Int. J. Cancer: 78, 106–111 (1998)
`r 1998 Wiley-Liss, Inc.
`
`Publication of the International Union Against Cancer
`Publication de l’Union Internationale Contre le Cancer
`
`TARGETING TUMOR CELLS VIA EGF RECEPTORS: SELECTIVE TOXICITY
`OF AN HBEGF-TOXIN FUSION PROTEIN
`Lois A. CHANDLER1*, Barbara A. SOSNOWSKI1, John R. MCDONALD1, Janet E. PRICE2, Sharon L. AUKERMAN1, Andrew BAIRD1,
`Glenn F. PIERCE1 and L.L. HOUSTON1
`1Selective Genetics, Inc., San Diego, CA, USA
`2University of Texas M.D. Anderson Cancer Center, Houston, TX, USA
`
`Over-expression of the epidermal growth factor receptor
`(EGFR) is a hallmark of numerous solid tumors, thus provid-
`ing a means of selectively targeting therapeutic agents.
`Heparin-binding epidermal growth factor (HBEGF) binds to
`EGFRs with high affinity and to heparan sulfate proteogly-
`cans, resulting in increased mitogenic potential compared to
`other EGF family members. We have investigated the feasibil-
`ity of using HBEGF to selectively deliver a cytotoxic protein
`into EGFR-expressing tumor cells. Recombinant fusion pro-
`teins consisting of mature human HBEGF fused to the plant
`ribosome-inactivating protein saporin (SAP) were expressed
`in Escherichia coli. Purified HBEGF-SAP chimeras inhibited
`protein synthesis in a cell-free assay and competed with EGF
`for binding to receptors on intact cells. A construct with a
`22-amino-acid flexible linker (L22) between the HBEGF and
`SAP moieties exhibited an affinity for the EGFR that was
`comparable to that of HBEGF. The sensitivity to HBEGF-L22-
`SAP was determined for a variety of human tumor cell lines,
`including the 60 cell lines comprising the National Cancer
`Institute Anticancer Drug Screen. HBEGF-L22-SAP was cyto-
`toxic in vitro to a variety of EGFR-bearing cell
`lines and
`inhibited growth of EGFR-over-expressing human breast
`carcinoma cells in vivo. In contrast, the fusion protein had no
`effect on small-cell lung carcinoma cells, which are EGFR-
`deficient. Our results demonstrate that fusion proteins com-
`posed of HBEGF and SAP exhibit targeting specificity and
`cytotoxicity that may be of therapeutic value in treating a
`variety of EGFR-bearing malignancies. Int. J. Cancer 78:106–
`111, 1998.
`r 1998 Wiley-Liss, Inc.
`
`Numerous reports have demonstrated the ability of chimeric
`molecules containing growth factor and toxin moieties to selec-
`tively target and inhibit the growth of cells bearing their cognate
`growth factor receptors. These chimeras, termed mitotoxins or
`oncotoxins, bind to the surface of target cells via a growth factor
`receptor, are internalized by receptor-mediated endocytosis and
`trigger cell death. Several plant proteins that are toxic to eukaryotic
`cells as a result of catalytic inactivation of ribosomes have been
`identified (Stirpe et al., 1992). The eukaryotic cytotoxicity of these
`plant ribosome-inactivating proteins (RIPs) has been exploited to
`create potent mitotoxins. For example, basic fibroblast growth
`factor (FGF2) has been linked to saporin (SAP) to target FGF
`receptor-expressing cells (McDonald et al., 1996). This FGF2-SAP
`chimera exhibits targeting specificity to FGF receptors and high
`cytotoxicity in vitro and in vivo for transformed cells (Beitz et al.,
`1992), lens epithelial cells (Behar-Cohen et al., 1995) and smooth
`muscle cells (Farb et al., 1997).
`The epidermal growth factor (EGF) ligand family includes EGF,
`transforming growth factor alpha (TGF-␣), amphiregulin, heparin-
`binding EGF (HBEGF) and related polypeptides that stimulate
`approx. 170-kDa transmembrane glycoprotein receptors with tyro-
`sine kinase activity. Up-regulated expression of EGFRs and EGFR
`ligands has been implicated in numerous pathological conditions
`and is presumed to be one means by which cells acquire a selective
`growth advantage. For example, EGFR over-expression has been
`demonstrated in numerous tumor cell
`lines and solid human
`tumors, including glioblastoma, melanoma and lung, breast and
`bladder carcinomas, and is associated with reduced survival rates in
`human breast and bladder cancers (Davies and Chamberlin, 1996).
`
`Accordingly, EGFRs are an attractive target for the delivery of
`therapeutic agents.
`HBEGF was first identified as a 22-kDa secreted product of
`cultured human macrophages and has since been identified in a
`variety of tissues (Raab and Klagsbrun, 1997). HBEGF is derived
`from a transmembrane-anchored precursor from which proteolysis
`generates a family of 8- to 10-kDa biologically active proteins.
