁2₁2₁ space group in the presence of Mg²⁺ and spermine. Its structure is thus independent of the crystallization conditions which have been used.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The first detailed structure of a DNA oligonucleotide in the B-form determined by single crystal x-ray diffraction methods was published in this journal in 1982 (1). The structure of a bromine derivative was compared with that previously determined (2, 3) for the native dodecamer at different temperatures. Its features were analyzed in detail in several other publications (4-7). Many other derivatives from the same oligonucleotide and related sequences, alone and in association with drugs, have also been studied. Coordinates and references may be found in the Nucleic Acid Database (8). Most of them are in practically identical unit cells in the P2₁2₁2₁ space group. The original work (1-3) was a landmark in the study of DNA because it confirmed unequivocally the double helical structure of B-form DNA and, at the same time, showed many features of conformation as a function of sequence. More recently, the same dodecamer has been crystallized under various ionic conditions (9-13). A high resolution (1.4 Å) was obtained (9) in the presence of Na⁺/Mg²⁺/spermine, giving a structure essentially similar to those first reported by Dickerson and co-workers (1-3). Given the higher resolution of that structure (9), we used it for comparison with our results.

Most of the work described in the previous paragraph was carried out with crystals obtained in the presence of Mg²⁺ and spermine. Recently we discovered (14) that, in the presence of a high concentration of Ca²⁺ ion, this oligonucleotide could crystallize in the R3 space group. Here we report the results we have obtained under the latter crystallization conditions at a higher resolution than that previously reported (14). Our results show that the central sequence GAATTC is practically identical in both cases, whereas the conformation of the terminal CGC/GCG sequences is much more variable because of the interaction with neighbor molecules in the crystal.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Crystallization-- The crystal was grown by the vapor diffusion hanging drop method. The crystallization solution contained 0.4 mM dodecamer (NH₄⁺ salt), 300 mM CaCl₂, 10% MPD (2-methyl-2,4-pentanediol), and 20 mM cacodylic acid (pH 7.0). The concentration of monovalent cations (Na⁺ from the buffer and NH₄⁺ from the dodecamer) was considerably smaller than the Ca²⁺ ion concentration. It was equilibrated against a reservoir of 48% MPD at 20 °C. A 0.8 × 0.2 × 0.1 mm³ crystal was obtained in about two months. In some cases, the crystallization solution also contained either a basic peptide or protamine. However, they are neither found in the crystal structure nor essential for crystallization.

X-ray Diffraction Data Collection and Processing-- X-ray diffraction data were collected on a 345 mm MAR Research image plate scanner using the beamline X11 in the DESY synchrotron radiation station in Hamburg, Germany. An Oxford Cryo-systems Cryostream was used to flash cool the crystal at 120 K under a nitrogen vapor stream. A high resolution (1.20 Å) dataset was collected with long exposure, and a second dataset at low resolution (2.65Å) was collected with short exposure to avoid saturation of some strong low resolution reflections that are important for the interpretation of the electron density map during the structural refinement. Both datasets were processed with the program DENZO (15) and then reduced and merged with the program SCALEPACK (15). Crystal data are listed in Table I.

Structural Refinement-- Our previously determined R3 structure (14) of the title compound at low resolution (3.0 Å) was used as the initial model to start refinement with CNS 0.3 (16). First a rigid body refinement was performed with 8.0-3.0-Å resolution data. Then alternate positional and B-factor refinements were carried out while extending the resolution of the data as well as adding calcium ions and water molecules stepwise. It was noticed during data processing that the diffraction of the crystal is rather anisotropic, which is probably because of the nature of crystal packing and the disordered terminal cytosine residues. It diffracts to 1.2 Å in the z direction but to a lower resolution in perpendicular directions. In view of this situation, the data were cut to 1.45 Å to avoid overestimating the resolution. Overall anisotropic B-factor and bulk solvent corrections were introduced throughout the refinement. The two terminal cytosine residues were not detected in the electron density map and were thus excluded from the refinement. At this point, the R factor was 0.252 and R free 0.262. Then the model was further refined with SHELXL (17). Anisotropic overall scaling of the data and diffuse solvent corrections were applied. Phosphorus atoms and calcium ions were refined anisotropically. The R free test with 10% of the reflections was used to monitor the progress of the refinement, leading to R factor 0.210 and R free 0.247. Finally the model was refined with the complete dataset, resulting in R factor 0.211. The structural refinement statistics are also listed in Table I. Given the apparent resolution of the data, a lower R factor was expected, but the anisotropy of the data as well as the disorder of the cytosines probably prevented further improvement. The R factor could be substantially reduced to 0.157 by using anisotropic atomic B factors for all atoms, but the quality of our data did not warrant such an approach as the free R factor was only reduced to 0.231.

