Background: Various protocols for estimation of telomere length in individual cells by flow cytometry using fluorescence in situ hybridization of fluorescently labeled peptide nucleic acid (PNA) probes (Flow-FISH) have been described. Combined analysis of telomere length and cell phenotype, however, remains difficult because few fluorochromes with suitable emission spectra tolerate the harsh conditions needed for DNA denaturation during hybridization of the telomere-specific PNA probe. We overcame these problems and developed a method for measuring telomere length in cell subsets characterized by the expression of two surface antigens.
Methods: Alexa Fluor 488 and Alexa Fluor 546 were used for cell surface staining. Antigen-antibody complexes were covalently cross-linked onto the cell membrane before Flow-FISH. Cells were hybridized with a PNA probe conjugated to cyanine 5 (Cy5). Hoechst 33342 (HO342) was added for determination of cellular DNA content. For assay standardization, we added an aliquot of a single batch of 1,301 cells to each sample as an internal control before hybridization with the PNA probe. Samples were prepared in duplicate and analyzed on a standard three-laser BD LSR flow cytometer. For assay validation, the same samples were analyzed in parallel to correlate the percentage of telomere length of the sample versus 1,301 control cells to the mean size of terminal restriction fragments (TRFs) of DNA as determined by Southern gel analysis.
Results: The method permitted clear identification of lymphocyte subsets in samples hybridized for Flow-FISH, with subset frequencies comparable to those of untreated samples. At a concentration of 10 nM, the Cy5-labeled telomere-specific PNA probe produced a bright fluorescence signal well separated from background. Addition of HO342 in low concentration did not interfere with Cy5 telomere fluorescence, produced adequate DNA histograms, and permitted clear identification of cell phenotype. The probe concentration of 10 nM also proved optimal for inclusion of 1,301 control cells for assay standardization. Telomere length estimations by the current method correlated highly with TRF calculations by Southern gel hybridization (r(2)= 0.9, P = 0.0003). Application of our protocol to the analysis of human CD8CD28 lymphocyte subsets showed that CD8(+bright)CD28(-) lymphocytes generally exhibit shorter telomeres than CD8(+bright)CD28(+) cells. These data concurred with previous results of telomere shortening in CD8(+)CD28(-) T cells that were obtained by using different techniques.
Conclusions: The multiparameter Flow-FISH protocol permitted rapid determination of differences in telomere length in subpopulations characterized by two surface markers without prior cell separation.
Copyright 2002 Wiley-Liss, Inc.