Histochemistry (1992) 97:329-333
Histochemistry © Springer-Verlag 1992
Labeling with fluorescent carbocyanine dyes of cultured endothelial and smooth muscle cells by growth in dye-containing medium B. Ragnarson 1, L. Bengtsson 1'2, and A. H~egerstrand 1 i Department of Anatomy, Karolinska Institutet, Box 60400, S-104 01 Stockholm, Sweden 2 Department of Thoracic and Cardiovascular Surgery, Karolinska Hospital, Stockholm, Sweden Accepted October 10, 1991
Summary. We describe a method for labeling cultured endothelial cells (ECs) and smooth muscle cells (SMCs) by letting the cells grow for three days in culture medium containing a low concentration of the fluorescent carbocyanine dyes DiI and DiO. We show that good labeling can be obtained with considerably lower concentrations (2.5 gg/ml) than has previously been described. With optimal concentration the labeling is very strong and seems to label all membranous structures in the cells. It was possible to clearly distinguish differentially pre-labeled cells both in coculture and seeded on denaturated vascular grafts. The cells remain fluorescent for more than seven days and may be passaged with retained proliferative capability. We suggest that DiI/DiO-labeling using dye-containing medium may be used for several cell types and is applicable in tissue culture and in the detection of implanted cells in vivo.
Introduction The labeling of cells and biological structures in tissue specimens is a widely-used technique in biological science. Different techniques either label or visualize in situ components of the tissue, as in tissue stains such as hematoxylin-eosin or immunohistochemical techniques using specific antibodies. Other methods are based on detectable substances transported through the tissue, as in autoradiographic and other tracing techniques. Both groups of labels usually require fixation and further processing of the tissue prior to visualization. The second group often requires application of the marker in living tissue for transportation to take place. In the last few years a new kind of marker, the carbocyanines, have come into use in cell and tissue research (Honig and Hume 1986; Godement 1987; Honig and Hume 1989; von Bartheld 1990). These carbocyanines Correspondence to:
B. Ragnarson
are a family of lipid-soluble fluorescent dyes with some advantages: 1) They label cells non-selectively. 2) They are transported in all kinds of cells independently of any specific transportation mechanism, in both living and fixed cells. 3) They are themselves fluorescent. These properties make carbocyanines suitable for staining and visualizing living stained cells in vitro, thus making them particularly suitable for cell culture studies. We are at present interested in the lining of denaturated porcine heart valves and bovine vascular grafts with cultured adult human ECs. At the moment the use of these materials is limited due to severe degeneration and calcification causing graft failures. The cause of the degeneration is partly an immunological reaction against the tissue, although the commonly used glutaraldehyde preservation is known to decrease antigenicity (Spray and Roberts 1977; Heinzerling R 1982). An adverse effect of glutaraldehyde preservation of the tissue is the death of smooth muscle cells (SMCs) and fibroblasts and loss of endothelium (Dunn and Mamon 1985), exposing a non-regenerative matrix with a possibly antigenic and thrombogenic surface of the vascular/valvular substitute. Introducing autologous cells such as endothelial, smooth muscle and fibroblasts could be one way of making the bioprosthesis more functional especially in regard to the risk for thrombus formation and calcifying degeneration. To be able to study the persistance of the endothelial-cell lining after the vascular bioprosthesis has been subjected to in vivo conditions, we were interested in a labeling technique exclusive to seeded cells. This excludes immunohistochemical techniques since cell-specific antibodies label preexisting and ingrowing cells as well as seeded cells and, furthermore, primary antibodies are commonly raised in mice, which makes the differentiation of cell types in a single tissue section impossible. We initially tried the dissociation labeling technique previously described (Honig and Hume 1986; Honig and Hume 1989), but the ECs in suspension did not survive this treatment well enough to be able to adhere to the tissue culture plastics or to the bioprosthetic material.
330 This was p r o b a b l y due to the high c o n c e n t r a t i o n o f carb o c y a n i n e stock solution, resulting in a relatively high c o n c e n t r a t i o n o f e t h a n o l in the m e d i u m in c o m b i n a t i o n with m e c h a n i c a l a n d e n z y m a t i c d a m a g e d u r i n g centrifug a t i o n a n d replating. We therefore investigated if we c o u l d label cells w i t h o u t d i s s o c i a t i o n by g r o w i n g t h e m in a m e d i u m c o n t a i n i n g a low c o n c e n t r a t i o n o f the dye. This w o u l d a v o i d m e c h a n i c a l a n d e n z y m a t i c d i s r u p t i o n a n d also lessen the p o t e n t i a l adverse effects o f high conc e n t r a t i o n s o f e t h a n o l a n d o f the dye itself. F l u o r e s c e n t s t a i n i n g u s i n g different c o n c e n t r a t i o n s of D i I a n d D i O was e x a m i n e d a n d the a p p e a r a n c e o f cells in coculture as well as seeded o n t o a d e n a t u r a t e d biological v a s c u l a r m a t e r i a l was investigated.
