Rhodamine B, a Fluorescent Probe for Acidic

0022-1554/96/$3.30

Vol. 44, No. 3. pp. 267-274. 1996 Printed in US.A

The Journal of Histochemistry and Cytochemistry Copyright 0 1996 by The Histochemical Society, Inc.

Original Article

I Rhodamine B, a Fluorescent Probe for Acidic Organelles in Denervated Skeletal Muscle' FREDRIK VULT VON STEYERN,2 JAN-OWE JOSEFSSON, and SVEN TAGERUD Department of Pbamaco/ogy, University of Lund, Lund, Sweden. Received for publication April 17, 1995 and in revised form September 27, 1995; accepted October IO, 1995 (5A3646).

We describe a very efficient method for fluorescent labeling of acidic structures in denervated skeletal muscle with rhodamine B. Rhodamine B at 50 nglml gave selective and distinct segmental labeling of denervated muscle fibers after 5-min incubation at room temperature. Labeling was also achieved at 4'C. The labeling was disrupted by the ionophores monensin and nigericin, suggesting a labeling confmed to acidic structures. Rhodamine B CO-localizedwith the lymmotropic dye 4x1 Tracker Green and a marker for endocytosis (fluorescein isothiocpnate-labeled dextran). Rhodamine B, which is highly lipophilic, showed pHdependent fluorescence emission in saturated aqueous Noctanol. Tetramethylrhodamineshowed similar charactetistics for labeling of denervated muscle fibers and pH-

Introduction Endocytotic and lysosomal activities in skeletal muscle increase after denervation (cf. Tigerud and Libelius, 1984; Libelius et al., 1978; Bird, 1975). Uptake of endocytosismarkers occurs mainly in small segments of the muscle fibers located in the region of the denervated endplate, where activities of lysosomal enzymes are also highest (Elmquist et al., 1992; Libelius and Tigetud, 1984). This localization of high endocytotic activity to the region around the denervated endplate has been determined by its distribution in the denervated mouse hemidiaphragm (Libelius and Tigerud, 1984), by the high occurrence of postsynaptic folds in ultrastructural studies of segments with high endocytotic activity (Tigerud et al., 1986b). and by co-staining for acetylcholine esterase (Elmquist et al., 1992).Ultrastructural studies with horseradish peroxidase (HRP)have shown accumulation of endocytosed marker in large membrane-limited bodies of lysosomal nature in the muscle fibers (Tagerud et al., 1986a,b; Libelius et al., 1979). An extensive proliferation of the transverse tubule system suggests that en-

'

Supported by grants from the Medical Faculty, University of Lund. Sweden and by Crafoordska Stiftclsen, Lund, Sweden. Correspondence to: Fredrik Vult von Steyern, Dept. of Pharmacology, U. of Lund, Solvegatan 10, S-223 62 Lund, Sweden.

dependent fluorescence in N-octanol. The carboxyl group present in these two compounds appears important, because structurally related compounds that either lack this group or have it estetified failed to label denervated muscle fibers and showed no pH-dependent fluorescence in N-octanol. The results suggestthat rhodamine B labels acidic organelles belonging to the endosomal/lysosomalsystem of denervated skeletal muscle fibers. Nevertheless, it failed to label such organelles in a number of mammalian cell types other than denervated skeletal muscle fibers. JHistochem Cytochem 44267-274, 1996) KEY WORDS: Denervation; Endosomes; Fluorescent labeling; Lysosomes; Mouse; Rhodamine B Skeletal muscle.

docytotic uptake occurs from these structures (Tigerud et al., 1986b; Libelius et al., 1979). When uptake of rhodamine B isothiocyanate(NE)-labeled dextran was used for studies of endocytotic activity in denervated muscle fibers, segments with high endocytotic and lysosomal activities were labeled even when incubated at 4°C. Gel filtration of the RIX-dextran preparation revealed that it contained unconjugated fluorescent material (Elmquist et al., 1992),most likely representing rhodamine B or rhodamine B isothiocyanate.Rhodamine B has been previously used as a general purpose probe for lipids Uette and Ziomek, 1994; Marinetti, 1962) and could conceivably enter cells by passive diffusion, thus accounting for the labeling that occurred at 4°C. However, rhodamine B is not known as a dye for vital staining in cell biological systems. The present work describes an extremely efficient labeling of acidic organelles in denervated skeletal muscle fibers with rhodamine B. Neutral red and acridine orange have been used for comparison because these substances have previously been used for vital staining of lysosomes (Koenig, 1965; Allison and Young, 1964). Fluorescein isothiocyanate-labeleddextran (FIX-dextran) and the lysosome probe Lyso Tracker Green have been used for double labeling experiments. A number of substances structurally related to rhodamine B have also been studied to investigate which features of this molecule are most important for the labeling to occur. We have also examined a number of other mammalian cells for structures labeled with rhodamine B. 267

