Generation and Characterization of an Anti-CD19 Single-Chain Fv

Oct 15, 1997 - Duo Wang,‡ Quanzhi Li,‡ Wendy Hudson,‡ Erica Berven,‡ Fatih Uckun,§ and John H. Kersey*,‡,|. University of Minnesota Cancer ...
0 downloads 0 Views 294KB Size
Download PDF
Bioconjugate Chem. 1997, 8, 878−884

878

Generation and Characterization of an Anti-CD19 Single-Chain Fv Immunotoxin Composed of C-Terminal Disulfide-Linked dgRTA† Duo Wang,‡ Quanzhi Li,‡ Wendy Hudson,‡ Erica Berven,‡ Fatih Uckun,§ and John H. Kersey*,‡,| University of Minnesota Cancer Center, Biotherapy Institute, and Departments of Laboratory Medicine/ Pathology and Pediatrics, University of Minnesota, Minneapolis, Minnesota 55455. Received May 9, 1997X

Our laboratory utilized two methods to produce the anti-CD19 immunotoxin containing a singlechain Fv (scFv) FVS191 and a ricin A chain (RTA). The first method produced the recombinant protein FVS191CDRTA from a fusing gene containing sequences encoding FVS191, catheptsin D proteinase digestion site (CD), and RTA. FVS191CDRTA did not show CD19 antigen binding and cytotoxic activity. The second method generated a disulfide-linked FVS191cys-dgRTA from a FVS191cys, the FVS191 with an additional C-terminal cysteine, and a deglycosylated RTA (dgRTA). The formation of FVS191cys-dgRTA is efficient; up to 70% of the proteins participating in the reaction had formed FVS191cys-dgRTA when the molar ratio of FVS191cys to dgRTA was 1:1. A competitive ELISA assay indicated that FVS191cys-dgRTA and the parental monoclonal antibody B43 possessed comparable CD19 binding abilities. The protein synthesis inhibition assay revealed that FVS191cysdgRTA was toxic to CD19 positive cell lines, but it was less potent than the intact antibody-conjugated B43-dgRTA, which had an IC50 ) 2 × 10-11 M. 125I-Labeled FVS191 and 125I-labeled B43 were internalized by Nalm-6 cells at 37 °C as demonstrated by internalization studies; this result indicates that cross-linking of CD19 antigen is not required for the endocytosis of CD19 and raises the possibility that the lower cytotoxity of FVS191cys-dgRTA is not due to the monovalent binding of CD19 by FVS191cys-dgRTA. Our study with anti-CD19 scFv immunotoxin indicates that the formation of a disulfide-linked scFv immunotoxin is an alternative to the recombinant method of producing scFv immunotoxin.

INTRODUCTION

Immunotoxins are cytotoxic molecules designed to eliminate populations of cells that display specific cell surface antigens or markers. An immunotoxin has two functional components: one is the binding domain, which targets the specific marker of a cell; the other is a toxin domain, which kills the target cell. The binding domains of immunotoxins can be a monoclonal antibody (mAb), Fab, F(ab′)2, Fv, or single-chain Fv (scFv). The popular toxic domains of immunotoxins include the ricin A chain (RTA), Pseudomonas exotoxin B chain (1-5), or diphtheria toxin A chain (6). ScFv is a protein fragment composed of immunoglobulin heavy and light chain with a peptide linker connecting the heavy and light chain. A scFv-immunotoxin is a chimeric molecule containing a scFv and a toxin. There are two methods that can be used to generate a scFv immunotoxin. The conventional method is to produce a recombinant fusion protein by genetically linking a scFv gene with a toxin gene. Several scFv immunotoxins have been made using this method, such as B3(Fv)-PE38KDEL (1, 2), e23(Fv)-PE38KDEL (3), antiTac(Fv)-PE38 (4), and BR96 sFV-PE40 (5). All of these scFv immunotoxins have been shown to possess antitumor * Address correspondence to this author at Box 86 Mayo, 420 Delaware St. S.E., Minneapolis, MN 55455 [e-mail [email protected]; telephone (612) 625-4659; fax (612) 624-3069]. † This work is supported in part by an Outstanding Investigator Grant Award to J.H.K. (CA 49721) from the National Cancer Institute. ‡ University of Minnesota Cancer Center. § Biotherapy Institute. | Departments of Laboratory Medicine/Pathology and Pediatrics. X Abstract published in Advance ACS Abstracts, October 15, 1997.