i.e. it works, at least in theory, with all cells. Secondly, it works both in living and in fixed tissue (Godement 1987). DiI is a green crystalline powder which is brilliant red in solution. It fluoresces in bright red-orange and (dissolved in ethanol) its maximum absorbtion and peak emission wavelengths are 546 and 563 nm, respectively (Sims et al. 1974). This makes it possible to view DiI-fluorescence with filters used for TRITC. With FITC filters DiI fluorescence is still visible, although much fainter and more yellowish. If FITC filters are combined with a blocking (highpass) filter, the DiI fluorescence is not visible. DiO is a red-brownish, crystalline powder which is brick-red in solution. It fluoresces in bright green and (dissolved in ethanol) its maximum absorbtion and peak emission wavelengths are 489 and 499 nm, respectively (Sims et al. 1974). DiO fluorescence is thus easily viewed with FITC filters whereas no visible DiO fluorescence is seen using TRITC filters.
Materials and methods
Labeling procedures
Endothelial cell culture ECs from 3-5 cm long residual segments of the great saphenous vein, taken from patients undergoing open heart surgery, were isolated and cultured. Briefly, the veins were cannulated and rinsed with 5 ml of Minimal Essential Medium (MEM; Gibco, Grand Island, New York, USA) and subsequently filled with a collagenase solution (0.1%; Worthington, Freehold, New Jersey, U S A ) i n MEM and incubated at 37°C for 20 min. Harvested cells were centrifuged ~it 800 x g for 5 min and then resuspended and cultured as described elsewhere (H~egerstrand et al. 1992). After reaching confluence, the ECs were detached with a 0.1% trypsin/0.02% EDTA (1:1) solution and subcultivated as described for primary cultures. For experiments 5-7th passage cells were used.
Smooth muscle cell culture Primary cultures of SMCs were derived from explants of human great saphenous vein, after removal of ECs. The cells were grown in F-12 Ham's medium with the addition of 10% newborn calf serum (NCS, Gibco) and antibiotics (penicillin 50 gg/ml and streptomycin 50 U/ml). For experiments 4-8th passage cells were used.
ECs and SMCs were both plated in separate rows on two 24-well clusters, one without and one containing coverslips, and cells were cultured as described. After 2 days of culture in their respective media, the medium was changed to culture medium supplemented with DiI for ECs and with DiO for SMCs. For each dye a dilution series was made. A stock solution of 0.25% (w/v) DiI was first prepared by dissolving DiI crystals in 99.5% ethanol. The solution was then sonicated briefly and sterile-filtered (0.22 gm filter pore size) to minimize the content of undissolved crystals (modified after Honig and Hume 1986). This solution was then added to the culture medium for ECs with final concentrations ranging from 20 ng/ml to 40 gg/ml. A stock solution of 0.3% (w/v) DiO was made by dissolving DiO-crystals in a 1:9 mixture of dimethyl sulfoxide (DMSO) and 99.5% ethanol. The solution was then sonicated for approximately 10 min and sterile-filtered (0.22 gm filter pore size). This solution was added to culture medium for SMCs with final concentrations ranging from 20 ng/mt to 40 gg/ml. The cells were cultured in dye-containing medium for three days. The culture medium was renewed with an identical medium after two days. After a total of three days in dye-containingmedium, the cells were fixed with a solution of 4% paraformaldehyde in phosphate buffer (pH 7.0). To test the viability of cells and longevity of labeling, ECs labeled with 10 gg/ml DiI were kept in dye-free culture medium for one week. The coverslips were mounted on slides, coated with glycerol in PBS and covered with coverslips. They were then viewed in a Nikon epifluorescence microscope.
Properties of carbocyanine dyes Cocultures of endothelial and smooth muscle cells Carbocyanines are bilaterally symmetric molecules, with a positively charged "head", containing two conjugated rings, to which are attached two uncharged" tails" of long hydrocarbon chains (Honig and Hume 1986). The "head" determines the colour of the dye and also gives it fluorescent properties (Sims et al. 1974). It differs slightly between different carbocyanines, providing them with different colours and fluoresceiat properties (Sims et al. 1974), and several different carbocyanine dyes are commercially available (Molecular Probes, Eugene, Ore., USA) with DiI and DiO being the most widely used. The "tails" make carbocyanines extremely lipid-soluble and virtually insoluble in water. They dissolve in all lipophilic components in tissues, predominantly in the phospholipid membrane structures of cells, where the "tails" become inserted into the phospholipid bilayer and so dissolve the carbocyanines in the membrane. The carbocyanines are not only dissolved in the phospholipid membranes, they are also "transported" in the membranes by tangential diffusion (Derzko and Jacobson 1980; Vaz et al. 1982). A cell membrane of a cell partly in contact with the dye would, in time, become completely stained. Although this method of transport is slow (Vaz et al. 1982), it has two distinct advantages. Firstly, it is a non-selective means of transportation,
A note was made of six of the wells with the best labeled cells over the dilution range, which in the case of DiI was: 40 gg/ml; 10 gg/ml; 2.5 gg/ml; 625 ng/ml; 156.2 ng/ml; 39.1 ng/ml, and in the case of DiO: 40 ~tg/ml; 10 gg/ml; 2.5 gg/ml; 625 ng/ml; 156.2 ng/ml; 39.1 ng/ml. The cells in these wells were selected from the cluster without coverslips and the two cell types were plated together in matching concentration-pairs as described above into new wells containing coverslips. These "mosaic" cultures will here be named after the concentrations of DiI/DiO, i.e. 40/40; 10/10; 2.5/2.5; 625/625; 156.2/156.2; 39.1/39.1. After one day in culture the cells were fixed and further processed as described above. All slides were viewed with TRITC filters for DiI and FITC filters solely or in combination with a blocking filter for DiO.