`Like other members of the EGF family, HBEGF binds to EGFRs
`and triggers receptor tyrosine kinase activity and cell proliferation.
`Unlike EGF and TGF-␣, HBEGF also interacts with cell-surface
`heparan sulfate proteoglycans, thus increasing its mitogenic poten-
`tial (Raab and Klagsbrun, 1997).
`Because of its increased potency compared to other EGF family
`members, we examined the feasibility of using HBEGF to target
`cytotoxic agents to EGFR-bearing cells. We report here that novel
`mitotoxins employing mature human HBEGF as the targeting
`agent and SAP as the cytotoxic agent can effectively compete with
`EGF for receptor binding and show selective cytotoxicity, in vitro
`and in vivo, to tumor cells bearing EGFRs.
`
`MATERIAL AND METHODS
`Plasmid construction
`All enzymes were obtained from Boehringer-Mannheim (India-
`napolis, IN). DNA fragments were purified from agarose gels using
`the Geneclean II kit (Bio 101, Vista, CA). PCR and sequencing
`primers were synthesized on a Cyclone Plus DNA Synthesizer
`(Millipore, Bedford, MA). All PCR-generated fragments were
`sequenced using Sequenase Version 2.0 (Amersham, Arlington
`Heights, IL).
`PCR was used to generate 2 overlapping HBEGF fragments
`from a plasmid containing cDNA encoding human HBEGF
`(pJMU2–1; generously provided by Dr. J. Abraham, Scios Nova,
`Mountain View, CA; GenBank accession number M60278). Inter-
`nal NcoI sites were mutated without changing the amino acid
`composition of the corresponding protein. The 5’ HBEGF fragment
`spans codons 13–129 in the HBEGF precursor and was generated
`with the ‘‘sense’’ primer 58-CTG GCT GCA GTT CTC TCG
`GCA-38, which contains a PstI site, and the anti-sense primer
`58-AGC CCG GAG CTC CTT CAC ATA TTT GCA TTC TCC
`GTG GAT GCA GAA-G-38. The 38 HBEGF fragment spans from
`codon 124 to downstream of the HBEGF-coding region and was
`generated using the sense primer 58-GTG AAG GAG CTC CGG
`GCT CCC TCC TGC ATC TGC CAC CCG GGT TAT CAT GGA
`GAG AGG-38 and the anti-sense primer 58-ATA TAG AAT TCT
`GTC TTC TCA GAG GTA-38, which contains an EcoRI site. The 2
`PCR-generated fragments overlapped at a SacI site and were
`ligated into the PstI and EcoRI sites of the vector pGEM-4
`(Promega, Madison, WI). Using this plasmid as a template, the
`region encoding mature HBEGF (amino acids 73–149 in the
`precursor) was amplified by PCR using the sense primer 58-CTG
`
`*Correspondence to: Selective Genetics, Inc., 11035 Roselle Street, San
`Diego, CA 92121, USA. Fax: (619) 625–0050.
`E-mail: loisc@selectivegenetics.com
`
`Received 3 March 1998; Revised 27 April 1998
`
`IMMUNOGEN 2157, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`TARGETING TUMOR CELLS WITH HBEGF
`
`107
`
`GAC CAT ATG AGA GTC ACT TTA-38, which adds an NdeI site
`and a methionine codon to the 5’ end of HBEGF, and an anti-sense
`primer 58- ATA TAC CAT GGC TGG GAG GCT CAG CCC ATG
`ACA-38, which introduces an NcoI site immediately downstream
`of the coding region for HBEGF. The plasmid pZ1B contains the
`gene encoding the mitotoxin FGF2-SAP cloned into the prokary-
`otic expression vector pET11a (Novagen, Madison, WI) (McDon-
`ald et al., 1996). FGF2 sequences were specifically removed from
`pZ1B by digestion with NdeI and NcoI and replaced with the
`HBEGF sequence. The resulting plasmid encodes a fusion protein
`(designated HBEGF-SAP) that consists of 78 amino acids of
`HBEGF, an Ala-Met spacer encoded by the NcoI site and 253
`amino acids of SAP, has a predicted m.w. of 37.6 kDa and an
`isoelectric point of 9.6.
`Complementary single-stranded oligos were synthesized (Mid-
`land Certified Reagent Company, Midland, TX) and annealed to
`generate a linker with NcoI ends that encoded (Gly-Gly-Gly-Gly-
`Ser)x4. The sequence of the sense oligo was 58-CTA TAC CAT
`GGG CGG CGG CGG CTC TGG CGG CGG CGG CTC TGG
`CGG CGG CGG CTC TGG CGG CGG CGG CTC TGC CAT
`GGT ATA-38. The plasmid encoding HBEGF-SAP was digested
`with NcoI, which cuts between the HBEGF and SAP moieties, and
`the NcoI-digested linker was ligated into the plasmid. Sequencing
`of the resulting plasmid revealed a 22-amino-acid linker encoding
`Ala-Met-Gly4-Ser-Gly2-Ser-Gly4-Ser-Gly4-Ser-Ala-Met (L22). The
`divergence from the intended linker sequence was likely caused by
`a secondary structure formed during primer annealing because of
`sequence redundancy and high GC content of the primers. HBEGF-
`L22-SAP is 353 amino acids long, with a predicted m.w. of 38.9
`kDa and an isoelectric point of 9.6.