S1043-1802(97)00071-2 CCC: $14.00

activity in vitro and in animal models bearing human tumor xenografts. The second method of generating a scFv immunotoxin is to build a scFv with C-terminal cysteine and then covalently link the scFv-cys with a toxin through a disulfide bond (7). The construction of anti-CD19 scFv immunotoxins using the above two methods is described in this paper. CD19, as a pan B-cell antigen, is an ideal target for immunotoxin therapy of B-lineage leukemia and lymphomas (8, 9). Our laboratory has previously constructed an anti-CD19 scFv, FVS191, from a hybridoma producing mAb B43 (10). FVS191cys, a FVS191 containing a C-terminal cysteine, has also been constructed (11). RTA inactivates the 60S ribosomal subunit of eukaryotic ribosomes and has been used for construction of antiCD19 immunotoxins (12). The RTA gene has been cloned (13); the DNA sequence indicates that RTA has two cysteines. Therefore, RTA can be used to form both recombinant and disulfide-linked immunotoxins. The recombinant anti-CD19 scFv immunotoxin we produced contained a cathepsin D cleavable peptide (CD) (14); therefore, this protein was named FVS191CDRTA. Deglycosylated ricin A chain (dgRTA) was used to form a disulfide bond with FVS191cys to generate FVS191cysdgRTA. The characterization of these two proteins is also described in this paper. MATERIALS AND METHODS

Construction of FVS191CDRTA. FVS191 had been previously cloned in our laboratory (10). To construct the FVS191CDRTA, a DNA fragment containing the sequences encoding cathepsin D proteinase sensitive peptide and RTA was inserted in frame at the 3′ end of FVS191 in pFVS191. The nucleotide sequence encoding cathepsin D proteinase sensitive peptide is ACCTGTTCTTCCTGAACCTGTTCTACCTGA; the RTA gene was © 1997 American Chemical Society

Anti-CD19 scFv Immunotoxin

Bioconjugate Chem., Vol. 8, No. 6, 1997 879

Figure 1. Gene structure of FVS191CDRTA. The fusion gene of FVS191CDRTA is composed of nucleic acids encoding FVS191, cathepsin D sensitive peptide, and RTA. The expression of FVS191CDRTA is under the T7 promoter. The expression vector is derived from pET3b (Novagen).

obtained from a genomic clone of ricin (13). The structure of the expression vector of pFVS191CDRTA is shown in Figure 1. Production and Purification of FVS191CDRTA. pFVS191CDRTA plasmid DNA was transformed into Escherichia coli BL21(DE3) (Novagen, Madison, WI). The transformed bacterial cells were grown in SOB medium (20 g of tryptone, 5 g of yeast extract, 0.5 g of NaCl, 5 g of MgSO4‚7H2O, per liter) at 37 °C. When the absorbance of A600 of the bacterial culture reached 0.65, production of FVS191CDRTA was induced with 1 mM isopropyl β-Dthiogalactopyanoside (IPTG) for 1.5 h at 37 °C. FVS191CDRTA was expressed in bacterial cells as inclusion bodies. The method used to isolate FVS191CDRTA inclusion bodies was the same as that described by Wang et al. (11). Briefly, the harvested cell pellets from the above cell culture were suspended in 50 mL of inclusion body separation (IBS) buffer (0.1 M KCl, 0.02 M Tris-HCl, 0.005 M EDTA, 0.1% Nonidet P-40, pH 8.0) by sonication. Lysozyme was added into the cell lysate to a final concentration of 0.2 mg/mL. After incubation at room temperature for 1 h, the cell lysate was frozen at -80 °C. When thawed at room temperature, the cell lysate was sonicated for 10 min and centrifuged at 17000g for 30 min at 4 °C. The pellets were suspended in 50 mL of IBS buffer and added with sodium deoxycholate to a final concentration of 2%. The mixture was stirred at room temperature for 1 h and then centrifuged at 17000g for 30 min at 4 °C. The inclusion bodies were washed with IBS buffer and water. FVS191CDRTA was refolded according to the method of Buchner (15) with minor modifications. The inclusion bodies of FVS191CDRTA were dissolved in a denaturing buffer [0.1 M Tris, 6 M guanidine hydrochloride, 0.3 M dithioerythreitol (DTE), 0.002 M EDTA, pH 8] at room temperature for 2 h. The insoluble materials were removed by centrifugation at 30000g for 30 min. The concentration of solubilized proteins was adjusted to 20 mg/mL with denaturing buffer. Renaturation of FVS191CDRTA was carried out by a rapid 100-fold dilution of the denatured protein into a refolding buffer (0.1 M Tris-HCl, 0.5 M L-arginine, 0.008 M GSSG, 0.002 M EDTA, pH 8) at 10 °C and an incubation at the same temperature for 48 h. Other refolding methods were also used to refold FVS191CDRTA. Reduced FVS191CDRTA in the above denaturing buffer was dialyzed against phosphatebuffered saline (PBS) buffer containing 0.4 M L-arginine (PBS-arginine buffer) or a buffer composed of 20 mM Tris and 100 mM urea, pH 7.4 (Tris-urea buffer). After refolding, FVS191CDRTA was dialyzed against PBS to eliminate redox agents and concentrated by ultrafiltration using a YM10 membrane (Amicon, Beverly, MA). The final product of FVS191CDRTA was purified by FPLC using a Superdex 75 (HiLoad 60) column (Pharmacia).