Lining of vascular bioprostheses with ECs and SMCs Glutaraldehyde in segments of bovine internal mammary artery (BIMA, Biocor Hospital, Brasil) with a length of 4 cm and a diame-
331
a
b
// /
C
,2 /
Fig. 1 a-d. Fluorescence micrograph showing cells in "mosaic" cocultures of DiI-labeled ECs and DiO-labeled SMCs from the great saphenous vein (a-c) and cocultured cells lining a vascular bioprosthesis (d). Single arrows denote ECs and double arrows denote SMCs. With TRITC filters (a) only the DiI-labeled ECs are visible. Note the typical shape of the cells and the intense staining of the plasma membrane and cytoplasmic granules. Using FITC filters and blocking filter (b) only the DiO-labeled SMCs are visible. Note
the clear difference in morphology compared to (a) and the similar type of staining. With a double-exposure using TRITC filters and FITC filters plus a blocking filter (c) both cell types are clearly visible and distinguishable. The micrograph showing the vascular bioprosthesis (d) is double-exposed using the same filter combination. Note the cells lining the luminal surface, with both celltypes clearly distinguishable. Scale bars in (a) and (d) denote 50 gm
332 ter of 8 mm were eluated in 100 ml of PBS for three weeks at room temperature. The PBS was changed every third day. DiI/ DiO-labeled ECs/SMCs (105 cells/eraz) were allowed to adhere to the luminal surface of a piece of a bioprosthesis placed in a 2-cm2 culture well. The bioprosthesis was kept under culture conditions for 24 h to allow attachment and spreading of the cells. Cryostat tissue sections (14 gm) of seeded vascular bioprostheses were mounted on glass slides, coated with glycerol in PBS, and covered with coverslips. All sections were viewed in a Nikon epifluorescence microscope, as described above.
Results
DiI-Iabeling of ECs The appearance of DiI-labeled ECs was striking at all magnifications used. The labeling was bright red and of the same nature at all concentrations, although the intensity appeared to be concentration-dependent. At 40, 20 and 10 gg/ml the plasma membrane was completely outlined and in the cytoplasm numerous labeled granules of different sizes were present. At 5 and 2.5 gg/ml the labeling seemed to have diminished slightly, although it was still intense. When the concentrations were lowered further the labeling intensity diminished successively. At the lower concentrations, 1.25 gg/ml down to 19.5 ng/ml the outline of the plasma cell membrane was no longer complete and cytoplasmic granules had become faint. The DiI-labeled ECs cultured in dye-free medium for one week were still intensely labeled, and when split 1 : 3 they were able to grow to confluence in culture medium and retained their typical cobblestone morphology.
DiO-labeling of SMCs The appearance of DiO-labeled SMCs was also striking. They were bright green with the same kind of labeling as the DiI-labeled ECs. As with DiI-labeled cells the labeling was concentration-dependent and of the same type at all concentrations. Intense labeling was obtained at 40, 20 and 10 gg/ml, and was gradually weaker at concentrations below 10 gg/ml.
Cocultures of endothelial and smooth muscle cells The endothelial and SMCs in the six " m o s a i c " cocultures showed the same cellular features and the same kind of labeling as did the cells of the single-cell type cultures of the corresponding dye-concentration. The concentration-pairs most suitable for double-labeling were 40:40, 10/10 and 2.5/2.5. Among these the staining intensity did not diminish appreciably. The staining of the 625/625 mosaic was clearly weaker, but nevertheless possible to use. " M o s a i c s " made at lower concentrations, i.e. 156.2/156.2 and 39.1/39.1, had a labeling too weak to be useful for double-staining. We found the 2.5/2.5 concentration-pair (Fig. 1 a-c) to be suitable for these cells, since it was the lowest concentration still
yielding an intense and useful degree of labeling, thus putting the least possible strain on the cells during staining.
