Figure 1.
Realistic illustration of the interior of an E. coli bacterial cell by D.S. Goodsell (TSRI).
It shows the spatial organization of the cell and its crowding: 1/The nucleoid (in yellow/orange) is made of DNA containing the genetic information and of proteins interacting with the DNA (to repair it for instance) 2/Some of these proteins read the genetic information and transfer it (see the red circle) into the cytoplasm (in blue/purple) where it serves to build new proteins performing most of the tasks inside the cell 3/Some of these proteins go back to the nucleoid, others remains in the cytoplasm, still others go to the membrane (in green). Other elements of this organization are the cytoskeletons (the blue filament in the upper left corner of the illustration), long and filamenteous “molecular motorways” that track various proteins from one pole of the cell to the other one. About genome organization in E. coli, the DNA is condensed into a compact structure called the nucleoid and is organized into macrodomains [25]. However the fine spatial organization of the DNA inside a macrodomain is still largely debated. This illustration is an original photograph of David Goodsell's painting. It differs from the version published in [26]–[27]. With kind permission of David S. Goodsell.
Figure 2.
Partial reproduction with GraphiteLifeExplorer of the E. coli illustration shown figure 1.
This multicomponent scene contains several ribosomes and proteasomes, many tRNA and elongation factors, various enzymes, a RNA polymerase transcribing the DNA, a replication fork, a RecA filament and two RecBCD complexes, several DNA binding proteins.
Figure 3.
To create DNA the user draws a path which can be adjusted in the three directions of space thanks to control points. The whole atomic structure is built upon the path. Closed structures can be created as shown by the circular A-DNA model.
Figure 4.
GraphiteLifeExplorer is equipped with a tool allowing to draw complicated shapes (like wrapping of DNA around a nucleoparticle for example) quickly and easily with a few number of control points. Note that the twist of the DNA can be locally or globally adjusted allowing one particular atom of the DNA to face one particular residue at the surface of the protein without changing the whole conformation (see figure 5).
Figure 5.
From top to bottom: 1/The initial DNA strand 2/A base pair is selected and 3/Twisted. All pairs follow the twist 4/Another pair is selected and 5/Twisted. The DNA is unwound between the two selected pairs 6/More pairs have been twisted.
Figure 6.
Visualization of DNA simulation results.
The comprehension of the (dynamic) configuration adopted by the DNA, of its packing into the cell is one of the major goals of biology. Numerical simulations address the topology of longer and longer DNA structures like plasmids. Here is shown the compaction of 28,000 DNA base pairs resulting from the action of fifty transcription factors (data shown courtesy of Ivan Junier; see [28] for details about the self-avoiding worm-like chain numerical simulation). Left image: The DNA is displayed as a thick line. Middle image: The DNA is displayed as a helicoidal double ribbon when the user zooms in. Right image: A transition between the helicoidal double ribbon rendering and the atomic rendering occurs when the user brings the camera closer to the object.
Figure 7.
Reconstruction process of a complex in GraphiteLifeExplorer.
Case of the bacterial DNA repair protein MutL: (a) the N-terminal (NTD) and the C-terminal domain (CTD) of MutL are imported under the form of a full-atom representation from their respective PDB file. (b) The CTD is moved manually relatively to the NTD. (c) A surface appearance is given where residues 331 of the NTD and 432 of the CTD are colored in blue to help the modeling of two 40 nm long amino-acid linkers connecting the two domains. Although their amino-acid sequence is known, the linkers are disordered and therefore missing in the attempts to experimentally solve the 3D structure of the whole MutL. Here, the linkers have been built as a simple tube of fixed length and in agreement with AFM images [29]. Some residues at the surface of the N-ter domain have been colored in red to show where another repair protein called MutH interacts with MutL. PDB code N-ter: 1B63, PDB code C-ter: 1x9z (both subunits have been reconformed as suggested in [30]).
Figure 8.
Connecting nucleic acid structures.
Missing DNA between two nucleosomal complexes containing DNA (PDB code 1KX5) is modeled with GraphiteLifeExplorer (upper image). The “welding” between modeled DNA and DNA from crystallography is carried out in a simple manner with Web 3DNA [10]. Whole DNA is shown as an isosurface in GraphiteLifeExplorer (lower image). Molecular dynamics would then be necessary to take relaxation effects into account.
Figure 9.
Top: Simple lighting from the viewpoint. Bottom: Ambient occlusion without lighting.
Figure 10.
Upper left image: This 3D scene has been made in GraphiteLifeExplorer and exported (upper right image) in Maya thanks to Molecular Maya (http://www.molecularmovies.com/toolkit/) and (lower images) in the Blender 3D tool thanks to the ePMV plugin [31]. In Blender, a script written by L. Autin (Scripps) superimposes an inverse kinematic armature to the linkers. This mechanical articulation greatly helps and facilitates the animation of the proteins from a first position (right figure) to a second position (left figure): thanks to the IK chain, moving the pink and green domains results in a reconformation of the linkers, pushed or pulled like a chain, in the deformation of the geometry (the tubes representing the linkers) attached to the linkers, and in the displacement of the CTD domain (green-blue/yellow). The model can serve as an interactive data-constrained thinking tool to help a lab contemplate plausible dynamics for this system. Note that the inverse kinematic joints of such an armature can be set at each alpha-carbon of any backbone if the linkers is no more a tube but an aminoacid based linker modeled with tools like Phyre2 [32] and be combined with physics solvers hosted in the high-end 3D packages.
Figure 11.
A uniform, rotation minimizing frame sequence along a cubic Bézier curve. The normal and binormal vectors are shown yellow.
Figure 12.
Casting a ray towards an atom.
Figure 13.
The structure corresponding to the PDB code 2IUU is imported and an isocontoured surface is created showing its shape. This shape is shown at three different precision levels from coarsed-grain (A, B) to molecular skin (C). Specific residues at the surface are highlighted in red to help the connection of this hexameric protein with six linkers.