January 1, 2013 (Vol. 33, No. 1)
Andreas J. van Agthoven, Ph.D.
Fabrice Malergue, Ph.D.
No-Wash Solution Designed to Improve Quantitative Analysis
Is it necessary to put question marks at the end of intracellular flow assay? Just now that the most exciting reports on cell signaling and phospho-protein detection are coming out? Wasn’t DNA analysis one of the first diagnostic applications in flow?
The answer to the last question is “yes”, but the small molecular stains are replaced by fluorescent antibodies, and these large molecular structures are causing a problem. At the surface of a cell, we are relatively certain to reach a target, and even there the diagnostic quality procedures of an antibody binding assay are difficult. Going intracellular adds a number of problems to this—like denaturing conditions, hindrance of macromolecule entry, and cellular decomposition.
Many companies are offering solutions for intracellular flow with antibodies and many examples are shown, but little quantitative analysis is performed. An example of such an analysis will be discussed in this article.
While immunofluorescence flow cytometry is a relatively new technique, and one that is used for research purposes in a majority of cases, it is also quickly acquiring diagnostic applications in the study of leukemia and lymphoma, solid tumors, and in immunotherapy, thanks to the massive expansion of monoclonal antibody technology. In most of these cases, the physiological target is located on the surface of the cell and reflects a state of differentiation.
Intracellular antigens of interest for cell signaling such as phosphoproteins are, however, functional markers, not markers of differentiation. Precise detection of this type of molecule in cells of interest is necessary for choosing the most appropriate phosphokinase inhibitor drug for the treatment of an individual’s cancer or, more generally, for phosphokinase inhibitor drug discovery.
There is no reference value for phosphoproteins; their expression can vary in a minute under the influence of activators or inhibitors. Moreover, phosphorylated epitopes are often not available for antibody recognition. Strong protein denaturing procedures are necessary to expose the phosphorylation sites. For the purpose of quantitative analysis it is better to study a less cumbersome type of assay before entering the field of the phosphoproteins.
Cell Preparation and Permeabilization
Measurement of intracellular antigens using antibodies has, except for rare cases, not yet entered into routine analysis. Especially cumbersome is the combination of surface labeling and intracellular labeling in a multicolor assay.
Technically, the challenge is to remove the impermeable lipid bilayer of the cell without disturbing the intermittent and underlying nonlipid structures. Techniques that create temporary holes in the membrane, used in the transfection of genetic material and in which cells can be maintained alive, are not suitable for intracellular antibody labeling experiments.
Lipid-dissolving reagents are small organic molecules or detergents. The critical step in a procedure using these reagents is the fixation of the remaining cellular structure, which can be achieved by chemical cross-linking or protein aggregation of the type that would make a protein precipitate if in solution.
The art of making the cell permeable for an antibody reagent lies in using the chemicals involved as economically as possible. Methanol, which combines the properties of an aggregating reagent and of a lipid extracting reagent, is a good example of a reagent useful in cell permeabilization. The lipid-extracting detergent lauroyl sarcosinate, which can be a protein-aggregating reagent under conditions of moderate acidity and low salt, is another example.
For the moment, the technology of permeating the cells and exposing epitopes and especially phospho-epitopes is complicated and requires the use of poorly controlled procedures that include many washing steps. It is expected that these procedures will be optimized, simplified, and ultimately automated. It is essential for a quantitative measurement to use a technique that provides free access to 100% of the antigen. No-wash is preferred, since washing steps and the partial loss of antigens are a main source of variability.
Quantitative Glycosylated Hemoglobin (HbA1c) Intracellular Flow Assay
The glycosylated hemoglobin assay is well known and routinely used in diagnostics, and is therefore an ideal reference model to validate certain procedures used in flow intracellular assays. Many reagents and protocols are available for these assays, most of them developed for some specific application.
In one specific model using lauroyl sarcosinate as described in this article, it is shown that the glycosylated hemoglobin molecule is freely accessible for an antibody. In a competitive experiment comparing a mixture of intracellular HbA1c and HbA1c in solution, the reactivity of the antibody with freely soluble HbA1c could be compared with the reactivity in the permeated red cells. The antibody was at nonsaturating conditions and similar amounts of HbA1c from the same source were used. The cells bound 50% of the antibody, providing proof that no hemoglobin was lost from the cells and of the free access of antibody into the cell.
In this optimal setting of the permeating conditions allowing free access to the antigen, we can expect a coherent dose-response relationship between antigen concentration and antibody.
The concentration of glycosylated hemoglobin in red blood cells increases with the age of a particular cell and in the case of diabetes, with the glucose concentration in the blood serum. Red blood cells express large variations of glycosylated hemoglobin content (Figure 1). The youngest red cells (reticulocytes) contain the lowest percentage of glycosylated hemoglobin and can be seen in the flow cytometric analysis of Figure 2, in which the nucleic acid-containing cells have a low staining for glycosylation. (FC 500 flow cytometer, Beckman Coulter Life Sciences.)
In this assay, the hemoglobin content of the red blood cells can be derived from the side-scatter parameter of the cells in a flow cytometer. Knowing the mean value of the hemoglobin content of the cells and knowing the mean of the amount of glycosylated hemoglobin in the blood cells, the percentage of glycosylation of the hemoglobin can be measured.
Existing HbA1c assays can be used as a reliable predicate. Over 100 blood samples from diabetic and normal individuals were analyzed for hemoglobin glycosylation (HbA1c) percentage with three different methods: affinity column, immune-analysis, and HPLC. The correlation with the cell-by-cell flow cytometry analysis shows a slightly curved logarithmic relationship (Figure 3).
This is not surprising. Antibody binding immunoassays very often show such a relationship. It shows, however, that the initial observation of no hindrance of antibody entering the cell is not true at higher concentrations of antigen. The relationship is curved in spite of saturating amounts of antibody and of internal control cells.
The Future of Quantitative Intracellular Flow Cytometry Procedures
What does this teach us for intracellular flow analysis? The glycosylated hemoglobin example may be a rare exception. Little quantitative data exist for other flow assays. For the purpose of linearity, the antibody has to be shown to react freely with the antigen, for example, by a competition assay. For the purpose of sensitivity, the degradation of epitopes due to the handling of cells has to be minimal. For the purpose of reproducibility, internal control cells and a no-wash procedure limit variations in the assay.
The no-wash procedure has been implemented thanks to the high concentration of red blood cells; it will be challenging to transpose this to the analysis of white blood cells. Moreover, if very stringent procedures are to be used, as in the case of phospho-epitope labeling, a no-wash procedure could be even more difficult. Perhaps, after optimization of the protein aggregation technology, a no-wash solution and subsequent automation could be envisaged.
Cell permeabilization procedures as they exist today need improvement. The results of the glycosylated hemoglobin assay might inspire hope, although for each antigen or group of antigens an appropriate adaptation has to be expected. A predicate method exists for the glycosylated hemoglobin assay. For the phospho-protein assay, unless Western blot analysis would be accepted as a predicate method, no other assay would be available.