Estimating Cellular Abundances of Halo-tagged Proteins in Live Mammalian Cells by Flow Cytometry.

Accurate abundance measurements of cellular proteins are required to achieve a quantitative and predictive understanding of any biological process inside the cell. Existing methods to determine absolute protein abundances are labor-intensive and/or require sophisticated experimental and computational infrastructure (e.g., fluorescence correlation spectroscopy (FCS)-calibrated imaging and quantitative mass spectrometry). Here we detail a straightforward flow cytometry-based method to measure the absolute abundance of any Halo-tagged protein in live cells that uses a standard mammalian cell line with a known number of Halo-CTCF proteins recently characterized in our lab. The protocol only comprises a few steps. First, a cell line expressing the Halo-tagged protein of interest is grown and labeled side-by-side with our standard line. Then, average fluorescence intensities are measured by conventional flow cytometry analysis and finally a simple calculation is applied to estimate the absolute number of the Halo-tagged protein of interest per cell. Once the protein of interest has been endogenously tagged with HaloTag, which we routinely achieve by Cas9-mediated genome editing, the presented protocol is fast, convenient, reproducible, cost-effective and readily accessible.

2018; Walther et al., 2018;Holzmann et al., 2019). However, it again requires advanced microscopy equipment and computational infrastructure. Electrophoresis-based methods (e.g., traditional SDS-PAGE or more advanced capillary systems followed by staining/blotting (Chen et al., 2013;Bennett et al., 2017) are less complex, but still rely on accurate cell counting and/or depend on the availability of a pure recombinant protein to use as a calibrator.
We have recently developed a new flow cytometry-based absolute quantification method that can be easily applied to any Halo-tagged protein in the cell. This was part of a larger effort to quantify nuclear transcription factors and architectural 3D-genome regulators like CTCF and cohesin (Cattoglio et al., 2019), with the ultimate goal of deciphering the spatiotemporal regulation of transcription and its interplay with 3D genome organization. The method has several advantages over existing ones: it employs a human U2OS-derived cell line standard that is easy to culture; it only requires a flow cytometer, an instrument available in most institutions; it has single-cell resolution; it can be reproducibly repeated multiple times with little effort and cost; and, it involves very basic data analysis.
We focused on the HaloTag protein-fusion platform (Los et al., 2008) because of its popularity and versatility, with applications in a broad range of experimental systems (England et al., 2015), and since it is currently the preferred choice for live-cell single-molecule imaging (Presman et al., 2017). In fact, we first generated what is here used as a standard cell line with the intent of studying CTCF nuclear dynamics by single-molecule super-resolution microscopy (Hansen et al., 2017). This is a U2OS human clonal cell line (clone C32) with the endogenous CTCF locus homozygously engineered via Cas9mediated genome editing to express a Halo-CTCF protein. As such, its expression level is stable and reflects endogenous CTCF. We realized that clone C32 could become a quantification standard for any Halo-tagged protein if we could estimate and cross-validate the absolute protein copy number per cell of Halo-CTCF in this U2OS cell line. We achieved this through a combination of in-gel fluorescence and FCS-calibrated imaging, obtaining consistent results across the two orthogonal methods performed in two different labs (~104,900 ± 14,600 and 114,600 ± 10,200 Halo-CTCF proteins per U2OS interphase cell, respectively) (Cattoglio et al., 2019). We took the mean of the two methods, 109,800 CTCF proteins per cell, as our best and final cross-validated estimate. This number can be now easily used to infer absolute abundances of other Halo-tagged proteins by simply growing, labeling and measuring fluorescence intensities of the cell line of interest side-by-side with our C32 standard, as we have previously demonstrated using Sox2 and TBP in mouse embryonic stem cells (Teves et al., 2016 andCattoglio et al., 2019) (see also the data re-analysis in Table 1). After subtracting the background fluorescence of unstained cells, the mean fluorescence intensity of the cell line of interest is divided by the C32 standard fluorescence and multiplied by 109,800 to obtain the absolute abundance of the Halotagged protein under study ( Figure 1). Although we focus exclusively on endogenously tagged cell lines here, the same pipeline can also be deployed to calculate abundances of exogenously expressed Halotagged proteins. 3 www.bio-protocol.org/e3527  Subtract the background fluorescence of unlabeled controls and use the adjusted fluorescence values to calculate the absolute abundance (n) of the protein of interest (POI). 109,800 is the average number of Halo-CTCF molecules we estimated in the U2OS C32 cell standard.
The quantitative approach described here comes with a few limitations. It does not allow direct measurement of the concentration of the protein of interest, nor its quantification in different cellular compartments, it may not be robust for very lowly or very highly expressed proteins (below ~10,000 and above 10-20 million proteins/cell, respectively), and it is obviously restricted to Halo-tagged proteins. 4 www.bio-protocol.org/e3527 Apart from these few drawbacks, our quantification protocol is simple and yet powerful. For instance, although we have only estimated average cellular protein numbers so far, the method can be used in principle to measure protein abundances in single cells, since the flow cytometer outputs fluorescence values for each individual cell. Another advantage of using flow cytometry as a readout is that the Halo- In summary, we here detail a rapid, straightforward and convenient method that will allow other researchers to estimate absolute cellular abundances of their Halo-tagged protein of interest with only a few hours of hands-on work.

