Influence of cytotoxicity enhancers in combination with human ... - Nature

Leukemia (1999) 13, 241–249  1999 Stockton Press All rights reserved 0887-6924/99 $12.00 http://www.stockton-press.co.uk/leu

Influence of cytotoxicity enhancers in combination with human serum on the activity of CD22-recombinant ricin A against B cell lines, chronic and acute lymphocytic leukemia cells PJ van Horssen2, YVJM van Oosterhout2, S Evers2, HHJ Backus2, MGCT van Oijen2, R Bongaerts2, T de Witte1,2 and FWMB Preijers2 1

Department of Hematology and 2Central Hematology Laboratory, University Hospital St Radboud, Nijmegen, The Netherlands

Despite the strong in vitro activity of some immunotoxins (ITs), clinical application did not result in complete cure. The outcome of therapy may be improved by combining ITs with ITcytotoxicity enhancing agents. We studied the effect of various agents that influence the intracellular routing of ITs on the activity of the anti-B cell IT CD22-recombinant (rec) ricin A. In protein synthesis inhibition assays the carboxylic ionophores monensin and nigericin enhanced the activity of the IT 117- and 382-fold, respectively, against the cell line Daudi, and 81- and 318-fold, respectively, against the cell line Ramos. IT activity to Daudi and Ramos was enhanced to a lesser extent by the lysosomotropic amines chloroquine (14- and 11-fold, respectively) and NH4Cl (nine- and 10-fold, respectively). However, the combination of NH4Cl and chloroquine induced more than an additive effect (145- and 107-fold, respectively). Cytotoxicity was not influenced by brefeldin A, all-trans retinoic acid (ATRA), verapamil and perhexiline maleate. Bacitracin enhanced the IT cytotoxicity in contrast to the other protease inhibitors aprotinin, leupeptin and soybean trypsin inhibitor, albeit enhancement was weak (two-fold). The enhancers exerted only a negligible effect on bone marrow progenitor cells. We recently developed a flow cytometric cytotoxicity assay in which cell elimination can be assessed. In order to detect enhancement in this assay, we used 5 × 10−11 M IT (approximately the 50% protein synthesis inhibiting dose (ID50)). This concentration killed 41% of the Daudi cells and 42% of the Ramos cells. In the presence of 10 nM monensin the IT killed 74% and 99% and in the presence of 10 nM nigericin 96% and 99% of the Daudi and Ramos cells, respectively. At 10−8 M, CD22-rec ricin A eliminated malignant cells originating from three patients with B-CLL (0.42 log) and two with B-ALL (0.19 log) patients. Cytotoxicity to malignant cells was enhanced by NH4Cl, chloroquine, monensin and nigericin. The combination of NH4Cl and chloroquine enhanced the activity most effectively (up to 2.06 log). To determine the applicability of the IT in combination with enhancers in vivo we investigated the effect of human serum. Human serum inhibited IT activity which could not be restored by monensin and nigericin because of complete inhibition of these enhancers by serum. In contrast, chloroquine partially restored the activity of CD22rec ricin A in the presence of human serum. We conclude that monensin, nigericin and the combination of NH4Cl and chloroquine can be used instead of NH4Cl to potentiate CD22-rec ricin A activity in purging autologous bone marrow transplants contaminated with malignant B cells. Chloroquine might be a promising enhancer of CD22-rec ricin A for treating patients in vivo. Keywords: CD22-rec ricin A; human serum; cytotoxicity enhancers

Introduction Hematological malignancies are difficult to cure by chemotherapy, radiotherapy and/or surgery. Bone marrow transplantation has improved the outcome of treatment in many

Correspondence: PJ van Horssen, Department of Hematology and Central Hematology Laboratory, University Hospital St Radboud, PO Box 9101, 6500 HB Nijmegen, The Netherlands; Fax: 31 24 354 2080 Received 4 March 1998; accepted 19 October 1998

patients but relapses occur frequently in high risk patients. IT therapy seems an excellent additional treatment in view of the potent and specific killing of malignant cells in vitro as well as in vivo.1,2 ITs are constructed by linking chemically or genetically a cell binding moiety to a toxin. The commonly used toxins belong to the ribosome-inactivating proteins3,4 and monoclonal antibodies (MoAbs) are usually used as targeting moiety. CD22 MoAbs bind to the B-lymphocyte cell adhesion molecule (Bl-CAM) that is expressed on the surface of B cells from the pro-B cell stage to the mature B cell stage.5 Upon binding, the Bl-CAM/IT complex is rapidly internalized, whereafter the antigen is re-expressed.6,7 This makes Bl- CAM an optimal target for ITs. Several investigators have shown that CD22 ITs have a strong anti-B cell potency in vitro as well as in vivo.6,8–14 These findings resulted in clinical phase I/II studies.15–18 Though achieving some complete responses, treatment of these patients did not result in complete cure. Combining IT with enhancing agents may improve efficacy. Usually, enhancers influence the intracellular processing of the IT. There are several groups of IT enhancers such as carboxylic ionophores, calcium channel antagonists, lysosomotropic amines, calmodulin antagonists and protease inhibitors.19 Enhancement by a particular agent may depend on the toxin,20 on the target cell and on the intracellular routing of the IT which is, in turn, influenced by the antigen and the MoAb21 that is used. Since ITs are to be used in vitro as well as in vivo, we investigated enhancers of CD22-rec ricin A cytotoxicity suitable for both applications. The lysosomotropic amines NH4Cl and chloroquine inhibit intracellular IT degradation by neutralizing the endosomal and lysosomal pH, which results in postponement of activation of lysosomal proteases.22 NH4Cl also induces a strong increase of newly synthesized lysosomal proteases released from the cell in proportion to the amount delivered to the lysosomes.23 NH4Cl is toxic to patients and can therefore only be used to potentiate IT activity ex vivo.24 Chloroquine has been used successfully to treat rheumatoid arthritis and various parasitic infections such as malaria. Chloroquine has been administered in combination with an IT for treating a B-CLL patient.25 A 300 mg intramuscular dose of chloroquine administered on day 2 and 3 resulted in a peak level of 1 ␮m on day 3. No conclusions could be drawn from this study due to the differences in IT dosages between the patients treated with the combination of IT and chloroquine and IT alone.25 The carboxylic ionophores, monensin and nigericin, have been studied extensively.26–30 These agents also neutralize the lysosomal pH but at higher concentrations than required for IT potentiation.31 This indicates that the mechanism of enhancement by carboxylic ionophores differs from that by lysosomotropic amines.31,32 Monensin induces distortion and swelling of endosomes and the Golgi and multivesicular bodies22 which may counteract the release of IT from the cell.

