INDUCIBLE OF TUMOUR CELLS 1395

Nile Blue Inducible Fluorescence of Tumour Cells

A. L. BASTOS and DANTE MARQUES

Servio de Patologia Celular Calouste Gulbenkian, Instituto Portugues de Oncoldgia Francisco Centil, Lisboa 4, Portugal

(Z. Naturforsdi. 27 b, 1395—1398 [1972] ; received June 6, 1972)

Inducible fluorescence, tumor cells

An oxazine dye, Nile Blue sulfate, induces a fluorescent reaction in cytoplasmic granules (NBIFG) of living or fixed tumour cells in the same manner as reported before for Thiazine dyes. The NBIFG correspond to phase contrast positive granules when cells are viewed by phase contrast microscopy. Fluorescence disappears from NBIFG in a matter of 2 —4 days and the bodies turn deep blue. These granules have a succinate dehydrogenase (SDG) activity and are negative for acid phosphatase, peroxidasic activity and porphyrin. The cytological findings support the assumption that Nile Blue sulfate forms a salt linkage with unsaturated fatty acids of NBIFG which also show an oxi-reductive activity. The molecular nature of the (s) is (are) unknown.

We have previously shown that thiazine dyes can Material and Methods induce a fluorescent reaction in cytoplasmic granu- les of tumour cells when freshly stained by these Tumour Cells dyes and examined by fluorescence microscopy 2. The procedures (supravital and in vivo 1 The fluorescent reaction is called primary when staining) were carried out using transplantable tumour cells of mice, Sarcoma 37 tumour cells and Ehrlich the granules are visualized immediately after the tumour cells growing in ascitic form. Cells were ob- staining; secondary, when the granules become tained from tumour-bearing mice 5 to 10 days after fluorescent only after exposure of the cells to light. cell transplantation. The experiments were carried out The primary reaction is observed only with a parti- while the tumour was maintained during 378 conse- cutive passages. cular batch of Toluidine Blue and the secondary reaction with all Thiazine dyes so far tested, in- Nile Blue Staining cluding samples obtained from several manufactu- A modification of L i 11 i e's3 technique for de- rers. monstration of fatty acids was used as follows: — The dye inducible fluorescent (DIF) granules are 1. Fix tumour cells in 4% saline at 4 C refractile, which makes their microscopic recogni- for 1 hours. tion easy (i. e. Interference microscopy); cytochemi- 2. Remove the excess fixative by cell centrifugation for cal reactions for (Oil Red 0, Sudan Black 6-10 min at 1000 rpm. and reduction of osmium tetroxide) are positive. 3. Resuspend the cells in an aqueous solution of Nilft Blue (0.005% w/v) and stain for 30 min. Sudanophilia and osmiophilia usually indicate 4. Mount the cell sediment in . the presence of unsaturated fatty acids. Since the 5. Examine by fluorescence microscopy. presence of these highly oxidizable lipids could help to explain the cytochemical nature of the DIF Fixatives reaction, L i 11 i es Nile Blue procedure was used to To study the effect of the following fluids demonstrate unsaturated fatty acids. were used: Neutral formalin at 10% and 15%, Car- noy's fluid, Alcohol-Ether, 2.5% glutaraldehyde in The observations now reported show that Nile cacodylate HCl buffer. Blue sulfate (samples from several manufacturers) also gave the DIF reaction in the same granules Control of pH, dye concentration, temperature and stained by Thiazine dyes, both in living and fixed staining time cells. a) pH: The cells were stained in Nile Blue solution at pH 0.9 -12. The pH was buffer adjusted with NaOH Requests for reprints should be sent to Dr. A. L. BASTOS, (saturated solution). Above pH 9 the colour of the Servio de Patologia Celular Calouste Gulbenkian, Instituto solution changes from blue to pink and precipitates. Portugues de Oncologia Francisco Gentil, Lisboa 4, Portu- b) Dye concentration and tempera- gal. ture: The dye concentration in the staining solution 1396 A. L. BASTOS AND D. MARQUES was varied in tenfold dilutions from 0.5% to 0.00005% perchloric acid-mercapto ethylamine UV method for (w/v). The ratio of the cell volume to the volume of porphyrins 7. staining solution was 1 : 10. The temperature of the The above cytochemical preparations were then staining solution ranged from 0 °C to 100 °C. stained by Nile Blue sulfate to demonstrate the pre- c) Staining time: The staining times used sence of DIF granules and slides examined both by were als follows: - 0', T, 5', 10', 15', 30', 60', 2 hrs., light and fluorescence microscopy. 12 hrs., 24 hrs., 48 hrs. Nile Blue Compounds tested Results Nile Blue Sulfate [C.I. (913)51180] British Drug Houses, Poole, England, Batch N. 2868090; Matheson The fluorescence microscopic appearance of vi- Coleman & Bell, Norwood, Ohio, USA, Batch N. tally and supravitally stained and fixed tumour cells 481208; Chroma-Gesellschaft Schmid & Co., Stuttgart, are shown in Figs. 1,2* and 3. The Nile Blue indu- Germany, Batch unknown. All samples gave the fluo- rescent reaction. y

