Chapter 18
` Design, Development, and Characterization
`of Recombinant Immunotoxins Targeting
`HER2/neu
`
` Yu Cao and Michael G. Rosenblum
`
` Background
`
` The human epidermal growth factor receptor 2 (HER2), also known as ErbB2,
`c-erbB2, or HER2/neu, was initially discovered in 1985 by two independent labora-
`tories [ 1, 2 ] . HER2/neu is a 185 kDa (1,255 aa) transmembrane receptor encom-
`passing an intracellular tyrosine kinase domain and an extracellular ligand binding
`component [ 3– 5 ] . Extensive clinical studies have shown that overexpression of
`HER2/neu is found in 20–40% of patients with breast, ovarian, endometrial, gastric,
`bladder, prostate, and lung cancers. Studies clearly demonstrate that HER2/neu over-
`expression correlates with the prevalence of metastatic spread of many tumors and
`is generally considered to be a poor prognostic indicator [ 6– 9 ] .
` Since HER2/neu overexpression by tumor cells is quite speci fi c, therapies directed
`against this receptor have rapidly gained recognition for their selectivity and ef fi cacy
`in the clinical setting. While targeting of HER2/neu with humanized antibodies such
`as trastuzumab (Herceptin; Genentech) has proven to be an effective approach for the
`treatment of HER2/neu-overexpressing breast cancers, there are a signi fi cant number
`of patients with HER2/neu-positive tumors who do not respond or who acquire resis-
`tance to this therapy [ 10– 13 ] . Therefore, there is a need for novel therapeutic
`approaches using HER2/neu not only as a target for interfering with the growth fac-
`tor signaling component but also for receptor-mediated delivery of cytotoxic agents.
` Immunotoxins are a novel approach for the development of highly speci fi c,
`targeted agents and which generally employ a powerful class of protein toxins
` [ 14, 15 ] . These include plant toxins such as ricin [ 16– 25 ] , saporin [ 26– 29 ] , and
`gelonin [ 30– 32 ] , which inactivate ribosomes, and single-chain bacterial toxins
`
` Y. Cao • M. G. Rosenblum (*)
` Immunopharmacology and Targeted Therapy Laboratory, Department of Experimental
`Therapeutics , MD Anderson Cancer Center , Houston , TX 77030 , USA
`e-mail: mrosenbl@mdanderson.org
`
`G.L. Phillips (ed.), Antibody-Drug Conjugates and Immunotoxins: From Pre-Clinical
`Development to Therapeutic Applications, Cancer Drug Discovery and Development,
`DOI 10.1007/978-1-4614-5456-4_18, © Springer Science+Business Media New York 2013
`
`319
`
`IMMUNOGEN 2016, pg. 1
`Phigenix v. Immunogen
`IPR2014-0676
`
`

`
`320
`
`Y. Cao and M.G. Rosenblum
`
` Diphtheria toxin (DT) [ 33 ] and Pseudomonas exotoxin (PE) [ 34– 43 ] , which ADP
`ribosylate elongation factor 2 (EIF2). Anti-HER2/neu immunotoxins have been
`created initially by chemically conjugating an antibody to a whole protein toxin
`or, for more selective activity, using a protein toxin devoid of its natural binding
`domain [ 19, 23, 30, 44 ] . Technical advances in antibody engineering now enable
`us to produce various antibodies or antibody fragments in Escherichia coli , and as
`a result, HER2/neu-speci fi c antibodies and engineered fragments thereof have
`been developed to deliver various toxins to HER2/neu-positive tumor cells [ 45–
` 51 ] . Various anti-HER2/neu immunotoxins which have been developed or are cur-
`rently under evaluation are described in Table 18.1 .
`
` Antibody–Drug Conjugates: Promise and Problems
`
` Antibody-based therapeutics is of growing signi fi cance for cancer therapy. To date,
`two of the most promising strategies to enhance the antitumor activity of antibodies
`are antibody–drug conjugates (ADCs) and antibodies (or fragments) chemically
`conjugated or genetically fused to various toxins (immunotoxins).
` One successful application of the ADC approach is Trastuzumab–DM1. This is
`a covalent conjugation of trastuzumab with the maytansinoid DM1—a highly toxic
`derivative of the antimitotic drug maytansine. The therapeutic potential of
`Trastuzumab–DM1 has been extensively investigated in both in vitro and in vivo
`models of trastuzumab sensitive and insensitive breast cancers [ 52– 56 ] . It has dem-
`onstrated remarkable activity in phase I and II studies in which it was given to
`patients harboring trastuzumab-insensitive breast tumors [ 57, 58 ] . Furthermore,
`several other ADCs using anti-HER2/neu antibodies have been developed and have
`shown potent antitumor activity [ 59, 60 ] . Despite successful reports, it is important
`to note that this strategy has some limitations. The fi rst is the limited reproducibility
`of chemical conjugation due to the fact that there are numerous coupling sites on an
`antibody molecule. Secondly, chemically modi fi ed antibodies have demonstrated a
`greater tendency to aggregate, especially when multiple drug molecules are conju-
`gated to a single antibody. Furthermore, it is challenging to remove remaining
`unconjugated antibodies from the ADC mixture. Finally, the emergence of multi-
`drug resistance (MDR and MRP) mechanisms in tumors from heavily treated
`patients may engender cross-resistance to ADCs.