Downloaded from jhc.sagepub.com by guest on September 30, 2016

WLT VON STEYERN, JOSEFSSON, TAGERUD

268

Materials and Methods Experiments on Muscle. All experiments were carried out on mouse hemidiaphragm muscle, which has a well-defined innervation pattern (Waser and Liithi, 1956) and after denervation exhibits high endocytotic and lysosomal activities in the region of the denervated endplates (Elmquist et al., 1992; Libelius and Tigerud, 1984). Adult male NMRI mice. weight about 30 g when taken to the experiments, were used and the left hemidiaphragm was denervated under diethyl ether anesthesia. A unilateral thoracotomywas performed and the left phrenic nerve was lifted from the thoracic cavity using a glass hook. The nerve was pulled out through the incision and cut. The mice were kept in a separate animal department under professional supervision and fed a standard laboratory diet with access to tapwater ad libitum. Most experiments were performed on muscles denervated for 8 days to 4 weeks, but two muscles denervated for 6 and 13 months, respectively, were also used in this study. For in vivo studies, substances to be injected were dissolved in 0.2-0.5 ml physiological saline and injected into a tail vein. The mice were sacrificed by cervical dislocation and the diaphragm was dissected and placed in oxygenated Ringer’s solution, p H 7.2-7.4 (composition in m M NaCl 135; KCI 5; CaC12 2; MgC12 1; Na2HP04 1; NaHCO3 IS; glucose 11; bubbled with 95% 0 2 and 5% C02) at room temperature (RT). The experimental manipulations were approved by the University of Lund Ethical Committee for Animal Experiments. Muscles were incubated at room temperature (about 22°C) or at 4’C with the dyes in continuously oxygenated Ringer’ssolution, inspected, and photographed in a Nikon Diaphot inverted microscope equipped for epifluorescence. Two different filter combinations were used: B-2A (excitation filter of 450-490 nm, dichroic mirror at 510 nm, and barrier filter of 520 nm) and G-2A (excitation filter of 510-560 nm, dichroic mirror at 580 nm, and barrier filter of 590 nm). Tetrodotoxin was usually added to a final concentration of 1 pM before microphotography to arrest fibrillatory activity in the denervated muscle fibers. HRP labeling of denervated hemidiaphragms was performed by injecting mice with 20 mg HRP Two hours later, muscles were dissected, mounted on Sylgard resin holders, and washed overnight in continuously oxygenated Ringer’s solution at 4°C. Muscles were fixed and stained for peroxidase activity as previously described (Vult von Steyern et al., 1994).

to determine the concentrations of rhodamine B. All fluorescence measurements were made using a Perkin-Elmer LS-5B spectrofluorophotometer. Dyes structurally related to rhodamine B were screened by a similar method. The dyes were dissolved in the buffer at pH 3.0 or pH 7.8, equilibrated with an equal volume of octanol, and fluorescence in the octanol and buffer phases was measured at an excitation wavelength (of540 nm and an emission wavelength of 590 nm. Some experiments were also performed to study the fluorescence properties of rhodamine B in organic solvents of various polarities and water content. The solvents used were butanol, octanol, and heptane. Chemicals. FIX-dextran, average molecular weight 3900, rhodamine 6G chloride, monensin, nigericin, horseradish peroxidase (Type 11). and tetrodotoxin were obtained from Sigma Chemicals (St Louis, MO). Rhodamine B (>99.1% pure) was from Radiant Dyes Chemie (Wermelskirchen, Germany). Neutral red chloride was from the British Drug Houses (Poole, UK). Tetramethylrhodamine was from Research Organics (Cleveland, OH). Acridine orange, Lyso Tracker Green DND-26. rhodamine B hexyl ester chloride, rhodamine 123, and tetramethylrosamine chloride were from Molecular Probes (Eugene, OR).

