Enhancement of the Cytotoxicity of Liposomal Ricin by the Carboxylic Ionophore Monensin and the Lysosomotropic Amine NH 4 Cl in Chinese Hamster Ovary Cells

Abstract

Ricin was encapsulated in negatively charged liposomes and its effect on the cytotoxicity was compared with native ricin in Chinese hamster ovarian (CHO) cells. The cytotoxicity of ricin, as measured by a marker protein synthesis (incorporation of ³H-leucine), was reduced markedly (300-fold) following encapsulation in liposomes. Lactose, a potent inhibitor of ricin cytotoxicity, had no effect on the binding, internalization, and cytotoxicity of liposomal ricin, indicating that liposomal ricin enter into mammalian cells by an alternative route, bypassing galactose-mediated endocytic pathway. Both monensin (a carboxylic ionophore) and NH₄Cl (a lysosomotropic amine) markedly enhances the cytotoxicity of liposomal ricin, indicating endocytotic uptake of liposomal ricin. The degree of potentiation of the cytotoxicity of liposomal ricin by both monensin and NH₄Cl was significantly higher (441- and 51-fold) as compared to native ricin (62.5- and 12.5-fold). The extent of exocytosis of free ricin was found to be much higher as compared to liposomal ricin; on the other hand, the extent of degradation of free and liposomal ricin was identical. Consequently, the intracellular level of liposomal ricin was increased to 3.5-fold. This higher level of intracellular liposomal ricin may allow more efficient ricin A-chain release into the cytosol under the influence of NH₄Cl and monensin. Monensin-induced potentiation of liposomal ricin was prevented by brefeldin A, indicating that in the presence of monensin, the liposomal ricin was efficiently routed through the Golgi apparatus en route to the cytosol. Thus, liposomal ricin in combination with monensin may have potential application for selective elimination of malignant cells.

Ricin is a heterodimeric cytotoxic protein consisting of disulphide-linked A- and B-chain (Olsnes and Phil 1982). It is known to inhibit the synthesis of protein in eukaryotic cells by inactivation of 60S ribosomal subunit due to the cleavage of specific adenine residue from 28S rRNA mediated by the N-glycosidase activity of the A-chain (Endo and Tsurugi 1987; Endo et al. 1987). The toxin, after binding to the galactose residues on cell surface through the B-chain, is internalized by endocytosis (Sandvig and van Deurs 1996). After endocytosis, the ricin is transported to the Golgi apparatus (Gonatas et al. 1975; and van Deurs et al. 1986) mainly to the trans-golgi network (TGN) (van Deurs et al. 1988) en route to endoplasmic reticulum (ER) (Wales, Roberts, and Lord 1993; Rapak, Falnes, and Olsnes 1997; Wesche, Rapak, and Olsnes 1999), before the release of the A-chain into the cytoplasm. However the potential application of ricin as an anticancer agent is limited due to the nonselectivity of the B-chain. Hence, attempts have been made to prepare immunotoxins (ITs) and hormonotoxins by cross linking the ricin A-chain with monoclonal antibodies (mAbs) directed against tumor cell–specific antigens (Thrush et al. 1996; Kreitman and Pastan 1998) and hormones whose receptors are overexpressed specifically on tumor cell surface (Singh et al. 1989). These heteroconjugates are tumor cell specific but are less efficacious due to the suboptimal entry of the A-chain into the cytoplasm, in the absence of facilitation afforded by the B-chain (Olsnes and Sandvig 1985).

Negatively charged and antibody-tagged liposomes have been reported to enter into mammalian cells by endocytosis and reach an acidic pH compartment (endosome) (Straubinger et al. 1983; Allen, Hansen, and Zalipsky, 1995). These liposomes have been employed to deliver ricin and ricin A-chain in tumor and ricin-resistant normal mammalian cells (Dimitriadis and Butters 1979; Gardas and Macpherson 1979; Wantanbe et al. 1989). However, the cytotoxicity of ricin is significantly reduced following its entrapment in liposomes (Dimitriadis and Butters 1979). The lysosomotropic agent NH₄Cl and the carboxylic ionophore monensin are known to enhance the cytotoxicity of ricin and ricin A-chain–based ITs (Ghosh et al. 1985; Ghosh and Wu 1988; Raso and Lawrence 1984; Casellas et al. 1984; Ramakrishnan, Bjorn, and Houston 1989; Carriere et al. 1985). It has also been reported that brefeldin A, a macrolide antibiotic that specifically inhibits transport of proteins from the Golgi apparatus to the endoplasmic reticulum, also significantly modulated the cytotoxicities of ricin and ricin-based ITs (Yoshida et al. 1991; Hudson and Grillo 1991; Sandvig et al. 1991). However, there has been no report on the effect of these agents on the cytotoxicity of liposomal ricin either under in vitro or in vivo condition. In the present study, we have compared the cytotoxicity of ricin in free and liposomal form and compared the effect of various agents such as monensin, NH₄Cl, and brefeldin A on the cytotoxicity of ricin in free and liposomal form in Chinese hamster ovarian (CHO) cells.

