Conformational Ensembles from Experimental Data and Computer Simulations

Conformational Ensembles from Experimental Data

and Computer Simulations

Poster Abstracts

60-POS

Board 20

Elucidating Structure and Conformational Changes with Cross-Linking/Mass

Spectrometry (CLMS)

Michael Schneider

Mahmoud Mabrouk

, Kolja Stahl

, Oliver Brock

, Juri Rappsilber

3,2

Technische Universität Berlin, Berlin, Germany,

Technische Universität Berlin, Berlin,

Germany,

University of Edinburgh, Edinburgh, United Kingdom.

Protein structure and dynamics are studied typically outside the native context needed for correct

function and possibly also folding. Cross-linking/mass spectrometry (CLMS) is a maturing

structure analysis tool, poised to overcome this limitation. Under native conditions, cross-linkers

can convert proximity into chemical bridges that can be observed experimentally. CLMS can

reveal the structure of individual proteins, probe protein assemblies, and deliver proteome-wide

information. Our labs recently added novel chemistry to drastically improve the density of data

obtained.

Leveraging CLMS data computationally remains challenging since it contains noise and

averaged data. We recently demonstrated that using high-density CLMS data, which increases

the amount of cross-links by an order of magnitude, enables the de novo reconstruction of the

human serum albumin domains. The key to this success is the complementary integration of

experiment and computation: We were able to utilize CLMS data with a degree of noise that

would normally be rejected using a noise intolerant conformational space search algorithm. Our

method is novel in two key aspects: first, it models cross-link constraints with a modified

Lorentz function allowing it to be robust to wrongly predicted cross-links. Second, it leverages

cross-links to guide the search towards promising regions which are then searched using an

unbiased all-atom energy function.

Combining high-density cross-linking with quantitative experimental setups hold a high potential

to elucidate protein dynamics. We have recently made use of CLMS to reveal subtle

conformational changes in the complement protein C3 and its activated cleavage product C3b.

Yet, the development of computational methods for interpreting this data is still in its early

stages. We see the upcoming conference as an ideal chance to disseminate our findings to a panel

of experts and discuss the future of this novel type of experimental data.