Results L$ophilicity of Rhodamine B and Effect of p H on Fluorescence Emission The partition ratio octanollcitricacid-phosphate buffer for rhodamine B was 100-200 (somewhat lower at pH 3.0 than at pH 7.8), which means that more than 99% of the rhodamine B was redistributed to the octanol phase. This partition ratio is similar to the value reported by Araie and Maurice (1987) for octanollPBS. The fluorescence emission in octanol showed a pH dependence in the sense that at acidic pH the emission,, for rhodamine B was shifted towards a longer wavelength and the peak fluorescence emission was increased (Figure 1). When fluorescencewas measured with excitation at 546 nm and emission at 590 nm, simulating the conditions in the microscope, this resulted in a 2.5-fold increase in fluorescence at pH 3.0 compared to that at pH 7.8. No significant pH dependence of the fluorescencewas observed in the buffer phase. Among the other dyes screened (Figure 2), only tetramethylrhodamine showed a similar increase in fluorescence in the octanol phase at acidic pH (results not shown).

Labeling of Other Mammalian Cells. Other mammalian cells examined for rhodamine B labeling were mouse connective tissue cells (present in the diaphragm preparations), MDCK (Madin-Darby canine kidney) cells, mouse peritoneal macrophages, and non-neural cells from embryonal rat spinal cord. Cells were incubated for 10-30 min in normal Ringer’s solution (MDCK cells, mouse peritoneal macrophages, and the non-neural cells from embryonal rat spinal cord were also incubated in each respective culture medium) at RT or at 37°C with rhodamine B present at various cono centrations ( ~ O - ~ o ong/ml).

Labeling of Denervated Muscle Fibers

Lipophilicity of Rhodamine B and Effect of pH on Fluorescence Emission. N-octanol was saturated with a citric acid-phosphate buffer (pH 3.0 or 7.8) and 1 volume of this saturated aqueous octanol was mixed with 1 volume of the buffer containing rhodamine B (1.0 pglml), shaken vigorously, and then left to equilibrate. After centrifugation, (300 x g, 2 min), the two phases were collected separately and fluorescence was measured. To simulate the conditions used in the fluorescence microscope (excitation filter 510-560 nm; mercury lamp with spectral line at 546 nm; barrier filter 590 nm), an excitation wavelength of 546 nm, which is also the absorption,,, of rhodamine B, was used, and emission was measured at 590 nm. A partition ratio (PR) was determined by relating the rhodamine B concentration in the buffer phase before (CO1 pg/ml) and after (Ci) mixing with octanol, using the formula PR = (C&l)/Ci. Calibration curves constructed in buffer saturated with octanol, at pH 3.0 and pH 7.8, were used

Incubation of denervated hemidiaphragms for a short period of time in the presence of low concentrationsof rhodamine B resulted in very selective labeling of small segments of the muscle fibers located near the center of the muscle (Figures 3a and 3b). This is similar to the distribution and appearance of uptake of HRP in denervated hemidiaphragm (Figure 3c) (see also Vult von Steyern et al., 1994; Libelius and TPgerud, 1984). We have routinely used rhodamine B at 50 nglml (about 100 nM) and 5-10 min incubation at RT (20-22°C). The same intensity of labeling was attained after 30-min incubation at 4°C. The labeling was resistant to washing, and the labeled segmentscould still be observed after prolonged washing at 4°C (48 hr). When muscles were incubated at 37°C for 30 min, a concentration of rhodamine B as low as 5 nglml (about 10 nM) was sufficient for clear labeling of the segments.