`
`Expression and purification of HBEGF-SAP fusion proteins
`For expression, plasmid DNAs were transformed into the
`Escherichia coli strain BL21(DE3) (Novagen), which contains
`chromosomal copies of the T7 RNA polymerase gene linked to an
`IPTG-inducible lacUV promoter. Transformed cells were grown in
`a 7L Applikon (Foster City, CA) fermenter in complex batch media
`as previously described for FGF2-SAP (McDonald et al., 1996).
`The culture was induced with 0.1 mM IPTG (isopropyl ␤-D-
`thiogalactopyranoside) at an A600 of 85. Cells were harvested by
`centrifugation (8,000 g, 10 min) 4 hr after induction, and the culture
`paste was stored at –80°C until ready for processing.
`The HBEGF-SAP fusion proteins were purified using steps
`similar to those employed for purification of recombinant FGF2-
`SAP (McDonald et al., 1996). Briefly, bacterial pellets were
`resuspended in 3–4 volumes of 10 mM sodium phosphate, pH 6,
`containing 10 mM EDTA, 10 mM EGTA and 50 mM NaCl, then
`passed 3 times through a microfluidizer (Microfluidics, Newton,
`MA) at 18,000 lb/in2. Lysates were subjected to expanded-bed
`adsorption chromatography using Streamline cation-exchange resin
`(Pharmacia, Piscataway, NJ). Partially purified fusion proteins
`were then purified to homogeneity by a combination of anion-
`exchange (Q-Sepharose, FF), cation-exchange (SP-Sepharose, HP)
`and hydrophobic interaction (Phenyl-Sepharose, HP) chromatogra-
`phies before buffer exchanging the purified proteins into 10 mM
`sodium citrate, pH 6, containing 0.1 mM EDTA and 0.14 M NaCl
`by size-exclusion chromatography (Sephacryl S100, HR). The final
`materials were over 98% pure, as evaluated by SDS-PAGE,
`reverse-phase HPLC and size-exclusion HPLC (data not shown).
`All purified proteins were stored at –80oC.
`
`Human tumor cell lines
`The cell lines employed in this study were obtained from 3
`sources. The 60 cell lines comprising the NCI Anticancer Screen
`have been described previously (Monks et al., 1991). The colon
`lines N87 and SW948 and the ovarian lines OVCAR2 and
`OVCAR10 were generously provided by Dr. L.M. Weiner (Fox
`Chase Cancer Center, Philadelphia, PA). N87, OVCAR2 and
`OVCAR10 cells were maintained in RPMI 1640. SW948 cells
`were maintained in L-15 medium. All other cell lines were obtained
`
`from the ATCC (Rockville, MD) and maintained according to the
`conditions specified by the ATCC. MDA-MB-468 cells were
`maintained in medium containing 5% FBS, and all other cells were
`maintained in medium containing 10% FBS (Hyclone, Logan UT).
`
`Protein synthesis inhibition assays
`The protein synthesis-inhibitory activities of the fusion proteins
`and of native SAP were determined by measuring effects on in vitro
`translation of luciferase RNA. SAP was purified from the seeds of
`Saponaria offıcinalis as previously described (Stirpe et al., 1983).
`Reagents for the assay were obtained from Promega unless
`specified otherwise. Briefly, samples were serially diluted in 20
`mM Tricine, pH 7.8 (Sigma, St. Louis, MO), and 5 µl of diluted
`protein were combined with 5 µl of reaction mix (50 µg/ml
`luciferase RNA, 0.1 mM amino acid mixture minus leucine, 0.1
`mM amino acid mixture minus methionine) and 15 µl of rabbit
`reticulocyte lysate. Samples were incubated for 1 hr at 30°C and
`diluted 1:11 in 20 mM Tricine, pH 7.8; 10 µl were transferred in
`duplicate to a Dynatech (Chantilly, VA) 96-well plate. Using a
`Dynatech ML3000 luminometer that was pre-warmed to 30°C, 50
`µl of luciferase assay reagent were injected per well and relative
`light units were determined at room temperature with a 1-sec delay,
`a 5-sec integration and medium gain.
`
`Receptor-binding studies
`The relative affinities of the fusion proteins for the EGFR were
`determined by binding competition assays. A431 cells were plated
`onto 24-well plates at 16,000 cells per well. The next day, plates
`were put on ice for 10 min, then the media was aspirated and the
`cells were washed with 0.5 ml of ice-cold binding buffer (DMEM
`containing 1 mg/ml BSA and 50 mM N,N-bis[2-hydroxy-ethyl]-2-
`aminoethane sulfonic acid). Sample wells were treated in duplicate
`with 200 µl of binding buffer, 50 µl of test protein (or binding buffer
`for controls) and 50 µl (0.075 µCi) of [125I]-EGF (Amersham).
