Vol. I. 425-434, April 1995
`
`Clinical Cancer Research 425
`
`An Immunoconjugate of Lys3-Bombesin and Monoclonal Antibody
`
`
`
`
`
`22 Can Specifically Induce FC'yRI (CD64)-dependent Monocyte­
`
`and Neutrophil-mediated Lysis of Small Cell Carcinoma
`of the Lung Cells1
`
`Jian Chen, Michael Mokotoff, Jie-Hua Zhou,
`Michael W. Fanger, and Edward D. Ball2
`Division of Hematology/Bone Marrow Transplantation, University of
`Pittsburgh Medical Center and Pittsburgh Cancer Institute, Pittsburgh,
`Pennsylvania 15213 (J. C., J-H. Z., E. D. B.); Depanment of
`Pharmaceutical Sciences, University of Pittsburgh, Pittsburgh,
`Pennsylvania 15213 [M. M.]; and Depanment of Microbiology,
`Danmouth Medical School, Lebanon, New Hampshire 03756
`(M. W.F.]
`
`ABSTRACT
`Small cell carcinoma of the lung (SCCL) accounts for
`25% or all lung cancers and has a very poor prognosis. It is
`known that SCCL cells produce gastrin-releasing peptide, a
`peptide which has similar biological actions to that or bom­
`besin, an amphibian counterpart of gastrin-releasing pep­
`tide, and express high affinity cell surface bombesin/gastrin­
`releasing peptide receptors. These receptors can serve as
`targets for specific immunotherapy. Cell surface receptors
`for the Fe portion or IgG (FcyR) are a family of molecules
`that can mediate a variety of immune reactions, including
`tumor cell cytotoxicity. We hypothesized that an immuno­
`conjugate of bombesin and a mAb directed to the high­
`affinity FcyRI (mAb 22) should be able to trigger specific
`cytotoxicity against SCCL cells. In this article, we report the
`construction of this immunoconjugate and demonstrate its
`capacity to redirect immune effector cells toward SCCL
`cells and el.icit lysis or these target cells. The immunoconju­
`gate stained the majority or cells from four SCCL cell lines
`and reacted with FcyRI on activated monocytes and neu­
`trophils. After preincubating monocytes and neutrophils
`with recombinant 'I interferon to enhance the expression or
`FcyRI on the cell surface, we demonstrated that 60-98% of
`SCCL cells could be lysed in the presence of the immuno­
`conjugate in a chromium release assay. Tumor cell lysis was
`observed over a wide range of immunoconjugate concentra­
`tions, was dependent on the ratio of E:T cells, and could be
`blocked by the addition or either parental molecule or the
`immunoconjugate. Bispecific molecules redirecting immune
`effector cells to target SCCL cells may have clinical appli­
`cation in the therapy of SCCL.
`
`INTRODUCTION
`The prognosis for SCCL J remains poor despite recent
`advances in the management of certain other cancers. An esti­
`mated 42,500 cases of SCCL were diagnosed in the United
`(1, 2).
`
`States in 1993, accounting for 25% of all lung cancers
`
`
`Although the majority of SCCL patients respond lo chemother­
`
`
`
`apy, the response is of short duration (3, 4). Median survival for
`
`with limited-stage disease is 14-18 months, and for
`patients
`
`
`patients with extensive disease is 9-11 months. Only 15-25%
`survive for more than 2 years ( 1-4).
`
`of limited-stage patients
`
`
`Many combinations of chemotherapeutic agents have been used
`
`
`clinically without much improvement in the long-term survival
`
`
`of rate. Recent efforts have been focused on the development
`
`
`
`other therapeutic strategies for the treatment of SCCL, including
`
`
`immunological and hormonal therapy (5, 6) and intensive che­
`
`
`
`motherapy with autologous bone marrow or peripheral stem cell
`support (7-9).
`BN is a 14-amino acid peptide which was initially isolated
`
`from the skin of the frog Bombina bombina (IO). The mamma­
`
`lian analogue of BN, GRP. contains a COOH-terminal hep­
`
`
`
`tapeptide sequence identical to that of BN (11). The majority of
`human SCCL cell lines produce GRP and express
`a single class
`II has also been
`
`
`of high-affinity receptors for BN/GRP (12-16).
`
`
`
`demonstrated that GRP functions as an aurocrine growth factor
`
`for human SCCL cells. Blockage of this autocrine growth stim­
`
`ulatory activity of GRP in SCCL has been the focus of several
`A mAb against GRP and a number of BN/GRP receptor
`studies.
