Iterative restaining as a pivotal tool for n-color immunophenotyping by slide-based cytometry


  • Part of this work was presented at the 10th Leipziger Workshop “Systems Biology and Clinical Cytomics,” April 7–9, 2005, Leipzig, Germany.



Slide-based cytometry (SBC) allows to “ask a cell a second time.” We used this tool for detailed immunophenotyping of peripheral blood leukocytes (PBLs).


PBLs primarily stained for CD-markers and DNA were immobilized on a glass slide and analyzed by laser scanning cytometry. Then, iterative restaining was applied for a second and a third analysis. Based on the cells' fixed location, analyses were merged on a single-cell level.


We analyzed six virtual immunostainings by “recycling” the PE-channel for four CD-markers.


Iterative restaining might prove to be a pivotal tool for n-color immunophenotyping exclusive to SBC concepts. © 2006 International Society for Analytical Cytology

Although there are many similarities between flow and slide-based cytometry (SBC), there is one fundamental difference: if performed properly, no cell is lost in slide-based assays, as cells can be not only structurally fixed but also mechanically immobilized on the slide. This means that cells do not change their position on the slide. In previous publications, this attribute was used to simply counter-stain cells with conventional cytological dyes after fluorescence analysis for morphological documentation (1). It can also be used to gain supravital information before and after fixation of cells cultured on the slide in apoptosis studies (2). To obtain this combined information, the data sets of the two analyses are merged to one new composite file on a cell-to-cell basis based on the cells' x- and y-coordinates on the slide. We developed an assay that exploits this feature unique to SBC so as to perform immunophenotyping with virtual colors.


1° Staining and 1st Analysis

All steps are performed at room temperature in the dark. Peripheral blood (20 μl) collected in EDTA tubes are incubated with 2.5 μl of each of the following antibodies: CD45-APC (Caltag, Hamburg, Germany), CD14-PE (DAKO-Cytomation, Glostrup, Denmark), and CD3-FITC (DAKO-Cytomation). After 15 min incubation, 1 ml of FACS lysing solution (BD Bioscience, Heidelberg, Germany) is added. After 15 min incubation, cells are spun down and the pellet is washed with phosphate-buffered saline (PBS; Gibco-BRL, Paisley, Scotland, UK) and spun again. After decanting, the pellet is resuspended and pipetted onto a conventional glass slide. After air drying, slides are rehydrated and incubated with 100 μl of 25 μl/ml 7-AAD in PBS supplemented with 0.5% BSA for 20 min in a humidified chamber. Slides are washed and covered with 40 μl of glycerol/PBS 75%/25% supplemented with 2.5 μg/ml 7-AAD (Sigma-Aldrich, St. Louis, MO). Cells are then analyzed in the laser scanning cytometry (LSC; CompuCyte, Cambridge, MA) with triggering on the DNA.

2° and 3° Staining with 2nd and 3rd Analysis

After the 1st analysis, the cover slip is removed and cells are incubated with 2.5 μl CD4-PE (Caltag) in 90 μl PBS for 15 min. Cells are washed, covered again, and the 2nd analysis is performed in the LSC with identical settings. In the same way, the 3° staining with CD8-PE and CD19-PE (both, Caltag) with the subsequent 3rd analysis is performed. After completing the analyses, slides are stained with HE and single cells of the various subpopulations are relocalized and micrographs are taken to document the cells' morphology.

Data Analysis

The data sets of the 1st, 2nd, and 3rd analysis are combined using the “merge” function of the proprietary WinCyte®-software (CompuCyte). This means, that data from two analyses are overlaid based on the x- and y-coordinates of the events. It is a prerequisite that (a) cells do not move on the slide, and (b) the slide is positioned into the microscope stage in a highly standardized and reproducible way.


We observed that the cell-loss resulting from the repeated detachment of the coverslip summed up to 5% but was not cell-type specific. Detachment of the coverslip was achieved by placing slides upright in a container filled with PBS. If coverslips were manually withdrawn from the slides, merging was virtually impossible due to cell loss and cell movement on the slide.

