RESULTS AND DISCUSSION

1
ANISOTROPIC REFINEMENT OF THE STRUCTURE OF
THERMOASCUS AURANTIACUS XYLANASE AT 1.14Å
RESOLUTION Susana Teixeiraa, Leila Lo Leggiob,
Richard Pickersgillc and Christine Cardina
aChemistry Department, University of Reading,
Whiteknights, Reading RG6 6AD, England bCentre
for Crystallographic Studies, Chemical Institute,
University of Copenhagen, Universitetsparken 5,
DK-2100 Copenhagen, Denmark cMolecular and
Cellular Biology, Queen Mary, University of
London, Mile End Road, London E1 4NS,England
Chemistry Department Faculty of
Sciences University of Reading

• RESULTS AND DISCUSSION
• Refinement of the primary sequence.
• Sequence discrepancies were found against the
  available sequence (Bousson Parriche, 1999,
  unpublished). Clear electron density maps,
  supported by multiple sequence alignement,
  allowed for corrections to be made for residues
  245 (Proline) and 217 (Glycine). The terminal
  glutamine is not seen in the maps. If it is
  present it is either statically disordered or has
  considerable thermal motion. ? See alignement
  below and figure on the left
• Protonation state of the glutamates at the
  active site
• A distance of 5.5Å was observed between OE2 of
  the acid/base Glu131 and OE2 of the nucleophile
  Glu237. This is consistent with the retention of
  the anomeric configuration during xylan
  hydrolysis. In the active enzyme and in the
  absence of xylan, the acid/base glutamate should
  be protonated while Glu237 should be deprotonated
  (Sinnot, 1990 Davies Henrissat, 1995). This
  was shown through a block-diagonal unrestrained
  refinement (with SHELXH), using one block for all
  atoms and retaining the positional parameters.
  The C-O distances and corresponding standard
  uncertainties confirmed the expected protonation
  states ? Mechanism scheme on the left
• Side-chain disorder
• The clear improvement of the quality of the maps
  after ADPs were introduced
  made possible a more precise fitting of
  side-chain disorder and solvent modelling.
  Alternate conformations previously not modelled
  were introduced for 10 residues. Double
  conformations for residues 223 and 235 modelled
  by Lo Leggio et al. were removed. The most
  striking case is tryptophan 275, where alternate
  conformations were modelled. This residue is
  thought to be very important in substrate binding
  to TAXI by closing in on xylan (Lo Leggio et al.,
  in preparation). ? See structure figures
• Quality of the model/Anisotropy
• Judging by the overall improvement of the final
  statistics the anisotropic model seems to fit the
  data better, as expected considering the mean
  anisotropy of TAXI. The mean anisotropy is 0.54
  for the protein and 0.52 for the solvent, with
  standard deviations of 0.16 and 0.12,
  respectively. The distribution of anisotropy
  among the protein atoms of TAXI was calculated
  with the program PARVATI. It shows a deviation
  from the typical more anisotropic distribution
  curve. This same behaviour has been observed
  before (Merritt, 1999). It is a fact that the
  number of structures refined with ADPs is
  increasing. The growing number of high quality
  structures will surely bring important
  information to the parameterisation used in
  refinement programs and to future studies on
  anisotropy and refinement protocols. ? See
  anisotropic displacement figure and refinement
  statistics.

INTRODUCTION Thermoascus aurantiacus Xylanase I
has a (ba)8 TIM-barrel fold and belongs to the
family 10 of glycosyl hydrolases (Banner et al.,
1975 Jenkins et al., 1995 Pickersgill et al.,
1998 Coutinho Henrissat, 1999). Interest in
such enzymes is due to their potential and actual
practical applications. Understanding the
structure of xylanases and how it correlates to
their function is important to support studies
aiming at improving and using the properties of
these enzymes in practical applications. TAXI
catalyses the hydrolysis of xylan, which
represents the major group of hemicelluloses. It
is an interesting xylanase, as it has been shown
to have a high degree of thermal stability, high
activity and high specificity towards xylan.
These characteristics may be very useful in
future applications in the pulp and paper
industries, where reduction in the use of
chlorine as a bleaching agent is being imposed by
environmental regulations.
Retaining mechanism of the glycosidic bond
hydrolysis.
REFERENCES Banner, D.W., Bloomer, A.C., Petsko,
G.A., Phillips, D.C., Pogson, C.I., Wilson, I.A.,
Corran, P.H., Furth, A.J., Milman, J.D., Offord,
R.E., Priddle, J.D., Waley, S.G. (1975). Nature
255, 609-614. Coutinho, P.M., Henrissat, B.
(1999). Carbohydrate-Active Enzymes server at
URL http//afmb.cnrs-mrs.fr/pedro/CAZY/db.html D
avies, G., Henrissat, B. (1995). Structure 3,
853-859. Ito, K., Ikemasu, T., Ishikawa, T.
(1992). Biosci. Biotechnol. Biochem. 56,
906-912. Jenkins, J., Lo Leggio, L., Harris, G.,
Pickersgill, R. (1995). FEBS Letters 362,
281-285. Lo Leggio, L., Kalogiannis, S., Bhat,
M.K., Pickersgill, R.W. (1999). PROTEINS
Structure, Function and Genetics 36,
295-306. Merritt, E.A., (1999). Acta Cryst. D55,
1109-1117. Pickersgill, R., Harris, G., Lo
Leggio, L., Mayans, O., Jenkins, J. (1998).
Biochemical Society Transactions 26,
190-198. Sheppard, P.O., Grant, F.J., Oort,
P.J., Sprecher, C.A., Foster, D.C., Hagen, F.S.,
Upshall, A., McKnight, G.L., O'Hara, P.J.
(1994). Gene 150 (1), 163-167. Sinnot, M.L.
(1990). Chem. Rev. 90, 1171-1202. Wu, S.C.,
Kauffmann, S., Darvill, A.G., Albersheim, P.
(1995). Mol. Plant Microbe Interact. 8,
506-514. Teixeira, S., Lo Leggio, L.,
Pickersgill, R., Cardin, C., (2001). Acta Cryst,
in press at the time of the poster printing.
ACKNOWLEDGEMENTS ST is grateful to the Chemistry
Department of the University of Reading for
financial support, and to Dr. E. Merritt and Dr.
G. Sheldrick for their invaluable advice. LL
thanks the Danish National Research Foundation
for financial support, Dr. Anne Mølgaard and
Henning Osholm Sørensen for helpful discussions.