Journal oflmmunological Methods, 43 (1981) 261--267 261 Elsevier/North-Holland Biomedical Press

PROPIDIUM IODIDE AS A NUCLEAR MARKER IN IMMUNOFLUORESCENCE. I. USE WITH TISSUE AND CYTOSKELETON STUDIES

C.D. OCKLEFORD i, BAE-LI HSI 2, j. WAKELY l, R.A. BADLEY 3, A. WHYTE 4 and W. PAGE FAULK 2,*

I Department of Anatomy, University of Leicester Medical School, Leicester; 2 Blond MeIndoe Centre for Transplantation Biology, East Grinstead; 3 Biosciences Division, Unilever Research, Colworth Laboratory, Sharnbrook, Bedford; and 4 Department of Pathology, University of Cambridge, Cambridge, U.K.

(Received 3 October 1980, accepted 10 January 1981)

Some examples are given of immunofluorescence with tissue sections and microtubular cytoskeletons of cultured cells where the fluorescent dye propidium iodide (PI) has been used as marker of nuclei. The emission wave length of PI is longer than that of fluores- cein, making it possible to use several different and commonly available filter combina- tions. The use of nuclei as positional indicators is often a more suitable method than phase microscopy combined with immunofluorescence because of low background illu- mination against which morphology is viewed, circumventing the need for often expen- sive phase optics.

INTRODUCTION

Propidium iodide (PI) (Fig. 1), like its analogue ethidium bromide, has been known for some time as a marker of nucleic acids (Krishan, 1975; Kolesnikov et al., 1979; Krishan and Ganapathi, 1979; Kuypers et al., 1979). Also like ethidium bromide, it does not penetrate living cell membranes, bu t is endocytosed when used as a vital dye (Kuypers et al., 1979; Okudaira et al., 1979). PI has an excitat ion spectrum between 400 and 540 nm and gives an emission spectrum between 560 and 620 nm, but when intercalated between nucleic acid bases its fluorescence increases up to 20-fold (Hudson et al., 1969; Krishan et al., 1978). Used alone it complexes with ribonucleo- protein and thus binds to r ibosomes and nucleoli as well as to nuclear and mitochondrial DNA, and in combinat ion with ribonuclease digestion it can be used as a DNA specific cytochemical probe (Crissman and Steinkamp, 1973; Irvin and Bagwell, 1979). Given the increased range of available spe- cific ant ibody probes for immunofluorescence, there is now a useful niche for a fluorescent nuclear counterstain. We have used propidium iodide as a

* Reprint requests: Professor W. Page Faulk, Blond McIndoe Centre for Transplantation Biology, Queen Victoria Hospital, East Grinstead, Sussex RH19 3DZ, U.K.

0 022-1759/81/0000--0000/$02.50 © Elsevier/North-Holland Biomedical Press

262

,.~-~" N H 2

i

NH 2 I \

Fig. 1. Structure of propidium iodide.

N ÷ ~ CH3

/ ' ~(c2Hs 12 I -

nuclear marker in several immunohistological studies (Galbraith and Faulk, 1979; Matter and Faulk, 1980} but the details of its use in studies of tissue sections and the distribution of microtubules have not been reported. We present here some examples of its usefulness in these circumstances as well as several different optical systems in which it can be employed.

MATERIALS AND METHODS

Human placental sections Normal human placentae were collected immediately after delivery and

placed in plastic bags on ice until snap frozen. Tissue blocks (1 cm3} were ob- tained from the middle of the central cotyledon and snap-frozen in liquid ni t rogen~ooled isopentane. Sections (4.5 pm) were prepared using a cryostat (Bright Instrument Co., Ltd. , Huntingdon, U.K.) and pre-washed in phos- phate-buffered saline (PBS) for 20 min, before being reacted with rabbit anti-human trophoblast serum and a fluorescein isothiocyanate (FITC) con- jugate of sheep anti~abbit immunoglobulin (Wellcome Reagents, Becken- ham, Kent, U.K.) as described in detail by Faulk et ai. (1979).

