K2::
`Int. J. Cancer: 60, 137-144 (1995)
`0 1995 Wiley-Liss, Inc.
`EGF RECEPTOR AND P ~ ~ ~ " ' ~ ~ - ~ - S P E C I F I C SINGLE-CHAIN ANTIBODY
`
`TOXINS DIFFER IN THEIR CELL-KILLING ACTIVITY ON TUMOR
`CELLS EXPRESSING BOTH RECEPTOR PROTEINS
`Winfried WELS' ', Roger BEERLI', Peter HELLMAN" 6 , Mathias SCHMIDT', Barbara M. MARTE', Elena S. KORNILOVA1~7,
`Armin HEKELE', John MENDELsOHN4, Bernd GRONER2 and Nancy E. HYNES'
`'Fnednch Miescher Institute, P. 0. Box 2543, CH-4002 Basel, Switzerland; 21nstitute for Experimental Cancer Research,
`Tumor Biology Center, Breisacher Strasse I 17, D- 79106 Freihurg, Germany; 'Institute for Genetics, Keniforschirngszentnrm
`Karlsnihe, P 0. Box 3640, D-76021 Karlsnihe, Germany; and 4Memonal Sloan-Kettenng Cancer Center, 1275 York Avenue,
`New York, NI'10021, USA.
`
`-2--
`+@@&
`
`~
`
`~
`
`t
`
`:
`
`~
`
`~
`
`~
`
`:
`
`"
`
`Many human tumors over-express erbB-2 and EGF receptors.
`The membrane localization of these receptor tyrosine kinases
`make them appropriate targets for directed tumor therapy. We
`have used recombinant DNA technology to produce single-
`chain antibody exotoxin A (scFv-ETA) fusion proteins which
`specifically bind the erbB-2 and EGF receptors. The scFv
`portion is composed of the heavy- and light-chain variable
`domains of monoclonal antibodies which recognize the extracel-
`Mar portion of each receptor. We have previously described
`the anti-tumor activity of the bacterially produced scFv(FRP5)-
`ETA directed to the erbB-2 receptor. In this paper we describe
`the characteristics of scFv(225)-ETA, a protein which binds the
`EGF receptor. The bacterially produced recombinant protein
`binds to the receptor with high affinity and inhibits the in vitro
`growth of the EGF receptor over-expressing tumor cell lines
`A43 I and MDA-MB468. Combination treatment with scFv-
`(FRP5)-ETA and scFv(225)-ETA led to an additive inhibitory
`effect on the in vitro growth of A431 cells. SKBR3 cells
`expressing low levels of EGF receptor but high levels of p 185e*n-2
`were not affected by scFv(225)-ETA treatment but were sensi-
`tive to scFv(FRP5)-ETA. Stimulation of SKBR3 cells and HC I I
`RI # I I mouse mammary epithelial cells expressing the human
`erbl-2 with EGF led to an increase in scFv(FRP5)-ETA activity,
`showing that the EGF-induced activation of erbB-2 can potenti-
`ate the action of the erbB-2-directed toxin. Treatment of
`athymic nude mice with scFv(FRP5)-ETA and the combination
`of both scFv-ETA proteins led to the transient arrest of growth
`of established A43 I tumors. scFv(225)-ETA treatment alone
`was the most effective, leading to tumor shrinkage during the
`course of treatment, whereas treatment with the parental
`monoclonal antibody 225 led to retarded tumor growth.
`(i' 199.5 Wiley-Liss, fnc.
`
`Members of the growth-factor-receptor tyrosine-kinase fam-
`ily play an important role in the development of human
`malignancies. Many tumors of epithelial origin, including
`glioblastoma and cancers of the lung, breast, head and neck,
`and bladder express increased EGF receptor levels on their
`cell-surface membranes (reviewed in Gullick, 1991). The
`tumors with increased receptor expression sometimes display
`increased production of TGF-a, allowing receptor activation
`by an autocrine pathway (Derynck et al., 1987). The c-erbB-21
`
`neu gene coding for ~ 1 8 5 ~ ' ~ ~ - ~ / H E R 2 , another member of the
`sub-class-I family of growth factor receptors is amplified
`andlor over-expressed in a high percentage of human adeno-
`carcinomas arising at numerous sites, including breast, ovary,
`lung, stomach and salivary gland (reviewed in Hynes, 1993).
`Many clinical studies have shown that patients with tumors
`showing elevated expression of these receptors have a poorer
`prognosis. The elevated expression of the erbB-2- and EGF-
`receptor proteins on the membrane of tumor cells and their
`involvement in the transformation process make them appro-
`priate targets for directed therapy.