MATERIALS AND METHODS

Materials

Monensin, cholesterol, l-α-phosphatidic acid (PA), lactoperoxidase, RPMI-1640 (complete and deficient) medium, penicillin-G (potassium salt), streptomycin sulfate, and proline were obtained from Sigma Chemical (St. Louis, MO). Fetal calf serum (FCS) was obtained from Gibco BRL (Gaithersburg, MD). Brefeldin A (BFA) was obtained from Epicentre Technologies (Madison, WI). Soya phosphatidylcholine (SPC) was obtained from Natterman, Germany. Sepharose CL 6B, Sephacryl S-200, and Sephadex G-25 were obtained from Pharmacia Fine Chemicals (Uppsala, Sweden).

Ammonium chloride (NH₄Cl) was of Analar/Excelar grade from E. Merck Ltd. (Bombay, India). ³H-leucine (153 Ci/mmole) was obtained from NEN Research Products, (Boston, MA). Na¹²⁵I (7.2 mCi/μg iodine) was obtained from BARC (Bombay, India).

Cells

CHO Pro⁻ cell line is auxotrophic for proline was generously provided by Dr. Henry C. Wu (Uniformed Services University of the Health Sciences, Bethesda, Maryland). It was maintained in RPMI-1640 supplemented with 10% FCS, penicillin (100 U/ml), and streptomycin (100 μg/ml). These were grown in a humidified incubator at 37°C, 5% CO₂ and 95% air atmosphere.

Purification of Ricin

Ricin was purified from the seeds of Ricinus communis by affinity chromatography on cross-linked guargum column following the published procedure by Appukuttan, Surolia, and Bachhawat (1977), followed by gel permeation chromatography on Sephacryl S-200.

Radioiodination of Ricin

Ricin was radiolabelled with Na¹²⁵I by lactoperoxidase according to the method reported earlier (Madan and Ghosh 1992b).

Preparation of Liposomal Ricin

The lipids, soya phosphatidylcholine/cholesterol/phosphatidic acid (40 μmoles total lipids) in the molar ratio 49.5:40.5:10 were dissolved in chloroform in a 100-ml round bottom flask. The chloroform was evaporated to dryness at 37°C, under reduced pressure by using rotary evaporator (Wheaton). The thin film so formed was desiccated for 1 h, followed by hydration with 1 ml phosphate-buffered saline (PBS), 20 mM, pH 7.2, containing ricin (6 mg/ml) and trace amounts of ¹²⁵I-ricin as aqueous phase marker for overnight at 4°C. Next day, the liposomes were sonicated for 45 min at 30°C in a bath type sonicator (Branson). The liposomal ricin was separated from free ricin by affinity chromatography on Sepharose CL 6B column pre-equilibrated with PBS (20 mM, pH 7.2) at 25°C. Briefly, 1 ml liposomal suspension (40 μmole total lipid in 1 ml) was loaded on Sepharose CL 6B column (1 × 43 cm). The column was eluted first, with 1 bed volume of PBS, followed by, 1 bed volume of 0.1 M lactose (in PBS). The presence of liposomes in the fractions thus obtained was determined by checking the turbidity at 450 nm and measuring the radioactivity in each fraction in LKB 1275 mini gamma counter. It was observed that elution of the column with PBS gave only one peak corresponding to liposomal ricin as both the turbidity and the radioactivity was maximum in the peak (Figure 1). It was found that 1.5% of the total amount of ricin added during the preparation of liposomes was associated with the liposomal fraction. The rest, 98.5% of free ricin bound to the column, was subsequently eluted completely with 0.1 M lactose as shown in Figure 1. The mean diameter of the liposomes measured from the volume distribution curves produced by particle analyzer was found to be 120 nm.