Enhancement of CD22-rec ricin A PJ van Horssen et al

242

The protease inhibitors bacitracin, aprotinin, leupeptin and soybean trypsin inhibitor may enhance IT activity by reducing lysosomal IT degradation. The calcium channel blocker verapamil has been shown to enhance pseudomonas exotoxin, as well as ricin A IT activity.33 However, this drug can only be used in vitro due to cardiac toxicity induced by the effective concentrations of 2–20 ␮g/ml.34,35 All-trans retinoic acid (ATRA) enhances IT activity by disrupting the Golgi apparatus and increasing the cytosolic routing of proteins.36 The fungal metabolite brefeldin A influences the intracellular routing of internalized molecules and can enhance or block ricin A IT intoxication depending on the concentration used.37–39 Thus it is unsuited for therapy. In this study, we investigated the enhancement of the cytotoxicity of the anti-B cell IT CD22-rec ricin A using cell lines and B-CLL and B-ALL cells originating from patients. The aim was to select an agent that can be used in combination with the IT to purge bone marrow ex vivo, as well as treat patients suffering from B cell lymphoproliferative diseases. Serum has been described to inhibit the IT enhancing activity of monensin.30,40 Therefore the effect of the enhancers in combination with serum was assessed. Materials and methods

Enhancers Chloroquine (diphosphate salt), NH4Cl (cell culture), monensin, nigericin, leupeptin, perhexiline maleate, bacitracin, aprotinin, soybean trypsin inhibitor and brefeldin A were obtained from Sigma Chemical (St Louis, MO, USA). All-trans retinoic acid was generously provided by Hoffmann-La Roche (Basel, Switzerland). Verapamil was obtained from Knoll (Amsterdam, The Netherlands).

Cell lines The B-lineage cell lines Daudi and Ramos (Burkitt lymphomas) were maintained in the log phase by culturing them in medium consisting of RPMI 1640 (Flow Labs, Irvine, UK) supplemented with 10% heat-inactivated fetal calf serum (Integro, Zaandam, The Netherlands), 2 mm glutamine, 50 U/ml penicillin and 50 ␮g/ml streptomycin (Flow Labs) in a humidified incubator with 5% CO2 in air at 37°C. Cells were maintained in log phase of cell growth.

were enriched from peripheral blood and bone marrow by Ficoll–Paque (Pharmacia Biotech, Uppsala, Sweden) centrifugation and cryopreserved with 10% dimethylsulphoxide (DMSO) (Merck, Darmstadt, Germany).

Monoclonal antibody The murine MoAb CD22 (CLB-B-ly/1; IgGl) was purchased from the CLB (Amsterdam, The Netherlands). CD22 was produced in a hollow fiber bioreactor and was purified from the supernatant by affinity chromatography using staphylococcal protein A-Sepharose (Pharmacia).

CD22 binding assay Ramos cells (105) were washed and incubated with various concentrations of CD22 (5 × 0−12 to 1 × 10−7 m) in 100 ␮l 0, 50 or 100% human serum in PBS for 30 min at 4°C. After washing the cells were incubated with 5 ␮l goat-anti-mouse IgG(H+L)F(ab⬘)2-FITC (American Qualex, San Clemente, CA, USA) (prediluted 1:50) in 100 ␮l 20% human serum for 30 min at 4°C. Samples were washed twice and analyzed by flow cytometry (Coulter Epics Elite; Coulter, Hialeah, FL, USA) for relative fluorescence intensity.

Immunotoxin preparation CD22 was coupled to rec ricin A chain (generously provided by Zeneca Pharmaceuticals, Macclesfield, UK) by means of 4-succinimidyloxylcarbonyl-␣-methyl-␣-(2-pyridyldithio)toluene (SMPT; synthesized by the Department of Organic Chemistry, University of Nijmegen, The Netherlands) using the method of Thorpe et al.41 Briefly, 10 mg CD22 in 1 ml 25 mm sodium borate (pH 9) was treated with a 10-fold molar excess of SMPT (110 ␮l of 1 mg/ml SMPT solved in N,N-dimethylformamide (DMF, Merck)) for 1 h at 20°C. Free SMPT and DMF were removed by G25 gel filtration. To introduce a free sulfhydryl group ricin A was incubated in 50 mm DTT for 30 min at 20°C. DTT was then removed by gel filtration on the G25 column. The modified MoAb was incubated with a 2.5 molar excess ricin A for 72 h at 20°C resulting in a coupling ratio (ricin A:MoAb) of 1.1. Free ricin A was separated from the IT by gel filtration on a Sephacryl S200 High Resolution (Pharmacia) column.

Patient samples Samples originated from one patient with T-acute lymphoblastic leukemia (ALL), two with B-ALL and three with B-chronic lymphocytic leukemia (B-CLL) (Table 1). Mononuclear cells Table 1

Patient no. 1 2 3 4 5 6

Patient samples

Disease

CD22+ cells (%)

Sample

B-CLL B-CLL B-CLL B-ALL B-ALL T-ALL

79 74 84 81 93 ND

PB PB PB BM BM PB

PB, peripheral blood; BM, bone marrow; ND, not determined.

