BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//University of Liverpool Computer Science Seminar System//v2//EN
BEGIN:VEVENT
DTSTAMP:20260922T121212Z
UID:Seminar-MIF-1445@lxserverM.csc.liv.ac.uk
ORGANIZER:CN=Othon Michail:MAILTO:Othon.Michail@liverpool.ac.uk
DTSTART:20241209T140000
DTEND:20241209T150000
SUMMARY:MIF Series
DESCRIPTION:Elspeth F. Garman: Two different challenges in biological structure determination: radiation damage and metal identification\n\nFollowing on from Professor Wlodawer’s 18th November 2024 salutory MIF++ talk on a smorgasboard of infelicities in Protein Data Bank (PDB) structures, in this talk I will describe two additional challenges in macromolecular structure determination. The PDB is a critical resource for researchers worldwide. In 2021, there were on average 1.86 million downloads per day in the US alone which suggests that over 350,000 models downloaded per day may not contain the correct metal.\nThe first of these challenges is that of radiation damage (RD) artefacts induced by X-rays absorbed by the crystalline sample during the experiment. The diffracted intensity spot fading caused by RD progression during the experiment has long been used as a metric for monitoring the state of both room temperature [1] and cryo-condition [2] samples. Recent developments in micro-electron diffraction (µED) have noted similar pathologies [e.g. 3]. Being aware of this and other RD symptoms is important [4], since even for samples held at 100 K, RD is a limiting problem. It can prevent structure determination and the effects can mislead the experimenter when interpreting the relevant biology of the structure. Pertinent to this challenge, I will describe our single metric, Bnet, by which the level of damage in a single PDB entry from a cryo-cooled crystal (i.e. not accompanied by a dose series) can be assessed [5], and which we have used to assess RD in 93,978 deposited 100 K structures.\nSecondly, I will address the issue of accurately identifying metal atoms bound to protein structures. Metalloproteins comprise over one-third of proteins, with approximately half of all enzymes requiring metal to function. Identifying the bound metal and its environment is a prerequisite to understanding biological mechanism. However, there are no routine analysis methods with the sensitivity and quantitative accuracy to do this unambiguously. We have previously developed microProton Induced X-ray Emission (PIXE) as a tool for quantifying metals in liquid and crystalline proteins using the known sulphur content (methionines and cysteines) as an internal standard [6]. We have now automated this method to permit unattended high throughput analysis of many samples, validating the approach by using it to analyse three distinct sets of 30 proteins identified as metalloproteins in the PDB. In all three sets, we found that over half of the metals had been misidentified in the deposited PDB models. Some of the PIXE-detected metals not seen in the models were explainable as artefacts from promiscuous crystallization reagents. For others, using the correct metal improved the structural models and identified new functionality. This has profound implications for those using the models, whose understanding of them may therefore be flawed.\nReferences:\n[1] C.C.F. Blake, D.C. Phillips In Proceedings of the Symposium on the Biological Effects of Ionising Radiation at the Molecular Level (Vienna: International Atomic Energy Agency), (1962) pp. 183–191.\n[2] A. Gonzales, C. Nave, Acta Cryst. D (1994) 50, 874-877.\n[3] J Hattne Methods Mol. Biol. (2021) 2215, 309-319.\n[4] EF Garman, M Weik, (2023) Current Opinion Struct. Biol. 2, 102662.\n[5] KL Shelley, EF Garman (2022) Nature Communications 13,1314- 1325.\n[6] GW Grime, EF Garman (2023) Nuclear Inst. & Methods in Physics Research, B 540, (2023) 237–245.\n\nhttps://www.csc.liv.ac.uk/research/seminars/abstract.php?id=1445
LOCATION:
END:VEVENT
END:VCALENDAR