`We have described monoclonal antibodies (MAbs) directed
`to the extracellular portion of the erbB-2-receptor protein
`which inhibit tumor-cell growth in vitro and irz vivo (Hanverth
`et al., 1992, 1993). Antibody domains derived by recombinant
`DNA technology can be coupled to cytotoxic reagents in
`
`order to enhance their tumoricidal potential. We have con-
`structed an erbB-2-specific single-chain antibody toxin, scFv-
`(FRPS)-ETA, which displays potent in vitro and in vivo
`tumor-cell-killing activity (Wels et a!., 1992a; Harwerth et al.,
`1993). The biological characteristics of the EGF-receptor-
`specific MAb 225 have been studied in great detail. MAb 225
`competes with EGF for binding to the E G F receptor, thereby
`blocking ligand-dependent receptor activation (Fan et al.,
`1993). Treatment with MAb 225 inhibits the growth of EGF-
`receptor-expressing tumor cells in vitro and in vivo (Masui et
`al., 1984; Ennis et al., 1989).
`We have constructed a recombinant single-chain immuno-
`toxin consisting of a scFv domain derived from the MAb 225
`and truncated Pseudomonas aeruginosa exotoxin A. Compari-
`son of the anti-tumor effects of this bacterially expressed
`scFv(225)-ETA with those of the similar erbB-2-specific scFv-
`(FRP5)-ETA on human tumor cells expressing various levels
`of the respective target receptors revealed differences in toxin
`sensitivity which cannot be simply attributed to the different
`expression levels of the receptor proteins.
`
`MATERIAL AND METHODS
`Cells and culture conditions
`The SKBR3 and MDA-MB468 human breast-tumor cells
`and the A43 1 human vulvar squamous-carcinoma cells were
`maintained in DMEM containing 8% FCS. Hybridoma cells
`producing the MAb 225 (IgG1, kappa) (Kawamoto et al., 1983)
`were grown in RPMI 1640 containing 20% heat-inactivated
`FCS, 4 mM glutamine, 1 mM sodium pyruvate and 14.2 mM
`P-mercaptoethanol. The HCll R l # l l cells expressing the
`human erbB-2 receptor (Hynes et al., 1990) were grown in
`RPMI 1640 supplemented with 8% FCS and 5 kg/ml insulin.
`
`5To whom correspondence and reprint requests should be sent, at
`Institute for Experimental Cancer Research, Tumor Biology Center,
`Breisacher Strasse 117, D-79106 Freiburg, Germany. Fax: (49) 761 206
`1599.
`
`hPresent address: Department of Anatomy, Hufelandstrasse 55,
`D-45122 Essen, Germany.
`
`'Present address: Institute of Cytology, Tichoretsky pr. 4, St.
`Petersburg, Russia.
`
`Abbreviafions: MAb, monoclonal antibody; ETA, exotoxin A; scFv,
`single-chain antigen-binding protein; DMEM, Dulbecco's modified
`Eagle's medium; FCS. fetal calf serum: EF-2. elongation factor 2;
`SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophore-
`sis; IPTG, isopropyl P-D-thiogalactopyranoside: ELISA, enzyme-
`linked immunosorbent assay; ICsu, 50% inhibitory concentration;
`BSA, bovine serum albumin.
`
`~ Received: July 14, 1994 and in revised form August 30.1994.
`
`IMMUNOGEN 2067, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`138
`
`WELS ETAL.
`
`cDNA synthesis and constnrction of scFv
`Total RNA was extracted from the 225 hybridoma cells by
`the acid-guanidium-thiocyanate-phenol-chloroform method
`(Chomczynski and Sacchi, 1987). First-strand cDNA synthesis,
`carried out using a cDNA synthesis kit (Pharmacia Biotech,
`Brussels, Belgium), was in a standard 33-pl reaction contain-
`ing 5 pg total RNA and 0.2 pg Notl-d(T)18 primer. For
`amplification of VH and VL domains using PCR, 5 pI of the
`first-strand cDNA reaction was used as a template in a PCR as
`described (Wels et al., 199%). For amplification of the VH
`domain, SO pmol each of the oligonucleotides VHlFOR
`5 ' - T G A G G A G A C G G T G A C C G T G G T C C C T T G -
`GCCCCAG-3' and VHIBACK 5'-AGGTSMARCTGCAG-
`SAGTCWGG-3' were used, for amplification of the VL kappa
`domain, SO pmol each of the oligonucleotides VKWlFOR
`S'-GTTAGATCTCCARYTTKGTSCS-3' and VKl BACK 5'-
`GACATTCAGCTGACCCAGTCTCCA-3' were uscd (M =
`A + C , R = A + G, S = C + G, W = A + T, Y = C + T, K =
`G + T). PCR products were digested with PstI and BstEII
`(VH) or PvuII and BglII (VL). MAb 225 VH cDNA was inserted
`into PstIiBstEII-digested plasmid pWW152, a derivative of
`the modified bluescript plasmid pWW15 (Wels et a/., 1992b)
`containing a sequence encoding the 15-amino-acid linker
`(GGGGS)3. Subsequently the 225 VL fragment was inserted 3'
`of the VH and linker sequences resulting in the scFv(225)-
`encoding plasmid pWW152-225. The scFv(FRP5) gene encod-
`ing an erbB-2-specific antibody domain was sub-cloned into
`pWW152 as a PstUXbaI fragment derived from pWW15-5
`(Wels et a/., 1992b). The plasmid pFLAG-1 (IBI Biochemicals,
`New Haven, CT, USA) was digested with HindIII and XbaI
`and a double-stranded DNA linker encoding 6 His residues at
`its 5' end and the original HindIII-, EcoRI- and XbaI-
`restriction sites of pFLAG-1 at its 3' end was inserted 3' of the
`FLAG epitope. The resulting plasmid, pSWS0, was digested
`with HindIII and XbaI and the scFv genes which were isolated
`from pWW152-225 and pWW1.52-5 as HindIIIiXbaI frag-
`ments were inserted yielding the scFv expression plasmids
`pSW50-5 and pSW50-225.