Protein synthesis inhibiting activity of CD22-rec ricin A Cells (105) in culture medium were incubated with various concentrations ranging from 10−13 to 10−8 m CD22-rec ricin A in the presence or the absence of varying concentrations of potential enhancers in a final volume of 200 ␮l. Before determining the potentiating activity of the enhancers, we incubated cells with a range of concentrations of each enhancer to define the maximum concentration in vitro, ie the concentration that did not inhibit protein synthesis for more than 30%. Protein synthesis inhibiting activity of CD22-rec ricin A was also assessed in the presence of pooled human serum collected from healthy donors. Cells were incubated in 96-well microtiter plates (Costar, Cambridge, MA, USA) in triplicate for 24 h at 37°C. Subsequently, 0.5 ␮Ci 3H-leucine


(Amersham, Buckinghamshire, UK) was added. After a further 24 h, cells were harvested and incorporated radioactivity was determined. Cytotoxicity was expressed as the percentage 3Hleucine incorporation inhibition of untreated cells corrected for non-specific incorporation determined by incubating cells in the presence of 1 mm cycloheximide (Boehringer, Mannheim, Germany).

Cell killing of cell lines by CD22-rec ricin A Cell killing was determined according to a method previously described.42 Briefly, cells were treated with 5 × 10−11 m CD22ricin A for 24 h in the absence or presence of varying concentrations of potential enhancers. The cells were then washed and incubated in culture medium for 3 days. Propidium iodide (PI; 2 ␮g/ml) was added to stain dead cells and calcein (2 ␮g/ml) to stain living cells 45 min before flow cytometric analysis (Coulter Epics Elite). Inert beads (5 × 104/ml, DNACheck, Coulter) were added to enable the quantification of cell killing. The % of killed cells was calculated using the formula: ((number of viable cells in control sample per 1000 beads − number of viable cells in treated sample per 1000 beads)/number of viable cells in control sample per 1000 beads) × 100%.

Cell killing of B-CLL and B-ALL cells by CD22-rec ricin A Patient cells (Table 1) were treated for 24 h with 10−8 m CD22ricin A in the absence or the presence of varying concentrations of potential enhancers. Thereafter the cells were washed and incubated in culture medium for 3 days. Samples were stained before flow cytometric analysis with combinations of CD3-FITC (Immunotech, Marseille, France), CD5FITC (Dako, Glostrup, Denmark), CD10-PE (Immuno Quality Products, Groningen, The Netherlands), CD19-PE+ (Dako), CD22-PE+ (Dako) and CD45-FITC (Immunotech) to allow detection of depletion of malignant, T and B cells, B-CLL was defined as (CD5+, CD19+, CD22+, Ig+), B-ALL as (CD19+, CD22+, CD45+−, CD10−, Ig+), and T cells as CD3+. Propidium iodide (2 mg/ml) was added 45 min before flow cytometric analysis to stain dead cells. Inert beads (5 × 104/ml) were added to enable the quantification of cell killing. The number of living cells per 1000 beads, that were counted simultaneously with the cells of a treated sample was compared with the number of living cells per 1000 beads in the control sample in order to calculate the factor of depletion. The log kill was calculated using the formula: log (number of viable cells in treated sample per 1000 beads/number of viable cells in control sample per 1000 beads).

Cytotoxic activity of CD22-rec ricin A to Daudi and Ramos cells The cytotoxic activity of CD22-rec ricin A to the B cell lines Daudi and Ramos as determined in the protein synthesis inhibition assay is expressed by ID50 values for Daudi and Ramos cells were 3.0 × 10−11 m and 5.2 × 10−11 m (data not shown). Enhancement of the activity of CD22-rec ricin A determined in the protein synthesis inhibition assay is expressed by the factor of enhancement (see Materials and methods). In order to verify whether protein synthesis inhibition is equal to cell elimination and to quantify this elimination we investigated viability after treatment with 5 × 10−11 m IT, approximately equivalent to the ID50 value, in a recently developed quantitative cell killing assay. This IT concentration killed 41% of the Daudi and 42% of the Ramos cells (Table 2, Figure 1).

Influence of calcium-channel antagonists on CD22-rec ricin A cytotoxicity The maximum concentration of verapamil (2 ␮m) tolerated in vivo did not enhance CD22-recombinant ricin A. Verapamil concentrations of 5 mm and higher were too toxic for Daudi and Ramos and were not tested for enhancement. Perhexiline maleate (50–200 ␮m) also did not enhance the activity of CD22-rec ricin A (data not shown).

Influence of brefeldin A and ATRA on CD22-rec ricin A activity Brefeldin A was assessed to enhance CD22-rec ricin A in concentrations ranging from 1 ng/ml to 1 ␮g/ml. Concentrations of up to 100 ng/ml did not enhance IT activity while concentrations of 100 ng/ml and higher as such reduced protein synthesis in Daudi and Ramos cells to less than 10%. Up to 50 ␮m ATRA did not enhance CD22-rec ricin A activity (data not shown).