`Recombinant human HBEGF was obtained from R&D Systems
`(Minneapolis, MN). Plates were incubated for 2 hr at 4°C with
`gentle agitation. Cells were then washed 3 times with cold binding
`buffer and lysed by the addition of 500 µl of lysis buffer (10 mM
`Tris, pH 8; 0.5% SDS; 1 mM EDTA). The radioactivity in each
`sample was determined using a Cobra II auto gamma counter
`(Packard, Meriden, CT).
`
`In vitro cytotoxicity assays
`HBEGF-L22-SAP was submitted to the National Cancer Institute
`Anticancer Screen (NSC D672640) to measure growth inhibition in
`a panel of 60 human tumor cell lines representing 9 tumor types
`(Monks et al., 1991). Cells were treated in duplicate over a range of
`concentrations from 0.012 to 120 nM, and after 48 hr a sulforhoda-
`mine B (SRB) assay was performed (Monks et al., 1991). All other
`cell lines were evaluated for sensitivity to HBEGF-L22-SAP, using
`the MTT colorimetric assay, as previously described (McDonald et
`al., 1996). Briefly, culture media were removed 24 hr after plating
`and duplicate wells were treated with fresh media containing
`serially diluted test samples. After an additional 48 hr, cell survival
`was estimated by the ability of live cells to reduce MTT.
`Absorbance was read at 550 nm on a Molecular Devices (Sun-
`nyvale, CA) plate reader using Softmax Pro software.
`
`Inhibition of tumor growth in vivo
`MDA-MB-468 human breast cancer cells (2 x 106) in 0.1 ml of
`PBS were injected into the thoracic mammary fatpad of female
`nude mice. Tumor growth was monitored weekly, using calipers to
`measure 2 orthogonal diameters. When mean tumor diameters
`reached 5 mm (day 52), HBEGF-L22-SAP (10 ng/g in 0.05 ml PBS)
`was injected s.c. adjacent to the tumors twice weekly for 4 weeks.
`The treated animals showed no signs of drug-induced toxicity.
`Tumor growth was monitored for up to 35 days after the last
`treatment. At the end of the experiment, mice were killed and
`residual tumors weighed. Student’s t-test was used to evaluate the
`statistical significance of differences in tumor weight. Mice were
`maintained in specific pathogen-free conditions in a facility
`
`IMMUNOGEN 2157, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`108
`
`CHANDLER ET AL.
`
`approved by the AALAC (Rockville, MD). Animal care and
`experimental procedures were in accordance with the regulations
`and standards of the USDA, DHSS and NIH.
`
`RESULTS
`Expression and purification of recombinant
`HBEGF-SAP mitotoxins
`A bacterial expression system was employed for the production
`of HBEGF-SAP fusion proteins. The initial HBEGF-SAP fusion
`protein consisted of a 78-amino-acid isoform of HBEGF, an
`Ala-Met linker and the coding sequence of SAP. Because the
`receptor-binding domain is in the C-terminal half of HBEGF, we
`reasoned that receptor binding, and thus activity of the mitotoxin,
`may be improved by introducing a long, flexible linker between the
`HBEGF and SAP moieties. Therefore, DNA encoding a flexible
`linker consisting of AlaMetGly4SerGly2SerGly4SerGly4SerAlaMet
`(L22) was inserted between the sequences encoding HBEGF and
`SAP. E. coli fermentation resulted in accumulation of appropriately
`sized proteins that were readily detectable in crude lysates by
`Western blotting with anti-SAP anti-sera (data not shown). Fusion
`proteins were purified from the soluble fraction of bacterial lysates
`using a straightforward strategy that yielded highly pure material
`(see ‘‘Material and Methods’’).
`
`In vitro activities of HBEGF-SAP mitotoxins
`HBEGF-SAP and HBEGF-L22-SAP were tested for the ability to
`inhibit protein synthesis in a cell-free system (Fig. 1). Native SAP
`exhibited an IC50 of 7 pM in this assay, and although significantly
`less active than native SAP, HBEGF-SAP showed a dose-
`dependent inhibition of protein synthesis with an IC50 of 0.16 nM.
`HBEGF-L22-SAP exhibited activity comparable to that of HBEGF-
`SAP (IC50 ⫽ 0.15 nM), suggesting that the flexible linker had no
`effect on the RIP activity of the fusion protein.
`A431 is a human epidermoid carcinoma cell line that expresses
`high numbers of cell-surface EGFRs (approx. 3 x 106/cell).
