Movies (Molecular Dynamics, Microglial Motion)

Movies showing random-accelerated molecular dynamics (RAMD) simulations of NMDA receptor agonist unbinding, conventional molecular dynamics (MD) simulations of agonist movement in the NMDA receptor glutamate binding pocket, microglial motion in vitro, microglial motion in vivo, and custom Matlab routines for measuring the wavefront of microglial motion in vivo approaching a laser injury.  Microglial express GFP under the CX3CR1 promotor.

The automated Patch Clamp Recording station developed at Georgia Tech by Dr Craig Forrest and colleagues is now implemented in our laboratory as a collaborative effort between Drs. Traynelis, Jenkins, Forrest, Perszyk, and Yip. The system can make whole cell recordings autonomously without changing the pipette.

Random accelerated molecular dynamics (RAMD) simulation of glutamate dissociation from the cleft of the isolated GluN2B agonist binding domain.  Glutamate is in yellow, water oxygens in red, and side chains in cyan. When atoms are close enough to form a hydrogen bond, the bond is shown.

Molecular dynamics (MD) simulation of glutamate bound in the GluN2A NMDA receptor agonist binding pocket (Erreger et al., 2005, J Neurosci 24: 7858. Glutamate and key contact residues are in cyan, polypeptide C-alpha backbone is in orange. When atoms are close enough to form a hydrogen bond, the bond is shown.

Molecular dynamics (MD) simulation of homoquinolinate bound in the GluN2A NMDA receptor agonist binding pocket (Erreger et al., 2005, J Neurosci 24: 7858. Glutamate and key contact residues are in cyan, polypeptide C-alpha backbone is in orange. When atoms are close enough to form a hydrogen bond, the bond is shown.

 Molecular dynamics simulation of a GluN1/GluN2B model of the agonist binding domains and transmembrane pore derived from Karakas & Furukawa (2014). The pre-M1 residues are shown as stick structures. GluN1 is orange, GluN2 is blue

 Molecular Dynamics simulation of glutamate-bound GluN2B agonist binding domain opening. The unbiased simulation was initiated from a partially detached state of the ligand (obtained with Random Accelerated Molecular Dynamics")   (Wells et al., 2018)

 CX3CR1GFP/+ mice were imaged with two-photon microscopy over a 10-min period in the absence of tissue damage to establish baseline motility. Microglial processes extend and retract with no net directionality in both control and LPS-treated (2 mg/kg i.p.) mice. 

 Resting microglia in CX3CR1GFP/+ mice (Control) respond quickly to laser-induced tissue damage by extending their processes to the damaged site. In contrast, activated microglia in CX3CR1GFP/+ mice that were injected with 2 mg/kg LPS (LPS) respond to tissue damage at a slower rate.

Time-lapse recordings were analyzed with a custom-written Matlab code. The code splits the space in 36 sectors (10° increments) and detects fluorescent microglial processes in binary black-and-white movies in each sector. It then tracks the movement of the fluorescence in each sector at each consecutive time point. 

 Microglia from actin-GFP mice are cultured in Matrigel and imaged with a confocal microscope to obtain multiple optical planes through the cell in a time-lapse mode. For the resting microglial cell shown here, application of 20 µM ATP induces process extension that can be seen in 3D reconstructions of the cells performed in Bitplane Imaris. 

 GFP-expressing microglia maintained in culture in Matrigel adopt a complex process-bearing morphology characteristic of microglia in vivo. Microglia were optically sectioned using confocal imaging, and reconstructed in 3D as a function of time to produce a time-lapse movie. Application of ATP throughout the bath induces rapid and reversible process extension in all directions, which is associated with increase in process velocity, process length, and the surface area-to-volume ratio of the cell. Process extension occurs as a result of ATP activation of P2Y12 receptors that are expressed in resting microglia; we cannot detect expression of A2A adenosine receptors in resting microglia (Orr et al., 2009).

GFP-expressing microglia maintained in culture in Matrigel adopt a complex process-bearing morphology characteristic of microglia in vivo. Application of 100 ng/mL LPS leads to an activated phenotype with shorter processes and increased expression of pro-inflammatory mediators (IL-1β, COX-2, MCP-1, and others). Microglia were optically sectioned using confocal imaging, and reconstructed in 3D as a function of time to produce a time-lapse movie. Application of ATP throughout the bath induces rapid and reversible process retraction in activated microglia. This is associated with a decrease in process velocity, process length, and the surface area-to-volume ratio of the cell. Multiple lines of evidence suggest that process retraction results from rapid ATP breakdown to adenosine, which activates A2A receptors. A2A receptors are absent in resting microglia, but are upregulated during LPS treatment, while P2Y12 receptors are downregulated. The switch in surface receptor expression leads to ATP-induced retraction via adenosine activation of A2A receptors (Orr et al., 2009).

Baseline microglial motility was evaluated using confocal time-lapse imaging in slices from CX3CR1-GFP mice, which express GFP only in microglia. Imaging was performed in 200 um live slices of the subtantia nigra before induction of injury. The approximate border between the substantia nigra pars compacta (SNc) and pars reticulata (SNr) is shown with a line. A total of 31 optical sections (spaced 1 um apart) were taken and collapsed into 2–dimensions for analysis. The duration of the movie was 20 minutes. This movie shows random baseline movement of microglial processes under resting conditions (Gyoneva, Traynelis, unpublished data)

Microglial motility was evaluated using time-lapse confocal imaging in slices from CX3CR1-GFP mice, which express GFP only in microglia. Imaging was performed in 200 um live slices of the subtantia nigra immediately after mechanical injury was produced by driving a 20 um stainless steel rod 150 um into the slice at 100 um/sec. The approximate border between the substantia nigra pars compacta (SNc) and pars reticulata (SNr) is shown with a line. The site of injury is shown as an ellipse. A total of 31 optical sections (spaced 1 um apart) were taken and collapsed into 2–dimensions for analysis. The duration of the movie was 20 minutes. Note that cells close to the injury migrate towards it, whereas cells distant to the injury do not move. Also note the cell marked by the arrow, which shows particularly clear process extension toward the injured site (Gyoneva, Traynelis, unpublished data)

Evaluation using 2-photon in vivo imaging of the response of microglia to laser-induced injury in the brain of live anesthetized mice. Imaging was performed through a window produced by thinning the skull to ~50 um. The site of injury is visible in the middle of the image as a circle of bright autofluorescence from the damaged tissue. Cells rapidly extend processes towards the site of injury (Gyoneva, Traynelis, Davalos, Akassoglou, unpublished data)