`
`Construction, expression and pirrification of scFv-ETA proteins
`pFLAG-1 was digested with SalI and treated with the
`Klenow enzyme to create blunt ends; the linearized fragment
`was digested with XbaI. A truncated Pseudomonas ETA gene
`lacking the cell-binding domain Ia, was isolated from pWW20
`(Wels et al., 199241) by EcoRI cleavage, Klenow fill-in and
`subsequent XbaI digestion. This blunt-ended XbaI fragment
`was inserted into the blunt-ended XbaI pFLAG-1 vector. The
`resulting plasmid, pSG100, was digested with HindIII and
`XbaI, and a double-stranded DNA linker encoding 6 His
`residues was inserted in frame 5' of the ETA sequences
`yielding pSW200. DNA fragments containing the erbB-2- and
`EGF-receptor-specific scFv genes, scFv(FRP5) and scFv(225),
`including the ompA signal peptide, the FLAG epitope and the
`N-terminal His-encoding sequences from pSW.50 were isolated
`from pSW50-5 and pSW50-225 by NdeI and XbaI digestion
`and inserted into NdeIiXbaI-digested pSW200. For expres-
`sion of the scFv-ETA fusion proteins, the resulting plasmids
`pSW202-5 and pSW202-225 were transformed into E. coli
`strain CC118 (Manoil and Beckwith, 1985). A single colony
`was grown overnight at 37°C in LB medium containing 0.6%
`glucose and 100 pg/ml ampicillin. The culture was diluted
`30-fold in the same medium, grown at 37°C to an OD550 of 0.5
`and induced 45 min at 37°C with 0.5 mM IPTG. Cells were
`harvested by centrifugation at 4000g for 15 min at 4°C and the
`cell pellet from 1 I of culture was lysed by freeze/thaw in 0.5 ml
`of 100 mM Tris-HCI, pH 8.0, 0.1 mg/ml lysozyme, 0.3 mM
`PMSF and 10 pg/ml DNAse I. PBS (15 ml) containing 8 M
`urea were added and incubated for 30 min at RT. The lysate
`was clarified by ultracentrifugation at 45,000 g for 30 min at
`2°C and the supernatant was dialyzed against PBS and clarified
`
`by Centrifugation. scFv-ETA proteins were purified via binding
`of the 2 His clusters to Ni2+ loaded chelating sepharose
`(Pharmacia Biotech) followed by clution with a step gradient
`of 50 to 400 mM imidazole in PBS. Fractions containing the
`scFv-ETA proteins were pooled, imidazole was removed by
`dialysis against PBS, and the proteins were concentrated by
`ultrafiltration through a YMTlO membrane (Amicon, Beverly,
`MA). Purified proteins were analyzed by SDS-PAGE and
`quantitated by densitometry after Coomassie staining in com-
`parison with BSA standards.
`scFv(225)-ETA binding assay
`The binding of scFv(225)-ETA to the EGF receptor on
`MDA-MB468 and SKBR3 cells was measured by ELISA
`(Wels et al., 1992a). Cells were grown on 96-well microtiter
`plates, fixed with 2% formaldehyde in PBS and blocked with
`3% BSA in PBS. scFv(225)-ETA (100 pl) at concentrations
`ranging from 1.2 nM to 150 nM was added to each well and the
`plates were incubated for 1 hr at 37°C. Unbound scFv(225)-
`ETA was removed and the cells were washed and incubated
`for 1 hr at 37°C with 100 pl of rabbit anti-ETA serum, then
`incubated with 100 pl of goat anti-rabbit IgG coupled to
`alkaline-phosphatase (Sigma, St. Louis, MO). The specifically
`bound scFv(225)-ETA was detected by 30-min incubation at
`37°C with a solution of 1 M Tris-HCI (pH 8.0) and 0.4 rng/ml
`p-nitrophenylphosphate disodium (Sigma). The absorbance at
`405 nm was measured.