`HBEGF-SAP and HBEGF-L22-SAP were tested for the ability to
`compete with [125I]-labeled EGF for binding to EGFRs on human
`A431 cells (Fig. 2). While HBEGF-SAP competed for receptor
`
`binding, though not as effectively as HBEGF, HBEGF-L22-SAP
`was nearly as effective as recombinant HBEGF. Therefore, inser-
`tion of a long, flexible linker between HBEGF and SAP is
`favorable for EGFR binding. Taken together, the RIP and receptor-
`binding assays demonstrate that while the HBEGF and SAP
`
`FIGURE 2 – Receptor-binding analysis of HBEGF-SAP mitotoxins.
`The ability of rHBEGF and the HBEGF-SAP mitotoxins to compete
`with [125I]-EGF for receptor binding on A431 cells was determined as
`described in ‘‘Material and Methods’’. The plotted values are the
`means ⫾ SD (n ⫽2), and the data are representative of at least 2
`independent experiments.
`
`FIGURE 1 – Inhibition of protein synthesis in vitro by HBEGF-SAP
`mitotoxins. The inhibition of luciferase RNAtranslation in a cell-free system
`was measured as described in ‘‘Material and Methods’’. The plotted values
`are the means ⫾ SD (n ⫽2) and the data are representative of 2 independent
`experiments. The calculated IC50 values are saporin, 0.007 nM; HBEGF-
`SAP, 0.16 nM; HBEGF-L22-SAP, 0.15 nM.
`
`FIGURE 3 – In vitro cytotoxicity of HBEGF-SAP mitotoxins to A431
`cells. The cytotoxic activities of HBEGF-SAP and HBEGF-L22-SAP to
`human A431 epidermoid carcinoma cells were compared to that of SAP
`alone. The plotted values are the means ⫾ SD (n ⫽2), and the data are
`representative of at least 2 independent experiments. The calculated
`IC50 values are HBEGF-SAP, 0.7 nM; HBEGF-L22-SAP, 0.3 nM.
`
`IMMUNOGEN 2157, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`TARGETING TUMOR CELLS WITH HBEGF
`
`109
`
`moieties retain their individual functions, steric hindrance can
`compromise receptor binding.
`The linker in HBEGF-L22-SAP most likely improves receptor
`binding by reducing the steric constraints on the C-terminal
`receptor-binding domain of HBEGF. We therefore reasoned that
`the effectiveness of the HBEGF-SAP mitotoxins might be im-
`proved by reversing the orientation of the protein moieties, thereby
`making the C-terminus of HBEGF more readily available for
`receptor binding. Plasmids that encode SAP-HBEGF fusion pro-
`teins with and without the long, flexible linker were constructed.
`These fusion proteins were highly unstable, as shown by Western
`blots developed with anti-SAP anti-sera (data not shown), and one
`of the reactive bands co-migrated with free SAP. Consistent with
`breakdown to free SAP, these materials were nearly as active as
`native SAP in the cell-free protein synthesis inhibition assay (data
`not shown). Due to their marked instability, the SAP-HBEGF
`molecules were not further characterized.
`HBEGF-SAP and HBEGF-L22-SAP were also evaluated for
`cytotoxicity to A431 cells (Fig. 3). HBEGF-SAP had a significant,
`
`TABLE I – CYTOTOXICITY OF HBEGF-L22-SAP TO HUMAN TUMOR CELL LINES1
`
`Cell line
`
`IC50 (nM)
`
`Cell line
`
`IC50 (nM)
`
`Epidermoid
`*A431
`*KB
`Bladder
`*EJ6
`*HT1197
`*J82
`*RT4
`*T24
`*TCCSUP
`Leukemia
`SR
`RPMI 8226
`HL-60 TB
`MOLT-4
`K-562
`CCRF-CEM
`Non-small cell lung
`A549
`NCI-H226
`NCI-H23
`NCI-H522
`NCI-H460
`NCI-H322
`HOP-92
`HOP-62
`EKVX
`Prostate
`DU-145
`PC-3
`Melanoma
`UACC-257
`SK-MEL-5
`UACC-62
`SK-MEL-28
`M14
`MALME-3M
`LOX-IMVI
`Renal
`A498
`SN-12C
`UO-31
`TK-10
`RXF-393
`CAKI-1
`ACHN
`786-0
`
`0.3
`3.9
`
`10
`1
`1.6
`1.5
`17
`5
`
`3.1
`10.3
`15.1
`⬎73
`4.6
`⬎73
`
`7.8
`0.1
`12.7
`7.4
`13.9
`4.7
`4.7
`13
`⬎73
`
`10.1
`0.04
`
`5.8
`0.3
`1.8
`0.8
`0.2
`0.2
`1.5
`
`0.05
`5.6
`4.2
`59
`1.2
`6.3
`2.6
`1.1
`
`Breast
`T47D
`BT-549
`MDA-N
`MDA-MB435
`MDA-MB231
`HS578T
`MCF-7/ADR
`MCF-7
`Colon
`KM-12
`HCT-15
`COLO-205
`SW-620
`HCT-116
`HCC-2998
`HT-29
`*N87
`*SW948
`Ovarian
`SKOV-3
`OVCAR-8
`IGROV-1
`OVCAR-5
`OVCAR-4
`OVCAR-3
`*OVCAR2
`*OVCAR10
`Brain
`SNB-19
`SF-539
`SF-295
`SF-268
`U-251
`SNB-75
`*A172
`*DAOY
`*H4
`*Hs683
`*T98G
`*U87MG
`Small cell lung
`NCI-H345
`NCI-H510
`NCI-H526
`NCI-H69
`
`36
`3.2
`2.4
`1.7
`⬎73
`0.6
`5.8
`27
`
`32
`36
`2.3
`42
`0.6
`1.1
`1.9
`6.6
`⬎100
`
`⬎73
`1.2
`2
`1.3
`16.2
`0.4
`⬎100
`12.1
`
`4
`0.2
`1.6
`0.2
`0.1
`4.4
`12.4
`2
`0.3
`⬎100
`8.4
`1
`
`⬎100
`⬎100
`⬎100
`⬎100
`
`1IC50 values represent the concentration of mitotoxin required to
`achieve 50% inhibition of cell growth as determined by the MTT assay
`for cells with an asterisk and the SRB assay for cells without an asterisk
`(see ‘‘Material and Methods’’).