`
`antagonists have been shown to inhibit the growth of SCCL
`cells both in vitro and in vivo (5, 17-19). Since BN/GRP
`
`
`receptors have limited distribution in the body, they can serve as
`A novel approach of
`
`for specific immune reactions.
`targets
`
`
`
`
`immunotherapy targeting the BN/GRP receptors expressed on
`
`SCCL cell surface has been developed in our laboratory. We
`have made an immunoconjugate between a BN-like peptide and
`
`
`a mAb against human high-affinity fC'(R (FC'(RJ, CD64), which
`
`is expressed on the surface of human monocyres and activated
`neut roph ils. We hypothesized that this immunoconjugate
`should
`
`Received 8/26/94; accepted 12/12/94.
`1 This work was supponed by Grant CA37868 from the National Cancer
`Institute. NIH, Bethesda, MD.
`2 To whom requests for reprints should be addressed, at Division of
`Hematology/Bone Marrow Transplantation, Montefiore University Hos·
`pital, 200 Lothrop Street, Pittsburgh, PA 15213.
`
`3 The abbreviations used are: SCCL, small cell carcinoma of the lung;
`BN, bombesin; GRP, gastrin-releasing peptide; Fc-yR, Fe 'f receptor;
`ADCC, antibody-dependent cell-mediated cytotoxicity; SATA, N-suc­
`cinimidyl S-acetylthioacetate; SulfcrSMCC. sulfosucci.nimidyl 4-(N-male­
`imidomethyl) cyclohexane-1-carboxylate; R·HPLC, reversed phase
`
`HPLC: SH, free sullbydryl group; rlFN·'f, human recombinant IFN-'(;
`MFI, mean fluorescence intensity; G-CSF, granulocyte-colony-stimu­
`lati.ng factor; PBA, phosphate-buffered saline-bovine serum albumin­
`sodium azide.
`
`Downloaded from
`
`clincancerres.aacrjournals.org
`
`on August 31, 2015. © 1995 American Association for Cancer
`Research.
`
`1 of 11
`
`BI Exhibit 1016
`
`

`

`426 lmmunoconjuga1c of Lys�·BN and mAb 22 Can Induce Lysis of SCCL
`
`A c - S·CiizCOO - N�
`
`0
`\\
`
`SATA
`
`./f
`0
`
`L y s3 ·B ll
`I
`NH·CO-CHz·S-Ac
`
`11Ah22
`NH 2
`
`H �- c o-OCH z - N�
`
`0
`"
`
`mAbZ2
`
`l-BN
`
`Ly
`
`ls
`
`//
`0
`�Ab22-Lys-BN immunoconju9at�
`Fi�. I Schema of 1hc chemical conjugalion be1ween Lys·'·BN and mAb 22 or i1s F(ab'h fragmen1s.
`
`5-CH2CONH
`
`ratio of I 0: I . After 30 min of reaction at room temperature, the
`
`cells toward SCCL be able to redirect these immune effector
`
`
`
`
`
`cells and elicit specific ADCC against these cells.
`
`
`
`Lys3-BN-SATA conjugate was separated from nonreacted
`Lys3-BN and SATA by R-HPLC on a Vydac Cl8 analytical
`MATERIALS AND METHODS
`
`
`
`column. The R-HPLC eluent containing Lys'-BN-SATA was
`
`
`adjusted to pH 7.0 by adding I M sodium phosphate (pH 8.0).
`
`
`
`Cell Lines. SCCL cell lines, NCl-H69, NCl-H345, and
`
`
`
`The free sulfhydryl group was generated by deacetylation with
`
`
`SHP· 77, were maintained in RPM! 1640 medium (GIBCO·
`
`hydroxylamine at 4°C for 2 h. A second R-HPLC was per­
`with 10% FCS. 2 mM
`
`BRL, Grand Island, NY) supplemented
`
`
`
`
`formed to separate Lys3-BN-SH. The fraction containing Lys3-
`
`
`L·glutamine. 100 units/ml penicillin, and JOO µg/ml streptomy·
`
`
`
`
`BN-SH was collected and neutralized to pH 7.0. The presence of
`
`
`
`cin (GIBCO-BRL) at 37°C in a humidified atmosphere with 5%
`
`
`groups could be determined via reaction with
`free sulfhydryl
`
`
`C02• Another SCCL cell line, DMS273 (20), was maintained in
`
`
`Ellman's reagent. At the same time. mAb 22 or F(ab'h frag­
`
`
`Waymouth"s MB 752/1 medium (GIBCO-BRL) supplemented
`
`
`
`a to produce ments of mAb 22 was reacted with Sulfo-SMCC
`with 10% FCS.