Conformational parameters were calculated with NEWHELIX93 (courtesy of Dr. R. Dickerson, University of California, Los Angeles).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Packing-- In the R3 space group, the dodecamers crystallize as infinite parallel columns as shown in Fig. 1. The columns are stabilized by an interaction between consecutive dodecamers as demonstrated in Fig. 2. This interaction was first described by Spink et al. (18) for decamers and since then has been found in other cases (19, 20). The terminal guanine of each dodecamer penetrates into the minor groove of the neighbor duplex and forms a set of hydrogen bonds that stabilize the columns of duplexes shown in Fig. 1. A Ca²⁺ ion that is associated with a guanine from a neighbor column is also involved in this interaction, as shown in Fig. 2. On the other hand, the terminal cytosines are disordered and cannot be detected in the electron density map, as found in other cases in which this interaction is present (18-20).

View larger version (62K): [in this window] [in a new window]   Fig. 1.   Stereopair of the unit cell. The nine dodecamers in the unit cell are shown. Calcium ions are indicated as black dots. Each asymmetric unit contains four different ions, two in the minor groove and another two involved in the interaction between different duplexes.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   End to end interaction of duplexes. The stereopair shows the two terminal Watson-Crick base pairs of consecutive duplexes and the terminal guanines that interact with the minor groove of the next duplex. A Ca2+ ion and an associated water are shown as black dots. A guanine from a neighbor column of duplexes also interacts with this Ca2+ ion. Dotted lines indicate hydrogen bonds and ionic interactions.

Neighbor duplexes along each column do not form pseudocontinuous helices. There is a rotation of about 70.5° between the terminal base pairs. Because of this rotation, stacking of the neighbor C·G base pairs is very limited, although the planes of the base pairs are rather parallel. The overall helical axis of each duplex is approximately parallel to the z direction of the crystal, with an angle of 2.0°, but the axes are laterally displaced in neighbor duplexes by about 3.8 Å.

The interaction between neighbor columns is stabilized by the presence of two types of Ca²⁺ ions that occur at regular intervals. One of them lies on the ternary axis of the unit cell and stabilizes the interaction of three duplexes at the level of phosphate 5, as shown in Fig. 3. Another Ca²⁺ ion stabilizes the interaction between phosphates 12 and 24 at the ends of neighbor duplexes, as shown in Fig. 2. The latter interaction is also apparent at the lower end of Fig. 1.

View larger version (9K): [in this window] [in a new window]   Fig. 3.   Interaction between neighbor duplexes. The stereopair shows the Ca2+ ion (large dot) in the 3-fold axis that interacts with the phosphates of three duplexes and with three water molecules (small dots).

It should be pointed out that there are occasionally short contacts (<2.5Å) between some of the water molecules. Some of them may correspond to disordered calcium ions. Because there is no extra electron density, we have treated them as water molecules.

Minor Groove Structure-- The central region of the duplex has a spine of hydration (5, 6, 9) identical to that found in the P2₁2₁2₁ structure, as it has been discussed in detail elsewhere (21). At both ends of the spine of hydration, there are two heavily hydrated Ca²⁺ ions. In contrast, in the case of the P2₁2₁2₁ structure, crystallized in the presence of Mg²⁺, no divalent cations are found in the minor groove. Thus the minor groove of the duplex is fully occupied, with guanines at both ends (shown in Fig. 2) followed by Ca²⁺ ions and a spine of hydration in the central region.

An additional characteristic feature of the P2₁2₁2₁ structure is the narrow minor groove at the level of the central AATT sequence. As shown in Fig. 4, this feature is also present in the structure reported here. It is also apparent from the figure that the R3 structure is more symmetric, as it will be shown in more detail below.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Minor groove width. The figure shows that the minor groove width is reduced in the central region of the duplex, both in the P212121 (left) and R3 (right) structures. The numbers give phosphate-phosphate distances in Å. Note that the R3 structure is more symmetric, opening the minor groove at both ends.