Assessment of Cytotoxicity of Ricin in Free and Liposomal Forms In Vitro Using CHO Pro⁻Cells

The cytotoxicity of ricin in free and liposomal form was determined from the observed inhibition of ³H-leucine incorporation into protein in cell cultures exposed to various concentrations of the toxin. Cells were plated in 24-well plates at the cell density of 8 × 10⁵ cells/well in 1 ml RPMI-1640 medium containing 10% FCS, penicillin (100 U/ml), and streptomycin (100 μg/ml), 18 h prior to the experiment. The monolayer cultures were washed twice with 1 ml Dulbecco’s balanced salt solution (DBSS) and incubated in 0.9 ml RPMI-1640 medium containing penicillin and streptomycin for 1 h at 37°C. Cells were incubated with different concentrations of ricin in free and liposomal form for 4 h at 37°C and then washed with DBSS twice and incubated with ³H-leucine (0.5 μCi/ml) for 1 h at 37°C in a leucine-free medium. The monolayers were then fixed with 3% (w/v) perchloric acid and 0.5% (w/v) phosphotungstic acid twice, washed with DBSS twice, and dissolved in 0.5 ml of 0.5 N NaOH. A 50-μl aliquot of the solubilized cell extract was transferred to a scintillation vial containing 5 ml scintillation cocktail, neutralized with 25 μl 1 N HCl, and counted in an LKB 1209 Rack beta liquid scintillation counter. For the determination of the effects of monensin, lactose, NH₄Cl, and brefeldin A, cells were preincubated with these agents for 1 h, followed by their incubation with ricin in free or liposomal form as described above.

Binding and Internalization of ¹²⁵I-Ricin in Free and Liposomal Forms

Binding and internalization of ¹²⁵I-ricin in free or liposomal form was assayed at 4°C and 37°C, respectively, as a function of concentration of ¹²⁵I- ricin. The cells at a density of 8 × 10⁵ cells/well in 24-well plates were incubated with RPMI-1640 medium containing varying concentration of ¹²⁵I-ricin in free and liposomal (6 μg ¹²⁵I-ricin = 1 μmole phospholipid/well) forms at respective temperatures for 3 h. Cells were washed thrice with either 0.1 M lactose or cold DBSS and solubilized with 0.1 N NaOH. The cell associated radioactivity was determined with the help of 1275 LKB mini gamma counter.

Assay of Degradation and Exocytosis of Intracellular Ricin

The degradation and exocytosis of ¹²⁵I-ricin in free and liposomal form by CHO Pro⁻cells were studied at 37°C as a function of time. Cells at a density of 8 × 10⁵ cells/well in 24-well tissue culture plates were incubated with RPMI-1640 medium containing ¹²⁵I-ricin in free (200 ng ¹²⁵I-ricin/well) and liposomal (6 μg ¹²⁵I-ricin/μmole phospholipid/well) forms at 37°C for 2 h. After 2 h, the surface-bound ¹²⁵I-ricin was removed by washing the cells with 0.1 M lactose twice. Cells in duplicate wells were dissolved with 0.1 N NaOH to determine the amount of internalized ¹²⁵I-ricin at 2 h, i.e., the 0 time point. The cells in the remaining wells were further incubated with fresh RPMI-1640 medium containing 1 mM lactose for various time intervals (5, 15, 30, and 60 min) at 37°C. After different time intervals, the culture medium of each well was collected separately and the cells were washed twice with DBSS and then solubilized with 1 ml of 0.1 N NaOH. The amount of 10% trichloroacetic acid (TCA) soluble and precipitable radioactivity, released in culture medium and that associated with cells was determined as described earlier (Madan and Ghosh 1992a).

Kinetics of Inhibition of Protein Synthesis by Ricin in Free and Liposomal Forms

The cells at the density of 8 × 10⁵ cells/ml/well were plated in 24-well plates, in RPMI-1640 medium containing 10% FCS, penicillin (100 U/ml), and streptomycin (100 μg/ml), 18 h prior to the experiment. The monolayer cultures were washed twice with DBSS and incubated with leucine free RPMI-1640 medium containing ricin in free (100 ng/ml) or liposomal (10 μg/ml) form for various time intervals at 37°C. After different time intervals, the cells were washed twice with DBSS and incubated with leucine free medium containing ³H-leucine for 30 min. The inhibition of protein synthesis was measured as described earlier in Assessment of Cytotoxicity of Ricin in Free and Liposomal Forms In Vitro using CHO Pro⁻Cells.