`EGF-receptor-activation assay
`NE1 mouse fibroblasts expressing the human E G F receptor
`cDNA were grown for 16 hr in DMEM containing 0.5% FCS,
`then treated for 5 rnin at 37°C with 10 ng/ml E G F in the
`presence or absence of a 500-fold molar excess of competitor.
`Cell lysates were prepared using lysis buffer containing 200 p M
`sodium-orthovanadate and total proteins were separated by
`7.5% SDS-PAGE and electroblotted onto polyvinylidenedifluo-
`ride membranes as described (Harwerth et al., 1992). Phospho-
`tyrosine-containing proteins were detected with a specific
`MAb; the filter was then treated with peroxidase-coupled
`anti-mouse IgG and bound antibody was visualized using the
`ECL system (Amersham, Aylesbury, UK).
`Cell-killing assay
`The cell-killing activity of scFv-ETA proteins was measured
`with the Cell Titer 96 Kit (Promega, Madison, WI) exactly as
`described (Wels et al., 1992a).
`Sub-cellular distribution of EGF receptor in EGF-treated
`SKBR3 cells
`The kinetics of EGF-receptor internalization and its sub-
`cellular distribution were determined as described (Kornilova
`et al., 1992). Briefly, iodinated EGF [20 ng/ml ('"I)EGF,
`Amersham; specific activity 100 mCi/mg] was bound to surface
`E G F receptors on SKBR3 cells for 60 rnin at 4°C in a working
`medium (WM) consisting of DMEM, 0.1% BSA and 20 mM
`HEPES pH 7.3. The cells were washed with cold WM to
`remove unbound ('2SI)EGF and one plate was set aside at 4°C
`to determine the position of plasma membrane-bound EGF.
`EGF-receptor internalization was stimulated by placing cells
`in WM at 37°C. After 5,15 or 30 rnin one plate was placed on
`ice and the surface ('2sI)EGF was removed by washing for 3
`rnin with 0.2 M acetic acid, p H 4.5, containing 0.5 M NaCI. The
`sub-cellular fractionation procedure was carried out on 17%
`Percoll gradients as described (Kornilova et al., 1992).
`In vivo anti-tumor activiv
`In vivo anti-tumor activity of MAb 225 and the scFv-ETA
`proteins was tested using A431 epidermoid-tumor xenografts
`in athymic nude mice. Approximately 25 mg of tumor tissue
`was implanted S.C. in each mouse (5 mice/group). Six days later
`the mice received twice-daily i.p. injections of 5 pg of MAb225
`
`IMMUNOGEN 2067, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`EGF RECEPTOR AND p185 crhB-2-SPECIFIC IMMUNOTOXINS
`
`139
`
`or the scFv-ETA proteins. The control group received PBS.
`Tumor growth was followed as described (Wels et af., 19920).
`
`A
`pSW202-225
`
`xFv(Z25)
`
`-
`
`ETA 252613
`
`B
`
`SP I
`
`VH
`
`n
`
`ETA
`
`RESULTS
`Construction and expression of a gene encoding the chimeric
`scFv(225)-ET,4 protein
`A gene encoding a protein consisting of the scFv of the
`EGF-receptor-specific MAb 225 (Sunada et af., 1986) fused to
`domains 11, Ib and 111 of the Pseudomorias aeruginosa ETA was
`constructed in the pFLAG-1 expression vector. Domain Ia of
`ETA is responsible for cell recognition, but is not necessary for
`its enzymatic activity, which is the ADP-ribosylation of EF-2
`(Siegall et af., 1989). The bacterial expression vector pSW202-
`225, shown schematically in Figure la, has an IPTG-inducible
`tac promoter followed by sequences encoding the ompA signal
`peptide, the FLAG epitope, 6 His residues, the VH, linker, the
`VL, 6 His residues and the ETA domains 11, Ib and 111. Figure
`lh presents the sequence of the scFv(225) portion of the
`chimeric gene. The sequence of the toxin has been published
`(Gray et al., 1984). Likewise, 2 clusters coding for 6 His
`residues each were introduced 5‘ and 3’ of the scFv domain in
`the coding region of the erbB-2-specific scFv(FRP5)-ETA
`(Wels et af., 1992a), resulting in the expression plasmid
`psw202-5.
`The scFv( 225)-ETA and scFv(FRP5)-ETA fusion proteins
`were expressed in E. coli strain CC118. Total bacterial lysates
`were prepared in 8 M urea, the lysates were dialyzed against
`PBS, and the soluble scFv-ETA proteins were purified by
`binding to Ni2+ columns and elution with imidazole step
`gradients. Fractions containing the recombinant scFv-ETA
`proteins were pooled, imidazole was removed by dialysis and
`AAGTA-A----
`the proteins were concentrated by ultrafiltration. SDS-PAGE
`K Y A S E S I S G I P S R F S G S G S G
`analysis of the purified material revealed a purity of more than
`ccp(2
`r m ~
`r
`c
`r
`~ m A c
`
`TATIWAIXmAmAlTGT
`70% after a single round of Ni2+-affinity purification (data not
`T D F T L S I N S V E S E D I A D Y Y C
`shown). The yield of purified scFv(225)-ETA and scFv(FRP5)-
`PglII/BClI
` v3a%XT
`ETA from 1 1 of bacterial culture was between 1 and 1.5 mg.
