We get a huge amount of data from our flow cytometers. A guaranteed way to impress visitors unfamiliar with the technique is to start talking about the number of parameters we can measure simultaneously and the speed at which we can analyse millions of cells. But there are of course limitations. Traditional flow cytometry tells us very little about cell morphology or cell-to-cell interactions and although we can find out which combination of fluorochromes is on, or in, each cell, we have no idea if the structures they are labelling are close together and interacting or entirely separated. All these things could be looked at by microscopy instead but that means sacrificing cell number so results are less robust, especially where rare events are concerned. To try to get around these problems we introduced imaging cytometry into the lab a few years ago. We now also look after a second for one of our research groups.
What are they?
Essentially an imaging cytometer is the hybrid offspring of a flow cytometer and a microscope. It can analyse up to 12 fluorescent parameters (10 if using brightfield 9 if using both brightfield and scatter), giving you the same kind of information you’d get from any of our other bench top machines but with the addition of images of the cells. It does this using a CCD camera with a time delay integration (TDI) mode which tracks the cells and can give light and darkfield images as well as images of each fluorescent channel on up to 5000 events per second. The imaging cytometers aren’t as fast and don’t have as many fluorescent parameters as our top end benchtop machines but for many of our users the addition of the cell images more than makes up for that and sometimes dots on a screen just aren’t as convincing as actually seeing your cells.
Our machines
We have two imaging flow cytometers in the Crick Flow Lab. The Amnis Flowsight is a compact benchtop machine with 20x magnification and 9-12 fluorescent parameters that are detected by the camera, rather than by the PMTs used in conventional flow cytometers. It has blue (488nm), yellow (561nm), Red (642nm) and violet (405nm) lasers. Its big brother, the ImageStream X MKII, also has the 9-12 fluorescent parameters but comes with more powerful lasers and 20x or 60x magnification. The trade off of course is the price, with the ImageStream costing more and taking up more precious bench space. It also can’t run at full speed when using the 60x magnification.
Actually running samples on the Imaging Cytometers is fairly straightforward, especially if you’re already familiar with conventional flow cytometers. Because a camera, rather than individual PMTs, detects the fluorescent parameters you can’t adjust the voltage on each parameter. Instead you need to tweak the laser power to alter the signal and this alters all the other fluorescent signals being analysed off that laser. This may mean you need to put more effort into fluorochrome selection and the concentration used in your sample. Something that works on a Fortessa might not automatically transfer to the ImageStream. My first attempt at cell cycle staining on the image stream just produced 10 channels full of PI with the channel I was interested in totally saturated and unable to tell me anything. But with a little bit of thought and sample prep good multicolour staining can be achieved.
The trickier part is analysing the data. Just like with any image analysis system it is possible to derive many metrics based on the images themselves as well as the localisation and spread of fluorescence. This does mean that to get the most out this technology, you need people to devote a good deal of time to the analysis and we have found, when training users, that analysis has to be tailored to their specific projects.
Our work
Although imaging flow cytometry is a relatively new technique, we have done all sorts of things with it in the last few years. Amongst the applications where we have found imaging essential have been looking at cell morphology and cell size, cell division, co-localisation of signals, cell death, DNA damage, and small point fluorescence amongst other things.
Sometimes with quite surprising results. Here are a few of the papers that came out of this work:
Mobilisation of Ca2+ in T Cells responding to different Stimuli
Asymmetric Cell Division
More Asymmetric Cell Division
Kirsty