Effect of protease inhibitors on CD22-rec ricin A cytotoxicity The protease inhibitors bacitracin, leupeptin, aprotinin and soybean trypsin inhibitor were investigated for CD22-rec ricin Table 2 Cell killing by 5 × 1011 m CD22-rec ricin A in the absence and in the presence of enhancers

Toxicity to bone marrow progenitor cells Bone marrow was aspirated from healthy donors after informed consent was obtained. Erythrocytes and mature granulocytes were removed by gradient centrifugation on Ficoll (1.077 g/ml). Cells (106/ml) were incubated with IT, monensin, nigericin, NH4Cl or chloroquine for 24 h and then washed. The decrease of progenitor cells was determined in clonogenic assays of granulocyte/macrophage colony-forming cells (CFU-GM) and erythroid burst-forming cells (BFU-E) as described previously.43

243

Results

5 × 10−11 M IT 10 nM monensin 10 nM nigericin IT + 5 nM monensin IT + 10 nM monensin IT + 5 nM nigericin IT + 10 nM nigericin

Daudi (% of killed cells)

Ramos (% of killed cells)

41 13 14 45 74 86 96

42 15 19 75 99 98 99

The percentage of viable cells was determined by the quantitative cell killing assay. Assays were repeated twice. SDs were less than 11%.


244

Figure 1 Cytotoxic activity of CD22-rec ricin A against Ramos determined in the flow cytometric cell killing assay. The number of living cells (calcein+ PI−, window B4) per 1000 simultaneously analyzed beads (window D) was determined for non-treated cells (a), IT-treated (5 × 10−11 M) cells (b) and cells treated with IT in combination with 10 nM nigericin (c).

A cytotoxicity enhancing activity. The IT activity was enhanced two-fold by 500 ␮m bacitracin (Table 3) whereas higher concentrations (1.0 mm) were toxic. Aprotinin (1– 100 ␮m), leupeptin (1–200 ␮m) and soybean trypsin inhibitor (1–100 ␮m) had no effect on CD22-rec ricin A activity.

Enhancement of CD22-rec ricin A cytotoxicity by carboxylic ionophores The potency of monensin and nigericin to enhance CD22-rec ricin A activity was studied at concentrations of 5 and 10 nm. These concentrations were not toxic in the protein synthesis inhibition assay. Concentrations of monensin and nigericin of 25 nm appeared to be toxic in bone marrow progenitor clonogenic assays particularly to the erythroid progenitors (Table 4). In the quantitative cell killing assay 10 nm monensin and nigericin reduced the viability by up to 7% of Daudi and Ramos cells (data not shown). Nigericin potentiated the IT cytotoxicity more effectively than monensin (Table 3). At 10 nm, monensin enhanced the IT activity as determined by a reduction of the ID50 in the

protein synthesis inhibition assay, 117- and 81-fold in Daudi and Ramos cells, respectively. Ten n m nigericin enhanced the IT activity 382-fold in Daudi cells and 318-fold in Ramos cells, respectively. We also studied enhancement of elimination of target cells (Figure 1). In agreement with the results of the protein synthesis inhibition assay we observed in the quantitative cell killing assay that nigericin potentiated stronger the cytotoxicity of CD22-rec ricin A towards Daudi and Ramos cells than monensin (Table 2). At 5 × 10−1 m, CD22-rec ricin A killed 41% of the Daudi cells and 42% of the Ramos cells (Figure 1). In the presence of 10 nm monensin, the IT killed 74% of the Daudi cells and 99% of the Ramos cells and when exposed to the same concentration of nigericin, the IT killed 96% of the Daudi cells and 99% of the Ramos cells (Table 2). Human serum blocks the enhancing activity of monensin.30,40 We confirmed that enhancement by both monensin (Figure 2) and nigericin (data not shown) was completely blocked when the assay was performed in 50% human serum. We found that the activity of CD22-rec ricin A was also impaired by human serum. The ID50 of CD22-rec ricin A for Ramos was reduced 11-fold in the presence of 50% human


Table 3

245

Enhancement of the cytotoxicity of the immunotoxin CD22-rec ricin A

Enhancer

Daudi Factor of enhancement

2 ␮M verapamil 200 ␮M perhexiline maleate 25 ng/ml brefeldin A 50 ␮M ATRA 500 ␮M bacitracin 100 ␮M aprotinin 200 ␮M leupeptin 100 ␮M soybean trypsin inhibitor 6 mM NH4Cl 5 ␮M chloroquine 20 ␮M chloroquine 50 ␮M chloroquine 5 ␮M chloroquine + 6 mM NH4Cl 20 mM chloroquine + 6 mM H4Cl 5 nM monensin 10 nM monensin 5 nM nigericin 10 nM nigericin

1 1 1 1 1.9 1 1 1 9.3 1.3 2.5 14.0 59 145 54 117 89 382

Ramos Confidence level (95%)

0.8

4.4 0.4 1.0 7.1 37.5 73 31 37 49 111

Factor of enhancement 1 1 1 1 2.0 1 1 1 9.7 1.7 2.9 11.4 40 107 63 81 253 318

Confidence level (95%)

1.1

4.1 0.6 0.5 0.5 5.0 48 39 49 101 129

The factor of enhancement is defined as the ID50 of CD22-rec ricin A divided by the ID50 of CD22-rec ricin A in the presence of an enhancer. ID50s were determined in protein synthesis inhibition assays which were performed in triplicate. Assays were repeated twice.

Table 4

Toxicity of enhancers to bone marrow progenitor cells

% of control

5 ␮M chloroquine 20 ␮M chloroquine 6 mM NH4Cl 5 ␮M chloroquine + 6 mM NH4Cl 20 ␮M chloroquine + 6 mM NH4Cl 5 nM monensin 10 nM monensin 25 nM monensin 5 nM nigericin 10 nM nigericin 25 nM nigericin

CFU-GM

BFU-E

98 88 89 107 105 87 89 72 91 111 77

94 77 107 130 92 117 94 47 122 104 32

The toxicity of varying concentrations of enhancers to bone marrow progenitor cells was determined as described in Materials and methods. Assays were performed three times. s.d. ⬍ 10%.

serum and 58-fold in the presence of 100% human serum (Figure 3). Reduction in IT activity was not caused by reduced binding of CD22 as this was not affected by 50% and 100% human serum (Figure 4).