Appearance of labeled endothelial cells and smooth muscle cells lining vascular bioprostheses It was shown that labeling of ECs and SMCs (DiI and DiO respectively, both dyes at 5 gg/ml) in culture for three days was sufficient to enable the ceils to be visualized and distinguished after seeding on the wall of a vascular bioprosthesis (Fig. I d). At some parts of the cell layer fluorescence was observed with both TRITC and FITC filters and this was probably due to superimposed cells. Discussion
Carbocyanine dyes have proved extremely useful for neuronal tracing and developmental studies (Bhide 1990, personal communication; Godement 1987; Honig and Hume 1986; Tan 1990). They have also been used in cell cultures, especially for the culture of neurons of defined origin (Honig and Hume 1986). The technique described by Honig and Hume (1986) for staining cell cultures stains the cells by dissociating and suspending the cells for 0.5-1h in dye-containing media with a concentration of 12-40 gg/ml DiI or 100-170 lag/ml DiO. Our experience is that ECs cannot withstand such treatment. One reason for this may be that enzymatic dissociation and suspension culture makes the cells loose their attachment, which is important to this adhesion-dependent celltype. Another reason could be that the high concentrations of the dye solvent, and also of the dye itself, may be toxic to the cells if administered at the concentrations used in dissociation labeling previously described (12~40 gg/ml (DiI); 100-170 gg/ml (DiO), Honig and Hume 1986). Our approach was to add the dye directly to the culture medium, thus allowing the cells to keep their attachment to the plastic surface and thereby become possibly less vulnerable. The presence of higher serum concentrations during dye incorporation than in previous investigations (Honig and Hume 1986) is another possible explanation for the survival of ECs and SMCs. To reach good staining with a low concentration, however, the cells must be exposed to the dye for a longer period than previously described (Honig and Hume 1986). With a longer exposure time one can assume that the labeling intensity, or amount of labeling, is a product of exposure time and dye concentration. A beneficial side-effect of the prolonged exposure is that it gives the cell time to circulate the dye through its compartments, thus exposing the intracellular membranes to the dye, which we probably see as stained cytoplasmic granules. This, we believe results in a more complete staining of the cell. In this study we examined the labeling after three days in tissue culture medium and showed that a sixteenfold lower concentration of dye was sufficient to strongly
333 label the cells in culture. The labeling was prominent and also suitable for coculture studies and studies of the distribution of the different cell types after seeding onto vascular bioprostheses. It is suggested that this double labeling procedure m a y also be used for studies of m o r p h o l o g y or proliferaton of differently labeled cells in coculture. The staining will probably diminish as a function of resynthesized cellular m e m b r a n e structures, but this method is still likely to be of value for shortterm experiments. Furthermore, in studies of cell fusion, which is a matter of interest in muscle cell/myoblastresearch, this a p p r o a c h m a y be of help. It m a y also be a suitable method by which to detect cultured cells that are transplanted into living tissue. Depending on the time for renewal of plasma m e m b r a n e and proliferation rate in vivo, the dye m a y be detectable weeks after transplantation. In conclusion, carbocyanine labeling of cells by growth in dye-containing medium appears to be a simple and reliable method of labeling cells in culture and also of clearly distinguishing cells subjected to further cultivation, seeding or transplantation.
Acknowledgements. We wish to thank Mrs. H. Wikstr6m for her excellent technical assistance. This work was supported by the Swedish Society of Medicine, the Swedish Heart Lung Foundation and the Swedish Medical Research Council (project no. 553 and no. 07126)
References Bartheld CS von, Cunningham DE, Rubel EW (1990) Neuronal tracing with DiI: decalcification, cryosectioning, and photoconversion for light and electron microscopic analysis. J Histochem Cytochem 38: 725-733
Derzko Z, Jacobson K (1980) Comparative lateral diffusion of fluorescent lipid analogues in phospholipid multibilayers. Biochemistry 19:6050-6057 Dunn J, Marmon L (1985) Mechanisms of calcification of tissue valves. Cardiol Clin 3 : 385-396 Godement P, Vanselow J, Thanos S, Bonhoeffer F (1987) A study in developing visual systems with a new method of staining neurones and their processes in fixed tissue. Development 101:697-713 Heinzerling R, Stein P, Riddle J, Magilligan D, Jennings J (1982) Immunological involvement in porcine bioprosthetic valve degeneration: preliminary studies. Henry Ford Hosp Med J 30:146-151 Honig MG, Hume RI (1986) Fluorescent carbocyanine dyes allow living neurons of identified origin to be studied in long-term cultures. J Cell Biol 103:171-187 Honig MG, Hume RI (1989) DiI and DiO: versatile fluorescent dyes for neuronal labelling and pathway tracing. TINS 12:333341 H~egerstrand A, Gillis C, Bengtsson L (1992) Serial cultivation of adult human endothelium from the great saphenous vein. J Vasc Surg (in press) Sims PJ, Waagoner AS, Wang C, Hoffman JF (1974) Studies on the mechanism by which cyanine dyes measure membrane potential in red blood cells and phosphatidylcholine vesicles. Biochemistry 13 : 3315-3330 Spray T, Roberts W (1977) Structural changes in porcine xenografts used as substitutes for cardiac valves. Gross and histologic observations in 51 glutaraldehyde-preserved Hancock valves in 41 patients. Am J Cardiol 40:319-330 Tan H, Miletic V (1990) Bulbospinal serotoninergic pathways in the frog Ranapipiens. J Comp Neurol 292:291-302 Vaz WLC, Derzko ZI, Jacobson KA (1982) Photobleaching measurements of the lateral diffusion of lipids and proteins in artifical phospholipid bilayer membranes. In: Poste G, Nicolson G (eds) Membrane reconstitution. Elsevier North-Holland Biomedical Press, Amsterdam, pp 83-136
Histochemistry © Springer-Verlag 1992
Labeling with fluorescent carbocyanine dyes of cultured endothelial and smooth muscle cells by growth in dye-containing medium B. Ragnarson 1, L. Bengtsson 1'2, and A. H~egerstrand 1 i Department of Anatomy, Karolinska Institutet, Box 60400, S-104 01 Stockholm, Sweden 2 Department of Thoracic and Cardiovascular Surgery, Karolinska Hospital, Stockholm, Sweden Accepted October 10, 1991
Summary. We describe a method for labeling cultured endothelial cells (ECs) and smooth muscle cells (SMCs) by letting the cells grow for three days in culture medium containing a low concentration of the fluorescent carbocyanine dyes DiI and DiO. We show that good labeling can be obtained with considerably lower concentrations (2.5 gg/ml) than has previously been described. With optimal concentration the labeling is very strong and seems to label all membranous structures in the cells. It was possible to clearly distinguish differentially pre-labeled cells both in coculture and seeded on denaturated vascular grafts. The cells remain fluorescent for more than seven days and may be passaged with retained proliferative capability. We suggest that DiI/DiO-labeling using dye-containing medium may be used for several cell types and is applicable in tissue culture and in the detection of implanted cells in vivo.
