IN HOUSE ANOMALOUS DIFFRACTION DATA COLLECTION IS APLLICABLE FOR METAL IDENTIFICATION AND PHASING

Julie Bouckaert, Remy Loris, Lode Wyns

Ultrastructuur, Interuniversitair Instituut voor Biotechnologie, Vrije Universiteit Brussel, Paardenstraat 65, B-1640 Sint-Genesius-Rode, Belgium

Concanavalin A, a lectin isolated from the leguminous plant Canavalia ensiformis, has two metal ion binding sites that neighbour the carbohydrate binding site of the lectin. The metal ions have a structural role in that their binding is a necessity to establish the carbohydrate binding conformation of the lectin. Metal ions are bound sequentially¹. The transition metal binding site preferentially accomodates Mn²⁺, but also can accommodate Ni²⁺, Co²⁺, Zn²⁺ or Cd²⁺ and possibly Ca²⁺ (Mn-site). The second site is a calcium binding site (Ca-site), only 4.25 Å from the Mn-site, that also can bind Cd²⁺ or Mn²⁺.²

Interestingly, while Ni²⁺, Co²⁺ and Zn²⁺ only bind the Mn-site, Cd²⁺ and Mn²⁺ can bind in both the Mn-site and the Ca-site. For the crystallographic structure of CdCd Con A, Cd²⁺ can easily be distinguished from Ca²⁺ on the basis of its larger electrondensity (46 compared to 18 electrons). The structure was solved up to 2.35 Å. On the other hand, for a proof of the identity of the ion in both the Mn-site and the Ca-site of MnMn Con A, we have successfully exploited the anomalous dispersion characteristics of manganese.

Manganese has its K-absorption edge at the X-ray wavelength of 1.896 Å. The experiment was performed however at the wavelength of 1.5418 Å of Cu-Ka radiation of the most common in house X-ray generators. The image plate used is a MAR area detector. The inverse beam geometry of data collection was the only successful method for Con A MnMn. Collection of the Bijvoet pairs on the same image was not successful. The crystal was oriented on a crystallographic zone initially. After every 10° step, the crystal was turned 180° to collect the corresponding Friedel pairs. The data were processed using DENZO³ and scaled using SCALA⁴. An anomalous difference Patterson was calculated with the program HEAVY⁵. Single atom, two atoms and cross peak searches were validated. The two most probable peaks in the Patterson map are due to the two Mn²⁺ ions. The metal ion positions prove to be exact as shown by the crystal structure of Con A⁶. It is a remarkable result that a difference in intensity of the reflections of less than 3 electrons (2.8 electrons at 1.5418 Å), caused by the absorption of X-rays by Mn²⁺, can be distinguished. As a control, data were collected in an identical way for MnCa Con A, that has calcium bound in the Ca-site and thus only one manganese present. Its intensity difference is 1.5 electrons. This difference merges in the noise and the correct peaks are not found in the Patterson maps. It is evident that the 2.8 electron intensity difference (for MnMn Con A) is at the limit of success for the location of the anomalous scatterer. Charge coupled device detectors may partly overcome this signal to noise problem.

In anomalous difference Fourier maps calculated by FFT both the MnMn Con A and the MnCa Con A clearly display the Mn-ions as the highest peaks in the map. In MnMn Con A, the peak height ratio of the Mn²⁺ bound to the Mn-site compared to the second Mn²⁺ that is bound in the Ca-site, is 1.44. For comparison and proof, the Ca²⁺ bound in MnCa Con A appears as the second highest peak with a ratio of Mn²⁺ over Ca²⁺ of 2.23 (theoretically 2.18 at 1.5418 Å).

The phasing power of the MnMn ConA anomalous scattering derivative was examined using the CCP4 programs SCALEIT and MLPHARE⁷. The two Mn²⁺ ions are not sufficient for phasing Con A. The fact however that they show up unambiguously as the two highest peaks in the Patterson map, implies that in house anomalous diffraction data collection is possible even within a rather broad wavelength range around the absorption edge of the anomalous scatterer.

1. Bouckaert, J., Loris, R., Poortmans, F. and Wyns, L. Sequential structural changes upon zinc and calcium binding to metal-free concanavalin A. J.Biol.Chem. 271:16144-16150, 1996.
2. Shoham, M., Kalb, A.J. and Pecht, I. Specificity of metal ion interaction with concanavalin A. Biochemistry 12:1914-1917, 1973.
3. Otwinowski, Z. DENZO. In: edited by Sawyer, L., Isaacs, N. and Bailey, S. Warrington, UK: SERC, 1993, p. 56-62.
4. Evans, P.R. SCALA. Joint CCP4 and ESF-EACBM Newsletter 22-24, 1997.
5. Terwilliger, T.C. MAD phasing, treatment of dispersive differences as isomorphous replacement information. Acta Cryst. D50:17-23, 1994.
6. Deacon, A., Gleichman, T., Kalb(Gilboa), A.J., et al. The structure of concanavalin A and its bound solvent determined with small-molecule accuracy at 0.94 Å resolution. J.Chem.Soc.,Faraday Trans. 93:4305-4312, 1997.
7. Collaborative computing project. Number 4. Acta Cryst. D50:60-763.