General advice and considerations
General advice and considerations
Imaging mass cytometry uses antibodies raised against specific protein targets in an almost identical way to traditional immunohistochemistry - the only real difference is the way that the signal coming from each antibody is detected. Therefore, many of the best working practice that apply to designing effective immunohistochemistry experiments also apply to IMC. Understanding IMC as massively multiplexed immunohistochemistry is not a bad approximation, however there are several important considerations for using IMC:
The resulting data is quantitative. Because the exact abundance of each metal can be quantified, the data resulting from IMC is much more quantitative than traditional fluorophore-based approaches.
Good nuclear staining is critical. Getting the nuclear stain right is critical for using most downstream analysis pipelines (e.g. using CellProfiler and HistoCAT – the ‘off the shelf’ analysis pipeline). Thankfully, the iridium stain works really well and is very consistent. However, if you can’t identify individual nuclei by eye, either because the nuclei are too poorly stained or closely packed, it's very unlikely that automated cell segmentation will be able to do it either.
Antibodies are directly conjugated, so there is very little signal amplification. Because the primary antibodies are directly conjugated to metals, we don’t use secondary antibodies. The disadvantage of this is that secondary antibodies normally serve to amplify the signal. That means the sensitivity is usually reduced, and our signal might not be as strong as it might be in IHC. This isn’t much of an issue for highly expressed markers (e.g. CD markers for immune cells), but is worth considering if you normally need to use a lot of amplification or a high exposure to visualise by IHC/IG.
Consistency is key! Being as consistent as possible in your staining to reduce run-to-run variation in staining is very important. Whereas slight variation in staining would not be obvious in traditional fluorophore based approaches, it is immediately obvious in IMC as the data is quantitative. This means be precise even in steps which are usually forgiving in traditional IHC, for example wash steps.
It's more than just 'pretty pictures'. Although the Hyperion will certainly give you striking visuals, IMC primary strength is in the ability to quantify the pictures, getting spatially-mapped single-cell data that you can use for quantification. Downstream data analyses will be more effective if antibodies have a good 'signal to noise' - we want to optimise our staining to minimise background and have a strong specific signal.
IMC is compatible with frozen and paraffin-embedded sections. Preparation of tissue is almost identical to regular on-slide immunohistochemistry, and IMC will work for both frozen and paraffin-cut sections. The main consideration is that tissue should be cut thin enough so that only one layer of cells is present on the slide, otherwise nuclei could overlap - 5 to 10 μm is ideal, but this is highly tissue dependent! For example, thinner sections are better for spleen as it’s very densely packed with cells.
Paraffin-cut sections should ideally be freshly cut. Tissue sections stored for long periods of time lose their antigenicity, which could reduce the signal seen in IMC. Where possible, use freshly cut paraffin sections.
IMC is compatible with any species. The IMC has most extensively been used to study human tissue, though several groups at UoM are also using it to study mouse tissue.
Pre-conjugated antibodies are available from Fluidigm, or can be conjugated in-house. Fluidigm has a growing range of 'pre-validated' Hyperion antibodies for use in human formalin-fixed paraffin-embedded (FFPE) tissue. For those working with frozen mouse and human tissues, their Helios range of antibodies could also be trialled, and may work in some tissues. Details of the conjugation service provided by the facility can be found here.
Fluidigm have protocols available for staining frozen and paraffin-embedded tissue. These are several resources available on Fluidigms website, including protocols for frozen and paraffin-embedded tissues. These are good places to start for optimising staining, however you may already know optimal staining conditions based upon successful immunohistochemistry experiments. In theory, IMC should be compatible with most blocks/fixation/permeabilization methods.