Conformational Parameters-- A comparison of some of the conformational parameters is given in Fig. 5. It is clear that the central GAATTC sequence is practically identical in the two structures compared, using any of the parameters shown in Fig. 5. On the other hand, the three terminal base pairs at both ends of the dodecamer show a mixed behavior. For example, the values of roll angles are practically identical in both cases, whereas rise and twist follow similar but not identical trends. Propeller twist and some  values are rather different in both cases. These differences are undoubtedly related to the interactions between duplexes in the crystal, which occur mainly at the terminal base pairs.

View larger version (27K): [in this window] [in a new window]   Fig. 5.   Comparison of conformational parameters. The conformational parameters of the P212121 structure () are compared with those of the R3 structure (). At the left are shown base step parameters and at the right are base pair angles. Note that in the plots of the  angles, the two strands of each duplex are compared. In a symmetric structure, the two strands should have identical values.

Another feature of interest is the higher degree of symmetry of the R3 structure. This feature is most clear in the values of the propeller twist and in the  angles, related with the sugar pucker. Ideally, the two chains of the duplex should have identical  angles in all the positions of the dodecamer. In Fig. 5, it is clear that the plots of  values deviate much more in the P2₁2₁2₁ than in the R3 structure.

Bending-- In the P2₁2₁2₁ space group, the duplex is found to be bent (1-3, 9), although the bromine derivative has a straight helical axis (1). Bending is demonstrated by the normal vector plot presented in Fig. 6. In the same figure, it is shown that the R3 structure only has small distortions, it is practically straight. The overall helical axis calculated with NEWHELIX93 forms an angle of 2° with the z direction of the unit cell and is displaced about 2 Å from the 3-fold screw axis that relates consecutive duplexes in a column.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Normal vector plots. The P212121 (left) and R3 (right) structure are compared. The numbers correspond to consecutive base pairs. The circles correspond to normal vector angles of 5°, 10°, and 15°. The plots show that both ends of the duplex in the P212121 structure are bent, whereas the R3 structure only has a small symmetric distortion with a practically straight helical axis.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The results reported here for the d(CGCGAATTCGCG) dodecamer under very different packing constraints show that the central GAATTC sequence has a very stable conformation, very similar to that reported in the P2₁2₁2₁ structure several years ago (1-3). We can therefore conclude that crystal packing constraints do not have much influence on the conformational parameters of that sequence.

On the other hand, the conformation of the 3 terminal base pairs at both ends of the duplex is strongly influenced (Fig. 5) by the packing interactions between the terminal regions of the duplexes, which have been shown above for the R3 structure (Fig. 2). In the P2₁2₁2₁ structure, guanine-guanine interactions between different duplexes are also present, as they have been analyzed in detail elsewhere (7, 22).

In this context we should point out that the average twist of the R3 structure corresponds to about 10.5 base pairs/turn, very close to the average value in solution (23). As we have shown elsewhere (22), the constraints imposed on dodecamers in the P2₁2₁2₁ space group result in an average twist which must be close to 10 base pairs/turn. On the other hand, in the R3 space group, the stacking interactions among neighbor duplexes in a column are less extensive, and apparently the duplexes have enough freedom to acquire an average twist closer to that found in solution.

A final question to consider is why Ca²⁺ ions promote a different packing than the typical P2₁2₁2₁ space group found in most dodecamers crystallized up to now (8). Two reasons may be invoked. One is that Mg²⁺ ions, at least in the lower concentration used for crystallization of dodecamers, do not penetrate the narrow groove. Penetration of the narrow groove by Ca²⁺ makes it wider toward the end of the dodecamer (Fig. 4) and may prevent the guanine-guanine interaction characteristic of the P2₁2₁2₁ dodecamers (7, 22). The second reason is that the interaction among duplexes which are stabilized by the divalent cations are very different in the presence of Ca²⁺ (Figs. 2 and 3) than in the presence of Mg²⁺ (11).

	ACKNOWLEDGEMENTS

We thank Drs. Tereshko and Minasov for advice and discussion.

	FOOTNOTES

^* This work was supported in part by the DGICYT (Grant PB93-1067), the Generalitat de Catalunya (Grant 1997SGR-135), and the TMR/LSF program to the EMBL Hamburg Outstation (ERBFMGECT980134).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The coordinates have been deposited in the Nucleic Acid Database (http://ndbserver.rutgers.edu) under accession number BD0014.