RESULTS

Cytotoxicity of Liposomal Ricin in CHO Pro⁻Cells

The cytotoxic effect of ricin in free and liposomal form in CHO pro⁻ cells, as inferred from the incorporation of ³H-leucine into the protein, is shown in Figure 2. As can be seen from the Figure 2, the concentration of ricin in free and liposomal forms required for 50% inhibition of protein synthesis (ID₅₀) in CHO cells are found to be 50 and 15,000 ng/ml, respectively. This shows that cytotoxicity of ricin is significantly reduced (300-fold) following encapsulation in liposomes. In order to ascertain whether the cytotoxicity of liposomal ricin towards CHO pro⁻ cells is due to the ricin released from the aqueous compartment of liposomes following its binding with cell surface as reported earlier (Watanabe and Osawa 1989), the cytotoxicity of liposomal ricin was therefore checked in the presence of lactose (50 mM). Lactose is known to reduce the cytotoxicity of free ricin by binding to the galactose binding site of B-chain of ricin, thus inhibits the binding of free ricin on cell surface (Sandvig and van Deurs 1996). As shown in Figure 2, lactose had a marginal effect (1.2-fold increase in ID₅₀) on the cytotoxicity of liposomal ricin. On the other hand, it significantly reduced (150-fold increase in ID₅₀) the cytotoxicity of free ricin. This result indicates that liposome containing ricin enters into the cells in intact form.

Binding and Internalization of Free and Liposomal Ricin

In order to determine whether the reduction of cytotoxicity of liposomal ricin as compared to free ricin is due to the difference in binding and uptake, we examined the concentration dependent binding and uptake at 4°C and 37°C, respectively, of free and liposomal ricin. Ricin is known to bind and endocytose efficiently in CHO pro⁻ cells (Ghosh et al. 1985). As can be seen from the data in Figure 3, the binding of ¹²⁵I-ricin in free form is significantly higher as compared to liposomal ricin at 4°C. At 3 h, 15% of free ricin is found to be cell associated. Ninty percent of this bound ricin could be released by washing with 0.1 M lactose, indicating galactose-specific binding. On the other hand, at 4°C, only 0.3% of liposomal ricin associated with cells. The cell-associated liposomal ricin could not be released by washing with lactose. Further, the uptake studies of ¹²⁵I-ricin in free and liposomal form at 37°C reveal that the cell-associated radioactivity increased linearly with increasing concentration of ricin in free and liposomal forms. The data in Figure 3 show that at 37°C, 8% of free ricin added in the incubation mixture was associated with cells at 3 h, and 50% of this cell associated ricin could be released by washing with 0.1 M lactose. However 2% of liposomal ricin added in the incubation mixture was associated with cells at 3 h, which is nondissociable with 0.1 M lactose.

Effect of NH₄Cl on the Cytotoxicity of Liposomal Ricin

NH₄Cl, a lysosomotropic amine is known to elevate the intravesicular pH level (Ohkuma and Poole 1978) and enhances the cytotoxicities of ricin and ricin A-chain–based ITs (Ghosh et al. 1985; Ghosh and Wu 1988; Raso and Lawrence 1984; Casellas et al. 1984; Ramakrishnan, Bjorn, and Houston 1989; Carriere et al. 1985; Yoshida et al. 1991). The potentiation of cytotoxicities of ricin and ricin A-chain–based ITs by NH₄Cl has been attributed to altered endosomal/lysosomal pH, suggesting that a neutral or alkaline pH is optimal for ricin intoxication (Sandvig and Olsnes 1982). We have examined the modulatory influence exercised by NH₄Cl on the cytotoxicity of liposomal ricin in CHO cells. The results of the studies are presented in Figure 4A and B . NH₄Cl significantly enhanced the cytotoxicity of both free and liposomal ricin. A concentration of 20 mM NH₄Cl enhances the cytotoxicity of free and liposomal ricin by 12.85-and 51-fold, respectively. Our result for the first time shows that NH₄Cl markedly enhances the cytotoxicity of liposomal ricin.

