Harvard Tissue Atlas Harvard Medical School
Methods & Software

Cyclic Immunofluorescence

CyCIF: An iterative method for generating highly multiplexed images

The CyCIF Workflow

Depiction of CyCIF workflow. Workflow is described in the Experimental Process text below. Click to enlarge.

Experimental process

CyCIF is a multiplexed imaging method that uses cycles of staining, bleaching, and imaging to generate high-plex images. CyCIF can be applied to many sample types (see references below). A major advantage of our workflow is that it can generate and analyze whole slide images (WSI). WSIs enable deep spatial analyses and provide enough statistical power to make conclusions about spatial relationships that are impossible in small fields of view.

Here, we describe how to use CyCIF with formalin-fixed, paraffin-embedded (FFPE) human tissue sections, known as tissue-based (t-) CyCIF.

In the t-CyCIF method, a tissue sample is first pre-stained with fluorescent secondary antibodies to reduce auto-fluorescence caused by non-specific antibody binding. The sample is then bleached with hydrogen peroxide to inactivate the fluorophores. Next, the sample is stained with 1-4 primary antibodies that target proteins or molecules of interest, paired secondary antibodies, and a nuclear stain. After a few rinses, the sample is imaged using a fluorescence microscope.

Then the cycle begins again: the sample is bleached, re-stained with new antibodies and a nuclear dye, and imaged again.

Once the cycles of fluorescence imaging are complete, the slide can optionally be stained with hematoxylin and eosin (H&E), a standard stain used in the clinic for diagnosing disease. The slide is then imaged once more with a bright-field microscope.

Extracting and analyzing spatial features

The images from each CyCIF cycle are then assembled into a high-plex whole slide image, processed into single-cell data, and quantified for spatial features using the open-source MCMICRO pipeline. The single-cell data can be analyzed with tools like SCIMAP, then visualized and shared with the software Minerva. Visit Software to learn more.

Five images of 4-6 channels each are overlaid into a single mosaic image containing ~20 channels. Note: The mosaic image shown here contains data from only a small portion of a slide. A whole slide image would be made up of many image ‘tiles’ of small regions, which must be assembled into the full whole slide image. Click to enlarge.

How to cite CyCIF

If you use the CyCIF method, please reference:
Lin J-R, Izar B, Wang S, Yapp C, Mei S, Shah PM, Santagata S, Sorger PK. Highly multiplexed immunofluorescence imaging of human tissues and tumors using t-CyCIF and conventional optical microscopes. Elife. 2018 Jul 11;7. PMCID: PMC6075866.

Getting started with CyCIF

The CyCIF method is flexible, does not require expensive or specialized reagents, and can be customized to work with many tissue types, microscopes, and analysis pipelines. For these reasons, CyCIF works well for exploratory experiments and for probing specific biological hypotheses.

CyCIF can be used with human and animal tissues of many origins, although denser tissues are typically more robust and can withstand more cycles. If you’re new to CyCIF, we recommend using a reference sample, like tonsil, to optimize the procedure in your laboratory. Once you’ve achieved the expected staining results with tonsil, proceed to your tissue of interest.

Available protocols

CyCIF Protocol pre-CYCIF FFPE sample preparation

Designing your experiment

  • Multiplexing with CyCIF requires conjugated primary antibodies
    • Since CyCIF does not strip antibodies away from the tissue, staining with secondary antibodies should only be performed in the first cycle, when there will only be one primary antibody for a given host species bound to the tissue.
  • Not all fluorophores bleach as efficiently as others; AlexaFluor antibodies bleach particularly well in our experiments.

  • The buffer used for staining and blocking can impact the performance of antibodies and nuclear dyes
    • Some antibodies will perform better in certain buffers; some buffers will cause the nuclear stain to fade more quickly over cycles.
    • When validating antibody staining performance, we recommend testing your buffers and antibodies as pairs and using these results to design staining plans (for each experiment) that match antibodies with an appropriate buffer.
  • Background autofluorescence tends to fade across cycles
    • We recommend using higher signal antibodies in early cycles and lower signal antibodies in later ones.

We have successfully used hundreds of antibodies in many different tissues and include an antibody list as a supplementary file in most of our papers. Please refer to these files for more information about the specific antibodies that have worked in our laboratories. Keep in mind that antibodies perform differently in different experimental conditions and require validation.

Overview of the t-CyCIF method
Click to enlarge
Example CyCIF data from Lin et al., 2018
Click to enlarge

Selected Publications using CyCIF

Please visit Publications for a more complete list of Harvard Tissue Atlas Publications.