Influence of lysosomotropic amines on CD22-rec ricin A activity The lysosomotropic amines NH4Cl and chloroquine were investigated for CD22-rec ricin A activity enhancing potency. At 6 mm, NH4Cl inhibited 18% of protein synthesis. Fifty ␮m chloroquine reduced the protein synthesis of Daudi and Ramos to 93 and 70%, respectively. A concentration of 100 ␮m reduced the protein synthesis to less than 25% (data not shown), and was therefore not studied for enhancement of CD22-rec ricin A.

Figure 2 Enhancement of CD22-rec ricin A (쑗) cytotoxicity against Ramos by 5 nm (쏔) and 10 nm (쎲) monensin. Influence of 50% human serum (쎲) on CD22-rec ricin A cytotoxicity and on enhancement of cytotoxicity by 5 (쮿) and 10 nm (왖) monensin. Assays were performed in triplicate and repeated twice. s.d. ⬍ 10%.

Previously we showed that 6 mm NH4Cl enhanced the CD22-rec ricin A activity to eight different B cell lines between seven- and 93-fold.6 Chloroquine enhanced the activity of CD22-rec ricin A in a dose-dependent fashion (Table 3) with 50 ␮m chloroquine enhancing the IT activity 14-fold for Daudi cells and 11-fold for Ramos cells. Human serum reduced, but did not completely block the enhancement of CD22-rec ricin A by chloroquine (Figure 3). At 50 ␮m, chloroquine enhanced the IT cytotoxicity 2.6- and 1.7-fold in the presence of 50% or 100% human serum, respectively. The combination of 6 mm NH4Cl and either 5 or 20 ␮m


246

Figure 3 Influence of human serum on the cytotoxicity of CD22rec ricin A and on enhancement by chloroquine in the cell line Ramos. Cytotoxicity was determined in a protein synthesis inhibition assay in the presence of 0% (쑗), 50% (쏔) or 100% (왕) human serum and in the presence of 50 mm chloroquine and 0% (쎲), 50% (쮿) or 100% (왖) human serum. Assays were performed in triplicate and repeated twice. s.d. ⬍ 10%.

Figure 5 Enhancement of the cytotoxicity of CD22-rec ricin A (䊊) to Ramos by 6 mm NH4Cl (쎲), 5 ␮m (쏔) and 20 ␮m (왕) chloroquine and by the combinations of 6 mm NH4Cl with 5 ␮m (쮿) or 20 ␮m (왖) chloroquine. Cytotoxicity was determined in a protein synthesis inhibition assay. Assays were performed in triplicate and repeated twice. s.d. ⬍10%.

patients with B-CLL and two patients with B-ALL was determined by the flow cytometric cell killing assay (Table 5). B-CLL cells (mean killing 0.42 log) seemed more sensitive to the IT than B-ALL cells (mean killing 0.19 log). Normal B cells were also killed by the IT (Table 5). Malignant T cells and normal T cells were not affected (data not shown). The enhancers that potentiated IT activity to the cell lines Daudi and Ramos were assessed for IT enhancing activity to these malignant cells. The enhancers were hardly toxic for the cells (Table 5). NH4Cl (mean log kill 1.70 and 1.52), chloroquine (mean log kill 1.51 and 0.47), monensin (mean log kill 1.44 and 0.59) and nigericin (mean log kill 1.43 and 0.96) all enhanced the activity of the IT against the B-CLL and B-ALL cells, respectively. However, the combination of chloroquine and NH4Cl potentiated the IT activity against B-CLL and B-ALL cells to the greatest extent with a mean log kill of 2.06 and 1.85, respectively. Discussion

Figure 4 Influence of 50% (쏔) and 100% (왕) human serum on binding of the CD22 MoAb (쑗) to Ramos cells performed as described in Materials and methods.

chloroquine enhanced the activity CD22-rec ricin A against Daudi and Ramos cells by more than the sum of the enhancement by both agents individually (145- and 107-fold, respectively, Table 3, Figure 5). These combinations reduced the protein synthesis of Daudi and Ramos cells by only 18% (data not shown) and were not toxic in bone marrow progenitor assays (Table 4).

Activity of CD22-rec ricin A to B-CLL and B-ALL cells The activity of CD22-rec ricin A in combination with enhancers against malignant cells originated from three

ITs are a promising treatment modality for malignancies either as initial therapy or as adjuvant therapy. Efficacy of IT treatment may be improved by several approaches such as applying varying MoAbs and toxins44 or IT cytotoxicity enhancing agents.19,27,28,33,37 We studied the enhancing effect of various agents on the activity of CD22-rec ricin A on the cell lines Daudi and Ramos and B-CLL, B-ALL and normal B cells using protein synthesis inhibition assays and quantitative flow cytometric cell killing assays. In protein synthesis inhibition assays, 3H-leucine incorporation is measured. The correlation of protein synthesis inhibition with cell elimination remains questionable. Therefore an assay should be used in which cell kill is determined. Clonogenic assays of limiting dilutions should be suitable but are laborious, time-consuming and only applicable for in vitro proliferating cells. We therefore employed an assay we recently developed in which cell elimination was determined by flow cytometry by combining cell killing parameters and immunofluorescence.