` With the development of recombinant DNA technology, anti-HER2/neu immu-
`notoxins composed of antibodies (or fragments) and protein toxins have become a
`promising alternative approach for HER2/neu-positive tumors. Compared to the
`ADC approach, one attractive advantage of immunotoxins is that the targeting
`antibody and antitumor toxin can be produced directly as a single molecule, thus
`avoiding laborious chemical conjugation steps. In addition, the linkage between
`the toxin and targeting antibody is identical and exactly de fi ned in a given prepara-
`tion of recombinant immunotoxin, thereby promoting homogeneity of the fi nal
`product. Compared with chemical conjugates, genetically engineered immunotoxins
`
`IMMUNOGEN 2016, pg. 2
`Phigenix v. Immunogen
`IPR2014-0676
`
`

`
`18 Design, Development, and Characterization...
`
`321
`
`reast cancer [ 45 ]
`reast cancer [ 47 ]
`olon cancer [ 44 ]
`
` C
`
` B
`
` B
`
`olon cancer [ 33 ]
`pidermoid cancer [ 40, 193 ]
`
` E
`
` C
`
`ecombinant protein
`ecombinant protein
`hemical conjugation
`
` C
`
`ecombinant protein
`ndirect bridge-linking
`
` R
`
` I
`
`reast cancer [ 35, 39, 151 ]
`
` B
`
`iposome-mediated
`
` L
`
`chemical conjugation
`
`pidermoid cancer [ 37, 46 ]
`
` E
`
`ecombinant protein
`
` R
`
`pidermoid cancer [ 41, 51 ]
`
` E
`
`ecombinant protein
`
` R
`
`cancer [ 49 ] , breast cancer [ 48 ]
`
`astric cancer [ 38, 43 ] , epidermoid
`
` G
`
`ecombinant protein
`
`astric cancer [ 152 ] , ovarian cancer [ 192 ]
`
` G
`
`ecombinant gene
`
`delivery
`
`cancer [ 42, 50, 124 ]
`schwannoma cancer [ 191 ] , breast
`gastric cancer [ 84, 123, 188– 190 ] ,
`4, 184 ] , lung cancer [ 185– 187 ] ,
`
` 3
` [
`
`
`cancer [ 40, 183 ] , prostate cancer
`
`varian cancer [ 36, 122 ] , epidermoid
`ung cancer [ 25 ]
`
` L
`
` O
`
`ecombinant protein
`hemical conjugation
`
`cancer [ 16, 18, 19 ] , gastric cancer [ 22 ]
`reast cancer [ 17, 20, 21, 23, 24 ] , ovarian
`varian cancer [ 28 ]
`reast cancer [ 26, 29 ] , Melanoma [ 27 ]
`varian cancer [ 31 ] , breast cancer [ 32 ]
`varian cancer [ 30 ]
`umor type
`
` O
`
` O
`
` B
`
` O
`
` B
`
` T
`
`hemical conjugation
`ndirect bridge-linking
`hemical conjugation
`ecombinant protein
`hemical conjugation
`roduction
`
` P
`
` C
`
` R
`
` C
`
` I
`
` C
`
` C
`
` R
`
` R
`
` R
`
` R
`
` R
`
`uman anti-HER2 scFv
`uman anti-HER2 scFv
`urine anti-HER2 mAb
`and anti-EpCAM)
`urine-bispeci fi c scFv (anti-HER2
`umanized anti-HER2 mAb
`
` H
`
` M
`
` M
`
` H
`
`enterotoxin
`taphylococcal
`
` S
`
`iptheria toxin
`
` D
`
` H
`
`acillus Cyt2Aa1 toxin
`
` B
`
`umanized anti-HER2 Fab ¢
`
` H
`
`and anti-EGFR)
`urine-bispeci fi c scFv (anti-HER2
`(dsFv) 2
`urine bivalent anti-HER2 dsFv
`Fv fragments (dsFv)
`urine anti-HER2 disul fi de-stabilized
`
` M
`
` M
`
` M
`
`urine anti-HER2 scFv
`urine anti-HER2 Fab ¢
`
` M
`
` M
`
`seudomonas exotoxin
`
` P
`
`acteria
`
` B
`
` H
`
` M
`
` M
`
` M
`
`icin
`
` R
`
`aporin
`
` S
`
`lant
`oxin source
`able. 18.1 Immunotoxins developed for HER2/neu targeted therapy
`
`elonin
`oxin
`
` T
`
` G
`
` T
`
` T
`
` P
`
`urine anti-HER2 mAb
`urine anti-HER2 mAb
`urine anti-HER2 mAb
`uman anti-HER2 scFv
`umanized anti-HER2 mAb
`argeting device
`
` T
`
` H
`
`IMMUNOGEN 2016, pg. 3
`Phigenix v. Immunogen
`IPR2014-0676
`
`

`
`322
`
`Y. Cao and M.G. Rosenblum
`
`can be easily designed to enhance antitumor ef fi cacy. Finally, data suggests that the
`emergence of MDR cellular protection mechanisms in heavily pretreated patients
`may not impact cytotoxic effects of immunotoxins.
