Part of this work was presented at the 10th Leipziger Workshop “Systems Biology and Clinical Cytomics,” April 7–9, 2005, Leipzig, Germany.
Automated four-color analysis of leukocytes by scanning fluorescence microscopy using quantum dots†
Article first published online: 14 FEB 2006
Copyright © 2006 International Society for Analytical Cytology
Cytometry Part A
Volume 69A, Issue 3, pages 131–134, March 2006
How to Cite
Bocsi, J., Lenz, D., Mittag, A., Varga, V. S., Molnar, B., Tulassay, Z., Sack, U. and Tárnok, A. (2006), Automated four-color analysis of leukocytes by scanning fluorescence microscopy using quantum dots. Cytometry, 69A: 131–134. doi: 10.1002/cyto.a.20217
- Issue published online: 22 FEB 2006
- Article first published online: 14 FEB 2006
- Manuscript Accepted: 22 JUN 2005
- Manuscript Revised: 18 MAY 2005
- Manuscript Received: 27 APR 2005
- Carl Zeiss Vision GmbH, Hallbergmoos, Germany
- IZKF-Leipzig, Project Z10, Germany
- fluorescence mounting medium;
Scanning fluorescence microscope (SFM) is a new technique for automated motorized microscopes to measure multiple fluorochrome labeled cells (Bocsi et al., Cytometry A 2004, 61:1–8).
We developed a four-color staining protocol (DNA, CD3, CD4, and CD8) for the lymphocyte phenotyping by SFM.
Organic (Alexa488, FITC, PE-Alexa610, CyChrom, APC) and inorganic (quantum dot (QD) 605 or 655) fluorochromes were used and compared in different combinations. Measurements were performed in suspension by flow cytometer (FCM) and on slide by SFM.
Both QDs were detectable by the appropriate Axioplan-2 and FCM filters and the AxioCam BW-camera. CD4/CD8 ratios were highly correlated (P = 0.01) between the SFM and FCM.
Automated SFM is an applicable tool for CD4/CD8 ratio determination in peripheral blood samples with QDs. © 2006 International Society for Analytical Cytology
Flow cytometer (FCM) and Laser scanning cytometer (LSC,1) are standard instruments for quantitative multicolor phenotyping of leukocytes; spectral imaging confocal microscopy is optimal for morphological cytometric analysis (2). The Scanning fluorescence microscope (SFM) is a slide-based, software-controlled new technique for automated motorized microscopes equipped with a scientific digital camera. It can also measure multiple fluorescent labeling and is appropriate for multichannel cytometric analysis (3, 4). Current development of digital cameras and computer systems promise broad spreading of SFM technique.
In fluorescence microscopy, fluorescence emission is stimulated by high intensity UV or visible light. Light absorption induces an excited state of the dye (usually singlet-state), leading to fluorescence. However, the excited state of the dye may undergo chemical reactions such as photo-oxidation, leading to its destruction as evidenced by fading or bleaching of the fluorescence and subsequent decrease in signal intensity during measurement.
High-content analysis of biological specimens requires multicolor labeling for polychromatic (1) or hyperchromatic cytometry (5). The hyperchromatic analysis needs spectrally combinable labels with stable long-lasting fluorescence. Conventional organic fluorochromes bleach more or less rapidly in buffer during illumination (6). This is overcome by using antifading solutions as mounting media (7, 8).
Fluorescent semiconductor nanocrystals (quantum dots, QDs) offer new ways in multiparametric characterization of biological samples. In comparison with organic fluorophores, the QDs have unique optical properties such as, size (5–20 nm diameter, similar size as a typical protein), and composition tunable fluorescence emission from visible to infrared wavelength. In high absorption from UV across a wide range of the spectrum, narrow and symmetric emission spectra, intensive brightness, and photostability are the characteristics (9, 10). QDs are produced with appropriate mean emission for the well-proved fluorescence systems and filter combinations. This should enable combination of QDs in multicolor staining protocols with the other “classical” fluorochromes. QDs have recently been successfully covalently linked to reactive molecules such as peptides, antibodies, nucleic acids, or small-molecule ligands (9, 10). The development of high-sensitivity and high-specificity probes that lack the intrinsic limitations of organic dyes and fluorescent proteins is of considerable interest in many areas of biomedicine, ranging from molecular and cellular biology to molecular imaging and diagnostics.