`P
`T
`A
`?
`T
`A
`A
`M
`C
`A
`A
`C
`Q Q N N N W P T T F G A G T K L E I K A
`XbaI
`CMU
`Bindirzgpropenies of scFv(225)-ETA
`cr-mmnw-
`L E H H H H H H L E G G S L A A L T A H
`The affinity of the purified recombinant scFv(225)-ETA for
`252 .._
`-toxin A
`the EGF receptor was measured in an ELISA using MDA-
`MB468 and SKBR3 breast-tumor cells. MDA-MB468 cells
`FIGURE 1 - (a) The scFv(225)-ETA expression plasmid pSW202-
`have an amplified EGF-receptor gene and express approxi-
`225. The plasmid contains the IPTG inducible tac promoter (tac),
`the ompA signal peptide (SP), the synthetic FLAG epitope
`mately 1.5 x lo6 receptors per cell (Filmus et a/., 1987).
`(FLAG), a 6-residue His cluster (His), the PCR-amplified VH
`SKl3R3 cells which have an amplified c-erbB-7- gene (Hynes et
`cDNA of MAb 225, a sequence coding for a 15-amino-acid linker,
`al., 1989), but a single copy of the EGF-receptor gene, express
`the PCR-amplified VL cDNA of MAb 225, a second His cluster,
`approximately 9 x lo4 EGF receptors per cell (data not
`and the truncated Pseudornonas exotoxin A gene encoding do-
`shown). The cells were grown in 96-well dishes, scFv(225)-
`mains 11, Ib and I11 (amino acids 252-631 of ETA). (b) Partial
`ETA at concentrations ranging from 1.2 to 150 nM was added
`nucleotide and deduced amino-acid sequence of the scFv(225)-
`to the wells, and the plates were incubated at 37°C for 1 hr.
`ETA gene. The sequence shows: the ompA signal peptide bp
`Specifically bound protein was determined after incubation
`1-63); the FLAG epitope (bp 64-87); the first His cluster [bp
`91-108); the PstI/BstEII fragment encoding the VH domain (bp
`with a rabbit anti-ETA serum followed by goat anti-rabbit IgG
`124-462); the 15-amino-acid linker (bp 472-516); the PvuII/BglII
`coupled to alkaline phosphatase. The phosphatase reaction
`fragment encoding the VL domain (bp 523-835): the second His
`product was measured as absorbance at 405 nm and the results
`cluster (bp 847-864); a partial sequence of the fragment encoding
`are shown in Figure 2a. The apparent binding affinity of
`amino acids 252-613 of the Pseudornonas ETA protein (bp
`scFv(225)-ETA to the EGF receptor, measured as the half-
`868-900). The complementarity-determining regions (CDR) in the
`maximal saturation value, is 12 nM. The apparent binding
`deduced amino-acid sequence of the 225 VH and VL and the linker
`affinity for the parental MAb 225 was determined as 1 nM in a
`sequences are underlined. The additional nucleotides represent
`similar experiment (data not shown). The reported affinity of
`vector sequences used for cloning.
`the 225 Fab’ fragment to the E G F receptor is 5 nM (Fan er a/.,
`1993), therefore the binding affinity of the monovalent recom-
`binant protein is approximately 2-fold lower than that of the
`monovalent Fab’ fragment and 10-fold lower than that of the
`bivalent MAb.
`In a similar experiment, the binding of purified scFv(FRP5)-
`ETA to the erbB-2 receptor on SJSBR3 cells was measured.
`The apparent binding affinity of this version of the scFv(FRP5)-
`ETA containing 2 His clusters N- and C-terminal of the scFv
`domain specific for the erbB-2 receptor is 4.2 nM (data not
`shown). Both Ni2+-affinity-purified scFv(FRP5)-ETA and a
`
`
`P
`W
`A
`-
`A
`M K K T A I A I A V A L A G F A T V A Q
`HlndIII
`~
`S
`W
`Z
`G
`W
`C
`A
`T
`X
`T
`L
`C
`A
`A D Y K D D D D K L H H H H H H K L Q V
`PStI
`F L a
`
`Q L Q E S G P G L V Q P S Q S L S I T C
`
`P
`n
`r
`A
`n
`r
`T
`?