` We have provided general principles for development of anti-HER2/neu immu-
`notoxins, and current strategies to employ these molecules for directed cancer ther-
`apy are discussed focusing mainly on design optimization to improve antitumor
`ef fi cacy and off-target toxicity.
`
` Anti-HER2/neu Immunotoxins
`
` Construction of Recombinant Immunotoxins
`
` HER2/neu-overexpressing cancers are a model of disease for the development of
` rationally designed targeted therapies. The scienti fi c advances in understanding the
`role of HER2/neu function, the structural aspects of HER2/neu function, and the sig-
`naling partners and circuitry underlying tumorigenic HER2/neu signaling have afforded
`unique opportunities for rational drug design to target these pathways. The develop-
`ment and application of various HER2/neu-targeted therapies has bene fi ted greatly by
`a more advanced understanding of HER2/neu function and biology [ 61, 62 ] .
` Overall, strategies to enhance anti-HER2/neu immunotoxin potency include
`improvements to the af fi nity and speci fi city of targeting moiety, identi fi cation and
`incorporation of new and better toxins, reengineering known toxins for reduced
`immunogenicity, and designing novel linkers between toxins and targeting moieties
`to optimize toxin translocation to the cytosol [ 63– 66 ] . Numerous excellent reviews
`have previously compared the advantages and disadvantages of a variety of cyto-
`toxic proteins including bacterial, plant, and mammalian toxins successfully
`employed for the construction of immunotoxins [ 67, 68 ] . This review will address
`how linker design and antibody af fi nity affect immunotoxins in tumor-speci fi c tar-
`geted therapies.
`
` Peptide Linker Designs
`
` The development of various linkers which bridge disparate molecules such as small
`drugs conjugated to tumor-targeting carriers has been the subject of numerous stud-
`ies for the past few years [ 69, 70 ] . Based on numerous prior studies, the incorpora-
`tion and design of linkers is critical to the success of ADCs. Conceptually, an ideal
`linker must be stable in systemic circulation, while being ef fi ciently cleaved to allow
`rapid release of an active form of the drug once the construct has been internalized
`into the tumor cell target. To this end, a variety of linkers have been designed with
`different chemical structures and stabilities [ 71, 72 ] . Selection of an appropriate
`linker depends on the type of cancer and the required cytotoxic agent. Furin is a
`
`IMMUNOGEN 2016, pg. 4
`Phigenix v. Immunogen
`IPR2014-0676
`
`

`
`s==a-c:==o
`~;;;;;;;;;~~==·
`
`~:=.-..c:===·
`~=..-..:====·
`
`18 Design, Development, and Characterization...