Introduction The labeling of cells and biological structures in tissue specimens is a widely-used technique in biological science. Different techniques either label or visualize in situ components of the tissue, as in tissue stains such as hematoxylin-eosin or immunohistochemical techniques using specific antibodies. Other methods are based on detectable substances transported through the tissue, as in autoradiographic and other tracing techniques. Both groups of labels usually require fixation and further processing of the tissue prior to visualization. The second group often requires application of the marker in living tissue for transportation to take place. In the last few years a new kind of marker, the carbocyanines, have come into use in cell and tissue research (Honig and Hume 1986; Godement 1987; Honig and Hume 1989; von Bartheld 1990). These carbocyanines Correspondence to:
B. Ragnarson
are a family of lipid-soluble fluorescent dyes with some advantages: 1) They label cells non-selectively. 2) They are transported in all kinds of cells independently of any specific transportation mechanism, in both living and fixed cells. 3) They are themselves fluorescent. These properties make carbocyanines suitable for staining and visualizing living stained cells in vitro, thus making them particularly suitable for cell culture studies. We are at present interested in the lining of denaturated porcine heart valves and bovine vascular grafts with cultured adult human ECs. At the moment the use of these materials is limited due to severe degeneration and calcification causing graft failures. The cause of the degeneration is partly an immunological reaction against the tissue, although the commonly used glutaraldehyde preservation is known to decrease antigenicity (Spray and Roberts 1977; Heinzerling R 1982). An adverse effect of glutaraldehyde preservation of the tissue is the death of smooth muscle cells (SMCs) and fibroblasts and loss of endothelium (Dunn and Mamon 1985), exposing a non-regenerative matrix with a possibly antigenic and thrombogenic surface of the vascular/valvular substitute. Introducing autologous cells such as endothelial, smooth muscle and fibroblasts could be one way of making the bioprosthesis more functional especially in regard to the risk for thrombus formation and calcifying degeneration. To be able to study the persistance of the endothelial-cell lining after the vascular bioprosthesis has been subjected to in vivo conditions, we were interested in a labeling technique exclusive to seeded cells. This excludes immunohistochemical techniques since cell-specific antibodies label preexisting and ingrowing cells as well as seeded cells and, furthermore, primary antibodies are commonly raised in mice, which makes the differentiation of cell types in a single tissue section impossible. We initially tried the dissociation labeling technique previously described (Honig and Hume 1986; Honig and Hume 1989), but the ECs in suspension did not survive this treatment well enough to be able to adhere to the tissue culture plastics or to the bioprosthetic material.
330 This was p r o b a b l y due to the high c o n c e n t r a t i o n o f carb o c y a n i n e stock solution, resulting in a relatively high c o n c e n t r a t i o n o f e t h a n o l in the m e d i u m in c o m b i n a t i o n with m e c h a n i c a l a n d e n z y m a t i c d a m a g e d u r i n g centrifug a t i o n a n d replating. We therefore investigated if we c o u l d label cells w i t h o u t d i s s o c i a t i o n by g r o w i n g t h e m in a m e d i u m c o n t a i n i n g a low c o n c e n t r a t i o n o f the dye. This w o u l d a v o i d m e c h a n i c a l a n d e n z y m a t i c d i s r u p t i o n a n d also lessen the p o t e n t i a l adverse effects o f high conc e n t r a t i o n s o f e t h a n o l a n d o f the dye itself. F l u o r e s c e n t s t a i n i n g u s i n g different c o n c e n t r a t i o n s of D i I a n d D i O was e x a m i n e d a n d the a p p e a r a n c e o f cells in coculture as well as seeded o n t o a d e n a t u r a t e d biological v a s c u l a r m a t e r i a l was investigated.