269

ACIDIC ORGANELLES IN DENERVATED MUSCLE

7n 900

E

.z 5000

RI

0

23

R2

R3

400-

g1 300-

E

H2

Rhodamine 123

200 -

HCH&H,

Rhodamine 6G

Tetramcthylrhodamic

0 560 575 590 605 620

Tetnrmcthylrosamiie

-

(CH3)2

c 3

coocH,CH3

cooH

H

CH,

H

n

(m)

Figure 1. Effect of pH on fluorescence emission spectra and peak emission for rhodamine B in saturated aqueous N-octanol. Rhodamine B was dissolved in a citric acid-phosphate buffer (1 pglml) at pH 3.0 or pH 7.8. After centrifugation, the octanol phase was collected and fluorescence was measured in a spectrofluorophotometer with excitation at 546 nm. Values along the X-axis represent emission wavelength and values along the Y-axis are arbitrary units of emitted fluorescence. Arrow indicates wavelength of the barrier filter in the G-2A filter. 590 nm.

To compare the structures labeled with rhodamine B and those labeled with a marker for endocytosis, mice were injected with 50-100 mg of FITC-dextran (average MW 3900). This substance is taken up in denervated muscle at sites showing high endocytotic activity (Elmquist et al., 1992) but is relatively quickly excreted from the animal owing to its small size, thus reducing extracellularfluorescence. One hour after injection the animals were sacrificed and the diaphragm was dissected and labeled with rhodamine B. Although this lipid-soluble dye does not require endocytosisfor internalization, the segments labeled with rhodamine B were the same that were labeled with the endocytosismarker (Figure 4). This indicates that the structures labeled with rhodamine B belong to an endosomalllysosomal compartment in the muscle fibers. Segments with high endocytotic activity have recently been described in long-term (more than 6 months) denervated m u d e (Vult von Steyern et al., 1994). These segments have the same morphological characteristicsas those described for short-term denervated muscle fibers, but they are found in more peripheral parts of the fibers and are not associated with any endplate structures. In this study, similar segments in muscle fibers denervated for 6-13 months were labeled by rhodamine B (Figure 5) and appeared in peripheral parts of the fibers.

Figure 2. Molecular structures of rhodamine B and substances with related structures. Note the carboxyl group at RP in rhodamineBand tetramethylrhodamine, which makes these substances overall uncharged, whereas the other moleculeseither lack this group or have it esterified and therefore carry a positive charge.

Effects of Ionophores and Weak Bases on Rhodamine B Labeling in Denervated Muscle The concept that the structures labeled with rhodamine B in the denervated muscle fibers are part of the endosomalllysosomal system was further investigated by studying the effect of monensin and nigericin on rhodamine B labeling. These substances disrupt the low pH of acidic compartments by acting as Na+lH+and K+/H+ ionophores, respectively (cf. Tipper and Sundler, 1990; Wileman et al., 1984). As shown in Figure 6, incubation in the presence of 50 W M monensin (added to the Ringer's solution from a 50" stock solution in ethanol) reversed labeling by rhodamine B. Removal of monensin allowed the segments to become labeled again when they were reincubated with rhodamine B (Figure 6). In addition, nigericin removed rhodamine B fluorescence from the muscle (results not shown), but the process was slow and although the muscle fibers appeared to be in good shape they could not be relabeled with rhodamine B after removal of the drug. Ammonium chloride (30 mM) or methylamine (50 mM) was also used to increase the pH of the aqueous compartment of acidic organelles. However, these treatments failed to prevent labeling by rhodamine B when they were added to the muscle fibers before rhodamine B, and they also failed to reverse labeling when they were added after rhodamine B (results not shown).


VULT VON STEYERN, JOSEFSSON.

270

TAGERUD

Figure 3. Rhodamine B labeling of segments of denervated muscle fibers (a) at low and (b)at higher magnification. For comparison and illustration of HRP-labeled segments in a denervated hemidiaphragm, c is included. Hemidiaphragms denervated for (a,b)20 days and (c) 18 days. For rhodamine B labeling the muscle was cooled in Ringer's solution at 4°C for 10 min and then incubated for 30 min at 4°C in the presence of 50 nglml rhodamine B. The muscle was then rinsed in three changes of Ringer's s o b tion at 4OC. each wash lasting for 30 min. The labeled segments occur near the center of the muscle and therefore in the vicinity of the denervated endplates. Bars = 200 prn. Figure 4. Comparison of structures labeled with (b,d)rhodamine Band those labeled with (a&) FITC-dextran in the same muscle. Hemidiaphragm denervated 22 days. One hour before sacrifice the mouse was injected with 50 mg FITC-dextran (average MW 3900) IV. After dissection, the diaphragm was incubated in the presence of 50 nglml rhodamine B for 10 min at room temperature. (a$) Localization of FlTC-dextran using B-2A filter. Note presence of dot-like fluorescence in connective tissue cells and fluorescence in the muscle fibers. (b,d)Localization of rhodamine Busing G-2A filter. Note fluorescence in muscle fibers but absence of fluorescence in connective tissue cells. Bar = 200 um.