`
`dose-dependent cytotoxic effect on A431 cells (IC50 ⫽ 0.7 nM),
`and complete inhibition of cell growth was achieved at high
`concentrations. HBEGF-L22-SAP exhibited a small increase in
`toxicity to A431 cells (IC50 ⫽ 0.3 nM) compared to HBEGF-SAP.
`Therefore, even though HBEGF-SAP and HBEGF-L22-SAP had
`significantly different activities in the A431 receptor-binding assay
`(Fig. 2), the toxicity of the 2 mitotoxins is comparable. These
`results illustrate the potency of RIPs such as SAP in that internaliza-
`tion of very few molecules is sufficient to kill a cell (Yamaizumi et
`al., 1978). A431 cells were resistant to SAP alone (IC50 ⬎100 nM),
`demonstrating the dependence on HBEGF for internalization of the
`toxin and effectiveness of the mitotoxin.
`
`Cytotoxicity of HBEGF-L22-SAP to tumor cells in vitro
`The cytotoxicity of HBEGF-L22-SAP was tested on a panel of
`human cell lines representing tumors of 12 different cell types,
`including the 60 cell lines comprising the NCI Anticancer Screen
`(Monks et al., 1991). The results of this analysis are summarized in
`
`FIGURE 4 – Inhibition of MDA-MB-468 tumor growth by HBEGF-L22-
`SAP. MDA-MB-468 tumor cells were injected into the mammary
`fatpad of a nude mouse, and effects of s.c. injection of HBEGF-L22-
`SAP adjacent to the tumor were assessed. a: Growth curves of tumors
`in mice treated with s.c. injection of HBEGF-L22-SAP (10 ng/g, twice
`weekly for 4 weeks). Plotted values are mean tumor diameters in
`millimeters (bar ⫽ SD). Arrows indicate period of treatment. Closed
`circles, PBS alone (n ⫽ 6). Open circles, HBEGF-L22-SAP (n ⫽ 7). b:
`Mean tumor weights at end of experiment presented in (a). p ⬍ 0.05
`(Students t-test).
`
`IMMUNOGEN 2157, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`110
`
`CHANDLER ET AL.
`
`Table I. Each of the cell lines was also treated with native SAP and
`found to be largely unaffected (data not shown). Cell lines derived
`from small-cell lung carcinoma (SCLC) were uniformly resistant to
`HBEGF-L22-SAP, with IC50 values greater than 100 nM. In
`contrast, many of the cell lines representing the remaining 11 tumor
`types (i.e. melanoma, breast, bladder, colon, ovarian) exhibited
`IC50 values in the low nanomolar range.
`Inhibition of tumor growth in vivo by HBEGF-L22-SAP
`Since many breast tumor cell lines were sensitive in vitro, the
`effect of HBEGF-L22-SAP on MDA-MB-468 human breast cancer
`cells was evaluated in vivo. MDA-MB-468 cells are estrogen
`receptor-negative, grow in the mammary fatpad of female nude
`mice without estrogen supplementation and have an amplified
`EGFR gene, resulting in unusually high numbers (approx. 3.6 x
`106) of cell-surface EGFRs (Filmus et al., 1985). With peri-tumoral
`administration of HBEGF-L22-SAP, the mean tumor diameter and
`mean tumor weight were significantly less than in the control
`groups (Fig. 4). In fact, treated tumors showed arrest of tumor
`growth lasting longer than 3 weeks after cessation of treatment.
`
`DISCUSSION
`
`In recent years, the over-expression of cell-surface growth factor
`receptors in human disease has been exploited to develop novel
`therapeutic agents that
`target
`the diseased cell. For example,
`chimeric molecules containing growth factors fused to toxins
`(termed mitotoxins) are emerging as potentially powerful and
`versatile therapeutic agents. Because over-expression of EGFRs
`has been demonstrated in numerous solid human tumors and is
`associated with increased metastatic potential and poor prognosis,
`EGFRs are particularly attractive therapeutic targets (Davies and
`Chamberlin, 1996). Although normal cells do express EGFRs, the
`elevated number of receptors on tumor cells confers a degree of
`targeting specificity in that the tumor cells can bind proportionately
`more mitotoxin.