`
`323
`
`
`
`a
`C6.5-L-rGel
`
`C6.5-Fpe-rGel
`
`C6.5-Fdt-rGel
`
`C6.5
`
`C6.5
`
`C6.5
`
`L
`
`Fpe
`
`Fdt
`
`rGel
`
`rGel
`
`rGel
`
`b
`C6.5
`
`ML3-9
`
`MH3-B1
`
`B1D3
`
`SSSN------YFQH
`
`SSSN------YFQH
`
`TDRT------YFQH
`
`VH
`
`VH
`
`VH
`
`VH
`
`A--DS
`
`S--YT
`
`S--YT
`
`VL
`
`VL
`
`VL
`
`VL
`
`Kd
`16nM
`
`1nM
`
`0.12nM
`
`0.013nM
`
`TDRT------WLDN
`
`S--YT
`
`c
`
`C6.5/ML3-9/MH3-B1/B1D3-rGel cDNA
`
`scFv
`
`L
`
`rGel
`
`BspHI
`
`XhoI
`
`NcoI
`
`pET 32a
`(Novagen)
`
`Bacterial Expression Vector
`
` Fig. 18.1 Construction and preparation of scFv/rGel immunotoxins. ( a ) Schematic diagram of
`immunotoxin constructs containing scFv C6.5, peptide linker (L, Fpe or Fdt), and rGel toxin. ( b )
`Amino acid mutations and af fi nity parameters of the C6.5 and its mutants, ML3-9, MH3-B1, and
`B1D3. The listed amino acids for each scFv indicate mutations to the sequence and the substituting
`amino acids. Dashes indicate no changes from the original sequence. ( c ) Diagram of immunotoxin
`constructions containing scFv (C6.5, ML3-9, MH3-B1, or B1D3) and rGel
`
`cellular endoprotease and has been implicated in proteolytically activating large
`numbers of secreted proteins such as prohormones, growth factors, receptors, and
`viral glycoproteins. These proteins are synthesized as inactive precursors and must
`be proteolytically cleaved to become functionally mature. In previous studies, the
`inclusion of furin-cleavable linkers into fusion constructs containing ribotoxins,
`caspase3, or granzyme B (GrB) has demonstrated a signi fi cant improvement in
`speci fi c toxicity compared to constructs containing stable linkers [ 73, 74 ] .
` The incorporation of cleavable linkers for immunotoxins is essential since, in
`general, the toxin components are enzymatically inactive in the construct until
`intracellular release from their cell-targeting carriers [ 75, 76 ] . For recombinant
`gelonin (rGel)-based constructs, the enzymatic ( N -glycosidase) activity of the toxin
`is preserved in the intact fusion constructs, eliminating the absolute necessity for
`intracellular release of the rGel component. Nevertheless, we explored a variety of
`different linker strategies to determine whether intentionally cleavable linkers
`offered an advantage over linkers which were designed for fl exibility only.
`Illustrations of various immunotoxin constructs are shown in Fig. 18.1a . The initial
`rGel-based immunotoxins consisted of a fl exible linker (GGGGS, “L”) tethering
`the C-terminus of the human anti-HER2/neu single-chain antibody (scFv) C6.5 to
`the native rGel N-terminus. The C6.5/rGel construct was further engineered by
`incorporating two different enzymatically sensitive furin cleavage linkers between
`
`IMMUNOGEN 2016, pg. 5
`Phigenix v. Immunogen
`IPR2014-0676
`
`

`
`324
`
`Y. Cao and M.G. Rosenblum
`
`scFv and toxin components. The two furin-sensitive sequences designated “Fpe”
`(TRHRQPRGWEQL, 12 amino acid residues to PE273-284 sequence) and “Fdt”
`(AGNRVRRSVG, 10 amino acid residues to DT187-196 sequence), respectively.
`
` Tumor-Targeting scFv
`
` Previous studies identi fi ed a recombinant murine anti-HER2/neu scFv designated
`e23, and fusion constructs containing catalytic toxins such as PE and DT were
`shown to speci fi cally kill HER2/neu-expressing cells [ 33, 43 ] . A major drawback of
`such proteins is their potential for immune response after repeated administration.
`Repeated doses may cause hypersensitive reactions and lead to neutralization of the
`immunotoxins by antibodies directed against the nonhuman domains [ 77, 78 ] . The
`development of immunotoxins containing human or humanized components may
`circumvent these problems. Such immunotoxins may display reduced immunoge-
`nicity although antibodies to the toxin components may still limit prolonged therapy
` [ 79, 80 ] . We previously reported in vitro characterization and in vivo antitumor
`ef fi cacy studies of an immunotoxin composed of the human chimeric anti-HER2/
`neu antibody (BACH-250) chemically conjugated to rGel. The BACH-250/rGel
`conjugate demonstrated potent and speci fi c cytotoxicity against HER2/neu-overex-
`pressing human tumor cells in culture and against ovarian SKOV3 tumor xenografts
` [ 30 ] . However, the potential problems of diffusion of relatively large molecules
`such as full-length antibodies into solid tumors have been extensively addressed by
`Jain et al. [ 81 ] . The treatment of solid tumors presents a signi fi cant challenge since
`therapeutic antibodies must diffuse into the tumor against a hydrostatic pressure
`gradient and into disordered vasculature. These theoretical issues may not limit
`clinical response with antibodies alone since higher applied doses may circumvent
`these effects [ 81– 83 ] , but it may not be possible to overcome these limitations with
`higher doses of antibody–drug conjugates or immunotoxins because these agents
`frequently have a much narrower therapeutic window.