Figure 2. Schematic diagram of the formation of disulfidelinked FVS191cys-dgRTA. DgRTA was reduced with DTT and treated with DTNB prior to conjugation with DTT-treated FVS191cys.

Preparation of FVS191cys-dgRTA. The construction and production of FVS191cys have been described previously (11). The dgRTA was purchased from Inland Laboratory (Austin, TX). Prior to conjugation, dgRTA was buffer exchanged with PBS by dialysis. A schematic representation of forming disulfide-linked FVS191cys-dgRTA is shown in Figure 2. DgRTA (2 mg/ mL) was reduced with 2 mM dithiothreitol (DTT) for 1 h at room temperature and then separated from DTT using a PD10 column (Pharmacia). The DTT-treated dgRTA was immediately reacted with a 1/10 volume of 25 mM 5,5′-dithiobis(2-nitrobenzoic acid) (DTNB) in phosphate buffer (0.12 M K2HPO4, pH 7.2) at room temperature for 1 h. The DTNB-derivatized dgRTA was then isolated using a PD10 column. At the same time, FVS191cys in PBS buffer was reacted with 2 mM DTT at room temperature for 1 h. After the DTT had been removed by a PD10 column, FVS191cys was reacted with DTNBderivatized dgRTA at a molar ratio of 1:1 or 1:2. The conjugation was performed at room temperature for 2 h and at 4 °C overnight. The conjugation products were analyzed by a nonreducing SDS-PAGE and kept at 4 °C. The free sulfhydryl groups of DTT-treated FVS191cys were quantified using DTNB. After DTT-treated FVS191cys had been reacted with 25 mM DTNB in phosphate buffer (0.12 M K2HPO4, pH 7.2) at room temperature for 15 min, the absorbance at 412 nm was measured using a spectrophotometer. The molar concentration of the free sulfhydryl group (C) was calculated according to the equation C ) absorbance/14150 (mol/L) (16). The molar concentration of FVS191cys was determined according to the absorbance of FVS191cys at 280 nm. Preparation of B43-dgRTA. B43 was conjugated to dgRTA by the chemical cross-linker N-succinimidyl-3-(2pyridyldithio)propionate (SPDP, Pierce) using the method of Ghetie et al. (17) with minor modifications. Five milligrams of B43 mAb in 2 mL of conjugation buffer (0.05 M borate acid, 0.3 M NaCl, 0.5% butanol, pH 9.0) was reacted with 30 µL of SPDP (6 mg/mL, in diethylformamide) at room temperature for 30 min. The derivatized B43 was separated from SPDP using a PD10 column (Pharmacia) and dissolved in 2 mL of phosphate buffer (0.1 M sodium phosphate, 0.15 M NaCl, 0.005 M EDTA, pH 6.5). DgRTA in phosphate buffer was incu-