Experiments with Lysosomotropic Dyes and Dyes Structurally Related to Rhodamine B The lysosomotropic dyes neutral red (300 ng/ml for 20 min at RT) and acridine orange (250-5000 ng/ml for 10-30 min at RT) also labeled small segments of the denervated muscle fibers near the center of the muscle, similar to those labeled by rhodamine B (Figure 7). However, the labeling with these dyes was less distinct than with rhodamine B, and background fluorescence was more disturbing. Double labeling experiments were performed with the lysosome probe Lyso Tracker Green and rhodamine B. Most segments labeled by rhodamine B were also labeled by Lyso Tracker Green (Figure 8), but the muscle fibers appeared to deteriorate after incubation with this dye, and the background stainingwasdisturbing. Several substances related in structure to rhodamine B (Figure 2) were also tested. Tetramethylrhodamineat 50-100 nglml for 5-10 min at RT gave labeling similar to that seen with rhodamine B. This labeling, however, appeared to be less resistant to washing and disappeared almost completely after a 60-min wash at RT. Rhodamine B hexyl ester (50-1000 ng/ml), rhodamine 6G (20-1000 nglml), rhodamine 123 (0.2-10 pg/ml), and tetramethylrosamine (50-1000 ng/ml) gave no selective labeling of the kind observed with rhodamine B in denervated muscle.

Labeling of Other Ceiis Mouse connectivetissue cells, MDCK cells, mouse peritoneal macrophages, and non-neural cells from embryonal rat spinal cord were also incubated with rhodamine B. Lysosomes in these cells were readily labeled by acridine orange and neutral red, but only macrophages were somewhat labeled by rhodamine B (results not shown). This labeling, however, occurred only at high concentrations and did not resemble the type of labeling seen in denervated muscle, i.e., it was less distinct and was not obviously related to endosomalllysosomal structures.

Discussion After denervation of skeletal muscle, intense endocytotic activity is accompanied by activation of the lysosomal system in the endplate region of muscle fibers (Elmquist et al., 1992; Tagerud et al., 1986a,b; Libelius and Tagerud, 1984; Tagerud and Libelius, 1984). After long-term denervation (more than 6 months), similar segments with high endocytotic activity occur in more peripheral parts of the muscle fibers (Vult von Steyern et al., 1994). Irrespective of short- or long-term denervation, and hence location, these segments of denervated muscle fibers are here shown to become labeled in vitro by rhodamine B. This dye, at a concentration as low as 5 nglml (about 10 nM), gave a most distinct and brilliant labeling, with very little background fluorescence, in other parts of the muscle. We believe that this labeling by rhodamine B is caused by a rapid accumulation of this substance in endosomesllysosomes in the denervated muscle fibers, and this is strongly supported by the results of this study. Rhodamine B co-localized with the endocytosis probe FITCdextran and therefore labeled structures belonging to the endosomalllysosomal system of the muscle fibers. The lysosomotropic dyes acridine orange and neutral red were also found to label simi-