`The biological activity of mitotoxins relies on the ability of each
`component
`to perform its individual biological function. The
`recombinant HBEGF-SAP mitotoxins described in this study retain
`the ability of HBEGF to bind EGFRs and the ability of native SAP
`to inhibit protein synthesis. Together, the complex can kill eukary-
`otic cells upon internalization, and by separating the HBEGF
`domain from the SAP domain with a 22-amino-acid spacer, some
`improvement of function was realized. The intracellular mecha-
`nism by which the fusion proteins are metabolized is unknown.
`However, since SAP is highly resistant to proteolysis (Stirpe et al.,
`1983), it is tempting to speculate that upon entry into cells, HBEGF
`is proteolytically degraded to release a fully functional SAP moiety.
`
`In this study, cell lines derived from SCLC were unique in their
`uniform resistance to HBEGF-L22-SAP. This result is consistent
`with binding studies which have demonstrated few and even
`undetectable levels of EGFRs on human SCLC cell lines (Haeder et
`al., 1988). In contrast, HBEGF-L22-SAP was cytotoxic to cell lines
`representing all of the remaining 11 tumor types, with many of the
`cell lines having IC50 values in the low nanomolar range. The tumor
`cell lines that are sensitive to HBEGF-L22-SAP represent cancers
`generally considered to express EGFRs. A431 and KB epidermoid
`carcinoma cells, for example, have been shown to have 2 to 3 x 106
`and 1 to 2 x 105 EGFRs/cell, respectively, and to be sensitive to
`TGF-␣-PE (Siegall et al., 1989). Expression of EGFRs has also
`been demonstrated in human bladder cancer (Wood et al., 1992),
`primary human NSCLC (Rusch et al., 1993), primary human renal
`carcinomas (Petrides et al., 1990), human colon carcinoma cell
`lines (Radinsky et al., 1995), human ovarian tumors (Henzen-
`Logmans et al., 1992) and human gliomas (Ekstrand et al., 1991).
`the ability of HBEGF-L22-SAP to kill EGFR-
`In addition,
`expressing breast tumor cells in vivo (Fig. 4) further validates the
`use of HBEGF-containing mitotoxins for selective targeting to
`EGFR-bearing tumor cells.
`HBEGF is an attractive ligand for targeted drug delivery because
`of its high affinity for EGFRs, which are over-expressed in a variety
`of human malignancies and other proliferative disorders. The
`heparin-binding capacity of HBEGF facilitates binding to EGFRs,
`thus making HBEGF preferred over other EGF family members as
`a targeting ligand. Consistent with this observation, it has been
`demonstrated that fusing the heparin-binding domain of HBEGF to
`the amino terminus of TGF-␣-PE generates mitotoxins with
`improved cytotoxicity to proliferating smooth muscle cells (Mesri
`et al., 1993). Biologically active mitotoxins consisting of mature
`HBEGF fused to PE have also been generated (Mesri et al., 1994).
`SAP is an attractive alternative to the bacterial toxins, such as
`Pseudomonas exotoxin, because it has no high-affinity cell-binding
`domain and is highly resistant to proteolysis (Stirpe et al., 1983).
`The HBEGF-SAP mitotoxins described in this report represent a
`novel, potent and versatile class of mitotoxins with potentially
`widespread therapeutic applications.
`
`ACKNOWLEDGEMENTS
`
`We thank Mr. M. Ong, Ms. D. Pursell, Ms. T. Tetzke, Ms. M.
`Caton and Ms. G. Kiriakova for excellent technical assistance and
`Dr. Z. Parandoosh for critical review of the manuscript. The
`contributions of Drs. G.S. Johnson, T.G. Meyers and J. Johnson of
`the NCI Anticancer Drug Screening Program are also gratefully
`acknowledged.
`
`REFERENCES
`
`BEHAR-COHEN, F.F., DAVID, T., BUECHLER, Y., NOVA, M.P., HOUSTON, L.L.,
`POULIQUEN, Y.M. and COURTOIS, Y., Cytotoxic effects of FGF2-saporin on
`bovine epithelial lens cells in vitro. Invest. Ophthalmol. Vis. Sci., 36,
`2425–2433 (1995).
`BEITZ, J.G., DAVOL, P., CLARK, J.W., KATO, J., MEDINA, M., FRACKELTON,
`A.R., JR., LAPPI, D.A., BAIRD, A. and CALABRESI, P., Antitumor activity of
`basic fibroblast growth factor-saporin mitotoxin in vitro and in vivo. Cancer
`Res., 52, 227–230 (1992).