880 Bioconjugate Chem., Vol. 8, No. 6, 1997

bated with the derivatized B43 at room temperature for 1 h and at 4 °C overnight. The molar ratio of dgRTA to B43 in the reaction was 1:1. After conjugation, free dgRTA was removed by FPLC using a Superdex 75 column (Pharmacia) at a flow rate of 0.5 mL/min. The conjugated B43-dgRTA was separated from free B43 by an open column affinity chromatography. Briefly, Blue Sepharose gel (Supelco Separation Technology, Bellefonte, PA) was packed in a 1 × 20 cm column according to the packing manual of TSK-Gel Toyopearl column. The blue gel column was equilibrated with 50 mL of sodium phosphate buffer (50 mM, pH 7.5) at a flow rate of 2 mL/ min. Protein samples were passed through the Blue Sepharose column at a flow rate of 0.3 mL/min. The unconjugated B43 passed through the column; B43dgRTA, which bound to the column, was eluted with a buffer (0.05 M sodium phosphate, 0.5 M NaCl, pH 7.5) described by Knowles and Thorpe (18). ELISA Assay. B43 was conjugated to alkaline phosphatase (AP) using a commercial conjugation kit (Pierce, Rockford, IL). The conjugated B43-AP was used directly without further purification. The CD19 antigens used in the ELISA were isolated from CD19 positive Daudi cells using the methods of Siegall et al. (19). The optimal concentrations of CD19 antigen and B43-AP for the ELISA assay were determined by a series of experiments. Before the ELISA assay, the isolated CD19+ membrane proteins (80 µg/mL in H2O, 50 µL/well) were coated in ELISA plates (Immuron 4, Dynatech Laboratories Inc., Chantilly, VA). These plates were dried in a vacuum desiccator at 4 °C overnight and washed three times with PBS. During the ELISA assay, various amounts of competing proteins (FVS191CDRTA, FVS191cys-dgRTA, or B43 mAb) and a fixed amount of B43-AP were incubated inside the coated plates at room temperature for 1 h. The final dilution of B43-AP in the reaction was 1:2000. The plates were washed four times with PBS before incubation with 100 µL of substrate solution for alkaline phosphatase at 4 °C overnight. The optical density of each reaction well was measured at 405 nm using an ELISA reader. Cytotoxity Assay. Blin-1 or Nalm-6 cells were prepared to a final density of 1 × 106 cells/mL in RPMI 1640 medium containing 10% (w/v) fetal bovine serum (FBS), 1% L-glutamine, 1% penicillin/streptomycin, and 20 mM NH4Cl. One hundred microliters of the cells, along with 50 µL of immunotoxins diluted in the same medium, was plated into wells of the 96-well microtiter plates. The assay was performed in triplicate. After incubation at 37 °C for 48 h in an atmosphere of 5% CO2, the cells were pulsed with 1 µCi of [3H]leucine (Amersham, Life Science) in 50 µL of the medium for 4 h at 37 °C and frozen at -80 °C. After thawing, the cells were harvested onto GF/F glass microfiber filters (Whatman). Radioactivity was measured using a β scintillation counter (Beckman Instrument). The [3H]leucine incorporation of immunotoxin-treated cells was compared with that of nontreated cells to determine the percent of protein synthesis. Internalization Assay. FVS191 and B43 mAb were labeled with Na125I (New England Nuclear, Boston, MA) using the Iodo-Beads method (Pierce). Nalm-6, in RPMI 1640 medium supplemented with 10% FBS, was used in this assay. Ten micrograms of 125I-labeled FVS191 or 10 µg of 125I-labeled B43, with specific activity ∼5-7 µCi/ µg, was incubated with 1 × 108 Nalm-6 cells (5 mL) on ice for 1 h. After incubation, the cells were washed four times with cold, serum-free RPMI 1640 and then suspended in warm RPMI 1640 plus 10% FBS at 0.5 to 1.0 × 108 cells/mL. Aliquots of 5 × 106 cells were transferred to 1 mL sterile test tubes and incubated at 37 °C in a

Wang et al.

Figure 3. Formation and purification of FVS191CDRTA and FVS191cys-dgRTA. Protein samples were analyzed by a 12% nonreducing SDS-PAGE and visualized by Coomassie blue stain: (lane 1) FVS191CDRTA purified by gel filtration chromatography; (lane 2) FVS191cys used in the chemical conjugation with dgRTA; (lane 3) dgRTA; (lane 4) conjugation products when FVS191cys:dgRTA ratio is 1:2; (lane 5) conjugation products when FVS191cys:dgRTA ratio is 1:1; (lane 6) FVS191cys-dgRTA purified by gel filtration FPLC; (lane 7) DTT-reduced products of FVS191cys-dgRTA. The positions of molecular weight markers are labeled at the left side of the gel.

tissue culture incubator for various time intervals. At each time point, the samples were analyzed for the percents of cell-surface-bound, dissociated, internalized, and degraded FVS191 (or B43), on the basis of a total cellassociated radioactivity at time 0 according to the methods of Press et al. (20). The nonspecific binding of 125Ilabeled FVS191 or 125I-labeled B43 was assessed by incubating the 125I-labeled proteins with B43-blocked Nalm-6 cells. RESULTS