Microscopic equipment _,40 UJ A Wild M20 microscope with a dual illumination O CC system for fluorescence, bright field and phase contrast UJ was used. It was fitted with Xenon lamp XBO 150w CL for fluorescence and a 12V/l00w quartziodine lamp for bright field and phase contrast. O

Fig. 1. Fluorescence microphotograph of ascitic Sarcoma 37. Figs. 6 a and 6 b. Fluorescence microphotograph of formalin Tumour cells vitally stained by Nile Blue. Bright green (ultra- fixed Nile Blue stained cells. Fluorescent emission disappears violet excitation) granules occur in the cytoplasm of almost in two days becoming the granules deep blue in bright field, every cell. Nucleus are non-fluorescent. 800 x. Fig. 6 b. 1200 x. Figs. 7 a and 7 b. Fluorescence microphotograph of an acid Fig. 2. Sarcoma 37 cells supravitally stained by Nile Blue. phosphatase preparation of Nile Blue stained tumour cells. Green colour emission of cytoplasmic granules as in vital The NBIFG do not show acid phosphatase activity although staining. 800 x. there are at least eight darker stained cells showing acid phos- phatase in bright field, Fig. 7 b. 1200 x. Fig. 4 a. Formalin fixed Sarcoma 37 tumour cell stained by Figs. 8 a and 8 b. Fluorescence microphotograph of peroxidac- Nile Blue and observed under bright field. No cytoplasmic tic activity in Sarcoma 37 tumour cells stained by Nile Blue. granules are seen. 2000 x. The NBIFG show no activity when the erythrocytes and neu- trophil have a positive reaction in the same preparation, Fig. 4 b. Phase contrast microphotograph of the cell shown Fig. 8 b. 1200 x. in Fig. 4 a. Note the presence of more than ten phase contrast Figs. 9 a and 9 b. Fluorescence microphotograph of five Sar- positive granules and four macro nucleolus. coma 37 tumour cells stained for succinic dehydrogenase and DIF reaction. Note the distribution of succinic dehydrogenase Fig. 5 a. Phase contrast microphotograph of formalin fixed activity in the same cells photographed in bright field, Fig. Nile Blue stained Sarcoma 37 tumour cells showing the pre- 9 b. 1200 x. sence of cytoplasmic granules. 1200 x. Figs. 10 a and 10 b. Diffuse cytoplasmic porphyrin emission Fig. 5 b. Fluorescence microphotograph of the same field as (red) of tumour cells which is weak when compared to the Fig. 5 a. Note the size, form and distribution of fluorescent reaction in erythrocytes (arrows). When these porphyrin pre- granules. parations of tumour cells are stained by Nile Blue (10 b) DIF granules are present showing yellow emission with blue light * Fig. 2 s. Tafel S. 1396 a. excitation. 1200 x. A. L. BASTOS and D. MARQUES. Nile Blue Inducible Fluorescence of Tumour Cells (S. 1395)

Zeitschrift für Naturforschung 27 b, Seitel396 a Zeitschrift für Naturforschung 27 b. Seite 1396 b INDUCIBLE FLUORESCENCE OF TUMOUR CELLS 1397