` Over the past decade, a variety of different anti-HER2/neu recombinant antibody
`formats have been engineered which are suitable for diverse therapeutic applica-
`tions and include monovalent, bivalent, and multivalent derivatives; single- or dou-
`ble-chain formats; and covalently or noncovalently linked assemblies of antibody
`heavy (V H ) and light chain variable domains (V L ) [ 84– 89 ] . Among these, human
`scFv appear to be effective when utilized as targeting domains incorporated into
`chimeric fusion proteins. They consist of antibody V H and V L sequences genetically
`linked via a fl exible linker, and they lack constant regions and Fc domains, thereby
`preventing possible binding to normal tissues and cells via interaction with Fc
`receptors. We previously engineered a series of novel fusion proteins created from
`various human anti-HER2/neu scFv designated C6.5 and various af fi nity mutants
`(designated ML3-9, MH3-B1, and B1D3, created by site-directed amino acid sub-
`stitutions in the CDR3s). The af fi nities of the scFvs ranged from 10 -8 to 10 -11 M
`(Fig. 18.1b ) [ 90, 91 ] . Recombinant immunotoxins containing each scFv and rGel
`were constructed by overlapping PCR and were designated C6.5/rGel, ML3-9/rGel,
`MH3-B1/rGel, and B1D3/rGel, respectively (Fig. 18.1c ).
`
`IMMUNOGEN 2016, pg. 6
`Phigenix v. Immunogen
`IPR2014-0676
`
`

`
`18 Design, Development, and Characterization...
`
`325
`
` Functional Activity Analysis of Immunotoxins
`
` Tumor-antigen af fi nity and speci fi city of scFvs are important variables which may
`impact off-target tissue distribution and toxicity in vivo. These attributes have led to
`the commonly held concept that scFv must have high af fi nity in order to be thera-
`peutically relevant. However, studies by Adams et al. suggested that high-af fi nity
`scFv may be suboptimal vehicles and that lower-af fi nity scFv appear to diffuse more
`uniformly throughout the tumor interior [ 92, 93 ] . In addition, since the presence of
`shed tumor antigen has the potential to misdirect the targeted constructs through
`immune complex formation [ 64, 94 ] , higher-af fi nity scFv could potentially be less
`effective compared to lower-af fi nity constructs. Therefore, applying a series of
`rGel-containing fusion constructs composed of various linkers and scFv mutants
`with varying af fi nity to HER2/neu, we will examine the impact of construct design
`on in vitro cytotoxicity, pharmacodynamics, and antitumor ef fi cacy. Further inves-
`tigations included the effect of antibody af fi nity on behavior in the presence of sol-
`uble antigen, formation of immune complexes, and the coincident development of
`off-target toxicity.
`
` In Vitro Studies: Impact of Various Design Modi fi cations
`on In Vitro Immunotoxin Cytotoxic Activity
`
` Effects of Linker Design on Immunotoxin Potency and Functional Stability
`
` Based on C6.5/rGel containing the universal fl exible GGGGS linker, we introduced
`proteolytically cleavable linkers (Fpe and Fdt) and to examine whether this change
`would improve killing ef fi ciency. To investigate the susceptibility of various chime-
`ric toxins to proteolytic cleavage, puri fi ed fusions were subjected to proteolysis
`with recombinant furin. As indicated in Fig. 18.2a , the Fdt linker was the most sen-
`sitive to cleavage among all constructs tested. In contrast, cleavage of the molecule
`containing the Fpe linker was highly dependent on pH. As expected, the L linker
`was found to be comparatively resistant to intracellular protease action without
`regard to the pH.
` The intracellular release of rGel after endocytosis of various C6.5/rGel fusion
`constructs was next assessed in SKOV3 cells (Fig. 18.2b ). Although the maximal
`rGel release of different fusions was achieved at different time points, the absolute
`amounts of delivered rGel found in the cytosol were virtually identical. Therefore,
`this data con fi rms the observation that introduction of an unstable furin cleavage
`linker does not improve the intracellular rGel release of the constructs.