Materials and Reagents
Note: With the exception of the U2OS C32 cell standard, you can replace all the listed reagents, equipment and software with equivalent alternatives.  g. Distribute the cell suspension to 6-well microplates. Seed at least 2 wells per cell type: one to be left unlabeled to measure background fluorescence and the other to be stained with TMR. One ml of cells per well will give you enough cells to perform staining and flow cytometry the day after seeding (1 ml will contain ~250,000-500,000 U2OS cells when starting from a 60-80% confluent 100-mm plate). Use less if you need to delay the experiment (e.g., 0.5 ml to wait one additional day or 0.25 ml to wait a couple of days).

5.
In the upper menu bar, click "Experiment" and choose "New Experiment" from the dropdown menu.
6. Select a blank template. Your experiment will now appear as an opened book in the Browser window. Right-click on the experiment name and click "Rename" to assign a different name. 9 www.bio-protocol.org/e3527 8. Move to the Inspector window to setup the Cytometer Settings. In the "Parameters" tab, select and delete everything but FSC (forward scatter), SSC (side scatter) and PE-Tx-Red-YG (for TMR staining; the instrument will use a 561-nm laser as an excitation source and read the emission through a 610/20 band pass filter or similar (TMR emission max is ~585 nm). If you are using a different fluorophore, modify accordingly. Check "A" and "H" for the FSC and SSC parameters to record the signal area and height (this will allow you to distinguish single cells from doublets during analysis). For the PE-Tx-Red-YG parameter, check "A" and "Log", for the data to be in logarithmic scale. 14. With the histogram plot selected, move to the Inspector window and check "P1". This will only display the TMR signal from the live cells that you gated in 12.
15. Move back to the Browser window. 16. Expand the specimen icon by clicking on the "+" sign to expose the "Tube" icon. Right-click on it and rename it with your sample name (U2OS_C32_unlabeled).

and internal complexity (SSC). Live cells have larger FSC and SSC values than dead cells or cell debris, and normally constitute the majority of the population. Cell doublets/aggregates will have higher FSC and SSC values than single cells but should not represent the majority of cells,
if you have thoroughly resuspended the samples (you will exclude doublets at the analysis stage. See Step E5 below). 28. In the Acquisition Dashboard window click on "Next Tube" to create the next sample. You can rename it either right-clicking on it in the Browser window or from the "Tube" tab in the Inspector window. 29. Take the next sample and repeat Steps D18-D21 and D27-D28, until you have acquired and recorded all your samples.

Back to the
30. Perform any recommended cleaning procedure for your flow cytometer. 31. In the Browser window, export your data by right-clicking on the specimen name. Select "Export", "FCS files" and leave the default FCS3.0 format. Click "OK", when prompted choose a directory where to save your data and click "Save". 32. Quit the BD FACSDIVA TM software (from the top menu click first "File" and then "Quit").   c. In the workspace window select the "Live Cells" gate and apply it to all samples (simply drag and drop it on the above window on the "All Samples" group).

FlowJo. A license-free alternative is the Matlab code we used to analyze the data in our
d. Go back to the forward and side-scatter plot; move from one sample to the next using the arrow at the top-right of the window and adjust the "Live Cells" gate as needed, dragging it around. If necessary, you can also modify the gate shape ( Figure 2). 12 www.bio-protocol.org/e3527   Table 1 below we could not exclude   cell doublets as recommended in Step E5. a. Go back to the U2OS C32 unlabeled control's scatter plot and double-click anywhere inside the "Live Cells" gate. This will open a new scatterplot that only displays gated cells.

record FSC-H and SSC-H values. Thus, in the re-analysis of
b. Set the y-axis to FSC-H and the x-axis to FSC-A to exclude doublets. Single cells localize along the same diagonal, while doublets lie underneath (Figure 3). c. Create another free-polygon gate to only include cells along the top diagonal. Click "OK" to name it "Single Cells". d. In the workspace window select the "Single Cells" gate and apply it to all samples (simply drag and drop it on the above window on the "Live Cells" subgroup).
e. Adjust the "Single Cells" gate to each sample individually as you did in Step E4d for the "Live Cells" gate. 6. Obtain a mean fluorescence intensity (MFI) table a. Back to the workspace, select the Table Editor and click the "Edit" tab. Click "Add Column" and select "Mean" on the left, and the "Single Cells" Population and the "PE-Tx-Red-YG-A" Parameter from the dropdown menus on the right. b. Click OK and go back to the "Table Editor" tab. Click on "Create Table". This generates a table with the mean fluorescence intensities for all the samples and the unlabeled controls.
To download the table as an Excel file, from the dropdown menus to the right of the "Create Table button" choose "To File" and "Excel". Choose a "Destination" folder and hit "Create