V fluorescence microscopy. Graph 1 represents the cor- (* relation of phase-positive granules per cell to * * e fluorescent granules per cell. * A constant observation of Nile Blue stained O „ (pH 5) fixed cells is the disappearance of the # fluorescence in a matter of 2 — 4 days with the gra- * •' ' * * nules turning deep blue. § % Visual control of the preparations obtained i- > 0 under different staining conditions (Dye concentra- * * tion, temperature and staining time) showed that t fluorescence of NBIF granules (Excitation at 400) » was not affected by these variables. Figs. 7 to 10 illustrate simultaneously acid phos- • *\ phatase, peroxidactic activity, succinodehydro- > » genase, porphyrin and NBIF bodies in the same Fig. 3. Formalin fixed Sarcoma 37 tumour cells stained by preparations. These NBIF granules exhibit an asso- Nile Blue. Bright yellow emission of NBIFG with blue light ciated SDG activity (Figs. 9 a —9 b) with demon- excitation. 800 x. strable fluorescence in all cytochemical double cible fluorescent (NBIF) granules are identical to stained preparations. A clear differentiation is seen those shown in Fig. 1 A when tumour cells were between red porphyrin fluorescence and yellow emis- stained fresh with a particular batch of Toluidine sion from the NBIF granules (Figs. 10 a — 10 b). blue. No acid phosphatase or peroxidactic activities are detected in NBIF bodies. The fluorescent reaction is visualized immediately in living and formalin-fixed tumour cells provided that the latter are examined with blue light excita- Discussion tion (c. 400 nm). When these fixed cells are examined with UV light (c. 360 nm), no fluo- The microscopic data presented suggest that the rescence is seen. A fluorescent reaction develops in granules which fluoresce are identical to those these preparations in about 1—2 min, when the previously discussed in relation to Thiazine Dyes 2. cells are exposed to blue light, UV light or — in a Since lipids have been cytodiemically demon- shorter time — to a strong white light source. strated 8 in these Nile Blue fluorescent granules, it The conditions for light exposure are the same is interesting to note the good correlation between as those used for fluorescence microscopic observa- counting of phase contrast granules per cell and tions. After light exposure the granules are readily fluorescent granules per cell. It is to be noted that seen when preparations are observed under UV ex- Nile Blue was initially used to demonstrate the citation. The colour emission (green — UV excita- presence of fatty acids 3. DUNNIGAN suggests a salt tion) of these light exposed formalin-fixed Nile Blue linkage between the oxazine base and fatty acids stained preparations is identical to that obtained which is not dependent on critical staining condi- 9 from vital and supravital staining. tions . If this is so an acid-base reaction could ex- plain that NBIF granules can be stained at 0 °C Under bright field microscopic conditions, the in a few seconds and at dye concentrations of granules are practically invisible and colourless 0.00005%. using fixed cell preparations, but are really detected 1 using phase contrast equipment. Figs. 4 a** and 4 b Although in our earlier papers we stated that show the same cell under bright light and phase fluorescent cytoplasmic granules with affinity for contrast, using an automatic photographic exposure. basic dyes might be lysosomes, cytochemical studies reported elsewhere in a preliminary form10 and Figs. 5 a and 5 b serve to demonstrate that the above suggest that this is not the case. When double phase positive granules are fluorescent, as the same staining techniques are used (Nile Blue for DIF cells were photographed using phase contrast and granules and Gomori acid phosphatase for lyso- ** Figs. 4—10 see illustrations on pages 1396 a —b. somes) the DIF granules show no acid phosphatase. 1398 INDUCIBLE FLUORESCENCE OF TUMOUR CELLS 1398

In fact the DIF granules show succinate dehydro- lipids. The fact that the Chromodye observed is genase activity, suggesting that they may be related either blue or pink seems to indicate the presence to mitochondria. Perhaps the appearance of the of both these components 9. Chromodye reflects the existence of an oxidation- It is interesting to note that the light induced reduction process. fluorescence, for observation in UV, can be studied Nile Blue has been extensively used in Cyto- qualitatively as well as quantitatively using micro- chemistry as an indicator for acidic and neutral fluorometric equipment as mentioned elsewhere2'11.

1 A. L. BASTOS, D. J. MARQUES, J. F. M. NUNES, A. M. 6 T. BARKA and P. J. ANDERSON, Histochemistry Theory, TERRINHA, J. D. VIGÄRIO, A. D. CORREIA, and J. F. SILVA, Practice and Bibliography, p. 313, Harper and Row Z. Naturforsch. 23 b, 969 [1968]. Publishers, Inc., New York 1963. 2 D. J. MARQUES, A. L. BASTOS, A. M. BAPTISTA, J. D. 7 S. GRANICK and R. D. LEVERE, J. Cell Biol. 26, 167 VIGÄRIO, J. M. NUNES, A. M. TERRINHA, and J. A. F. [1965]. SILVA, Summary reports of the 3rd International Congress 8 A. L. BASTOS, D. MARQUES, and M. T. BATOREU, 119th of Histochemistry and Cytochemistry, New York, p. 171, Meeting of the Pathological Society of Great Britain and Springer-Verlag, New York 1968. Ireland. Oxford 1969. Abstract No 22. 3 R. D. LILLIE, Histopathologic Technic and Practical Histo- 9 M. G. DUNNIGAN, Stain Techn. 43, 249 [1968]. chemistry (Ed. R. D. LILLIE), 3rd Edn., p. 281, McGraw- 10 A. L. BASTOS. J. F. M. NUNES, J. D. CARDIGOS, D. MAR- Hill Book Company, New York 1965. QUES, J. A. F. SILVA, and M. C. 0. TEIXEIRA, 123rd 4 T. BARKA and P. J. ANDERSON, Histochemistry Theory, Meeting of the Pathological Society of Great Britain and Practice and Bibliography, p. 239, Harper and Row Ireland. Edinburgh 1971. Abstract N° 26. Publishers, Inc., New York 1963. 11 D. MARQUES and A. L. BASTOS, Exptl. Cell Res. 73, 525 5 K-i. HIRAI, J. Histoch. Cytodiem. 17 (9), 585 [1969]. [1973],