Placental sections were examined using a Zeiss Universal Microscope fitted with an epi-fluorescence condenser (III RS), an HBO 50 mercury-arc lamp and a 25× Planapochromat objective. As reported by Yeh et al. (1981) we have used 3 filter combinations as follows: (1) blue interference filter set 455--490 as the exciter filter with an FT 510 chromatic splitter as the beam splitter and a green interference filter set 520--560 to serve as the barrier filter; (2) blue interference filter set 455--490 as the exciter filter with an FT 510 chromatic splitter as the beam splitter and an orange LP 520 interfer- ence filter to serve as the barrier; or (3) a BR 546/7 green interference filter set as exciter with an FT 580 chromatic splitter as the beam splitter and a red LP 590 to serve as the barrier filter. It should be noted that the optical effects of filter combination No. 1 can be mimicked by simply placing a KP 560 barrier filter over the filter combination No. 2, and that the optical effects of filter combination No. 3 can be mimicked by substituting the orange LP 520 for the LP 590 barrier in combination No. 2, but the red fluorescence of PI with this combination is less intense than that obtained

263

with the suggested filter combination No. 3. A Zeiss MC63 camera equipped with Kodak Ektachrome 200 daylight professional film was used to obtain the photomicrographs.

Propidium iodide was purchased from Sigma (Poole, Dorset, U.K.) and a stock solution was prepared by dissolving 5 mg in 100 ml of 0.1% trisodium citrate (Na3C~HsO7 • 2H20). According to Fried et al. (1976) this solution lasts for 21 days if stored in lightproof conditions at 4°C, but we found that it was good for at least 3 months. The stock solution was diluted 1 : 3 in PBS and applied directly to the tissue sections as two drops following the first wash after the conjugate and left to incubate for 5 min at 4°C following which the slides were washed twice (20 min/wash) in excess PBS and mounted in v/v PBS-buffered glycerol. PI can also be mixed with conjugate by diluting the stock solution 1 : 40 with the working dilution of rabbit con- jugate. This solution is of use in indirect immunofluorescence procedures.

Human placental cells These were grown from 1st trimester placentae according to the method

of Loke and Borland (1970) and were prepared for immunofluorescence using FITC-conjugated goat anti-rabbit immunoglobulin bridged to affinity purified rabbit anti-tubulin according to Badley et al. (1978). The rabbit anti-tubulin serum was used at a concentration of 50/~g/ml and the FITC conjugate was used at a dilution of 1 : 30. The stock solution of PI was diluted 1 : 3 and applied directly to the cells as was done with the tissue sec- tions. The cells were examined with a Zeiss standard microscope fitted for epifluorescence illumination with filter combinations Nos. 1 and 3. Expo- sures were controlled by using a Zeiss MC63 photomicroscope at tachment and images were recorded on Agfachrome 50 L professional film or Kodak Ektachrome 400. Two exposures were made on each frame of film with filter combination sets changed between exposures.

RESULTS

Human placental sections When filter combination No. 1 was used, the red colour of propidium

iodide was blocked by the BP 520--560 filter, allowing study of the FITC reactions wi thout distraction by the PI reaction (Fig. 2A). In contrast, filter combination No. 2 revealed a brilliant green fluorescence of syncytiotropho- blastic membranes that encircled each chorionic villus, the nuclei of which were easily identified as bright red clusters because they had been reacted with PI thus highhghting the morphology in each microscopical field (Fig. 2B). For purposes of morphology, PI reactivity can be studied unobstructed by fluorescence by using filter combination No. 3 (Fig. 2C).

Cultured human placental cells The microtubular network of these cells as revealed by anti-tubulin anti-

264

body is complex and the circum-nuclear fluorescence is of ten so intense that it is not possible to discern the number or shape of nuclei. This is important because nuclear morphology can be used to identify one type of cell in this mixed populat ion, the details and interphase microtubule patterns of which will be described elsewhere (Ockleford and Whyte, in preparation). With the beam splitter set to allow the transmission of longer wave lengths (i.e., filter combinat ion 3). PI fluorescence could be observed wi thout FITC fluores- cence and the number and morphology of nuclei could be seen together with a faint perinuclear reaction (Fig. 3A}, but when double exposures were made by using filter combination Nos. 1 and 3, the results showed (Fig. 3B) that PI compared favourably with phase optics (Fig. 4A and B), but a com- bined phase/fluorescence system could still of course be used (Fig. 5A and B). Photomicrographs could be made through any of the filter combina- tions bu t if the intensity of one of the colours is faint and it is necessary to reproduce both on the same film, double exposures of the film with filter combinat ion Nos. 1 and 3 can be used to obtain the correct exposure t ime for both , or the same result can be obtained with a single exposure employ- ing filter combinat ion No. 2 as shown in Fig. 2B.

DISCUSSION

Many biochemical methods, including most immunofluorescence tech- niques, suffer from the common disadvantage of not being able to simulta- neously demonstrate specific histochemical reactions with familiar and recognizable morphological landmarks. It is thus of ten necessary to change optical systems, such as alternating be tween epi-illumination immunofluo- rescence and phase contrast, in order to be able to correlate the appearances

Fig. 2. Syncytiotrophoblastic and nuclear patterns in normal human placenta. A: filter combination No. 1 shows the pattern of anti-trophoblast reactivity as a brilliant green fluorescence of syncytiotrophoblast that encircles each chorionic villus. B: filter com- bination No. 2 reveals the same green fluorescence with red nuclei marking the trophoblas- tic, stromal and endothelial cells. C: filter combination No. 3 only identifies the red nu- clei. Magnification, x 390.