Table 5

247

Depletion (log kill) of B-CLL and B-ALL cells by CD22-rec ricin A (10−8 m) in the absence and the presence of enhancers

Treatment

IT (10−8 M) 6 mM NH4Cl 20 ␮M chloroquine 10 nM monensin 10 nM nigericin 6 mM NH4Cl + 20 ␮M chloroquine IT + 6 mM NH4Cl IT + 20 ␮M chloroquine IT + 10 nM monensin IT + 10 nM nigericin IT + 6 mM NH4Cl + 20 ␮M chloroquine

Patient No. 1

2

3

4

5

6

0.40 −0.06 −0.28 0.09 −0.21 0.45 1.52 1.48 ND 1.22 1.92

0.33 0.03 0.32 0.16 0.12 0.12 1.60 1.53 1.26 1.44 1.87

0.53 0.01 ND 0.10 0.15 1.14 1.99 n.d. 1.62 1.63 2.40

0.12 0.02 0.08 0.04 0.02 0.06 1.16 0.41 0.51 0.94 1.43

0.25 0.03 0.10 0.01 0.04 0.04 1.87 0.52 0.66 0.98 2.26

0.29 0.11 0.17 0.03 0.06 n.d. 1.17 0.94 0.99 1.08 1.66

The depletion of CD22+ B cells was determined by the quantitative cell killing assay. The activity of CD22-rec ricin A to B-CLL (1, 2, 3), B-ALL (4, 5) and normal B cells derived from a T-ALL patient (6) was investigated. Depletion (log kill) is determined by log (number of viable cells in treated sample per 1000 beads/number of viable cells in control sample per 1000 beads). ND, not determined.

CD22-rec ricin A was highly cytotoxic to Daudi and Ramos cells with ID50 of approximately 5 × 10−11 m as determined by the protein synthesis inhibition assay. In flow cytometric cell killing assays we determined that this concentration killed approximately 40% of these target cells. There was little agreement on enhancement between the two assays (Tables 2 and 3). Although 10 nm monensin and nigericin enhanced IT activity in the protein synthesis inhibition assay against Daudi cells to a greater extent than against Ramos cells, the former cells were killed less effectively after exposure to monensin and nigericin. More Daudi cells were killed by 5 × 10−11 m IT in the presence of 5 nm nigericin than in the presence of 10 nm monensin, although 10 nm monensin enhanced to a greater extent than did 5 nm nigericin when measured by the protein synthesis inhibition assay (Tables 1 and 2). Protein synthesis does not completely reflect the cytotoxic activity of ITs possibly because of the relatively long time it takes ITs to exert their action which may allow a cell to recover from incomplete or insufficient protein synthesis inhibition. Thus it is preferable to determine the effectiveness of ITs using an assay that measures cytotoxicity directly, as used in this study. Though ATRA, verapamil and brefeldin A have been shown to enhance the activity of ricin A ITs,34,36,37 these agents did not enhance CD22-rec ricin A activity. These varying results might be explained if ricin A follows a different intracellular routing after it is delivered to the cells by CD22 MoAb than occurs with other cell binding moieties. Alternatively, different target cells might exhibit different susceptibility to the different enhancers. Similar contradictory results have been obtained with NH4Cl.6 Monensin, nigericin, NH4Cl, chloroquine and bacitracin all enhanced the activity of CD22-rec ricin A. Nigericin was more potent than the combination of chloroquine and NH4Cl, which in turn was more potent than monensin and bacitracin. The combination of chloroquine and NH4Cl was more potent than both agents individually. In order to apply ITs in combination with enhancing agents in vivo we studied the enhancement in the presence of human serum. We confirmed the results of Candiani et al30 and Franceschi et al40 who showed that monensin is unsuitable for use in vivo because it forms complexes with serum components thereby losing its ability to enhance IT. Human serum also

abrogated the enhancing activity of nigericin. Furthermore we showed that human serum inhibited the activity of CD22-rec ricin A itself in a concentration-dependent fashion. We could not detect any degradation of the IT after incubation in 50% human serum for 24 h at 37°C (by means of immunoblotting; data not shown) yet only 10% of the activity was retained. This reduced activity can not be caused by instability of the disulphide bond between the MoAb and toxin of the IT in serum nor by reduced binding of the CD22 antibody as this was unaffected by human serum (Figure 3). We assume that CD22-rec ricin A is inhibited by complexation of the IT with serum components as described for monensin.30,40 Inhibition of activity might also be induced by covalent binding of ␣2macroglobulin to rec ricin A as has been described for deglycosylated ricin A chain and its ITs by Ghetie et al45 who found a substantial amount of complexes formed after 24 h. However, we could not detect these high molecular complexes with the immunoblotting assay used. Besides, the kinetics of CD22-rec ricin A activity are too fast6 to allow these complexes to form.45 This suggests that more factors contribute to the reduction of IT activity by serum. We assume that CD22rec ricin A activity is inhibited as a result of complexes formed with human serum components and that these impede internalization of this IT. The activity of chloroquine was partially blocked in the presence of human serum so we intend to pursue this further in mice models to assess its potential for improving the activity of CD22-rec ricin A in vivo. Cytotoxicity of CD22-rec ricin A against malignant B-CLL and B-ALL cells was studied in the flow cytometric cell killing assay. B-CLL cells seem to be eliminated more effectively by the IT than do B-ALL and normal B cells. IT activity to malignant B cells was enhanced by NH4Cl, chloroquine, monensin and nigericin but the combination of NH4Cl and chloroquine was more effective in contrast to enhancement of IT activity against cell lines. CD22-rec ricin A was more active in killing B cell lines than cells originating from patients which may be the result of the cell lines ability to proliferate in vitro. We conclude that monensin, nigericin or the combination of chloroquine and NH4Cl are very effective enhancers of CD22-rec ricin A and can be used to treat bone marrow contaminated with malignant B cells ex vivo.