`(
`A
`A
`T
`l
`P
`T V S G F S L T N Y G V H W V R Q S P G
`m
`
`ATAATACACTT
`
`P
`A
`T
`-
`K G L E W L G V I W S G G N T D Y N T P
`cDR2
`
`PAAA’IG
`mCAcA---
`F T S R L S I N K D N S K S Q V F F K M
`
`~
`
`~
`
`~
`
`A
`
`60
`-2
`
`~
`120
`19
`
`180
`39
`
`
`240
`59
`
`300
`79
`
`360
`99
`
`420
`119
`
`480
`139
`
`540
`159
`
`RCTAT
`
`P
`m
`A
`T
`A
`T
`V
`M
`W
`-
`N S L Q S N D T A I Y Y C A R A L T Y Y
`BstEII
`a x 3
`P
`-
`A
`T
`W
`G
`
`D Y E F A Y W G Q G T T V T V S S G G G
`
`G S G G G G S G G G G S D I Q L T Q S P
`PStI
`linker
`
`m
`m
`V I L S V S P G E R V S F S C R A S Q S
`o m
`660
`A T l W G O J A C A T A W I T X T A ~ C W A ~ T A
`I G T N I H W Y Q Q R T N G S P R L L I
`199
`
`600
`179
`
`720
`219
`
`780
`239
`
`840
`259
`
`900
`379
`
`previously described scFv(FRP5)-ETA protein purified via
`FLAG-affinity chromatography have very similar binding char-
`acteristics and cell-killing activity (Wels et d., 1992a), indicat-
`ing that the His clusters included in the molecule do not alter
`its biological activity.
`scFv(225)-ETA competes with EGF for receptor binding
`MAb 225 competes with E G F for receptor binding (Sunada
`et al., 1986). An experiment to show that the same is true for
`
`IMMUNOGEN 2067, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`140
`
`a
`I
`
`0.2
`
`0.0 r
`
`0
`
`-
`
`SKBR3
`
`I
`I
`1 0 0
`5 0
`scFv(225)-ETA nM
`
`I
`1 5 0
`
`1
`
`2
`
`3
`
`4
`
`
`
`rl p170 EGFR
`
`-
`
`+
`-
`
`+
`+
`-
`
`+
`-
`+
`
` EG F
` SCFV( 225)-ETA
` MAb 225
`FIGURE 2 - (a) Binding of scFv(225)-ETA to the EGF receptor
`of MDA-MB468 and SKBR3 breast-tumor cells. Cells were fixed
`with 2% formaldehyde and incubated with various concentrations
`of scFv(225)-ETA. The amount of specifically bound scFv(225)-
`ETA was measured, after incubation with rabbit anti-ETA fol-
`lowed by AP-coupled goat anti-rabbit IgG, as the absorbance at
`405 nm. Each point was determined in triplicate. The apparent
`bindihg affinity of scFv(225)-ETA to the EGF receptor on both
`cell lines is 12 nM. (b) Competition of EGF binding to the EGF
`receptor and inhibition of receptor activation by scFv(225)-ETA.
`NE1 mouse fibroblasts expressing human EGF-receptor cDNA
`were treated for 5 rnin at 37°C with 10 ng/ml EGF (lane 2), in the
`presence of a 500-fold molar excess of scFv(225)-ETA (lane 3) or a
`500-fold excess of MAb 225 (lane 4) or were mock treated (lane 1).
`Equal amounts of cell lysates were analyzed by SDS-PAGE and
`g FR
`phosphotyrosine *as detected by irnrnunoblotting with a s ecific
`anti-phosphotyrosine MAb. The position of the ~ 1 7 0 ~
`IS
`indicated.
`
`the recombinant scFv(225)-ETA protein was carried out by
`measuring inhibition of EGF-induced tyrosine phosphoryla-
`tion with an excess of scFv(22S)-ETA. Figure 2b shows that the
`increase in phosphotyrosine caused by 5-min incubation of
`NE1 mouse fibroblasts expressing a human E G F receptor
`cDNA with 10 ng/ml EGF (lane 2) was inhibited by a 500-fold
`molar excess of MAb 225 (lane 4) and scFv(225)-ETA (lane 3).
`In vitro toxicity and specificity of scFv(225)-ETA
`The cell-killing activity of scFv(225)-ETA was tested on
`MDA-MB468 cells using an enzymatic assay (Wels et a/.,
`1992~). The cells were incubated for 40 hr with 100 ngiml(1.5
`nM) of scFv(225)-ETA in the absence or presence of a
`100-fold molar excess of MAb 225, and cell viability was
`measured in comparison with PBS-treated cells. The results
`are shown in Figure 3. Approximately 44% of the cells were
`killed by scFv(225)-ETA alone. In the presence of an excess of
`
`WELS E T A L
`
`IZ0 1
`
`I
`
`p 80
`
`additions: u none
`scFv(225)-ETA
`scFv(225)-ETA + mAb FRP5
`scFv(225)-ETA + mAb225
`mAb 225
`
`0
`
`1
`
`
`MDA-ME468
`FIGURE 3 - Inhibition of scFv(225)-ETA cell-killing activity by
`competition with MAb 225. MDA-MB468 cells were incubated for
`40 hr with 100 ng/ml scFv(225)-ETA without the addition of
`competitor or in the presence of 10 pg/ml MAb 225 or control
`MAb FRPS. The effect of MAb 225 was also tested. The relative
`number of viable cells was determined with an enzymatic assay as
`the absorbance at 570 nrn as described in "Material and Meth-
`ods". Each point was determined in triplicate.