Production of FVS191CDRTA, FVS191cys-dgRTA, and B43-dgRTA. FVS191CDRTA was produced in E. coli as inclusion bodies. The optimal amount of FVS191CDRTA was observed when FVS191CDRTA was refolded in buffer containing GSSG. When FVS191CDRTA was refolded in PBS-arginine or Trisurea buffer, the majority of the protein aggregated; these refolding approaches were not used in future experiments. After the refolding process in GSSG buffer, aggregated FVS191CDRTA was separated; soluble FVS191CDRTA was purified by gel filtration chromatography and examined by SDS-PAGE. The purity of FVS191CDRTA is shown in Figure 3, lane 1. Figure 3, lane 2, shows the refolded-FVS191cys used to form FVS191cys-dgRTA. Prior to chemical conjugation, the number of DTT-reduced cysteine per FVS191cys was examined. When the concentration of DTT-treated FVS191cys was 3.8 × 10-6 M, the absorbance of DTNBderivatized FVS191cys was 0.045. According to the formula given under Materials and Methods, the concentration of free sulfhydryl group was 3.1 × 10-6 M; therefore, the ratio of FVS191cys concentration to free sulfhydry concentration is about 1:1. This result indicated that there was only one cysteine per FVS191cys that had been reduced by DTT. The formation of FVS191cys-dgRTA was demonstrated by a nonreducing SDS-PAGE. When FVS191cys (Figure 3, lane 2) was reacted with dgRTA (Figure 3, lane 3), a 58 kDa protein (Figure 3, lanes 4 and 5) was formed. When this 58 kDa protein was treated with DTT, it was separated into smaller proteins with the corresponding sizes of FVS191cys and dgRTA (Figure 3, lane 7). This 58 kDa protein is, therefore, the disulfide-linked FVS191cys-dgRTA. The observed 58 kDa protein (Figure 3, lanes 4 and 5) is not the dimeric dgRTA or the dimeric FVS191cys. The formation of dimeric dgRTA had been prevented by derivatizing dgRTA with DTNB prior to chemical conju-

Bioconjugate Chem., Vol. 8, No. 6, 1997 881

Anti-CD19 scFv Immunotoxin

Figure 4. Competitive ELISA to compare the CD19 binding ability of FVS191CDRTA, FVS191cys-dgRTA, and B43. Various amounts of FVS191CDRTA, FVS191cys-dgRTA, and B43 along with a fixed amount of B43-AP were incubated with CD19 antigens. The means of percentage activity of alkaline phosphatase associated with B43-AP were plotted against log molar concentrations of FVS191CDRTA (4), FVS191cys-dgRTA (b), and B43 (0).

gation. Under the conjugation condition, the formation of dimeric FVS191cys was slow (11); it could not compete with the formation of FVS191cys-dgRTA. The maximum yield of FVS191cys-dgRTA was reached when equal moles of FVS191cys and dgRTA FVS191cys were used in the conjugation. When the mole ratio of FVS191cys to RTA was 1:2, the yield of FVS191cysdgRTA was ∼40% as estimated by a densitometer (Figure 3, lane 4). When that ratio was 1:1, ∼70% of proteins formed FVS191cys-dgRTA (Figure 3, lane 5). FVS191cys-dgRTA was separated from FVS191cys and dgRTA by a gel filtration chromatography. The purity of isolated FVS191cys-dgRTA is demonstrated by Figure 3, lane 6; no FVS191cys or dgRTA was detected. The formation of covalently linked B43-dgRTA was confirmed by a nonreducing SDS-PAGE (data not shown). After purification by a gel filtration and affinity chromatography, >90% B43-dgRTA was found to have been conjugated with one dgRTA (data not shown). CD19 Binding of FVS191CDRTA and FVS191cysdgRTA. The CD19 binding activities of FVS191CDRTA, FVS191cys-dgRTA, and B43 were indicated by the relative activities of B43-AP in ELISA (Figure 4). FVS191cys-dgRTA (IC50 ) 1.33 × 10-9 M) had a similar CD19 binding ability with B43 (IC50 ) 1.3 × 10-9 M). FVS191CDRTA did not bind CD19 as indicated by the lack of inhibition of the activity of B43-AP; this result suggested that FVS191CDRTA was not properly refolded. Cytotoxic Activities of FVS191CDRTA, FVS191cys-dgRTA, and B43-dgRTA. FVS191CDRTA did not show cytotoxic activity to CD19+ Blin-1 and Nalm-6 cells as shown by Figure 5. To further evaluate the lack of activity of FVS191CDRTA, a cell-free transcription and translation system of rabbit reticulocyte lysates (Promega) was used. The results showed no evidence of catalytic activity of RTA in the reticulocyte lysates system (data not shown). However, FVS191cys-dgRTA inhibited protein synthesis of Blin-1 (IC50 ) 1.3 × 10-9 M) and Nalm-6 cells (IC50 ) 1.0 × 10-9 M) as shown by Figure 6. With the same cell lines, dgRTA (IC50 > 1 × 10-7 M) is 100-fold

Figure 5. Protein synthesis inhibition assay of FVS191CDRTA: (a) Blin-1 cells and (b) Nalm-6 cells; FVS191CDRTA (0), FVS191 (O), and dgRTA (b). The percentage of [3H]leucine incorporation was calculated using the immunotoxin-untreated cells as controls. Data were based on either duplicate or triplicate experiments. The standard deviation is indicated by the vertical bar.