248

Acknowledgements This work was supported in part by grants from the ‘Ank van Vlissingen Foundation’. We would like to thank Zeneca Pharmaceuticals for kindly providing recombinant ricin A, Ms E Rutten for performing the BM progenitor assays, Mr A Pennings for carrying out flow cytometric analysis and Dr P Donnelly for critically reading the manuscript. References 1 Preijers FW. Rationale for the clinical use of immunotoxins: monoclonal antibodies conjugated to ribosome-inactivating proteins. Leuk Lymphoma 1993; 9: 293–304. 2 Thrush GR, Lark LR, Clinchy BC, Vitetta ES. Immunotoxins: an update. Annu Rev Immuno 1996; 14: 49–71. 3 Olsnes S, Fernandez Puentes C, Carrasco L, Vazquez D. Ribosome inactivation by the toxic lectins abrin and ricin. Kinetics of the enzymic activity of the toxin A-chains. Eur J Biochem 1975; 60: 281–288. 4 Eiklid K, Olsnes S, Pihl A. Entry of lethal doses of abrin, ricin and modeccin into the cytosol of HeLa cells. Exp Cell Res 1980; 126: 321–326. 5 Brusa P, Pietribiasi F, Bussolati G, Dosio F, Arione R, Comoglio PM, Prat M, Cattel L. Blocked and not blocked whole-ricin-antibody immunotoxins: intraperitoneal therapy of human tumour xenografted in nude mice. Cancer Immunol Immunother 1989; 29: 185–192. 6 Van Horssen PJ, van Oosterhout YV, de Witte T, Preijers FW. Cytotoxic potency of CD22-ricin A depends on intracellular routing rather than on the number of internalized molecules. Scand J Immunol 1995; 41: 563–569. 7 Shih LB, Lu HH, Xuan H, Goldenberg DM. Internalization and intracellular processing of an anti-B cell lymphoma monoclonal antibody, LL2. Int J Cancer 1994; 56: 538–545. 8 Kreitman RJ, Hansen HJ, Jones AL, Fitzgerald DJ, Goldenberg DM, Pastan I. Pseudomonas exotoxin-based immunotoxins containing the antibody LL2 or LL2-Fab⬘ induce regression of subcutaneous human B cell lymphoma in mice. Cancer Res 1993; 53: 819–825. 9 Ghetie MA, Tucker K, Richardson J, Uhr JW, Vitetta ES. Eradication of minimal disease in severe combined immunodeficient mice with disseminated Daudi lymphoma using chemotherapy and an immunotoxin cocktail. Blood 1994; 84: 702–707. 10 Ghetie MA, May RD, Till M, UHR JW, Ghetie V, Knowles PP, Relf M, Brown A, Wallace PM, Janossy G. Evaluation of ricin A chaincontaining immunotoxins directed against CD19 and CD22 antigens on normal and malignant human B cells as potential reagents for in vivo therapy. Cancer Res 1988; 48: 2610–2617. 11 Ghetie MA, Richardson J, Tucker T, Jones D, UHR JW, Vitetta ES. Antitumor activity of Fab⬘ and IgG-anti-CD22 immunotoxins in disseminated human B lymphoma grown in mice with severe combined immunodeficiency disease: effect on tumor cells in extranodal sites. Cancer Res 1991; 51: 5876–5880. 12 Ghetie MA, Tucker K, Richardson J, Uhr JW, Vitetta ES. The antitumor activity of an anti-CD22 immunotoxin in SCID mice with disseminated Daudi lymphoma is enhanced by either an antiCD19 antibody or an anti-CD19 immunotoxin. Blood 1992; 80: 2315–2320. 13 Shen GL, Li JL, Ghetie MA, Ghetie V, May RD, Till M, Brown AN, Relf M, Knowles P, Uhr JW. Evaluation of four CD22 antibodies as ricin A chain-containing immunotoxins for the in vivo therapy of human B cell leukemias and lymphomas. Int J Cancer 1988; 42: 792–797. 14 Van Horssen PJ, Preijers FW, van Oosterhout YV, de Witte T. Highly potent CD22-recombinant ricin A results in complete cure of disseminated malignant B cell xenografts in SCID mice but fails to cure solid xenografts in nude mice. Int J Cancer 1996; 68: 378–383. 15 Amlot PL, Stone MJ, Cunningham D, Fay J, Newman J, Collins R, May R, McCarthy M, Richardson J, Ghetie V. A phase I study of an anti-CD22-deglycosylated ricin A chain immunotoxin in the treatment of B cell lymphomas resistant to conventional therapy. Blood 1993; 82: 2624–2633.

