Hi.In this video we present a full single quantum tracking experiment. A dedicated algorithm has been entirely conceived in collaboration with image ISIS experts. During this video, we will use the epidermal growth factor receptor as a model to describe the method of this technique.
The epidermal growth factor receptor is a transmembrane protein. This receptor will be tagged using a monoclonal antibody, which is reduced to keep only the A a B fragment. The A a B fragment is biotinylated, and after that, linked to the streptavidin of the quantum dot, the wall system will precisely recognize the EGF receptor.
Our goal is to provide a comprehensive view of cell surface receptors dynamics. Indeed, molecular trajectories can deviate from brot diffusion by being confined within nano domains. For instance, and this may be the signature of an underlying interaction with, for example, signaling partners or molecular scaffolds.
We aim at describing such events exhaustively by mapping them over the cell, which requires working at high SPECT temporal resolution together with high labeling density. We hence name this approach multi targett tracing. The first step of the experiment is the preparation of the sample.
We work on live cells with without antibiotics. Here we use cost seven cells, which endogenously express EGF receptors. After detaching the cell, we need to confirm precisely in order to have the same number of cells in h well of the lab.
A precise number of cells is very important to increase the predictability for the number of quantum dot per cell After dispensing the cell in H, well, you need to incubate the labate for 24 hours at 37 degrees with 7%of CO2. The next step is the preparation of quantum dot coupled to antibodies. Quantum dots are small fluorescent nanoparticles.
These particles have a very big interest here because they are very bright and stable compared to classical fluorescent probe and signal. Nose ratio is very important for this type of experiment. In this case, we use complete medium to saturate the strip ides around the quantum dots and to prevent aggregation.
After that, we had the quantum dots in the bioTE elated protein or antibodies of interest with one-to-one ratio in order to have statistically more monovalent quantum dots. In this experiment, we use a a b fragment against EGF receptor produced in the lab from monoclonal antibodies and which are targeted with biotin. During the incubation around 15 minutes, you can keep the control under aggregation and homogeneity using a shaker at 1, 200 rotation per minute and two five degrees.
When the mix is ready, you can add it on your cells. Remove the medium from the well very gently to prevent cell detachment on the lab tech and add the needed quantity of mix for your experiment. In this case, we will add 100 microliters of mix per well.
After addition of the mix, you can incubate your cell for 15 minutes. In this case at 37 degrees with 7%of CO2. After this incubation, you may find a high number of frequent twin dots in suspension on the medium.
This quantum dots should be removed before imaging. Because of the noise we bring. It's why we need to wash each, well several time in this case, five times to wash yourselves.
Use imaging medium with non autofluorescence. Here we use HBSS buffer with aps. Now your cells are ready.
It's time to go to the setup for acquisition. The setup is composed of four major part, an inverted microscope, a camera with eye sensitivity, a powerful mercury lamp, and an incubator to keep cell at 37 degrees. The light from the mercury lamp pass through a fiber and through a filter wheel before illuminating the sample.
The flu eSense of the sample is filtered and collected by an E-M-C-C-D camera on the left. We currently use 1.3 and 1.49 numerical aperture oil merchant of objectives. The central step of this experiment is the acquisition by itself.
Once the cells are in focus, we look at a representative one with a strong label. In quantum dots, we first acquire a single image in transmitted white light, which can further be used to check the cell aspect and special limit of the LA podia. We then acquire video typically at 36 millisecond rate, the fastest rate allowed in the full frame.
We use an electron multiplying CCD to reach single molecule sensitivity with a sufficient signal to north ratio at least above 20 decibel. Typically around 25 decibel, we usually acquire 300 frames. To evaluate a given dataset, you only need to provide the pass to the directory containing the video files.
Then typing the command, the text reconnects in MetLab or or the command MTT 23 I, which first displays a graphic interface listing. All the parameters used will start the fully automatized analysis. The core MTT analysis is performed over each frame, invoking three main tasks, first detection for each pixel for the presence or absence of a target quantum debt.
Then for each detection estimation of the target relevant parameters such as its pixel position, signal intensity, and so on. Last reconnection of the new targets. With the traces already built over the previous frames for every pixel.
Considering a local subregion, we compare two hypothesis presence, either of only noise or of a signal. With the point spread function, ModuLite has a hin peak. We use a threshold ensuring low enough false alarms with less than one perus detection per frame for every detected target.
We next perform a least square ghost nutrient feet to estimate the position width and height of the detected Goshen. This provides notably the spic cell position of the dye, typically around 10 to 20 nanometers accuracy for typical signal to noise ratios at high density peaks can often be too closed and strong peaks may hinder weakest ones to handle that. We deflate the detected peaks from the initial image running.
Again, detection on the residual can typically re skew 10%of the peaks. Reaching an almost exhaustive detection is very fruitful for accurate reconnection. Then the set of new targets is matched with the set of previous traces for this pupil.
In order to assign each trace to a target if possible, and handling possible blinking, we use all available statistical information obtain from the detection step. Hence, targets are not just assigned to the nearest trace. In case of conflicts, when traces may cross, meaning that the respective relevant regions of research of overlap, we will consider the statistical value of intensity, speed, width, and blinking both for the traces and for the targets.
This delivers the statistically optimal reconnection score. The strategy permits to avoid when possible, biasing the reconnection toward the nearest neighbors, meaning the slowest motion. We next use a function to evaluate possible transient confinement within the trajectories.
This function is inversely related to the local diffusion as established by Sexton Simpson and collaborators. Applying a threshold allows to define confined or not episodes. By iterating this overall traces, we can finally map membrane dynamics in terms of transient confinement, slowdown evidence, and this can be represented.
Alternatively, using the binary or discrete values of this confinement index, by default, MTT will automatically perform those tasks, saving for each video, the row parameters for each peak in an eski file, and for further advanced investigations. Typical results such as the map of traces over each cell and plots histograms for parameters of interest, peak intensities, signal to noise ratio, local deficient values, and so this aspect can as well be easily adapted to any dedicated investigation. This video will now summarize typical results obtained by MTT.
A substantial part of the work resides in elaborating the algorithm, which may need to be adapted to dedicated investigations such as the suffering modes of motion or interactions. But running MTT is very straightforward. Users should only optimize a few parameters such as the space and time limits.
When elaborating MTT, we aimed at entirely reconsidering the analytical options used for each task. We optimize the process along two challenging axis. First, dealing with high densities to get the best special information over the surface surface, and secondly, handling weak SNR, which typically allows working at low elimination and high speed confinement can be interpreted as the signature of an underlying interaction.
Membrane receptors may interact with the sub membrane cytoskeleton or pro lipid domains, for instance. Such events can be investigated through confinement variations in space and time. Dynamic measures may be compared to complementary approaches such as FRA F Cs, or freight.
The open source code is available for academic research. It can be downloaded from our webpage at CML dot univ if an S fr on our teams page, which provides a link for download. Interested industrials are also welcome to contact us.
We would like to specially thank the members of our team and of the common facilities for the support and fruitful discussions. This project is supported by funding from the CNRS in c MAA University Pac Region National de La, and Foundation Medical is supported by the league controller. Thanks for watching.