Effect of Monensin on the Cytotoxicity of Liposomal Ricin

Various lines of evidences suggest that after endocytosis, ricin is transported from endosomes to TGN (van Deurs et al. 1988; Sandvig and van Deurs 1999) and retrograde transport from TGN to ER via Golgi stack is essential for its cytotoxicity (Wales, Roberts, and Lord 1993; Rapak, Falnes, and Olsnes 1997; Wesche, Rapak, and Olsnes 1999). Monensin, a carboxylic ionophore, is known to alter morphology of Golgi complex, and interrupts intracellular transport of macromolecules (Tartakoff, Vassalli, and Detraz 1978; Mollenhauer et al. 1990). As a result, the cytotoxicity of ricin and ricin A-chain– based ITs is significantly enhanced in cultured cells (Ghosh et al. 1985; Ghosh and Wu 1988; Raso and Lawrence 1984; Casellas et al. 1984; Ramakrishnan, Bjorn, and Houston 1989; Carriere et al. 1985). It has been reported that monensin-induced enhancement of cytotoxicity of ricin is due to enhanced transport of ricin from cell surface to the Golgi complex (Wesche, Rapak, and Olsnes 1999). In order to ascertain whether monensin affects intracellular transport of liposomal ricin and has any effect on the cytotoxicity of liposomal ricin, we have examined the modulatory influence of monensin on the cytotoxicity of liposomal ricin in CHO cells. As can be seen from Figure 4A and B , monensin significantly enhanced the cytotoxicity of both free and liposomal ricin. At 50 nM monensin enhances the cytotoxicity of free and liposomal ricin by 62.5- and 441-fold, respectively. At this concentration monensin has no effect on endosomal/lysosomal pH, but it still induces morphological changes in the Golgi complex (van Deurs et al. 1987), suggesting that the Golgi complex is involved in liposomal ricin intoxication process as native ricin.

Degradation and Exocytosis of Intracellular ¹²⁵I-Ricin

We have noted that the degree of enhancement of the cytotoxicity of liposomal ricin by both NH₄Cl and monensin is significantly higher (51- and 441-fold) as compared to free ricin (12.5- and 62.5-fold). In order to ascertain as to whether the differential enhancing potency of these agents on the cytotoxicities of liposomal and free ricin are due to variation in the degradation and exocytosis of intracellular ricin endocytosed in free and liposomal forms, the kinetics and the extent of degradation and release into extracellular medium of intracellular ¹²⁵I-ricin was examined. Figure 5A shows the rate of release of intracellular ¹²⁵I-ricin as well as the amount of released labeled materials, which is TCA soluble at 37°C. It can be seen that there is a rapid release of radioactivity during the first 30 min from cells containing intracellular ¹²⁵I-ricin either in free or liposomal form. During subsequent hours the release proceeds at a constant though slower rate. The appearance of degraded toxin from both free and liposomal form seems to occur at a constant rate throughout the experiment and it represents only a small fraction of the total intracellular material. The rate of degradation of both free and liposomal ricin was found to be identical (Figure 5A). On the other hand, the rate of exocytosis of free ricin is significantly higher as compared to liposomal ricin (Figure 5A). Within 60 min 71% of intracellular free ricin was released into the medium, in contrast, only 29.5% of liposomal ricin was released. It was also observed that 15% of the total free ¹²⁵I-ricin released was found to be in the TCA soluble form, i.e., in degraded form; on the other hand, 34.5% of the total liposomal ricin released was found in TCA soluble fraction (Figure 5A). Consequently, the intracellular level of liposomal ricin was found to be significantly higher (3.5-fold) as compared to free ricin (Figure 5B). We have also measured the effect of monensin (50 nm and 10 μm), NH₄Cl, and brefeldin A on the release, intracellular level, and degradation of ¹²⁵I-ricin. None of these agents had any measurable effect on the intracellular level and release of free and liposomal ¹²⁵I-ricin (Table 1). However, NH₄Cl and monensin (10 μm) significantly reduced the degradation of ¹²⁵I-ricin in both free and liposomal form (Table 2).