Pretests showed reduced staining intensity for CD-antigens on peripheral blood leukocytes (PBLs) fixed and immobilized on glass slides. However, staining was still sufficient for the discrimination of positive versus negative subpopulations when the data of the 1st and 2nd analysis were plotted against each other: cells with unchanged fluorescence line up in a 45° diagonal, whereas cells with newly attached fluorochromes shift off that diagonal (see Fig. 1).

Figure 1.

Merged .fcs-file before and after restaining. A: Merged .fcs-file was created after 1° staining with control Ig (x-axis) and 2° staining for CD3-FITC, CD14-PE, and CD45-APC of PBLs (y-axis). Data were gated on single cells. Dot plots correlate 1° vs. 2° staining for the three fluorochromes. Cells with unchanged fluorescence line up along the diagonal (45°), whereas newly stained cells exhibit a different fluorescence in the 2nd analysis as compared to the 1st analysis and therefore shift off the diagonal in y-direction (arrows) (modified from (3), reprint permitted by the publisher).

Taking just one fluorochrome (i.e. PE) for iterative restaining in combination with two single stained dyes (i.e. FITC and APC) plus 7-AAD, we performed a seven “color” analysis. The assay defined neutrophils, monocytes, basophils, eosinophils (not shown), and lymphocytes, with further differentiation of lymphocyte subsets (see Fig. 2). HE staining plus relocalization allowed to document cell morphology (see Fig. 3).

Figure 2.

Merged .fcs-file of a triple stained specimen. PBLs were 1° stained for CD3-FITC, CD14-PE, and CD45-APC, followed by 2° staining for CD4-PE, and 3° staining for CD19-PE and CD8-PE. Data of all analyses were merged and gated on lymphocytes and monocytes. Upper line: For PE only, dot plots show 1° vs. 2° staining (left) and 2° vs. 3° staining (right). Note in the right plot the shift of CD4+ cells into the diagonal whereas newly stained CD8+ and CD19+ cells shift off the diagonal in y-direction. Lower line: Dot plots showing 1° FITC staining (CD3) vs. 1° (CD14), 2° (CD4), and 3° (CD19, CD8) PE staining. Data are compensated for spectral overlap. Note the gradual loss of CD14-PE fluorescence due to photobleaching. Labeling: a, monocytes; b, CD4+ helper T-cells; c, CD8+ cytotoxic T-cells; d, CD19+ B-cells; and e, CD19-CD3- cells, most likely NK-cells (modified from (3)).

Figure 3.

Relocalized single cells. Single cells from the electronic gates set in Figure 2 were relocalized after HE-staining and micrographs were taken with a digital camera (modified from (3)).


Our growing understanding of the immune system's complexity demands to develop polychromatic assays that have the capacity to characterize specific cell-subtypes with more accuracy (4–6). However, the more fluorochromes per assay are engaged, the more problems arise due to spectral overlap; electronic compensation is cumbersome (7). Flow-based assays have the additional drawback that the cells' morphology cannot be documented.

This can be circumvented by using SBC assays. This system not only allows to photographically document the cells' morphology but also to reduce the number of different dyes used at one given time point. Instead, the very same dyes can be “recycled” by iterative restaining; this feature is exclusive to SBC assays and might prove to be a pivotal tool for n-color immunophenotyping. It markedly reduces the problems otherwise experienced in polychromatic assays.

In our view, SBC systems are more suitable for polychromatic cytometry than FCM instruments. Even though up to 17 colors may be measured by four laser FCM (8), the relocation feature of SBC allows to combine even more colors. Combining different methodologies such as iterative restaining shown here, sequential photobleaching (9, 10), and novel dyes like quantum dots (11) can guide the way to single cell proteomics and toponomics (10). This can use virtually endless “colors” and was recently termed hyperchromatic cytometry (9). Further advantages as compared to standard flow assays are the minimized amount of reagents and specimen needed and the possibility to store slides as conventional cytological samples; this might be important in medicolegal issues.