Fig. 3. Nuclear microtubular patterns in cultured human placental cells. A: cells detected by their nuclear fluorescence with PI as viewed through filter combination No. 3. B: double exposure photomicrograph using filter combination Nos. 1 and 3 showing the green immunofluorescence of an anti-microtubular serum reacting with cytoplasmic mi- crotubules and nuclear fluorescence of PI. Magnification, × 2000.

Fig. 4. Comparison of fluorescence with phase contrast microscopy. A: immunofluores- cence as microtubular cytoskeleton (green) identified by antibody and fluorescence of nuclei (red-yellow) revealed by PI. B: phase contrast image of the same cells. Note the pattern of stress fibres is not identical with the pattern of microtubules. Magnification, × 2000.

F."

"rl

"11'

266

Fig. 5. Comparison of fluorescence with phase/fluorescence microscopy. A: cytoakeleton detail is well resolved in this double exposed photomicrograph due to minimal back- ground. B: cytoplasmic detail is lost against background in combined phase/fluorescence microscopy. Magnification, x 5000.

of particular substances in cells with structural features. These reciprocal motions dampen the efficiency of visual accommodation required for effi- cient appreciation of immunofiuorescence detail and often cause some doubt about the precise localization of structures in the examination of complex tissues. However, by combining the fluorescent histochemical properties of PI as a marker of cell nuclei with the fluorescence of fluorochrome-labelled antibodies, it is possible to demonstrate morphological detail and antigen localization at the same time. PI can also be used with fluorochrome-labelled ant ibody in membrane immunofluorescence for cellular identification and viability measurements (Yeh et al., 1981).

ACKNOWLEDGEMENTS

C.D. Ockleford acknowledges M.R.C. Project Grant G.979/939/SB, and thanks Professor MacVicar, Mrs. Stephanie Bulman, Miss Alison Cole and Miss Susan Day and the obstetricians and gynaecologists of Leicester Royal Infirmary for support. The work was also funded in part by the Juvenile Dia- betes Foundat ion, Medical Research Council, the East Grinstead Research Trust and Seward Laboratory, East Grinstead.

267

REFERENCES

Badley, R.A., C.W. Lloyd, A. Woods, L. Carruthers, C. Allcock and D.A. Rees, 1978, Exp. Cell Res. 117,231.

Crissman, H.A. and J.A. Steinkamp, 1973, J. Cell Biol. 59, 766. Faulk, W.P., C. Yeager, J.A. Mclntyre and M. Ueda, 1979, Proc. Roy. Soc. B 206,163. Fried, J., A.G. Perez and B.D. Clarkson, 1976, J. Cell Biol. 71,172. Galbraith, R.M. and W.P. Faulk, 1979, in: The Diabetic Pregnancy: A Perinatal Perspec-

tive, eds. I.R. Merktz and P.A.J. Adam (Grune and Stratton, New York) p. 111. Hudson, B., W.B. Upholt, J. Devinny and J. Vinograd, 1969, Proc. Natl. Acad. Sci. U.S.A.

62,813. Irvin, G.L. and C.B. Bagwell, 1979, Am. J. Surg. 138, 567. Kolesnikov, V.A., O.I.U. Sondore, G.E. Fedoseeva and A.V. Zelenin, 1979, Tsitologia 21,

1029. Krishan, A., 1975, J. Cell Biol. 66, 188. Krishan, A. and R.N. Ganapathi, 1979, J. Histochem. Cytochem. 27, 1655. Krishan, A., R.N. Ganapathi and M. Israel, 1978, Cancer Res. 38, 3656. Kuypers, H.G., M. Bentivoglio, D. Van der Kooy and C.E. Catsman-Berrevoets, 1979,

Neurosci. Lett. 12, 1. Loke, Y.W. and R.W. Borland, 1970, Nature 228, 561. Matter, L. and W.P. Faulk, 1980, in: Hypertension in Pregnancy, ed. J. Bonnar (M.T.P.

Press, Lancaster) p. 351. Okudaira, K., H. Nakai, H. Tetsuo, K. Takamichi, J. Kiyoaki, H. Yoshihiko and T. Juji,

1979, J. Clin. Invest. 64, 1213. Yeh, C.G., B. Hsi and W.P. Faulk, 1981, J. Immunol. Methods 43,233.