^§ Present address: Dept. of Biochemistry, Cornell University Medical College, 1300 York Ave., New York, NY 10021.

^¶ To whom correspondence should be addressed. Tel.: 34-934016688; Fax: 34-934017150; E-mail: subirana@eq.upc.es.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

1.	Fratini, A. V., Kopka, M. L., Drew, H. R., and Dickerson, R. E. (1982) J. Biol. Chem. 257, 14686-14707[Abstract/Free Full Text]
2.	Drew, H. R., Wing, R. M., Takano, T., Broka, C., Tanaka, S., Itakura, K., and Dickerson, R. E. (1981) Proc. Natl. Acad. Sci U. S. A. 78, 2179-2183[Abstract/Free Full Text]
3.	Drew, H. R., Samson, S., and Dickerson, R. E. (1982) Proc. Natl. Acad. Sci U. S. A. 79, 4040-4044[Abstract/Free Full Text]
4.	Dickerson, R. E., and Drew, H. R. (1981) J. Mol. Biol. 149, 761-786[CrossRef][Medline] [Order article via Infotrieve]
5.	Drew, H. R., and Dickerson, R. E. (1981) J. Mol. Biol. 151, 535-556[CrossRef][Medline] [Order article via Infotrieve]
6.	Kopka, M. L., Fratini, A. V., Drew, H. R., and Dickerson, R. E. (1983) J. Mol. Biol. 163, 129-146[CrossRef][Medline] [Order article via Infotrieve]
7.	Dickerson, R. E., Goodsell, D. S., Kopka, M. L., and Pjura, P. E. (1987) J. Biomol. Struct. Dyn. 5, 557-579[Medline] [Order article via Infotrieve]
8.	Berman, H. M., Olson, W. K., Beveridge, D. L., Westbrook, J., Gelbin, A., Demeny, T., Hsieh, S.-H., Srinivasan, A. R., and Schneider, B. (1992) Biophys. J. 63, 751-759[Free Full Text]
9.	Shui, X. Q., McFail-Isom, L., Hu, G. G., and Williams, L. D. (1998) Biochemistry 37, 8341-8355[CrossRef][Medline] [Order article via Infotrieve]
10.	Shui, X. Q., Sines, C. C., McFail-Isom, L., VanDerveer, D., and Williams, L. D. (1998) Biochemistry 37, 16877-16887[CrossRef][Medline] [Order article via Infotrieve]
11.	Tereshko, V., Minasov, G., and Egli, M. (1999) J. Am. Chem. Soc. 121, 470-471[CrossRef]
12.	Tereshko, V., Minasov, G., and Egli, M. (1999) J. Am. Chem. Soc. 121, 3590-3595[CrossRef]
13.	Minasov, G., Tereshko, V., and Egli, M. (1999) J. Mol. Biol., in press
14.	Liu, J., Malinina, L., Huynh-Dinh, T., and Subirana, J. A. (1998) FEBS Lett. 438, 211-214[CrossRef][Medline] [Order article via Infotrieve]
15.	Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307-326
16.	Brunger, A. T. (1998) Crystallography and NMR System (CNS), Version 0.3 , Yale University Press, New Haven, CT
17.	Sheldrick, G. M., and Schneider, T. R. (1997) Methods Enzymol. 277, 319-343
18.	Spink, N., Nunn, C. M., Vijtechovsky, J., Berman, H. M., and Neidle, S. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 10767-10771[Abstract/Free Full Text]
19.	Nunn, C. M., Garman, E., and Neidle, S. (1997) Biochemistry 36, 4792-4799[CrossRef][Medline] [Order article via Infotrieve]
20.	Abrescia, N. G. A., Malinina, L., Fernandez, L. G., Huynh-Dinh, T., Neidle, S., and Subirana, J. A. (1999) Nucleic Acids Res. 27, 1593-1599[Abstract/Free Full Text]
21.	Soler-López, M., Malinina, L., Liu, J., Huynh-Dinh, T., and Subirana, J. A. (1999) J. Biol. Chem. 274, 23683-23686[Abstract/Free Full Text]
22.	Tereshko, V., and Subirana, J. A. (1999) Acta Crystallogr. Sec. D 55, 810-819 [CrossRef][Medline] [Order article via Infotrieve]
23.	Rhodes, D., and Klug, A. (1980) Nature 286, 573-578[CrossRef][Medline] [Order article via Infotrieve]