`
`and Reagents. Anti-Fc-yRI (mAb 22; Ref.
`Antibodies
`
`
`
`
`maleimide-activa1ed antibody. The activated antibody was sep­
`
`
`
`
`from unreacted Sulfo-SMCC by centrifugation through a
`arated
`
`
`
`21), F(ab'h fragments of mAb 22 (F(ab'h). and FITC-labeled
`
`
`
`30 apparatus (Amicon, Beverly, MA). The final con·
`
`
`
`mAb 22 were obtained from Medarex, Inc. (Annandale, NJ).
`Centricon
`
`
`
`SCCL-1, an lgG2a mAb reactive with transferrin receptor (22).
`
`
`
`
`between Lys3-BN-SH and the activated antibody was
`jugation
`
`carried out by mixing an equal molar amount at room temper­
`
`
`
`
`was produced in this laboratory. Lys3-BN, a BN analogue with
`
`
`ature overnight. The unreacted Lys3-BN and other by-products
`
`
`
`similar binding affinity to the BN/GRP receptor (11, 13), and
`
`
`
`were removed by centrifugation through a Centricon 30 appa­
`
`
`
`hydroxylamine were purchased from Sigma Chemical Company
`
`
`ratus. The concentration of the immunoconjugate was quantified
`
`
`(St. Louis, MO). Conjugation chemicals, SATA and Sulfo­
`
`
`
`
`using a Bio-Rad DC protein assay (Bio-Rad Laboratories, Rich­
`
`
`SMCC, were obtained from Pierce Chemical Co. (Rockford,
`
`mond, CA) and its purity was checked by SOS-PAGE. The
`IL).
`Protein Conjugation. Lys·'-BN was freshly dissolved in
`
`
`
`
`scheme for the peptide conjugation is shown in Fig. I.
`lmmunonuorescence. SCCL cells were washed twice
`
`
`0.1 M sodium phosphate buffer (pH 7.4) containing 2.5 mM
`
`
`
`EDTA and the SATA was freshly dissolved in 100% dimeth­
`
`
`0.1 % BSA and 0.1 % sodium azide with ice-cold PBS containing
`
`ylformamide. The SATA was mixed with L-BN in a final molar
`
`
`
`amounts of the with different (PBA solution) and incubated
`
`Downloaded from
`
`clincancerres.aacrjournals.org
`
`on August 31, 2015. © 1995 American Association for Cancer
`Research.
`
`0
`Na03S //
`
`\\
`0
`
`Sulfo-Sl!CC
`
`+ 'CN-0 co-OCHz -N :n
`1
`
`0
`\\
`
`ff
`0
`
`2 of 11
`
`BI Exhibit 1016
`
`

`

`Clinical Cancer Research 427
`
`immunoconjugate at 4°C for I h in the presence of 100 µg/ml
`human lgG. The final concentration of immunoconjugate added
`was 5, 25, and 50 µg/ml per 5 x 106 cells. After washing three
`times with PBA solution, the cells were resuspended and incu­
`bated with FITC-labeled goat F(ab'h anti-mouse immunoglob­
`utin (Caltag Lab., South San Francisco, CA) for 30 min at 4°C.
`After washing, the cells were fixed in I% paraformaldehyde.
`Monocytes and neutrophils, before and after IFN--y stimulation,
`were stained directly with FITC-tabeled mAb 22 to evaluate the
`expression of Fc-yRI. All of the samples were analyzed by
`FACScan flow cytometry (Becton Dickinson, San Jose, CA).
`The number of antibody sites bound per cell was deter­
`mined by using a standard bead mixture containing beads with
`six different concentrations of FITC labeling (23). The standard
`beads were purchased from Flow Cytometry Standards Corpo­
`ration (San Juan, Puerto Rico; Quantum 24 and 25). The median
`fluorescent intensities of the six bead peaks were used to set a
`standard curve for determination of the amount of FITC/cell.
`Antibody/cell was calculated by dividing the FITC/cell by the
`flourescein:protein ratio for the standard bead mixture. Appro­
`priate isotype-matched negative controls were performed to
`subtract nonspecific binding of antibody.
`Stability or Lys-' ·BN x 22 lmmunoconjugate
`Binding to
`SCCL Cells. It has been reported that BN/GRP receptors are
`rapidly internalized upon binding SN or GRP (24). To deter­
`mine the stability of immunoconjugate binding, SCCL cells
`were incubated with 20 µg/ml immunoconjugate for 30 min and
`then washed to remove unbound immunoconjugate. The immu­
`noconjugate-bound cells were incubated at 37°C in fresh media
`for various times before the addition of FITC-labeled antimouse
`antibody at 4°C. The binding of the immunoconjugate to SCCL
`cells at the initiation of incubation at 37°C was defined as 100%.
