The data describing the dynamic behaviour of every molecule labelled in the cell (up to 40000 molecules) can be very difficult to display in a meaningful way. If we were to simply super-impose 40000 tracks onto the cell image, the detail would be lost. We therefore use sophisticated software to render the data in the way we want. The example to the right is acquired from living cultured neurones, where we have tracked >5000 single molecules moving around the cells. Each square in the background is 250 nm along each edge (ie 250 billionths of a metre). The coloured lines indicate the trajectories followed by each molecule (shown as red spheres, each ~100X larger than a molecule would be). The colour of each line changes according to the time since the start of the experiment (indicated by the large 'time bar' in the lower right corner).

Being able to localise the positions of single molecules in their tens of thousands is truly amazing. The information contained in these datasets is enormously valuable. Of course, cells are alive, and things inside the cells move around, sometimes quite quickly. We can extend molecular imaging approaches to visualise the movement of many 1000s of single photoactivatable fluorescent molecules, at the membrane of living cells.

In order to make this approach work, therefore, we need to somehow determine what is signal (ie the light from our molecules) and what is noise (from the detectors, the room lights etc). Fortunately we can access the years of accumulated expertise of physicist colleagues expert in the rational 'de-noising' of image signal data, and subsequent tracking of single particles. The example on the left is a single frame from such an experiment. The molecules are barely visible against the noisy background - however, the 'de-noising' algorithms can identify each molecule and deliver quantitative data describing each molecule's trajectory, along with all the associated quantification. These approaches deliver an astonishing volume of quantitative data describing the precise movements and locations of many tens of thousands of single molecules within a living cell. This type of approach requires close collaborative research between cell biologists, physicists and statisticians, and perfectly illustrates the strength of the thinking behind wider Life Science Interface Theme.

LSI Molecular imaging - sptPALM

Images and data copyright LSI Laboratory

We can now visualise and track single molecules, inside living cells, with nano-scale precision. Biology often occurs on a very fast timescale (in the order of microseconds) however, and molecules, particularly proteins, rarely act in isolation. Proteins are the machines of the cell and assemble into macro-molecular structures in order to perform their tasks.

Introduction        PALM       Single particle tracking PALM        FCS        FLIM        GSD Microscopy

In order to do this, we need to image the samples very quickly (the reason for this is explained here). The problem is that each single molecule can emit only around 2000 photons (that's not a lot of light), and imaging at a high frame rate increases the apparent noise from the detectors in the system. Combined with tracking approaches, this is referred to as single particle tracking PALM (sptPALM).

In addition to the 'super-resolution' imaging approaches described here, we also employ bio-physical spectroscopic techniques to quantify dynamic protein interactions and movements, again inside living cells, but this time with microsecond time resolution.