Effect of Brefeldin A on the Enhanced Cytotoxicity of Liposomal Ricin in Monensin and Ammonium Chloride–Treated Cells

It has been demonstrated that after endocytosis native ricin is transported from the endosome to trans-Golgi network (van Deurs et al. 1988) and retrograde transport from TGN to ER via trans-/mid-cis-cisternae is essential for its efficient release into cytosol and optimum expression of cytotoxicity (Wales, Roberts, and Lord 1993; Rapak, Falnes, and Olsnes 1997; Wesche, Rapak, and Olsnes 1999). Brefeldin A has been reported to inhibit vesicular transport from ER to Golgi (Doms, Russ, and Yewdell 1989). In the presence of BFA, the Golgi apparatus disintegrate and cis-/medial and trans-Golgi resident proteins redistribute into the ER (Lippincott-Schwartz et al. 1989, 1990). Consequently, it also blocks the retrograde vesicular transport of vesicle from the Golgi to ER without affecting the transport of endosome to TGN (Sandvig et al. 1991). We have noted that monensin significantly enhances the cytotoxicity of liposomal ricin, suggesting the involvement of the Golgi apparatus in the intoxication process of liposomal ricin–like native ricin. To examine further whether monensin affects the routing of liposomal ricin through the Golgi apparatus and retrograde transport from TGN to ER is essential for its cytotoxicity, the effect of BFA on monensin-induced potentiation of cytotoxicities of liposomal and native ricin was examined. The results presented in Table 3 indicate that BFA completely blocks the potentiation of ricin cytotoxicity by monensin. We have also noted that BFA inhibited the cytotoxicities of both liposomal and native ricin in absence of monensin. These results indicate that liposomal ricin routes through the Golgi apparatus like native ricin in the absence of monensin, albeit slowly. This result also indicates that monensin significantly stimulate transport of the liposomal ricin to the Golgi apparatus or to ER en route to cytosol, resulting in the enhancement of the cytotoxicity. We have also observed that in the presence of BFA, alkalinization of endosome by NH₄Cl in CHO cells does not facilitate the intoxication of CHO cells by ricin in free and liposomal forms (Table 3). These results suggest that an elevated pH in endosome per se is not sufficient for the translocation of ricin in free and liposomal form froms in neutral or alkaline endosome. Most likely, liposomal ricin may have to reach the Golgi apparatus for its cytotoxic action.

Effect of Monensin on the Kinetics of Inhibition of Protein Synthesis by Liposomal Ricin

We and various investigators have reported previously that the lag period (the time interval between the binding of toxin and onset of inhibition of protein synthesis) is shortened by treatment of cells with monensin and lysosomotropic amines (Ghosh et al. 1985; Ghosh and Wu 1988; Raso and Lawrence 1984; Casellas et al. 1984; Ramakrishnan, Bjorn, and Houston 1989; Carriere et al. 1985). To examine whether monensin reduces lag period of liposomal ricin intoxication, we have studied the kinetics of protein synthesis inhibition by ricin in free and liposomal forms in the presence of monensin. As shown in Figure 6, at 100 ng/ml free ricin, a lag period of 1 h is observed in CHO pro⁻ cells with T₅₀ (time required to achieve 50% reduction in protein synthesis) of 3 h. Monensin (50 nM) reduces the lag period of free ricin from 1 h to 30 min and T₅₀ to 1.6 h. However, at 10 μg/ml liposomal ricin, a lag period of 4 h, with T₅₀ 8 h, is observed. However, monensin reduces the lag period of liposomal ricin from 4 h to 30 min with T₅₀ 2.7 h. Hence, the extent of reduction of lag phase of liposomal ricin intoxication by monensin is significantly higher (4 times) as compared to free ricin. This result clearly implied that monensin is able to bring about an efficient and enhanced release of ricin A-chain from liposomal ricin located in an intracellular compartment into the cytosol, leading to rapid onset of inhibition of protein synthesis.

DISCUSSION

During the last two decades, extensive studies have been carried out to ascertain the possible use of liposomes (lipid vesicle) as a delivery vehicle for directing important hydrophilic as well as hydrophobic drugs and biologically active molecules both in vitro and in vivo (Gregoriadis and Florence 1993; Lasic 1998). It has also been demonstrated that liposome-entrapped material could be selectively directed to specific cells or tissue by attaching monoclonal antibodies, lectin, and carbohydrates on the liposomal surface (Allen, Hansen, and Zalipsky 1995; Ghosh, Bachhawat, and Surolia 1981; Ghosh, Das, and Bachhawat 1982; Spanger and Scherphof 1983). In this paper, we have investigated the possibility of using liposomes as a carrier for ricin and compared the cytotoxicities of ricin in free and liposomal forms in CHO cells. Our results showed that cytotoxicity of ricin is significantly reduced (300-fold) following its encapsulation in liposomes. This result support earlier observation that there is a marked reduction in the cytotoxicity of ricin following its encapsulation in liposomes (Dimitriadis and Butters 1979). Binding and internalization studies revealed that liposomal ricin is taken up by cells by an alternative route, bypassing galactose-mediated endocytic pathway, and there is a significant reduction in binding and uptake of ricin following its encapsulation into liposomes. This may be one of the possible basis for the reduction of cytotoxicity of liposomal ricin.