`Blood.
`
`Isolation or Effector Cells from Peripheral
`Monocytes were separated from Leuko-Packs obtained from the
`Pittsburgh Central Blood Bank. After Ficoll-Hypaque gradient
`centrifugation, mononuclear cells were washed twice with
`HBSS (GIBCO-BRL) containing I mM EDTA and then cultured
`in a flask with RPMI 1640 medium containing 10% FCS for 2 h
`at 37°C. The nonadhercnt cells were removed. The adherent
`cells were detached, collected, stained with anti-CDl4, anti­
`CD45, anti-CD3, anti-CD 13, and anti-CD56 (Becton Dickin­
`son), and analyzed by FACScan flow cytometry to determine
`the purity. The neutrophils were isolated from normal donors. A
`20-ml blood sample was centrifuged on Ficoll-Hypaque. The
`cell pellet was suspended in an equal volume of RPMI 1640
`medium. Four hundred µI of 5% dextran were added to each ml
`of blood sample. After sedimentation at I x g for I h at 4°C, the
`neutrophil-rich supernatant was collected. The purity of neutro­
`phils thus isolated was usually higher than 95%.
`or Immune Effect.or Cells. rIFN-y was
`Activation
`kindly provided by Genentech (San Francisco, CA). The con·
`centration of rlFN--y used in this study (200 units/ml) has been
`shown to saturate its receptors and to induce a maximum in·
`crease in the expression of Fc-yRI on the surface of monocytes
`and neutrophils (25). Isolated monocytes and neutrophils were
`incubated with rlFN--y in RPMI 1640 medium containing 10%
`FCS for 18 h at 37°C before the ADCC assay. The expression
`of Fc-yRI before and after rlFN--y incubation was determined by
`
`staining with FITC-labeled mAb 22 and analyzed by FACScan
`flow cytometry.
`ADCC Assay. The assay was performed in 96-well
`round-bottomed microtiter plates (Rainin Instrument Co., Woburn,
`MA). The target SCCL cells were washed once with RPMI 1640
`medium and incubated with 100 µCi sodium [�1Cr)chromate
`(New England Nuclear, Boston. MA) for I h at 37°C. After
`washing several times, cells were resuspended in RPM! 1640
`medium containing 10% FCS to a concentration of I x 10-'
`cells/ml. Activated monocytes or neutrophils serving as effector
`cells were suspended in RPMI 1640 medium in a final concen­
`tration of 2 x 107 cells/ml. Effector cells (100 µI) were added
`to the first row of wells and serial dilution was performed with
`an equal volume of RPM! 1640 medium. Then, 100 µI of target
`cells were added in the wells to yield a final E:T ratio of 100: I,
`50: I, 25: I, and 12: I. In a standard assay, 5 µg immunocongu­
`gate were added to yield a final concentration of 25 µg/ml. The
`mAb SCCL-1 was included in each assay as a positive control
`to measure the activity of the monocytes and neutrophils. Sev­
`eral other controls with E:T ratio of I 00: I were also incorpo­
`rated in each assay, including incubation of target cells with
`immunoconjugate alone, incubation of target and effector cells
`without any antibody, and incubation of target and effector cells
`with unconjugated mAb 22. In each assay, a IQ.fold excess of
`Lys3-BN and unconjugated mAb 22 along with the immuno­
`conjugate were incubated along with the target and effector cells
`to determine whether the tumor cell lysis could be blocked by
`either of the parental molecule. To determine whether the pres­
`ence of human immunoglobulin could interfere with the redi­
`rected tumor cell killing. human lgG at a final concentration of
`I mg/ml was added in some assays. After incubation at 37°C for
`4 h or 18 h, the microplates were centrifuged and the supernatant
`was collected for the determination of 51Cr release. Maximum
`lysis was achieved by the addition of 100 µ1 5% NP40 to 100 µI
`target cells. The percentage of cell lysis was calculated as:
`
`(experimental cpm - spontaneous release mean cpm)
`
`X 100/(maximum release mean cpm
`
`- spontaneous mean cpm)
`Spontaneous release of� 1 Cr from the target cells was less than
`10% of maximum release in a 4-h assay and less than 20% in an
`18-h assay. For dose-response assays, the immunoconjugate was
`serially diluted and added. The E:T ratio in those assays was
`100: I. Since the amount of immunoconjugate added in a stan­
`dard assay was 25 µg/ml, we defined the percentage of tumor
`cell lysis with that amount of immunoconjugate as 100% activ­
`ity. The tumor cell lysis achieved with diluted immunoconjugate
`was calculated according.ly.