27 1

lar segments, albeit less distinctly, in the same region of the muscle fibers, suggesting that rhodamine B labeled acidic structures. This was confirmed in experiments showing co-localization of rhodamine B and the lysosome probe Lyso Tracker Green. Unlike the lysosomotropic amines neutral red (pKa 6.7) (Merck Index, 1989) and acridine orange (pKa 9.8) (kber and Achtert, 1969), which are trapped in acidic aqueous compartmentsby protonation, rhodamine B is an overall uncharged, lipophilic compound with pKa values of 0.25 and 3.22 (Adamovich et al., 1979). The low pKa values suggest that the accumulation of the dye is not solely due to protonation in acidic compartments. However, compounds that inhibit acidification of lysosomes, such as monensin and nigericin, Na+/W and K+/H+ionophores, respectively ( c f . Tapper and Sundler, 1990; Wileman et al. 1984), discolored the segmentswhen they were added to labeled muscle fibers. This finding suggests that the accumulation and/or the intensity of fluorescence are dependent on acidic pH in the structures that are labeled by the dye. Despite this, we could not reverse labeling with either ammonium chloride or methylamine, which would have been expected if pH in the aqueous phase of the organelles was of main importance. It is possible that an interaction with rhodamine B in the lipid phase is necessary to remove the dye. An attractive explanation for the labeling observed with rhodamine B is that the dye accumulates in lipids in a lipid-aqueous interphase. This accumulation and/or the fluorescence properties of rhodamine B may be augmented by factors such as the polarity and water content of the lipids and the pH. The results of this study and earlier work have shown that rhodamine B is highly lipophilic (Araie and Maurice, 1987). However, in a nonpolar solvent an inner lactone is formed in rhodamine B, resulting in decolorization of the dye (Gutsze and Walerys, 1963; see also Drexhage, 1973). The present study shows that when rhodamine B is allowed to partition between a buffered aqueous phase and different organic phases, followed by fluorescence measurements in the organic phase, highest fluorescence is observed in polar organic solvents with high water content (results not shown). Therefore, fluorescence was higher in butanol than in octanol and was almost undetectable in heptane (watercontents 9.4 M, 2.3 M, and 2.3 mM, respectively) (Wolfenden and Radzicka, 1994). This illustrateshow the specific properties of the lipid phase can influence the accumulation and/or fluorescence properties of rhodamine B. A pH dependence of the absorption and fluorescence properties of rhodamine dyes that carry a free carboxyl group has been previously reported in ethanol and, to a much lesser extent, in water (see Drexhage, 1973). In the present study, a similar pH dependence of the fluorescenceproperties of rhodamine B was observed in saturated aqueous octanol. When excitation and emission wavelengths were chosen to mimic the conditions used for the morphological studies, this resulted in a 2.5fold increase in fluorescenceemission as the pH was decreased from 7.8 to 3.0. The same characteristics were seen with tetramethylrhodamine. Interestingly, only tetramethylrhodamine gave labeling of denervated muscle fibers resembling that of rhodamine B. The other compounds tested (tetramethylrosamine, rhodamine B hexyl ester, rhodamine 6G, and rhodamine 123) did not show a pHdependent increase in fluorescence in octanol and did not label acidic structures in denervated muscle fibers. These compounds differ from rhodamine B and tetramethylrhodamine in that they either lack the carboxyl group (tetramethylrosamine)or have the


WLT VON STEYER", JOSEFSSON, TAGERUD

272


273


carboxyl group esterified (rhodamine B hexyl ester, rhodamine 6G, and rhodamine 123). They therefore carry an overall positive charge, but despite this they appear to be able to cross cell membranes, because at least rhodamine 123 has been described as a very useful substance for labeling mitochondria in living cells (Johnson et al., 1980). The presence of the carboxyl group in rhodamine B and tetramethylrhodamine therefore seems to be critically important both for labeling of denervated muscle fibers and for pH-dependent fluorescence in saturated aqueous octanol. Although we have not examined in detail whether rhodamine B labels other mammalian cells in a manner similar to that seen in the denervated muscle fibers, we have reason to believe that there are variations among different cells concerning their susceptibility to labeling by the dye. When FITC-dextran is used in vivo as a marker for endocytosis, the labeled segments in the muscle fibers are intensely fluorescent, but connective tissue cells in the central tendon and in the pleural and peritoneal membranes that cover the diaphragm are also labeled (see Figure 4). When the segments in denervated muscle fibers are labeled with rhodamine B, however, no labeling of connective tissue cells is seen, suggesting that the dye is selective for a certain type of endosomes or lysosomes or that the dye does not readily permeate across the cell membranes of all cells. We also tried to label MDCK cells, mouse peritoneal macrophages, and non-neural cells from embryonal rat spinal cord with rhodamine B. Of these cells, only macrophages showed some labeling, but only at relatively high concentrations, and this labeling did not resemble that seen in the denervated muscle fibers. Preliminary results on human monocyte-derived macrophages showed a clear labeling of round bodies that may be of endosomal/lysosomal origin, but this occurred, as for the peritoneal macrophages, only at high concentrations and after long incubation periods (40-60 min). Studies of rhodamine B in cultures of free-living amebae have shown that Mayorella sp. is intensely labeled at very low concentrations of the dye after only a few minutes incubation, whereas Amebaprotear needs about a 1000-foldhigher concentration to become stained unless the glycocalyx is altered by absorption of polycations (Vult von Steyern and Josefsson, unpublished observations). These observations suggest that differences among cells regarding their basal lamina composition and possibly their cell membrane composition determine whether or not rhodamine B will diffuse into the cell.