This report has demonstrated for the first time that the cytotoxicity of liposomal ricin is enhanced significantly by both NH₄Cl (20 mM) and monensin (50 nm). NH₄Cl and monensin (>1 μm) are known to inactivate lysosomal enzyme by raising intralysosomal pH, thereby preventing degradation of ricin and immunotoxins and resulting in the enhancement of the cytotoxicities of ricin and ricin A-chain–based immunotoxins (Ohkuma and Poole 1978; Raso et al. 1984; Casellas et al. 1984; Ramakrishnan, Bjorn, and Houston 1989). It has also been reported that in the presence of either of these substances, ITs accumulated in an enlarged endocytic vesicles and entry into the lysosomes was slowed down (Carriere et al. 1985; Manske, Buchsbaum, and Vallera 1989). The enhancement of cytotoxicity of liposomal ricin by NH₄Cl is not entirely unexpected, as it has been reported by several investigators that negatively charged liposomes enter into mammalian cells by endocytosis and liposomal contents encounter the acidic pH in the endosomes (Straubinger et al. 1983). It has also been demonstrated that NH₄Cl and chloroquine inhibit significantly the degradation of liposomal protein (Dijkastra, Galen, and Scherphof 1984). Similarly, our results showed that NH₄Cl inhibit significantly the degradation of liposomal ricin, suggesting that NH₄Cl either slowed down the transfer of liposomal ricin to lysosomes or prevents its degradation by inactivating the lysosomal enzymes, possibly by increasing the intralysosomal pH. This may be the basis for the enhancement of the cytotoxicity of liposomal ricin by NH₄Cl. We have also noted that lactose, a potent inhibitor of cytotoxicity of ricin, has no effect on the cytotoxicity of liposomal ricin. All these results indicate that liposomal ricin enter into mammalian cells by endocytosis and encounter acidic pH in the endosomes; however, neutral or alkaline pH is optimal for its release into the cytosol.

A variety of data suggest that ricin and ricin-based ITs route through the Golgi apparatus on the way to cytosol. A number of investigators have observed ricin and ricin-based ITs in the Golgi (Gontas et al. 1975; van Deurs et al. 1986, 1987, 1988; Youle and Colombatti 1987). Most convincing evidence of Golgi involvement has come from the use of potentiator of IT and ricin toxicities. The carboxylic ionophore monensin interrupts intracellular traffic at a concentration too low to affect endosomal or lysosomal pH (Tartakoff et al. 1978; Tartakoff 1983; Mollenhauer et al. 1990) and enhance the cytotoxicities of ricin and ricin-based ITs (Ghosh et al. 1985; Ghosh and Wu 1988; Raso and Lawrence 1984; Casellas et al. 1984; Ramakrishnan, Bjorn, and Houston 1989; Carriere et al. 1985). The site of intracellular arrest has been studied by several methods and in all cases the principle site of arrest is the Golgi complex. This makes monensin unique in its mechanism of action.

To explore the possibility of a Golgi role in the intoxication process of liposomal ricin, we have examined the modulatory influence of monensin on the cytotoxicity of liposomal ricin. We observed a dramatic enhancement of cytotoxicity of liposomal ricin by monensin (50 nm), suggesting that the Golgi apparatus is involved in liposomal ricin intoxication process as native ricin. It appears therefore, that transport of liposomal ricin from endosome to the Golgi apparatus is essential for its release into the cytosol and optimal expression of its cytotoxicity. Native ricin contains a B-chain that is thought to use a galactose-binding function, intracellularly, to route the toxin through the Golgi apparatus to reach the cytosol (Newton et al. 1995). Retrograde transport of ricin from TGN via trans-/medial/cis-Golgi to ER is essential for its release into the cytosol (Wales, Roberts, and Lord 1993; Rapak, 1997; Wesche, Rapak, and Olsnes 1999). The mechanism by which liposomal ricin routes through the Golgi apparatus is not obvious; however, role of ricin B chain cannot be ruled out.