less toxic than FVS191cys-dgRTA. FVS191cys, however, did not show any cytotoxic activity to Blin-1 and Nalm-6. B43-dgRTA was also potent to Blin-1 (IC50 ) 2 × 10-11 M) and Nalm-6 (IC50 ) 3 × 10-11 M) (Figure 7). B43 alone was not cytotoxic to the above two cell lines. Internalization of FVS191 and B43 by Nalm-6 Cells. Both FVS191 and B43 were internalized by CD19 positive Nalm-6 cells as shown by Figure 8. The internalizations of B43 occurred in the first hour of incubation, and the amount of internalized FVS191 increased from time 0 to the fourth hour of incubation. DISCUSSION

In this study, we have constructed and compared two different kinds of anti-CD19 scFv immunotoxins: FVS191CDRTA and FVS191cys-dgRTA. The disulfidelinked FVS191cys-dgRTA, but not the FVS191CDRTA, was cytotoxic to CD19 positive cells. The results demonstrating that FVS191CDRTA did not bind to CD19 positive cells and did not inactivate translation in a cellfree transcription and translation system strongly sug-

882 Bioconjugate Chem., Vol. 8, No. 6, 1997

Figure 6. Protein synthesis inhibition assay of FVS191cys-dgRTA: (a) Blin-1 cells and (b) Nalm-6 cells; FVS191cysdgRTA (0), FVS191 (O), and dgRTA (b). The percentage of [3H]leucine incorporation was calculated using the immunotoxinuntreated cells as standards. Data were based on either duplicate or triplicate experiments. The standard deviation is indicated by the vertical bar.

gest that FVS191CDRTA was not correctly refolded. It is not clear why FVS191CDRTA could not be refolded with the method used in this study; the amino acid sequence, the size, and the chimeric nature of FVS191CDRTA may contribute to the difficulty of refolding. In comparison to FVS191CDRTA, FVS191cys-dgRTA was produced using the refolded FVS191cys; therefore, the formation of FVS191cys-dgRTA avoids the difficulty of renaturation of the whole immunotoxin. The chemical conjugation provides an alternate method when a recombinant scFv immunotoxin cannot be refolded or when the yield of a recombinant scFv immunotoxin is extremely low. Our study shows that disulfide-linked scFv immunotoxin can be produced relatively efficiently; ∼70% of FVS191cys and dgRTA can form FVS191cys-dgRTA. The yield of refolded FVS191cys from 1 L of cell culture is from 1 to 0.5 mg using the method described in this study. The amount of FVS191cys may limit the largescale production of FVS191cys-dgRTA. One way to increase the yield of FVS191cys could be to express

Wang et al.

Figure 7. Protein synthesis inhibition assay of B43-dgRTA: (a) Blin-1 cells and (b) Nalm-6 cells; B43-dgRTA (0), B43 (2), and dgRTA (b). The percentage of [3H]leucine incorporation was calculated using the immunotoxin-untreated cells as controls. Data were based on either duplicate or triplicate experiments. The standard deviation is indicated by the vertical bar.

FVS191cys in the methylotropic yeast Pichia pastoris (21). An anti-CD7 scFvcys fragment, 3A1F, has been reportedly produced (60 mg/L) in P. pastoris (22). The disulfide-linked scFv immunotoxins are homogeneous and smaller in size compared with intact antibodyconjugated immunotoxin. The conjugation of FVS191cys to dgRTA is one to one, unlike the antibody-conjugated molecules which may have more than one dgRTA. FVS191cys-dgRTA is only one-third of the size of B43dgRTA; thus, it may penetrate tumor tissue better (2326). FVS191cys-dgRTA has a lower cytotoxic activity than B43-dgRTA. Several variables can influence the potency of an immunotoxin, such as the affinity of the immunotoxin to the target, the epitope on the target antigen recognized by the immunotoxin, the rate of endocytosis, and intracellular routing of the immunotoxin (27). The ELISA and internalization assay described in this paper were carried out to identify the factors influencing the cytotoxity of FVS191cys-dgRTA. The ELISA result reveals no differences of CD19-binding ability between FVS191cys-dgRTA and B43 (Figure 4). This indicates that the addition of a disulfide bond to FVS191cys does

Anti-CD19 scFv Immunotoxin

Bioconjugate Chem., Vol. 8, No. 6, 1997 883

dgRTA and B43-dgRTA is due to their differences in intracellular routing. The efficacy of an immunotoxin in vivo is a combination of several effects, such as the stability and the tumor penetration ability of an immunotoxin. Whether FVS191cys-dgRTA will have an efficacy in vivo needs to be studied further. In summary, we compared two different methods to generate scFv immunotoxin and demonstrated that the formation of a disulfide-linked protein is an effective approach to produce anti-CD19 scFv immunotoxin. To our knowledge, FVS191cys-dgRTA is the first anti-CD19 scFv immunotoxin showing cytotoxic activity. However, more work is necessary to develop anti-CD19 scFv immunotoxins with improved cytotoxity. One possible approach is the development of an immunotoxin made of dimeric scFv. An immunotoxin with dimeric scFv potentially retains the advantages of bivalent binding to antigen and a smaller size in comparison to intact antibody immunotoxin; the reduced size of an immunotoxin should enhance its tumor penetration. LITERATURE CITED