16 Sausville EA, Headlee D, Stetler Stevenson M, Jaffe ES, Solomon D, Figg WD, Herdt J, Kopp WC, Rager H, Steinberg SM. Continuous infusion of the anti-CD22 immunotoxin IgG-RFB4-SMPT-dgA in patients with B cell lymphoma: a phase I study. Blood 1995; 85: 3457–3465. 17 Vitetta ES, Stone M, Amlot P, Fay J, May R, Till M, Newman J, Clark P, Collins R, Cunningham D. Phase I immunotoxin trial in patients with B cell lymphoma. Cancer Res 1991; 51: 4052–4058. 18 Ghetie MA, Ghetie V, Vitetta ES. Immunotoxins for the treatment of B cell lymphomas. Mol Med 1997; 3: 420–427. 19 Jaffrezou JP, Laurent G. The intriguing link between modulation of both multidrug resistance and ligand-toxin conjugate cytotoxicity. FEBS Lett 1993; 323: 191–197. 20 Olsnes S, Sandvig K. Toxins. In: Pastan I, Willingham MC (eds). Endocytosis. Plenum Press: New York, 1985, pp 195–234. 21 Press OW, Martin PJ, Thorpe PE, Vitetta ES. Ricin A-chain containing immunotoxins directed against different epitopes on the CD2 molecule differ in their ability to kill normal and malignant T cells. J Immunol 1988; 141: 4410–4417. 22 Press OW, Desantes K, Anderson SK, Geissler F. Inhibition of catabolism of radiolabeled antibodies by tumor cells using lysosomotropic amines and carboxylic ionophores. Cancer Res 1990; 50: 1243–1250. 23 Von Figura K, Steckel F, Hasilik A. Juvenile and adult metachromatic leukodystrophy: partial restoration of arylsulfatase A (cerebroside sulfatase) activity by inhibitors of thiol proteinases. Proc Natl Acad Sci USA 1983; 80: 6066–6070. 24 Preijers FW, de Witte T, Wessels JM, de Gast GC, van Leeuwen E, Capel PJ, Haanen C. Autologous transplantation of bone marrow purged in vitro with anti-CD7-(WT1-) ricin A immunotoxin in T cell lymphoblastic leukemia and lymphoma. Blood 1989; 74: 1152–1158. 25 Laurent G, Pris J, Farcet JP, Carayon P, Blythman H, Casellas P, Poncelet P, Jansen FK. Effects of therapy with T101 ricin A-chain immunotoxin in two leukemia patients. Blood 1986; 67: 1680– 1687. 26 Madan S, Ghosh PC. Monensin intercalation in liposomes: effect on cytotoxicities of ricin, Pseudomonas exotoxin A and diphtheria toxin in CHO cells. Biochim Biophys Acta 1992; 1110: 37–44. 27 Griffin T, Rybak ME, Recht L, Singh M, Salimi A, Raso V. Potentiation of antitumor immunotoxins by liposomal monensin. J Natl Cancer Inst 1993; 85: 292–298. 28 Ramakrishnan S, Bjorn MJ, Houston LL. Recombinant ricin A chain conjugated to monoclonal antibodies: improved tumor cell inhibition in the presence of lysosomotropic compounds. Cancer Res 1989; 49: 613–617. 29 Colombatti M, Dell’Arciprete L, Chignola R, Tridente G. Carrier protein–monensin conjugates: enhancement of immunotoxin cytotoxicity and potential in tumor treatment. Cancer Res 1990; 50: 1385–1391. 30 Candiani C, Franceschi A, Chignola R, Pasti M, Anselmi C, Benoni G, Tridente G, Colombatti M. Blocking effect of human serum but not of cerebrospinal fluid on ricin A chain immunotoxin potentiation by monensin or carrier protein–monensin conjugates. Cancer Res 1992; 52: 623–630. 31 Casellas P, Ravel S, Bourrie BJ, Derocq JM, Jansen FK, Laurent G, Gros P. T-lymphocyte killing by T101-ricin A-chain immunotoxin: pH-dependent potentiation with lysosomotropic amines. Blood 1988; 72: 1197–1202. 32 Raso V, Watkins SC, Slayter H, Fehrmann C. Intracellular pathways of ricin A chain cytotoxins. Ann NY Acad Sci 1987; 507: 172–186. 33 Pirker R, Fitzgerald DJ, Willingham MC, Pastan I. Enhancement of the activity of immunotoxins made with either ricin A chain or Pseudomonas exotoxin in human ovarian and epidermoid carcinoma cell lines. Cancer Res 1988; 48: 3919–3923. 34 Pirker R, Fitzgerald DJ, Raschack M, Frank Z, Willingham MC, Pastan I. Enhancement of the activity of immunotoxins by analogues of verapamil. Cancer Res 1989; 49: 4791–4795. 35 Ozols RF, Cunnion RE, Klecker RW Jr, Hamilton TC, Ostchega Y, Parrillo JE, Young RC. Verapamil and adriamycin in the treatment of drug-resistant ovarian cancer patients. J Clin Oncol 1987; 5: 641–647. 36 Wu YN, Gadina M, Tao Cheng JH, Youle RJ. Retinoic acid disrupts


37 38 39 40

41

the Golgi apparatus and increases the cytosolic routing of specific protein toxins. J Cell Biol 1994; 125: 743–753. Hudson TH, Grillo FG. Brefeldin-A enhancement of ricin A-chain immunotoxins and blockade of intact ricin, modeccin, and abrin. J Biol Chem 1991; 266: 18586–18592. Prydz K, Hansen SH, Sandvig K, van Deurs B. Effects of brefeldin A on endocytosis, transcytosis and transport to the Golgi complex in polarized MDCK cells. J Cell Biol 1992; 119: 259–272. Sandvig K, Prydz K, Hansen SH, van Deurs B. Ricin transport in brefeldin A-treated cells: correlation between Golgi structure and toxic effect. J Cell Biol 1991; 115: 971–981. Franceschi A, Dosio F, Anselmi C, Chignola R, Candiani C, Pasti M, Tridente G, Colombatti M. Mechanisms involved in serumdependent inactivation of the immunotoxin enhancers monensin and carrier protein–monensin. Eur J Biochem 1994; 219: 469– 479. Thorpe PE, Wallace PM, Knowles PP, Relf MG, Brown AN, Watson GJ, Knyba RE, Wawrzynczak EJ, Blakey DC. New coupling

42

43

44

45

agents for the synthesis of immunotoxins containing a hindered disulfide bond with improved stability in vivo. Cancer Res 1987; 47: 5924–5931. Van Oosterhout YV, Preijers FW, Meijerink JP, Pennings AH, de Witte T. A quantitative flow cytometric method for the determination of immunotoxin-induced cell kill in marrow grafts. Prog Clin Biol Res 1994; 389: 89–95. de Witte T, Raymakers R, Plas A, Koekman E, Wessels H, Haanen C. Bone marrow repopulation capacity after transplantation of lymphocyte-depleted allogeneic bone marrow using counterflow centrifugation. Transplantation 1984; 37: 151–155. Theuer CP, Kreitman RJ, Fitzgerald DJ, Pastan I. Immunotoxins made with a recombinant form of Pseudomonas exotoxin A that do not require proteolysis for activity. Cancer Res 1993; 53: 340–347. Ghetie MA, Uhr JW, Vitetta ES. Covalent binding of human alpha 2-macroglobulin to deglycosylated ricin A chain and its immunotoxins. Cancer Res 1991; 51: 1482–1487.

249