The T-helper/T-cytotoxic (CD4+/CD8+) lymphocytes were selected as a model system. We developed a four-color fluorescence labeling method to demonstrate applicability of QDs in immunophenotyping by slide-based cytometry (SBC).
MATERIALS AND METHODS
EDTA anticoagulated blood samples from 28 healthy volunteers were stained by the whole blood method with CD4 Alexa488 (Caltag, Hamburg, Germany) and CD8 biotin (Caltag)/Streptavidin Qdot-605 or Qdot-655 (Quantum Dot Corp. Hayward, CA) or CD8 CyChrom (BD-Biosciences, San Jose, CA, USA). Aliquots were stained by CD4 PE-Alexa-610 (Caltag) and CD8 FITC (BD-Biosciences). For correct T-cell analysis, CD3 APC (Caltag) and nuclear counter-staining by Hoechst 33342 (Sigma-Aldrich GmbH, Deisenhofen, Germany) were additionally used. An aliquot was measured by FCM (FACScalibur, BD-Biosciences). The remaining suspension was transferred on glass slides embedded in antifade Fluorescent mounting medium (DakoCytomation, Glostrup, Denmark), ProLong™ antifade (Molecular Probes, Eugene, OR) or 50% Glycerin (Merck, Darmstadt, Germany), in PBS (Sigma-Aldrich). In spillover tests, CD8 biotin/streptavidin QD655, and CD14 PE-Cy5 (Caltag) labeling with comparable fluorescence intensity of labeled cells was used.
Digital slide images were acquired with an Axioplan 2 MOT (motorized) microscope (Carl Zeiss, Goettingen, Germany) equipped with 100-W Mercury lamp, a 12 bit Axiocam CCD camera (Carl Zeiss Vision GmbH, Hallbergsmoos, Germany), a motorized object desk and filter changer were controlled by SFM software (3DHistech Ltd, Budapest, Hungary). An entire smear area or at least 2,000 cells were scanned automatically at 20× magnification. Scanning sequence was Hoechst 33342, APC, QDs (CyChrom), Alexa488 (FITC) using the following fluorescence filter sets (Fs; all from Carl Zeiss, Axioplan-2 MOT): F1. G 365 nm, FT 395 nm, LP 420 nm (Hoechst 33342); F2. BP 575–625 nm, FT 645 nm, BP 660–710 nm (APC); F3. BP 546/12 nm, FT 560 nm, BP 575–640 nm (QD605/QD655 and PE-Alexa-610); F4. BP 450–490 nm, FT 510 nm, BP 515–565 nm (Alexa488 and FITC). Automated cell detection (based on Hoechst 33342 fluorescence) and CD3, CD4, and CD8 detection were performed; CD4/CD8 cell count ratio was calculated.
The Hoechst 33342 fluorescence signal (the strongest fluorescence) was used for focusing. After CCD camera dark current and Hg arc lamp excitation light inhomogeneity calibration, each of the microscopic fields of view was digitally recorded in different fluorescence channels. The multi-channel digital slides were evaluated using virtual microscopy and standard cytometry techniques as detailed elsewhere (3, 4). For method comparison, the Passing & Bablok linear regression procedure was used (11) (MedCalc software, www.medcalc.be).
Fluorescence signals could be well detected on FCM and SFM. Qdot605 (FL2), Qdot655 PE-Alexa-610, and CyChrom (all FL3) were well-measurable by 488-nm laser excitation (FCM) or with F3 in SFM. The QD staining needs to be made with diluted (20 nM) streptavidin conjugates to avoid unspecific staining of neutrophils (not shown). The unspecific staining of lymphocytes was not observed. With the increasing quantity of the streptavidin QDs (1, 5, and 10 μL), the fluorescence increased only with the specific staining and not in the negative population (not shown). The emission of QDs (as shown here for QD655) is known to be narrower than PE-Cy5 and the spectral overlap is lower. Consequently, QD fluorescence could be seen only in one channel (Fig. 1), giving better signal separation with lower demand for compensation. In this spillover test, QD labeled CD8 cells and PE-Cy5 labeled CD14 cells had the same fluorescence intensity.