After denervation, activities of acid hydrolases, including proteolytic enzymes such as cathepsin D, are increased in skeletal muscle, and this effect is most prominent in the endplate region (cf. Libelius and Tagerud, 1984). In addition, the activities of plasminogen activators (serine proteases) are increased in muscle after denervation (cf. Festoff et al., 1986), and locally generated plasmin has been reported to degrade the basement membrane of muscle (Hantai' and Festoff, 1987). Secretion of plasminogen activator and lysosomal enzymes from skeletal muscle is increased after denervation (Vult von Steyern and Josefsson, 1995). and this could conceivably lead to degradation of the extracellular matrix and the basal lamina, thereby perhaps increasing the permeability for rhodamine B. The results obtained in this study strongly suggest that rhodamine B is a very sensitive and efficient probe for organelles of endosomalllysosomal origin in denervated skeletal muscle fibers.However, further studies on other cells must be performed to determine why endosomesllysosomesin all cell types do not become labeled by rhodamine B.

Acknowledgments We thank Dr HAan Karlssonfor providing MDCK cells and Dr Roger Sund e r for providing mouse peritoneal macrophages.

Literature Cited Adamovich LP, Mel'nik W, Mchedlov-Petrosyan NO (1979): Equilibria of rhodamine B in aqueous salt solutions. Russ J Phys Chem 53:197 Allison AC, Young MR (1964): Uptake of dyes and drugs by living cells in culture. Life Sci 3:1407 Araie M, Maurice D (1987):The rate of diffusion of fluorophores through the cornea epithelium and stroma. Exp Eye Res 44:73 Bird JWC (1975): Skeletal muscle lysosomes. In Dingle JT, Dean RT, eds. Lysosomes in biology and pathology. Vol4. Amsterdam, North-Holland, 75 Drexhage KH (1973):Structure and properties of laser dyes. In Schifer FP, ed. Topics in applied physics: dye lasers. Vol 1. Berlin, Heidelberg, SpringerVerlag, 144 Elmquist S, Libelius R, Lawoko G, Tigerud S (1992): Dextrans as markers for endocytosis in innervated and denervated skeletal muscle. Muscle Nerve 15:876 Festoff BW, Hantai' D,SoriaJ, Thomaldis A, Soria C (1986): Plasminogen

Figure 5. Long-term (13 months) denervated hemidiaphragm with rhodamine B-labeled segments. Both segments shown occurred in very peripheral parts of the muscle. (b) The muscle has been cut free from the rib cage. The cut ends of the fibers are located close to the left edge of the photograph. Bar = 200 pm, Figure 6. Monensin reversibly disrupts rhodamine B labeling. Hemidiaphragm denervated 15 days. (a) Photograph taken after 5-min incubation with 50 n g h l rhodamine B at RT. (b) Same muscle after incubation with 50 WMmonensin for 30 min at RT. (c) Same muscle after 10-min wash followed by 5-min incubation in the presence of rhodamine 6 (50nglml) and 30-min wash in Ringer's solution, all at RT. All photographs were exposed for a similar period of time (a, 29 sec; b. 34 sec; c, 26 sec). Bar = 200 wm. Figure 7. Neutral red and acridine orange labeling of denervated muscle fibers. (a) Hemidiaphragmdenervated 24 days and incubated in the presence of neutral red (300 nglml, 20 min, RT). Photographed using a G-2A filter. (b) Hemidiaphragmdenervated 21 days and incubated in the presence of acridine orange (5 p g h l , 30 min, RT). Photographed using a G-2A filter. Bars = 100 pm. Figure 8. Segment of a hemidiaphragm denervated for 8 days and double labeled with rhodamine B and Lyso Tracker Green. The diaphragm was first incubated with Lyso Tracker Green (50 nM) for 40 min and then with rhodamine B (50nglml) for 10 min, both incubations FIT (a) Labeling by rhodamine B localized by using a G-2A filter. (b) Labeling by Lyso Tracker Green localized with a B-2A filter. Note correlation of labeled structures and the more distinct labelingby rhodamine B. The segment was photographed after only a 10-minwash after incubation in rhodamine B, and therefore background fluorescence was relatively high for the rhodamine B labeling. Bar = 100 pm.