` The linkers tethering the C6.5 scFv and the rGel toxin demonstrated a differ-
`ential sensitivity to protease action which may result in different clearance and
`metabolic kinetics in vivo [ 95, 96 ] . We next performed a stability study of the
`constructs in human plasma (Fig. 18.2c, d ). The binding activity and cytotoxicity
`
`IMMUNOGEN 2016, pg. 7
`Phigenix v. Immunogen
`IPR2014-0676
`
`

`
`a
`
`51-
`
`.. _
`
`~-----­
`
`ao-
`u-
`
`.. -.. ---
`b - 1 2 <t I U Jllh 1 2
`.. _
`•
`.. _
`25 - - -- ·---.. _ ~-- -
`
`.t I
`
`til<& Ill
`
`t J C I 11 Mil
`
`--
`
`-----------------
`
`
`~ - -
`t"N('lll
`
`-
`
`-
`
`11.1 • • • • 100 100
`
`c
`
`d
`
`60
`
`50
`
`~40
`c:
`~30
`!.2 20
`
`10
`
`0
`
`10
`
`60
`50
`40
`30
`20
`Incubation time (h)
`
`70
`
`80
`
`C6.5-l-tGel
`
`C6.5-l'pe-rGel
`
`C6 5-Fdl.fGel
`
`326
`
`
`
`Y. Cao and M.G. Rosenblum
`
` Fig. 18.2 Functional analysis of C6.5/rGel series immunotoxins in vitro. ( a ) Western blot analysis
`of furin cleavage of puri fi ed C6.5/rGel fusion constructs. ( b ) Western blot analysis of intracellular
`rGel release of C6.5/rGel fusions in SKOV3 cells. ( c , d ) Functional stability analysis of the fusions
`by whole-cell ELISA and cytotoxicity on SKOV3 cells. The proteins were incubated in human
`plasma at 37 °C for up to 72 h before test
`
`of the C6.5–Fdt–rGel construct was shown to be the least stable after incubation
`because of the instability of the Fdt linker. On the other hand, of all molecules
`tested, the C6.5–L–rGel demonstrated the highest degree of stability after plasma
`incubation. The two companion molecules with the furin-cleavable linkers Fpe or
`Fdt demonstrated signi fi cantly less stability in vitro.
` Based on our studies, introduction of a furin-cleavable linker between C6.5 and
`rGel did not result in improved intracellular rGel release and cytotoxic effects
`in vitro, despite showing more sensitivity to protease cleavage and greater intracel-
`lular release of the rGel component. On the other hand, the rGel-based molecules
`with the furin-cleavable linkers demonstrated signi fi cantly less functional stability
`in vitro. The enzymatic stability of the linker can seriously affect the pharmacoki-
`netics of immunotoxins and can apparently impact loss of targeting function and
`in vivo ef fi cacy. Therefore, we clearly demonstrated the highly individualized nature
`of some payloads and targeted constructs and that observations regarding rGel may
`not necessarily translate to other payloads.
`
`IMMUNOGEN 2016, pg. 8
`Phigenix v. Immunogen
`IPR2014-0676
`
`

`
`18 Design, Development, and Characterization...
`
`327
`
` The Impact of scFv Af fi nity on Immunotoxin Activity
`
` Although previous studies suggested that the binding af fi nity for antigen plays a
`pivotal role in the total concentration and penetration of scFv into tumors [ 92, 93 ] ,
`few companion studies have been conducted to determine whether scFv-based
`immunotoxins display the same behavior with regard to the relationship between
`af fi nity, tumor penetration, tumor residence, and ef fi cacy. Therefore, based on the
` fl exible GGGGS linker, we created a series of rGel-based immunotoxins from
`af fi nity mutants of the human anti-HER2/neu scFv C6.5.
` To ensure that immunotoxins retained antigen binding ability, the fusion proteins
`were compared in an ELISA-based binding assay using HER2/neu-positive (SKBR3,
`BT474 M1) and HER2/neu-negative (MCF7) cells. All the scFv/rGel constructs
`demonstrated speci fi c and signi fi cant ELISA binding to HER2/neu-positive cells
`with negligible binding to negative cells (Fig. 18.3a ). The equilibrium dissociation
`constants ( K d ) were calculated, and the af fi nities of immunotoxins for BT474 M1
`cells were found to be 53.13 nM (C6.5/rGel), 1.45 nM (ML3-9/rGel), 0.18 nM
`(MH3-B1/rGel), and 0.027 nM (B1D3/rGel). The correlation between the K d values
`of the scFvs and fusion constructs was found to be signi fi cant with a correlation
`coef fi cient of 0.939 ( p < 0.01), indicating that introduction of the rGel component
`did not affect the binding af fi nity of the scFv.
` We next examine the ability of various af fi nity scFv/rGel fusions to speci fi cally
`internalize into target cells. Immuno fl uorescence staining was performed on HER2/
`neu-positive and HER2/neu-negative cells after exposure to the constructs (Fig. 18.3b ).
`As quanti fi ed by relative fl uorescence (Fig. 18.3c ), the internalization ef fi ciency into
`HER2/neu-positive cells was shown to increase with increasing antibody af fi nity. For
`BT474 M1 cells, the relative fl uorescence intensities were 56.30 (C6.5/rGel), 73.69
`(ML3-9/rGel), 86.29 (MH3-B1/rGel), and 90.41 (B1D3/rGel). There was a good
`correlation between increases in apparent af fi nity and internalization ef fi ciency
`( r 2 = 0.8289; p < 0.01), indicating that ef fi cient binding to the cell surface appears to
`be primarily responsible for rapid internalization after cell exposure.
