| Scientific
Aims
My function at MCDB is two-fold. First, I took over the directorship
of the IVEM Lab from Prof. JR McIntosh, and secondly I will continue
my personal research that focuses on cryo-electron microscopy (cryo-EM)
based three-dimensional (3-D) reconstruction of large macromolecular
assemblies and cellular structures, whenever possible within the
in vivo context of an intact cell. These two responsibilities match
well together and as such there will be a strong interaction between
the two units. So far structural biology was often confined to in
vitro approaches, thereby reducing a complex biological system to
a very limited problem. While this will certainly still be the case
for some time for generating atomic-resolution data (X-ray, electron
crystallography or NMR spectroscopy), cryo-electron microscopy in
combination with new (tomographic) 3D reconstruction approaches
clearly has a perspective to investigate cellular structures in
vivo, not necessarily at atomic resolution, but at 2-3 nm detail
which will allow to recognize single protein domains within the
context of larger macromolecular assemblies and cellular organelles,
directly in the cell. With my background on microtubule structures
and associated proteins such as MAPS and molecular motors of the
Dynein and Kinesin families, a correlative microscopy approach will
be very useful for high-resolution studies on various dynamic aspects
in cells such as cell division (=> e.g. Cancer research), and
intracellular transport processes (e.g. in axons, neuronal tissues
=> Neuroscience). Apart from focussing on microtubule structures
alone, I am planning to further investigate the regulation of the
entire cytoskeleton, and in particular the functional connections
between microtubules, actin and intermediate filament structures
in an intact cell.
Preserving the cellular architecture will be a key-preparative issue,
while handling 3D data of large cellular structures will be the
key-interpretative issue for the future. This requires developing
procedures to handle large specimens and applying methods for cryo-EM
based 3-D reconstructions, which are beyond the conventional averaging
methods. To date the large majority of (cryo-) EM derived 3-D data
are derived from approaches, which attempt to average large numbers
of (seemingly) identical image elements and therefore either rely
on the accuracy of various kinds of symmetries expected to exist
within a macromolecular assembly (e.g. icosahedral viruses, helical
filaments or 2-D crystals), or at least expect individual particles
to be structurally identical (so called single particle averaging).
However, the biological default state of the vast majority of structures
from protein complexes to cellular structures is usually far from
this requirement. To this end we are moving towards tomographic
3-D data acquisition, which has been demonstrated by various pioneering
groups (Ellisman, Baumeister, McEwen, etc.) to be an extremely promising
technique. In principle tomography allows reconstructing in 3-D
any object that is suitable to be visualized in an EM. Once an initial
3-D dataset exists averaging may still be useful to further refine
pre-reconstructed particles and complexes at various conformational
states. On my previous post at EMBL I have just now succeeded to
collect funds for the purchase of one of the newest generation cryo-EM,
which uses liquid-helium technology for a better specimen conservation,
and a revolutionary new tilt-stage design for enhanced stability.
Both features are ideal to optimise tomography-based 3-D data acquisition.
Ranging in resolution typically from 1-3 nm, tomography is not the
method to generate data at atomic detail, but yields 3-D scaffolds,
which can be used to dock individual X-ray or NMR-data for structural
and functional interpretations at near-atomic detail. When going
in the other direction on the resolution scale, tomographical EM
data allows an advanced interpretation of light-microscopy and laser-confocal
3-D data, and therefore constitutes a very useful bridge to link
structural details from cells to atoms. Based on these perspectives
I would like to formulate our (maybe slightly over-ambitious) aim
for a distant future of being able to reconstruct highly complex
cellular structures (e.g. an entire spindle, nucleus, axon and synapse,
etc.) at near atomic detail, using tomography 3-D data as a basis
to combine all required information from various sources.
|