VULT VON STEYERN, JOSEFSSON, TAGERUD

274

activator in mammalian skeletal muscle: characteristics of effect of denervation on urokinase-like and tissue activator. J Cell Biol 103:1415 Gutsze A, Walerys H (1963): Reversible decolorization of rhodamine B in solutions due to changes in p H and polarity of the medium. Acta Phys Polon 23:581 HantaID. Festoff BW (1987): Degradation of muscle basement membrane zone by locally generated plasmin. Exp Neural 95:44 Jette J-F, ZiomekE (1944): Determination of lipase activity by a rhodaminetriglyceride-agarose assay. Anal Biochem 219:256 Johnson LV, Walsh ML, Chen LB (1980): Localization of mitochondria in living cells with rhodaminel23. Proc Natl Acad Sci USA 77:990 Koenig H (1965): The staining of lysosomes by basic dyes. J Histochem Cytochem 13:20 Libelius R, Josefsson J-0,Lundquist I (1979): Endocytosis in chronically denervated mouse skeletal muscle. A biochemical and ultrastructural study with horseradish peroxidase. Neuroscience 4:283 Libelius R, Lundquist I. Templeton W, Thesleff S (1978): Intracellular uptake and degradation of extracellular tracers in mouse skeletal muscle in vitro: the effect of denervation. Neuroscience 3:641 Libelius R, Tigerud S (1984): Uptake of horseradish peroxidase in denervated skeletal muscle occurs primarily at the endplate region. J Neural Sci 66:273 Uber G, Achtert G (1969): On the complex formation of acridine dyes with DNA. VII. Dependence of the binding on the dye structure. Biopolymers 8:595 Marinetti GV (1962): Chromatographic separation, identification, and analysis of phosphatides. J Lipid Res 3:l

Merck (1989): Merck index. 11th ed. Rahway, NJ, Merck Index Tigerud S, Jirmanova I, Libelius R (1986a): Biochemical and ultrastructural effects of chloroquine on horseradish peroxidase uptake and lysosomal enzyme activities in innervated and denervated mouse skeletal muscle. J Neurol Sci 75:159 Tigerud S. Libelius R (1984): Lysosomes in skeletal muscle following denervation. Time course of horseradish peroxidase uptake and increase of lysosomal enzymes. Cell Tissue Res 236:73 Tigerud S, Libelius R, Thesleff S (1986b): An ultrastructural study of the segmental uptake of horseradish peroxidase in the endplate region of denervated skeletal muscle fibres. J Neural Sci 75:141 Tapper H, Sundler R (1990): Role of lysosomal and cytosolic pH in the regulation of macrophage lysosomal enzyme secretion. Biochem J 272:407 Vult van Steyern F, Josefsson J-0(1995): Secretion of plasminogen activator and lysosomal enzymes from mouse skeletal muscle: effect of denervation. J Cell Physiol 164:555 Vult van Steyern F, Libelius R, Lawoko G, Tigerud S (1994):Endocytotic activity of mouse skeletal muscle fibres after long-term denervation. J Neural Sci 125:147 Waser PG, Liithi U (1956): Autoradiography of end-plates with carbon-14calabash-curarine I and carbon-14-decamethonium. Nature 178:981 Wileman T, Boshans RL, Schlesinger P,Scahl P (1984): Monensin inhibits recycling of macrophage mannose-glycoprotein receptors and ligand delivery to lysosomes. Biochem J 220:665 Wolfenden R, Radzicka A (1994): On the probability of finding a water molecule in a nonpolar cavity. Science 265936