i.e. it works, at least in theory, with all cells. Secondly, it works both in living and in fixed tissue (Godement 1987). DiI is a green crystalline powder which is brilliant red in solution. It fluoresces in bright red-orange and (dissolved in ethanol) its maximum absorbtion and peak emission wavelengths are 546 and 563 nm, respectively (Sims et al. 1974). This makes it possible to view DiI-fluorescence with filters used for TRITC. With FITC filters DiI fluorescence is still visible, although much fainter and more yellowish. If FITC filters are combined with a blocking (highpass) filter, the DiI fluorescence is not visible. DiO is a red-brownish, crystalline powder which is brick-red in solution. It fluoresces in bright green and (dissolved in ethanol) its maximum absorbtion and peak emission wavelengths are 489 and 499 nm, respectively (Sims et al. 1974). DiO fluorescence is thus easily viewed with FITC filters whereas no visible DiO fluorescence is seen using TRITC filters.
Materials and methods
Labeling procedures
Endothelial cell culture ECs from 3-5 cm long residual segments of the great saphenous vein, taken from patients undergoing open heart surgery, were isolated and cultured. Briefly, the veins were cannulated and rinsed with 5 ml of Minimal Essential Medium (MEM; Gibco, Grand Island, New York, USA) and subsequently filled with a collagenase solution (0.1%; Worthington, Freehold, New Jersey, U S A ) i n MEM and incubated at 37°C for 20 min. Harvested cells were centrifuged ~it 800 x g for 5 min and then resuspended and cultured as described elsewhere (H~egerstrand et al. 1992). After reaching confluence, the ECs were detached with a 0.1% trypsin/0.02% EDTA (1:1) solution and subcultivated as described for primary cultures. For experiments 5-7th passage cells were used.
Smooth muscle cell culture Primary cultures of SMCs were derived from explants of human great saphenous vein, after removal of ECs. The cells were grown in F-12 Ham's medium with the addition of 10% newborn calf serum (NCS, Gibco) and antibiotics (penicillin 50 gg/ml and streptomycin 50 U/ml). For experiments 4-8th passage cells were used.
ECs and SMCs were both plated in separate rows on two 24-well clusters, one without and one containing coverslips, and cells were cultured as described. After 2 days of culture in their respective media, the medium was changed to culture medium supplemented with DiI for ECs and with DiO for SMCs. For each dye a dilution series was made. A stock solution of 0.25% (w/v) DiI was first prepared by dissolving DiI crystals in 99.5% ethanol. The solution was then sonicated briefly and sterile-filtered (0.22 gm filter pore size) to minimize the content of undissolved crystals (modified after Honig and Hume 1986). This solution was then added to the culture medium for ECs with final concentrations ranging from 20 ng/ml to 40 gg/ml. A stock solution of 0.3% (w/v) DiO was made by dissolving DiO-crystals in a 1:9 mixture of dimethyl sulfoxide (DMSO) and 99.5% ethanol. The solution was then sonicated for approximately 10 min and sterile-filtered (0.22 gm filter pore size). This solution was added to culture medium for SMCs with final concentrations ranging from 20 ng/mt to 40 gg/ml. The cells were cultured in dye-containing medium for three days. The culture medium was renewed with an identical medium after two days. After a total of three days in dye-containingmedium, the cells were fixed with a solution of 4% paraformaldehyde in phosphate buffer (pH 7.0). To test the viability of cells and longevity of labeling, ECs labeled with 10 gg/ml DiI were kept in dye-free culture medium for one week. The coverslips were mounted on slides, coated with glycerol in PBS and covered with coverslips. They were then viewed in a Nikon epifluorescence microscope.
Properties of carbocyanine dyes Cocultures of endothelial and smooth muscle cells Carbocyanines are bilaterally symmetric molecules, with a positively charged "head", containing two conjugated rings, to which are attached two uncharged" tails" of long hydrocarbon chains (Honig and Hume 1986). The "head" determines the colour of the dye and also gives it fluorescent properties (Sims et al. 1974). It differs slightly between different carbocyanines, providing them with different colours and fluoresceiat properties (Sims et al. 1974), and several different carbocyanine dyes are commercially available (Molecular Probes, Eugene, Ore., USA) with DiI and DiO being the most widely used. The "tails" make carbocyanines extremely lipid-soluble and virtually insoluble in water. They dissolve in all lipophilic components in tissues, predominantly in the phospholipid membrane structures of cells, where the "tails" become inserted into the phospholipid bilayer and so dissolve the carbocyanines in the membrane. The carbocyanines are not only dissolved in the phospholipid membranes, they are also "transported" in the membranes by tangential diffusion (Derzko and Jacobson 1980; Vaz et al. 1982). A cell membrane of a cell partly in contact with the dye would, in time, become completely stained. Although this method of transport is slow (Vaz et al. 1982), it has two distinct advantages. Firstly, it is a non-selective means of transportation,
A note was made of six of the wells with the best labeled cells over the dilution range, which in the case of DiI was: 40 gg/ml; 10 gg/ml; 2.5 gg/ml; 625 ng/ml; 156.2 ng/ml; 39.1 ng/ml, and in the case of DiO: 40 ~tg/ml; 10 gg/ml; 2.5 gg/ml; 625 ng/ml; 156.2 ng/ml; 39.1 ng/ml. The cells in these wells were selected from the cluster without coverslips and the two cell types were plated together in matching concentration-pairs as described above into new wells containing coverslips. These "mosaic" cultures will here be named after the concentrations of DiI/DiO, i.e. 40/40; 10/10; 2.5/2.5; 625/625; 156.2/156.2; 39.1/39.1. After one day in culture the cells were fixed and further processed as described above. All slides were viewed with TRITC filters for DiI and FITC filters solely or in combination with a blocking filter for DiO.