This four-color staining is suitable for finding all leukocytes (FSC-SSC in FCM or DNA-Hoechst 33342 in SFM) and to identify within the T-cell (CD3-APC) population the Th (CD4-Alexa488) and Tc cells (CD8-biotin/streptavidin-QD605 or CD8-CyChrom) and determine their frequency (Fig. 2). The selected dyes gave strong signals so that exposition times shorter than 1 s/optical filter and image were sufficient. The B-W Axiocam can detect the light in the infrared region up to 1,000 nm so that QD655 and APC were also used. The gating cascade (CD3, CD4, and CD8) solved the problem that is originating from spectral overlap of APC and CyChrom in the FL4 channel and Hoechst 33342 and QD605 in the FL1 channel. This overlap was not so dominant in the case of very strong Hoechst 33342 and weaker QD605 fluorescence.
Passing and Bablok regression of CD4/CD8 ratios obtained by FCM and SFM (F(X) = 0.0577 + 0.9378x) shows that all measurement points are in the 95% confidence interval (Fig. 2). Cusum test for linearity did not show significant deviation from linearity (P > 0.10). This comparison indicates that there is no systemic bias between the two different methods.
Fluorescence of the dyes was affected by the mounting procedure. In contrast to the organic fluorescent dyes, the inorganic QD staining was very stable in glycerin-PBS, but fluorescence was lost shortly after mounting with antioxidant and free radical scavenger mounting media (DAKO and ProLong). This finding emphasizes the problem of combining organic and QD dyes.
To the best of our knowledge, this is the first paper showing that QD is feasible for use in SBC. Automatic four-color phenotyping and CD4/CD8 ratio determination is possible by using fully motorized Axioplan-2 microscope controlled by the SFM software. Three-color analysis was already shown for rare tumor cell detection among lymphocytes (4). Relation between results obtained by FCM and SFM was high. These results show that already available motorized fluorescent microscopes can be converted into scanning and measuring systems using SFM software. Nowadays its importance is further underlined by the increasing interest towards disease-mediated CD4/CD8 ratio alteration (12, 13).
QDs lost their fluorescence by mounting anti-fading medium. Antioxidants, free radical scavengers or other components of the mounting media probably initiate chemical reactions with the QDs. Possibly, directly the surface (what makes colloidal QDs water soluble and biocompatible) is damaged and QDs are released from the labeled cells or the whole QD is dissolved in the mounting medium leading to the loss of fluorescence. If this is the case, the QDs could be ruined, which means that in vivo usage of QDs (containing toxic elements) in strongly reactive microenvironments could poison the living organism. As QDs are produced with different saving layers (9, 10) more studies could answer these questions. Considering the necessity to use antifading agents for organic dyes, their combination with QDs could be difficult. Studies on bleaching effects of other antifading chemicals and compatibility with QDs are needed.
QDs with a wide range of bio-conjugation and with high quantum yields are now available commercially, so that it is no longer necessary for users to grow on their own (which takes quite a bit of practice) or to become lost in the endless discussions concerning the best way to render colloidal dots water-soluble and biocompatible. The range of biological experiments that these materials are employed in is growing rapidly, and this is one of the first commercial applications of modern nanotechnology.
Slide-based multi-fluorescence labeling system and automated SFM are applicable tools for the CD4/CD8 ratio determination in peripheral blood samples. Quantum dots are stable inorganic fluorescence labels that may be used as reliable high resolution dyes for cell labeling.
- 12Comparison of CD4 and CD8 lymphocyte counts in HIV-negative pulmonary TB patients with those in normal blood donors and the effect of antitubercular treatment: hospital-based flow cytometric study. Cytometry B Clin Cytom 2004; 61B: 20–26., , , .
- 13A multicenter evaluation of the PanLeucogating method and the use of generic monoclonal antibody reagents for CD4 enumeration in HIV-infected patients in Thailand. Cytometry B Clin Cytom 2005; 65B: 29–36., , , , , , , , , .