Dependence of leukemic cell autofluorescence patterns on the degree of differentiation Monica Monici,*† a Giovanni Agati, b Franco Fusi,† c Riccardo Pratesi,† d Milena Paglierani, e Valeria Santini f and Pietro Antonio Bernabei f a CEO – Centre of Excellence in Optronics, Largo Enrico Fermi, 6, 50125 Florence, Italy. E-mail: monici@ino.it b IFAC-CNR, Institute of Applied Physics “Nello Carrara”, Via Madonna del Piano, 50019 Florence, Italy c Department of Clinical Physiopathology, University of Florence, Viale Morgagni, 85, 50134 Florence, Italy d Department of Physics, University of Florence, Via Sansone, 1, 50019 Florence, Italy e Department of Human Pathology and Oncology, University of Florence, Viale Morgagni, 85, 50134 Florence, Italy f Haematology Unit, University of Florence, A. O. Carreggi, Viale Morgagni, 85, 50134 Florence, Italy Received 25th October 2002, Accepted 4th June 2003 First published as an Advance Article on the web 14th July 2003 The characterisation of leukemic cell autofluorescence during differentiation, induced by 12-O-tetradecanoylphorbol 13-acetate and all-trans retinoic acid, was performed by autofluorescence microspectroscopy and multispectral imaging autofluorescence microscopy. We have found that a dependence exists between the cell autofluorescence pattern and the degree of cell differentiation. When cells differentiate, their autofluorescence emission changes, following the morphological and functional rearrangement of cell structures. A decrease in emission intensity and a different distribution of endogenous fluorophores are observed. Thus, autofluorescence monitoring on living cells is a potentially useful tool for in vitro study of the differentiation processes. Furthermore, different maturation steps can be distinguished on the basis of the cell fluorescence pattern, leading the way to future application of the technique in diagnostics. Introduction It is known that cells contain molecules, engaged both in structural and functional processes, which become fluorescent under suitable excitation by UV-visible irradiation. The emission of these endogenous fluorophores is called auto- fluorescence or natural fluorescence, to distinguish it from the fluorescence obtained by adding exogenous markers. Studies have been performed in order to investigate the excitation bands and the corresponding autofluorescence emission patterns, and to identify natural fluorophores. The most important of these are proteins containing aromatic amino acids, 1 the reduced form of pyridine nucleotides, 2 flavins, 3 and lipopigments. 4 The possibility of exploiting the autofluorescence properties of cells for diagnostic applications has been investigated, too. Stübel 5 recognised the potential of autofluorescence analysis about ninety years ago. Indeed, cell autofluorescence analysis can be performed directly, avoiding preparative procedures that take a long time and cause cell death. Moreover, since the endogenous fluorophores play different and important roles in structural and functional processes in biological systems, information on both the morphological and functional cell state can be drawn from autofluorescence studies. Nevertheless, the opportunity to study and utilise the autofluorescence properties of cells has not been considered for many years, due to the difficulties in revealing and processing the signals. The autofluorescence of biological structures is a complex signal: it consists of emissions arising from intra- and † Istituto Nazionale per la Fisica della Materia (INFM), Sezione di Firenze. extracellular (in the case of tissues) fluorophores. Moreover, they usually show low quantum yields and broad, partly overlapping, excitation–emission spectra. Thus, assignment of the different spectral components is very difficult. 6 However, sophisticated spectroscopic techniques are now available 7 and, in the last few years, imaging detector tech- nology has made important advances, too. High sensitivity, low noise charge-coupled device (CCD) cameras allow the detection of low quantum yield autofluorescence signals at a level com- parable to the images obtained with high quantum yield exogenous markers. 8–10 Therefore, besides the spectroscopic techniques, advanced techniques based on multicolour fluor- escence imaging are now available for the analysis of auto- fluorescence signals in biological structures. 11,12 Multicolour imaging techniques combine spectral and spatial resolution, giving information on the fluorescence spectra and emission intensity of different fluorophores, as well as on their location at a subcellular level. In this way, identification of the different components is made easier. Thus, the possibility arises of utilising autofluorescence-based techniques, both in research and diagnostics, on single living cells as well as on tissue samples, with the technique and instrumental set-up most suitable for each different application being chosen on an individual basis. We approached the analysis of cell and tissue autofluor- escence by implementing and applying a relatively simple multicolour fluorescence imaging technique. 13 We found, in agreement with other authors, that the autofluorescence pattern reflects the intracellular structure organisation, and that a connection between autofluorescence and the metabolic state of a cell exists. 14,15 Recently, we demonstrated that peripheral blood white cells show different autofluorescence emission 981 Photochem. Photobiol. Sci., 2003, 2, 981–987 DOI: 10.1039/b306276g This journal is © The Royal Society of Chemistry and Owner Societies 2003