Lining of vascular bioprostheses with ECs and SMCs Glutaraldehyde in segments of bovine internal mammary artery (BIMA, Biocor Hospital, Brasil) with a length of 4 cm and a diame-
331
a
b
// /
C
,2 /
Fig. 1 a-d. Fluorescence micrograph showing cells in "mosaic" cocultures of DiI-labeled ECs and DiO-labeled SMCs from the great saphenous vein (a-c) and cocultured cells lining a vascular bioprosthesis (d). Single arrows denote ECs and double arrows denote SMCs. With TRITC filters (a) only the DiI-labeled ECs are visible. Note the typical shape of the cells and the intense staining of the plasma membrane and cytoplasmic granules. Using FITC filters and blocking filter (b) only the DiO-labeled SMCs are visible. Note
the clear difference in morphology compared to (a) and the similar type of staining. With a double-exposure using TRITC filters and FITC filters plus a blocking filter (c) both cell types are clearly visible and distinguishable. The micrograph showing the vascular bioprosthesis (d) is double-exposed using the same filter combination. Note the cells lining the luminal surface, with both celltypes clearly distinguishable. Scale bars in (a) and (d) denote 50 gm
332 ter of 8 mm were eluated in 100 ml of PBS for three weeks at room temperature. The PBS was changed every third day. DiI/ DiO-labeled ECs/SMCs (105 cells/eraz) were allowed to adhere to the luminal surface of a piece of a bioprosthesis placed in a 2-cm2 culture well. The bioprosthesis was kept under culture conditions for 24 h to allow attachment and spreading of the cells. Cryostat tissue sections (14 gm) of seeded vascular bioprostheses were mounted on glass slides, coated with glycerol in PBS, and covered with coverslips. All sections were viewed in a Nikon epifluorescence microscope, as described above.
Results
DiI-Iabeling of ECs The appearance of DiI-labeled ECs was striking at all magnifications used. The labeling was bright red and of the same nature at all concentrations, although the intensity appeared to be concentration-dependent. At 40, 20 and 10 gg/ml the plasma membrane was completely outlined and in the cytoplasm numerous labeled granules of different sizes were present. At 5 and 2.5 gg/ml the labeling seemed to have diminished slightly, although it was still intense. When the concentrations were lowered further the labeling intensity diminished successively. At the lower concentrations, 1.25 gg/ml down to 19.5 ng/ml the outline of the plasma cell membrane was no longer complete and cytoplasmic granules had become faint. The DiI-labeled ECs cultured in dye-free medium for one week were still intensely labeled, and when split 1 : 3 they were able to grow to confluence in culture medium and retained their typical cobblestone morphology.
DiO-labeling of SMCs The appearance of DiO-labeled SMCs was also striking. They were bright green with the same kind of labeling as the DiI-labeled ECs. As with DiI-labeled cells the labeling was concentration-dependent and of the same type at all concentrations. Intense labeling was obtained at 40, 20 and 10 gg/ml, and was gradually weaker at concentrations below 10 gg/ml.
Cocultures of endothelial and smooth muscle cells The endothelial and SMCs in the six " m o s a i c " cocultures showed the same cellular features and the same kind of labeling as did the cells of the single-cell type cultures of the corresponding dye-concentration. The concentration-pairs most suitable for double-labeling were 40:40, 10/10 and 2.5/2.5. Among these the staining intensity did not diminish appreciably. The staining of the 625/625 mosaic was clearly weaker, but nevertheless possible to use. " M o s a i c s " made at lower concentrations, i.e. 156.2/156.2 and 39.1/39.1, had a labeling too weak to be useful for double-staining. We found the 2.5/2.5 concentration-pair (Fig. 1 a-c) to be suitable for these cells, since it was the lowest concentration still
yielding an intense and useful degree of labeling, thus putting the least possible strain on the cells during staining.