` The cytotoxicity of the various scFv/rGel constructs was then tested against a
`panel of different tumor cell lines (Table 18.2 ). As expected, there appeared to be a
`good correlation ( r 2 = 0.7812; p < 0.01) between apparent af fi nity and IC 50 values.
`The highest targeting indices were found for the highest-af fi nity construct (B1D3/
`rGel). Therefore, in vitro, binding af fi nity appeared to mediate internalization
`ef fi ciency, and this appeared to directly impact the overall cytotoxic effects observed.
`Furthermore, against HER2/neu-negative cells, there was little or no speci fi c cyto-
`toxicity of the constructs compared to rGel itself.
` Based on scFv/rGel immunotoxins, we demonstrated that increasing af fi nity
`could improve cell binding ability, internalization ef fi ciency, and cytotoxic activity
`on HER2/neu-positive cells. However, fusion toxins with a tenfold increase in
`af fi nity did not show a corresponding improvement in either internalization or a
`concomitant improvement in cytotoxic effects. This suggests that the internalization
`rate of the construct may primarily be associated with HER2/neu receptor recycling
`and the rate of antigen endocytosis and this may primarily be unaffected by the
`
`IMMUNOGEN 2016, pg. 9
`Phigenix v. Immunogen
`IPR2014-0676
`
`

`
`~ 0.6 ""~ MCF7 (-)
`
`0.1
`o.
`0.6
`o.•
`0.3
`
`0 C6..,..
`....,.liiiGOI
`0
`A f*U.BllrGtl
`. .....
`• at~
`
`0.1
`
`3 0,5
`~ 0 ,4
`; 0.3
`~ 0.2
`~ 0.1
`0.0
`0.01
`
`0.1
`
`10
`
`100
`
`cone. nt.t
`
`0.1
`
`10
`
`100
`
`Cone. nt.t
`
`0.2
`
`0.1
`o.o
`0.01
`
`0.1
`
`10
`
`100
`
`cone. nM
`
`C6.5/rGel
`
`Ml3·9/rGel
`
`MH3·Bl/rGel
`
`8103/rGel
`
`rGel
`
`b
`
`SKBR3
`
`BT474 Ml
`
`MCF7
`
`a
`0,1
`o.•
`0.6
`o.•
`0.3
`
`0.2
`
`0.1
`
`0.0
`0 .0 1
`
`c
`
`328
`
`Y. Cao and M.G. Rosenblum
`
` Fig. 18.3 Characterization and comparison of different scFv/rGel immunotoxins. ( a ) Evaluation
`binding activity of the fusion constructs to HER2/neu-positive (SKBR3 and BT474 M1) and HER2/
`neu-negative (MCF7) cells by whole-cell ELISA. ( b ) Internalization of the immunotoxins on HER2/
`neu-positive and HER2/neu-negative cells. Cells were subjected to immuno fl uorescent staining
`with anti-rGel antibody (FITC-conjugated secondary), with propidium iodine nuclear counterstain-
`ing. ( c ) Quanti fi cation of internalization rate of the immunotoxins. The bar graphs were calculated
`from relative fl uorescence estimation, and the values are expressed as mean ± S.D ( n > 50)
`
`af fi nity of immunotoxin binding [ 97, 98 ] . Moreover, intracellular traf fi cking and
`distribution of the toxin component to the ribosomal compartment may be addi-
`tional critical factors which can de fi ne immunotoxin sensitivity [ 99 ] .
`
` Mechanistic Studies of Immunotoxin Cytotoxic Activity
`
` Studies of bacterial and plant immunotoxins have provided us with essential infor-
`mation regarding intracellular routing and translocation events [ 100, 101 ] , and this
`
`IMMUNOGEN 2016, pg. 10
`Phigenix v. Immunogen
`IPR2014-0676
`
`

`
`18 Design, Development, and Characterization...