Figure 8. Internalization assay: (a) internalization of 125Ilabeled FVS191 by Nalm-6 [the percentage radioactivities associated with cell-surface bound FVS191 (9), internalized FVS191 (4), dissociated FVS191 (b), and degraded FVS191 (0) were calculated on the bases of total CPM at each time point]; (b) internalization of 125I-labeled B43 by Nalm-6 cells [the percentage radioactivities associated with cell-surface bound B43 (0), internalized B43 (4), dissociated B43 (9), and degraded B43 (b) were calculated on the bases of total CPM at each time point and plotted against each time point].

not interfere with its antigen-binding activity. The internalization studies of FVS191 and B43 demonstrated that both FVS191 and B43 were internalized by Nalm6, although the internalization patterns were slightly different. This result suggests that cross-linking of CD19 is not required for CD19 endocytosis. The extent of internalization of FVS191cys-dgRTA needs further study. A study of anti-CD22-ricin A immunotoxin conducted by Horssen et al. (28) demonstrates that the cytotoxity of CD22-ricin A depends on intracellular routing rather than on the number of internalized molecules. Another study conducted by May et al. (27) also reveals that intracellular routing rather than cross-linking or rate of internalization determines the potency of immunotoxins directed against different epitopes of sIgD on murine B cells. These studies suggest that intracellular routing is an important factor in determining the potency of an internalized immunotoxin. It is very likely that the discrepancy in cytotoxic activities between FVS191cys-

(1) Brinkmann, U., Pai, L. H., FitzGerald, D. J., Willingham, M., and Pastan, I. (1991) B3(Fv)-PE38KDEL, a single-chain immunotoxin that causes complete regression of a human carcinoma in mice. Proc. Natl. Acad. Sci. U.S.A. 88, 86168620. (2) Seetharam, S., Chaudhary, V. K., FitzGerald, D., and Pastan, I. (1991) Increased cytotoxic activity of Pseudomonas exotoxin and two chimeric toxins ending in KDEL. J. Biol. Chem. 266 (26), 17376-17381. (3) Batra, J. K., Kasprzyk, P. G., Bird, R. E., Pastan, I., and King, R. C. (1992) Recombinant anti-erbB2 immunotoxins containing Pseudomonas exotoxin. Proc. Natl. Acad. Sci. U.S.A. 89 (13), 5867-5871. (4) Chaudhary, V. K., Queen, C., Junghans, J. P., Waldmann, T. A., FitzGerald, D. J., and Pastan, I. (1989) A recombinant immunotoxin consisting of two antibody variable domains fused to Pseudomonas exotoxin. Nature (London) 339, 394393. (5) Friedman, P. N., McAndrew, S. J., Gawlak, S. L., Chance, D., Trail, P. A., Brown, J. P., and Siegall, C. B. (1993) BR96 sFv-PE40, a potent single-chain immunotoxin that selectively kills carcinoma cells. Cancer Res. 53, 334-339. (6) Chauldhary, V. K., Gallo, M. G., FitzGerald, D. J., and Pastan, I. (1990) A recombinant single-chain immunotoxin composed of anti-Tac variable regions and a truncated diphtheria toxin. Proc. Natl. Acad. Sci. U.S.A. 87, 9491-9494. (7) Pauza, M. E., Doumbia, S. O., and Pennell, C. A. (1997) Construction and characterization of human CD7-specific single chain Fv-immunotoxins. J. Immunol. 158, 3259-3269. (8) Ling, N. R., Maclennan, and Mason, D. T. (1987) B-cell and plasma cell antigens: new and previously defined clusters Leucocyte Typing III. White Cell Differentiation Antigens (A. J. McMichael, Ed.) pp 302-335, Oxford University Press, Oxford, U.K. (9) Scheuerman, R. H., and Racila, E. (1995) CD19 antigen in leukemia and lyphoma diagnosis and immunotherapy. Leukemia Lymphoma 18, 140-151. (10) Becjek, B., Wang, D., Berven, E., Pennell, C. A., Peiper, S. C., Poppema, S., Uckun, F. M., and Kersey, J. H. (1995) Development and characterization of three recombinant single chain antibody fragments (scFv) directed against the CD19 antigen. Cancer Res. 55, 2346-2351. (11) Wang, D., Berven, E., Li, Q., Uckun, F., and Kersey, J. H. (1997) Optimization of conditions for formation and analysis of anti-CD19 FVS191 single-chain Fv homodimers (scFv′)2. Bioconjugate Chem. 8 (1), 64-70. (12) Conry, R. M., Khazaeli, M. B., Saleh, M. N., Ghetie, V., Vitetta, E. S., Liu, T., and LoBuglio, A. F. (1996) Phase I trial of an anti-CD19 deglycosylated ricin A chain immunotoxin