Appearance of labeled endothelial cells and smooth muscle cells lining vascular bioprostheses It was shown that labeling of ECs and SMCs (DiI and DiO respectively, both dyes at 5 gg/ml) in culture for three days was sufficient to enable the ceils to be visualized and distinguished after seeding on the wall of a vascular bioprosthesis (Fig. I d). At some parts of the cell layer fluorescence was observed with both TRITC and FITC filters and this was probably due to superimposed cells. Discussion
Carbocyanine dyes have proved extremely useful for neuronal tracing and developmental studies (Bhide 1990, personal communication; Godement 1987; Honig and Hume 1986; Tan 1990). They have also been used in cell cultures, especially for the culture of neurons of defined origin (Honig and Hume 1986). The technique described by Honig and Hume (1986) for staining cell cultures stains the cells by dissociating and suspending the cells for 0.5-1h in dye-containing media with a concentration of 12-40 gg/ml DiI or 100-170 lag/ml DiO. Our experience is that ECs cannot withstand such treatment. One reason for this may be that enzymatic dissociation and suspension culture makes the cells loose their attachment, which is important to this adhesion-dependent celltype. Another reason could be that the high concentrations of the dye solvent, and also of the dye itself, may be toxic to the cells if administered at the concentrations used in dissociation labeling previously described (12~40 gg/ml (DiI); 100-170 gg/ml (DiO), Honig and Hume 1986). Our approach was to add the dye directly to the culture medium, thus allowing the cells to keep their attachment to the plastic surface and thereby become possibly less vulnerable. The presence of higher serum concentrations during dye incorporation than in previous investigations (Honig and Hume 1986) is another possible explanation for the survival of ECs and SMCs. To reach good staining with a low concentration, however, the cells must be exposed to the dye for a longer period than previously described (Honig and Hume 1986). With a longer exposure time one can assume that the labeling intensity, or amount of labeling, is a product of exposure time and dye concentration. A beneficial side-effect of the prolonged exposure is that it gives the cell time to circulate the dye through its compartments, thus exposing the intracellular membranes to the dye, which we probably see as stained cytoplasmic granules. This, we believe results in a more complete staining of the cell. In this study we examined the labeling after three days in tissue culture medium and showed that a sixteenfold lower concentration of dye was sufficient to strongly
333 label the cells in culture. The labeling was prominent and also suitable for coculture studies and studies of the distribution of the different cell types after seeding onto vascular bioprostheses. It is suggested that this double labeling procedure m a y also be used for studies of m o r p h o l o g y or proliferaton of differently labeled cells in coculture. The staining will probably diminish as a function of resynthesized cellular m e m b r a n e structures, but this method is still likely to be of value for shortterm experiments. Furthermore, in studies of cell fusion, which is a matter of interest in muscle cell/myoblastresearch, this a p p r o a c h m a y be of help. It m a y also be a suitable method by which to detect cultured cells that are transplanted into living tissue. Depending on the time for renewal of plasma m e m b r a n e and proliferation rate in vivo, the dye m a y be detectable weeks after transplantation. In conclusion, carbocyanine labeling of cells by growth in dye-containing medium appears to be a simple and reliable method of labeling cells in culture and also of clearly distinguishing cells subjected to further cultivation, seeding or transplantation.
Acknowledgements. We wish to thank Mrs. H. Wikstr6m for her excellent technical assistance. This work was supported by the Swedish Society of Medicine, the Swedish Heart Lung Foundation and the Swedish Medical Research Council (project no. 553 and no. 07126)
References Bartheld CS von, Cunningham DE, Rubel EW (1990) Neuronal tracing with DiI: decalcification, cryosectioning, and photoconversion for light and electron microscopic analysis. J Histochem Cytochem 38: 725-733
Derzko Z, Jacobson K (1980) Comparative lateral diffusion of fluorescent lipid analogues in phospholipid multibilayers. Biochemistry 19:6050-6057 Dunn J, Marmon L (1985) Mechanisms of calcification of tissue valves. Cardiol Clin 3 : 385-396 Godement P, Vanselow J, Thanos S, Bonhoeffer F (1987) A study in developing visual systems with a new method of staining neurones and their processes in fixed tissue. Development 101:697-713 Heinzerling R, Stein P, Riddle J, Magilligan D, Jennings J (1982) Immunological involvement in porcine bioprosthetic valve degeneration: preliminary studies. Henry Ford Hosp Med J 30:146-151 Honig MG, Hume RI (1986) Fluorescent carbocyanine dyes allow living neurons of identified origin to be studied in long-term cultures. J Cell Biol 103:171-187 Honig MG, Hume RI (1989) DiI and DiO: versatile fluorescent dyes for neuronal labelling and pathway tracing. TINS 12:333341 H~egerstrand A, Gillis C, Bengtsson L (1992) Serial cultivation of adult human endothelium from the great saphenous vein. J Vasc Surg (in press) Sims PJ, Waagoner AS, Wang C, Hoffman JF (1974) Studies on the mechanism by which cyanine dyes measure membrane potential in red blood cells and phosphatidylcholine vesicles. Biochemistry 13 : 3315-3330 Spray T, Roberts W (1977) Structural changes in porcine xenografts used as substitutes for cardiac valves. Gross and histologic observations in 51 glutaraldehyde-preserved Hancock valves in 41 patients. Am J Cardiol 40:319-330 Tan H, Miletic V (1990) Bulbospinal serotoninergic pathways in the frog Ranapipiens. J Comp Neurol 292:291-302 Vaz WLC, Derzko ZI, Jacobson KA (1982) Photobleaching measurements of the lateral diffusion of lipids and proteins in artifical phospholipid bilayer membranes. In: Poste G, Nicolson G (eds) Membrane reconstitution. Elsevier North-Holland Biomedical Press, Amsterdam, pp 83-136