`DAVIES, D.E. and CHAMBERLIN, S.G., Targeting the epidermal growth factor
`receptor for therapy of carcinomas. Biochem. Pharmacol., 51, 1101–1110
`(1996).
`EKSTRAND, A.J., JAMES, C.D., CAVENEE, W.K., SELIGER, B., PETTERSSON,
`R.F. and COLLINS, V.P., Genes for epidermal growth factor receptor,
`transforming growth factor ␣, and epidermal growth factor and their
`expression in human gliomas in vivo . Cancer Res., 51, 2164–2172 (1991).
`FARB, A., LEE, S.J., MIN, D.H., PARANDOOSH, Z., COOK, J., MCDONALD, J.,
`PIERCE, G.F. and VIRMANI, R., Vascular smooth muscle cell cytotoxicity and
`sustained inhibition of neointimal formation by fibroblast growth factor
`2-saporin fusion protein. Circ. Res., 80, 542–550 (1997).
`FILMUS, J., POLLAK, M.N., CAILLEAU, R. and BUICK, R.N., MDA-468, a
`human breast cancer cell line with a high number of epidermal growth
`
`factor receptors, has an amplified EGF receptor gene and is growth
`inhibited by EGF. Biochem. biophys. Res. Comm., 128, 898–905 (1985).
`HAEDER, M., ROTSCH, M., BEPLER, G., HENNING, C., HAVEMANN, K.,
`HEIMANN, B. and MOELLING, K., Epidermal growth factor receptor expres-
`sion in human lung cancer cell lines. Cancer Res., 48, 1132–1136 (1988).
`HENZEN-LOGMANS, S.C., VAN DER BURG, M.E.L., FOEKENS, J.A., BERNS,
`P.M.J.J., BRUSSE´E, R., FIERET, J.H., KLIJN, J.G.M., CHADHA, S. and
`RODENBURG, C.J., Occurrence of epidermal growth factor receptors in
`benign and malignant ovarian tumors and normal ovarian tissues: an
`immunohistochemical study. J. Cancer Res. clin. Oncol., 118, 303–307
`(1992).
`MCDONALD, J.R., ONG, M., SHEN, C., PARANDOOSH, Z., SOSNOWSKI, B.,
`BUSSELL, S. and HOUSTON, L.L., Large-scale purification and characteriza-
`tion of recombinant fibroblast growth factor-saporin mitotoxin. Protein
`Expression Purif., 8, 97–108 (1996).
`MESRI, E.A., KREITMAN, R.J., FU, Y.-M., EPSTEIN, S.E. and PASTAN, I.,
`Heparin-binding transforming growth factor ␣ -Pseudomonas extotoxin A.
`J. biol. Chem., 268, 4853–4862 (1993).
`MESRI, E.A., ONO, M., KREITMAN, R.J., KLAGSBRUN, M. and PASTAN, I., The
`heparin-binding domain of heparin-binding EGF-like growth factor can
`
`IMMUNOGEN 2157, pg. 5
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`TARGETING TUMOR CELLS WITH HBEGF
`
`111
`
`target Pseudomonas exotoxin to kill cells exclusively through heparan
`sulfate proteoglycans. J. Cell Sci., 107, 2599–2608 (1994).
`MONKS, A. and 13 OTHERS, Feasibility of a high-flux anticancer drug screen
`using a diverse panel of cultured human tumor cell lines. J. nat. Cancer
`Inst., 83, 757–766 (1991).
`PETRIDES, P.E., BOCK, S., BOVENS, J., HOFMANN, R. and JAKSE, G.,
`Modulation of pro-epidermal growth factor, pro-transforming growth factor
`alpha and epidermal growth factor receptor gene expression in human renal
`carcinomas. Cancer Res., 50, 3934–3939 (1990).
`RAAB, G. and KLAGSBRUN, M., Heparin-binding EGF-like growth factor.
`Biochim. biophys. Acta, 1333, F179–F199 (1997).
`RADINSKY, R., RISIN, S., FAN, D., DONG, Z., BIELENBERG, D., BUCANA, C.D.
`and FIDLER, I.J., Level and function of epidermal growth factor receptor
`predict the metastatic potential of human colon carcinoma cells. Clin.
`Cancer Res., 1, 19–31 (1995).
`RUSCH, V., BASELGA, J., CORDON-CARDO, C., ORAZEM, J., ZAMAN, M.,
`HODA, S., MCINTOSH, J., KURIE, J. and DMITROVSKY, E., Differential
`expression of the epidermal growth factor receptor and its ligands in
`
`primary non-small cell lung cancers and adjacent benign lung. Cancer Res.,
`53, 2379–2385 (1993).
`SIEGALL, C.B., XU, Y.-H., CHAUDHARY, V.K., ADHYA, S., FITZGERALD, D.
`and PASTAN, I., Cytotoxic activities of a fusion protein comprised of TGF␣
`and Pseudomonas exotoxin. FASEB J., 3, 2647–2652 (1989).
`STIRPE, F., BARBIERI, L., BATTELLI, M.G., SORIA, M. and LA