The sensitization of cells to liposomal ricin by monensin is completely blocked by BFA. The block of the monensin potentiation by BFA suggests that monensin stimulates transport of the liposomal ricin through the cis-Golgi possibly to the ER, en route to the cytosol. These results are consistent with earlier report that monensin stimulate the transport of ricin and immunotoxins to the Golgi apparatus (Wesche, Rapak, and Olsnes 1999; Manske, Buchsbaum, and Vallera 1989). Similarly, Youle and Colombatti (1987) have shown that in the presence of retinoic acid immunotoxins and ribonucleases are efficiently routed through the Golgi apparatus (Wu et al. 1994, 1995). In the presence of BFA, the Golgi apparatus disintegrate and cis-/medial/and trans-Golgi resident proteins redistribute into the ER (Lippincott-Schwartz et al. 1989, 1990). Consequently, it blocks the retrograde vesicular transport of vesicle from the Golgi to ER without affecting the transport of endosome to TGN (Sandvig et al. 1991). It appears, therefore, that BFA blocks the retrograde transport of liposomal ricin from TGN to ER via Golgi cisternae en route to the cytosol, leading to inhibition of cytotoxicity of liposomal ricin.

Studies on the degradation and exocytosis of intracellular ¹²⁵I-ricin revealed that there is a rapid release of radioactivity during the first 30 min from cells containing intracellular ¹²⁵I-ricin either in free or liposomal form. The release then proceeds at a constant, but a slower, rate over the next few hours. The explanation for the initial rapid release is not obvious. However, a similar initial rapid loss of radioactivity from intracellular ¹²⁵I-ricin has also been reported from HeLa cells (Sandvig and Olsnes 1979). However, there has been no report on the extent of exocytosis of intracellular liposomal ricin. This report for the first time showed that the extent of exocytosis of liposomal ricin is much lower as compared to free ricin. On the other hand, the appearance of degraded toxin from both free and liposomal forms seems to occur at a constant rate throughout the experiment. The rate of degradation of both free and liposomal ricin was identical and very slow. The much lower degradation rates for ricin in free and liposomal form are consistent with the high resistance of ricin to several proteases (Sandvig and Olsnes 1979). Consequently, the intracellular level of liposomal ricin was found to be significantly higher (3.5-fold) as compared to free ricin. The observed higher intracellular level of liposomal ricin for a longer period of time as compared to free ricin may allow more efficient ricin A-chain release from liposomal ricin located in an intracellular compartment into the cytosol under the influence of NH₄Cl and monensin. This may explain why NH₄Cl and monensin brings about a higher degree of enhancement of cytotoxicity of liposomal ricin. The mechanism of differential release of intracellular ¹²⁵I-ricin in free and liposomal forms is yet to be elucidated.

A significant reduction in the lag period of liposomal ricin action by monensin suggests a more rapid and efficient release of ricin from intracellular compartment into the cytosol. This result is consistent with our earlier observation (Ghosh et al. 1985; Ghosh and Wu 1988; and Madan and Ghosh 1992a) and those reported by other investigators (Raso and Lawrence 1984; Casellas et al. 1984; Ramakrishnan, Bjorn, and Houston 1989; Carriere et al. 1985) that pretreatment of cells with carboxylic ionophore results in a reduction of lag period in the inhibition of protein synthesis by ricin and immunotoxins.

Liposomes as a delivery vehicle for ricin have an advantage over monoclonal antibody. For example, it is constituted from natural lipids, which are nonimmunogenic, and its structure could be altered with variety of ligands for malignant cell specific delivery. Furthermore, ricin in the aqueous compartment of liposomes is shielded from cytotoxicity and adverse immunological reaction in the circulation. The present study has added to our contention that by suitable alteration of liposomal lipid composition and tailoring of liposomal surface with appropriate ligand, it would be possible to direct liposomal ricin to specific tissues or cells for their selective elimination in combination with monensin. However, the hydrophobic nature of monensin has made it difficult to administer it in optimum doses under in vivo conditions in order to realize its full potential in the enhancement of toxicities of ricin and immunotoxin. We have shown for the first time that liposome can be used as a delivery vehicle for monensin under both in vitro and in vivo conditions (Madan and Ghosh 1992a; Vasandani, Madan, Ghosh 1992). Subsequently, it has been reported that liposomal monensin is very effective in the treatment of tumor in combination with immunotoxin (Griffin et al. 1993; Singh et al. 1994). Thus, liposomal ricin in combination with liposomal monensin may have potential application for selective elimination of malignant cells.