`
`the specific competitor, MAb 225, the cell-growth-inhibiting
`activity was reduced; whereas an excess of a non-specific MAb,
`FRP5, had no effect on the inhibiting activity of scFv(225)-
`ETA. The non-specific MAb FRPS alone had no effect on the
`growth of MDA-MB468 cells (data not shown). MAb 225 itself
`inhibits the growth of MDA-MB468 cells via blocking auto-
`crine stimulation by TGF-a (Ennis etal., 1989). Similarly, MAb
`225 alone at a concentration of 150 nM inhibits the growth of
`MDA-MB468 cells by 20% in this assay, indicating that the
`remaining growth-inhibiting activity observed with
`the
`scFv(225)-ETA/MAb 225 combination is probably caused by
`the excess of the parental MAb. The results also show that the
`cytotoxic fusion protein scFv(225)-ETA is more potent in
`inhibiting the in vitro growth of MDA-MB468 than a 100-fold
`molar excess of the original MAb 225.
`In vitro toxicity of scFv(225)-ETA and scFv(FRP5)-ETA
`We have described the properties of scFv(FRP5)-ETA a
`protein which specifically kills cells expressing the erbB-2
`receptor (Wels et a/., 1992~). Since many tumor cells express
`erbB-2 and EGF receptors, we tested the killing activity of a
`combination of scFv(225)-ETA and scFv(FRP5)-ETA on 3
`human tumor cell lines. A quantitative Western analysis for
`the relative level of erbB-2 and E G F receptor in SKBR3, A431
`and MDA-MB468 cells is shown in Figure 4. A431 cells express
`approximately 50-fold less erbB-2 receptor than SKBR3 cells,
`and MDA-MB468 cells express no detectable erbB-2 (Fig. 4a);
`A431 and MDA-MB468 cells express high amounts of EGF
`receptors and SKBR3 cells express approximately 50-fold less
`(Fig. 46).
`The in vifro toxicity of the scFv(225)-ETA and the scFv-
`(FRP.5)-ETA proteins and a 1:l combination of both were
`tested on the 3 human tumor cell lines. The cells were
`incubated for 40 hr with various concentrations of the scFv-
`ETA proteins, and the relative number of viable cells was
`determined using an enzymatic assay (Wels eta/., 1992~). The
`results are shown in Figure 5, and the ICso values are
`summarized in Table I. As anticipated, the A431 cells were
`very sensitive to scFv(225)-ETA with an ICsu of 5.8 ng/ml. The
`A431 cells were also as sensitive to the scFv(FRPS)-ETA as
`were the SKBR3 cells (ICso 33 vs. 34 ng/ml). For A431 cells,
`the combination of the 2 toxins was as active as the more
`potent (as a single modality) scFv(225)-ETA, suggesting an
`additive effect on these cells. MDA-MB468 cells were killed by
`scFv(22S)-ETA with an IC5" of 112 ng/ml and, as expected,
`
`IMMUNOGEN 2067, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`EGF RECEPTOR AND p185cr1'B >-SPECIFIC IMMUNOTOXINS
`
`141
`
`F~CURE 4 - Immunoblot of human tumor-cell extracts. The
`indicated amounts of protein from cellular extracts of SKBR3,
`A431 and MDA-MB468 human tumor cells were separated by
`7.5% SDS-PAGE and blotted onto PVDF membranes. (a) The
`erbB-2 protein was detected with the 21N anti-serum. (b) The
`EGF receptor was detected with the 15E anti-serum (Ennis et al.,
`1989). The position of the p18SCrbs-? and the p170EGF receptor and
`the relative level of expression are indicated.
`
`A
`C
`120 I
`
`erbB-2-directed toxin efficiently kills the A431 cells, in which
`the numbers of erbB-2 receptors are similar to those of EGF
`receptors in SKBR3 cells, the E G F receptor in SKBR3 cells
`was examined further.