`
`329
`
` Table 18.2 Comparative cytotoxicity of rGel-based fusion constructs against various tumor cell
`lines
`
` IC 50 (nM) (TI) *
` HER2/neu
` MH3-B1/
` ML3-9/
` Tumor
`Expression
`rGel
`rGel
` C6.5/rGel
`type
` Cell line
`level
` 2.7 (387)
` 5.0 (211)
` 6.4 (165)
` ****
` Breast
` SKBR3
` 3.9 (56)
` 10.9 (20)
` 18.9 (12)
` ****
` Breast
` BT474 M1
` 9.1 (106)
` 20.4 (47)
` 30.1 (32)
` ****
` Gastric
` NCI-N87
` 13.0 (41)
` 18.1 (29)
` 24.3 (22)
` ****
` Lung
` Calu3
` 155 (2)
` 149.9 (2)
` 145.8 (2)
` *
` Breast
` MDA MB231
` 260.6 (1)
` 246.7 (1)
` 246.9 (1)
` *
` Breast
` MCF7
` 153.9 (1)
` 126.9 (2)
` 61.4 (3)
` *
` Melanoma
` A375m
` 194.6 (1)
` 185.4 (1)
` 160.8 (1)
` *
` Cervical
` Me180
` *Targeting index (TI) represents IC 50 of rGel/IC 50 of immunotoxin
`
` B1D3/rGel
` 1.9 (567)
` 1.2 (177)
` 4.9 199)
` 10.0 (53)
` 204.8 (1)
` 266.9 (1)
` 173.9 (1)
` 213.1 (1)
`
` rGel
` 1061
` 219.8
` 965.6
` 531.1
` 297.3
` 247.4
` 207.3
` 222.5
`
`information can be employed in the design of optimized constructs. The mechanism
`of cytotoxicity of anti-HER2/neu immunotoxins generally involves inhibition of
`cellular protein synthesis, although the point of attack in this pathway can vary
`slightly depending on the toxin [ 102, 103 ] .
` The plant-derived toxins are primarily ribosome-inactivating proteins (RIPs) and
`are enzymes which depurinate rRNA, thus inhibiting protein synthesis. They may
`also depurinate other polynucleotide substrates [ 104– 106 ] . rGel is an N -glycosidase
`generating cytotoxic effects which are the direct result of protein synthesis inhibi-
`tion (Fig. 18.4a ). Although there have been numerous preclinical studies of rGel-
`based immunotoxins for the treatment of both solid and hematological tumors, the
`actual mechanisms behind the induction of cell death appear to vary depending on
`the cell type targeted [ 107– 109 ] . This occurs because the basic mechanism of pro-
`tein synthesis inhibition due to the rGel component causes the loss of different high-
`turnover proteins in cells derived from different tumor types. This fi nally results in
`cytotoxic patterns and mechanisms which can vary widely based on the proteins
`critical for survival of different cell types targeted.
` The cytotoxic effects mediated by our rGel-based immunotoxins were analyzed
`including apoptosis, necrosis, and autophagy in BT474 M1 cells. As shown in
`Fig. 18.4b , scFv/rGel fusions did not activate caspase-dependent apoptosis in tar-
`get cells, showing no cleavage of the caspase substrate PARP. We next assessed
`lactate dehydrogenase (LDH) release, a marker of abrupt membrane lysis, and
`found that exposure of BT474 M1 cells to immunotoxins did not induce necrotic
`cell death using this parameter (Fig. 18.4c ).
` We next examined whether the cytotoxic effects of these immunotoxins activate
`autophagic signaling. As shown in Fig. 18.4d , the ratio of LC3-II formation to the
` b -actin control was shown to be increased after treatment with the fusion constructs,
`demonstrating that autophagic fl ux was induced by rGel-based immunotoxins in
`BT474 M1 cells. In addition, autophagic induction by fusion constructs was further
`validated by the observed selective release of cellular high mobility group box 1
`
`IMMUNOGEN 2016, pg. 11
`Phigenix v. Immunogen
`IPR2014-0676
`
`

`
`a
`
`Tumor cells
`
`·····-·······
`
`Herl/neu homodimers
`Or heterodimers
`
`Turnover Proteins
`
`~~
`·- - __..,..
`~ Inhibition of
`Protein synthesis
`c
`
`0
`
`e
`_;////
`
`...,
`- ..
`- ;///)'
`- a •
`-------
`
`LC4 ~~-~ HMG81
`Lc•
`,_,.,. -
`
`C.lla
`
`Medium
`
`-t~/h-7 - t~//,{/
`,_., -
`
`--
`
`b
`
`14h
`
`d
`Nil
`
`LC•
`LC•
`
`.......
`
`330
`
`
`
`Y. Cao and M.G. Rosenblum
`
` Fig. 18.4 Cell-killing mechanism analysis of the immunotoxins on BT474 M1 cells. ( a ) A gener-
`alized view of cytotoxic mechanism of anti-HER2/neu scFv/rGel fusions in targeted cells. ( b )
`Western blot analysis of PARP cleavage after 24 and 48 h of scFv/rGel fusions treatment. ( c )
`Evaluation of necrosis by LDH release after treatment with fusion proteins or Triton X-100
`(mean ± S.D. from three replicates). ( d ) Analysis of LC3 after treated with scFv/rGel fusions for 24
`and 48h. ( e ) Analysis of cell extract and medium for HMGB1 protein after immunotoxins treat-
`ment for 48 h
`
`(HMGB1) (Fig. 18.4e ). Tumor cells undergoing autophagy can s