884 Bioconjugate Chem., Vol. 8, No. 6, 1997 in non-Hodgkin’s lymphoma: Effect of an intensive schedule of administration. J. Immunother. 18 (4), 231-241. (13) Halling, R. C., Halling, A. C., Murray, E. E., Ladin, B. F., Houston, L. L., and Weaver, R. F. (1985) Genomic cloning and characterization of a ricin gene from Ricinus communis. Nucleic Acid Res. 13 (22), 8019-8033. (14) Lojda, Z., Smidova, J., Barth, A., and Ueberberg, H. (1988) Are Z-Arg-Gly-Phe-Phe-Leu-MNA and Z-Arg-Gly-Phe-Phepro-MNA suitable substrates for the demonstration of cathepsin D activity? Histochemistry 88, 505-512. (15) Buchner, J., Pastan, I., and Brinkman, U. (1992) A method for increasing the yield of properly refolded fusion proteins: single-chain immunotoxins renaturation of bacterial inclusion bodies. Anal. Biochem. 205, 263-270. (16) Kuwata, K., Uebori, M., Yamada, K., and Yamazaki, Y. (1982) Liquid chromatagraphic determination of alkylthiols via derivatization with 5,5′-dithiobis(2-nitrobenzoic acid). Anal. Chem. 54, 1082-1087. (17) Ghetie, M. A., May, R. D., and Till, M. (1988) Evaluation of ricin a chain containing immunotoxins directed against CD19 and CD22 antigens on normal and malignant human B cells as potential reagents for in vivo therapy. Cancer Res. 48, 2610-2617. (18) Knowes, P. P., and Thorpe, P. E. (1987) Purification of immunotoxins containing ricin A chain and abrin A chain using blue sepharose CL-6B. Anal. Biochem. 160, 440-443. (19) Siegall, C. B., Chace, D., Mixan, B., Garrigues, U., Wan, H., Paul, L., Wolff, E., Hellstrom, I., and Hellstrom, K. E. (1994) In vitro and in vivo characterization of BR96 scFvPE40. J. Immunol. 152 (2), 2376-2484. (20) Press, O. W., Farr, A. G., Borroz, K. I., Anderson, K. A., and Martin, P. J. (1989) Endocytosis and degradation of monoclonal antibodies targeting human B-cell malignancies. Cancer Res. 49, 4906-4949.

Wang et al. (21) Cregg, J. M., Vedvick, T. S., and Raschke, W. C. (1993) Recent advances in the expression of foreign genes in Pichia pastoris. Bio/Technology 11, 905. (22) Eldin, P., Pauza, M. E., Hieda, Y., Lin, G., Murtaugh, M. P., Pentel, P. R., and Pennell, C. A. (1997) High-level secretion of two antibody single chain Fv fragments by Pichia pastoris. J. Immunol. Methods 201, 67-75. (23) Jain, R. K. (1989) Delivery of novel therapeutic agents in tumors: physiological barriers and strategies. J. Natl. Cancer Inst. 81, 570-576. (24) Fujimori, K., Covell, D. J., Flethcher, J. E., and Weinstein, J. N. (1989) Modeling analysis of the global and microscopic distribution of immunoglobulin G1, F(ab′)2, and Fab in tumors. Cancer Res. 49, 5656-5663. (25) Sung, C., Youle, R. J., and Dedrick, R. L. (1990) Phamacokinetic analysis of immunotoxin uptake in solid tumors: role of plasma kinetics, capillary, permeability, and binding. Cancer Res. 46, 3960-3978. (26) Covell, D. G., Barbet, J., Holton, O. D., Black, C. D., and Parker, R. J. (1986) Pharmacokinetics of monoclonal immunoglobulin G1, F(ab′)2, and Fab in mice. Cancer Res. 46, 3969-3978. (27) May, R. D., Wheeler, H. T., Finkelman, F. D., Uhr, J. W., and Vitetta, E. S. (1991) Intracellular routing rather than cross-linking or rate of internalization determines the potency of immunotoxins directed against different epitopes of sIgD on murine B cells. Cell. Immunol. 135, 490-500. (28) Horssen, P. J. V., Van Oosterhout, Y. V., Witte, D. E., and Preijers, F. W. M. B. (1995) Cytotoxic potency of CD22-ricin A depends on internalized molecules. Scand. J. Immunol. 41, 563-569. BC970071W