`For their activation, ETA and scFv-ETA fusion proteins
`require internalization and cleavage of the ETA portion by a
`cellular protease (Zdanovsky er al., 1993). We have reported
`that short-term treatment of SKBR3 cells with E G F leads to
`an increase in phosphotyrosine on erbB-2 and on other cellular
`proteins (Hanverth et nl., 1992), showing that the receptors are
`active kinases. The cleavage and activation of ETA occurs in
`the low-pH compartment of the endosomes (Zdanovsky et al.,
`1993). Activated E G F receptors normally pass through this
`compartment on the way to degradation in the lysosomes
`(Sorkin et al., 1988). The dynamics of EGF-receptor internal-
`ization in SKBR3 cells was examined by cellular-fractionation
`experiments (Kornilova et al., 1992). SKBR3 cells were incu-
`bated at 4°C with ("51)-EGF. Receptor internalization was
`promoted by raising the temperature of the cells to 37°C. At
`various times, cells were placed on ice, the surface-bound
`('?'I)-EGF was removed by an acid-salt wash, and the sub-
`cellular fractions were analyzed on Percoll gradients for the
`content of receptor-bound ( 1251)-EGF. The results are shown
`in Figure 6. Fifteen minutes after the cell temperature is raised
`to 37°C most of the internalized E G F receptors can be found
`in the heavy endosomes and lysosomes. After 30 min, a
`substantial portion of the receptors have been degraded. The
`results suggest that the inability of scFv(225)-ETA to kill
`SKBR3 cells is not due to the EGF receptor per se, since it is
`activated and internalized rapidly following ligand treatment.
`Effect of EGF on the activity of scFv(FW5)-ETA
`We have reported that E G F treatment accelerates erbB-2-
`receptor turnover in HCl1 cells (Kornilova et al., 1992). A431
`cells express TGF-a, which is transported to the cell surface,
`leading to autocrine stimulation of the E G F receptor (Van de
`Vijver et ul., 1991) which might in turn accelerate erbB-2
`turnover. To test whether activation of the EGF receptor can
`influence the activity of the erbB-2-directed scFv(FRP5)-ETA
`on cells which lack a TGF-a autocrine-stimulatory loop,
`SKBR3 and H C l l R l # l l cells were treated with increasing
`concentrations of scFv(FRP5)-ETA, with or without the addi-
`tion of 20 ng/ml E G F (SKBR3) or 10 ng/ml EGF (HC11
`R l # l l ) respectively. H C l l R l # l l cells are mouse mammary
`epithelial cells expressing human erbB-2 cDNA (Hynes et al.,
`1990). Cells were treated for 40 hr and cell viability was
`determined using an enzymatic assay as described. Figure 7
`shows that SKBR3 (upper panel) and H C l l R l # l l cells
`(lower panel) were, respectively, 1.8 and 3.1 times more
`sensitive to scFv(FRP5)-ETA in the presence of EGF, with
`ICjo values in this particular set of experiments of 38 vs. 70
`ng/ml for SKBR3 cells and 17 vs. 58 ng/ml for H C l l R l # l l
`cells. This suggests that the EGF transactivation of erbB-2
`stimulates toxin internalization or activation in these cells.
`scFv(225)-ETA and scFv(FRP5)-ETA suppress tumor growth
`The in vivo anti-tumor activity of scFv(225)-ETA was tested
`on A431 xenografts in nude mice. A431 tumor tissue (25 mg)
`was implanted S.C. into 5 groups of 5 mice each on day 0. Six
`days later, when the tumors had reached a size of approxi-
`mately 100 mm',
`treatment was begun. The mice received
`twice-daily i.p. injections of 5 pg of MAb 225, scFv(225)-ETA,
`scFv(FRP5)-ETA, or a combination of 2.5 pg of each scFv-
`ETA protein for a total of 10 days. Control mice received PBS.
`The results are shown in Figure 8. None of the treatments led
`to complete inhibition of tumor-cell growth. The scFv(225)-
`ETA-treated mice showed greater inhibition of tumor growth
`than MAb 225-treated mice (Fig. 8a). The onset of tumor-cell
`growth was delayed by 9 vs. 6.5 days. By day 25, when the
`experiment was terminated, the size of the tumors in the
`
`~
`
`~
`
`~
`
`~
`
`~
`
`~
`
`n
`
`
`
`n
`
`,
`
`,
`
`1
`
`10 100 1000
`
`1
`
`10 100 1000
`
`j MDAyMB468,
`Jsmy
`--
`~cFv(2251-ETA -
`
`1
`
`100 1000
`10
`ngt ml
`
`scFv(FRP5)-EXA
`--t scb-ETA comb.
`
`RCURE 5 - Inhibition of the growth of human tumor cell lines by
`recombinant scFv-ETA proteins. SKBR3 (a), MDA-MB468 (b)
`and A431 (c) tumor cells were incubated for 40 hr with the
`indicated concentrations of scFv(225)-ETA, scFv(FRP5)-ETA or
`a 1:l scFv(225)-ETA/scFv(FRPS)-ETA combination. The relative
`number of viable cells was determined with an enzymatic assay as
`the absorption at 570 nm as described in "Material and Methods".
`Each point was determined in triplicate in 2 independent experi-
`ments.
`
`TABLE I - I N VITRO TOXICITY OF scFv-ETA IMMUNOTOXINS
`ICil, nglml scFv-ETA irnmun~toxin'
`
`Cell line
`
`scFv(FRP5)-ETA
`scFv(2?5)-ETA
`> 1000
`34
`SKBR3
`108
`> 1000
`320
`MD