`Analysis. Each experimental result was ob­
`Statistical
`tained from triplicates and reported as the mean :! SD. Signif­
`icance level was determined by paired Student's r test when
`applicable.
`
`RESULTS
`Binding of the Lys3-BN x 22 Immunoconjugate to
`SCCL Cells. Unconjugated mAb 22 and its F(ab'h fragments
`did not react with SCCL cells. After conjugation with Lys3-BN,
`
`Downloaded from
`
`clincancerres.aacrjournals.org
`
`on August 31, 2015. © 1995 American Association for Cancer
`Research.
`
`3 of 11
`
`BI Exhibit 1016
`
`

`

`428 lmmunoconjugatc or Lys-'-BN and mAb 22 Can Induce Lysis or SCCL
`
`0
`0 10 20 30 •o 50
`
`100
`
`20
`
`! 80
`!: 60
`� •o
`.
`• Q.
`
`100
`
`! 80
`!: 60
`� 40
`.. :. 20
`
`100
`! 80
`
`OMS273 cell•
`
`�
`
`DMS273 cells
`
`300
`
`ii:
`::I
`
`� 200
`
`100
`
`H345 call•
`
`�
`
`0
`0 IQ 2 0 30 •o 50
`
`H345 coll•
`
`300
`
`200 ~ ii: ::I
`
`100
`
`0
`
`10 2 0 30 •o 50
`
`10 20 30 •O so
`
`H69 cello
`
`�
`
`H69 coll•
`
`300
`
`200 ? ii: ::I
`
`100
`
`MFI than the one prepared between the F(ab'h fragments of
`
`
`
`
`
`mAb 22 and Lys"-BN. The number of BN/GRP receptors has
`
`
`to be 1500 and 1000, respectively, for H345 and
`been estimated
`
`H69 cell lines (16, 24). We have not quantified the BN/GRP
`
`
`receptor numbers on DMS273 and SHP-77 cells. By flow cy­
`
`
`
`tomctric analysis, it appears that DMS273 cells express approx­
`
`
`imately the same number of BN/GRP receptors, while the
`
`
`SHP-77 cells have slightly fewer BN/GRP receptors when com­
`pared to H345 cells.
`of the Lys"-BN x 22 immunoconjugate
`to
`The binding
`
`
`
`normal peripheral lymphocytes and to two leukemia cell lines
`
`
`was also tested. The immunoconjugates did not bind to normal
`
`
`peripheral lymphocytes. The mAb 22 and its F(ab'h fragments
`
`
`stained 20-50% of cells from the HL-60 and NB4 leukemia cell
`
`
`lines with very dim fluorescence, consistent with the known
`
`
`expression of FcyRI on these cells. There was no change in
`fluorescence
`
`
`intensity when these cells were stained with the
`
`
`
`
`immunoconjugate, although the percentage of positive cells
`
`increased slightly.
`The number of immunoglobulin molecules per SCCL cell
`
`
`
`
`was calculated using a bead calibration method. H345 cells
`
`
`
`bound an average of 6116 immunoconjugate molecules, H69
`cells bound 5036. OMS 273 cells bound 94 73, and SHP-77 cells
`
`bound 2399. The number of bound immunoconjugates is greater
`
`
`than the published number of BN/GRP receptors per cell ( 14-
`
`
`16). Since there are three types of BN/GRP receptors present, it
`
`
`is possible that the immunoconjugate may bind to low-affinity
`
`
`BN/GRP receptors or cross-bind to other family members of
`GRP receptors.
`It has been reported that the BN/GRP receptors arc quickly
`
`
`
`
`
`internalized upon binding BN/GRP. To determine the stability
`of Lys"-BN X 22 immunoconjugate
`
`binding, cells from four
`(25 µg/ml)
`
`
`SCCL lines were incubated with a saturation amount
`0
`
`of immunoconjugate for 30 min and then washed to remove
`0 10 20 30 • o 50
`0 10 20 30 •o 50
`Concentretlon
`(•glml)
`Concentr•tlon
`(•g/rnl)
`
`
`unbound immunoconjugate. The cells were incubated at 37°C in
`Fig. 2 Flow cytometric analysis or four SCCL lines stained with the
`
`
`fresh media for various times before the addition of FITC­
`immunoconjugatc made from Lys-'-BN and mAb 22 (•) or F(ah'h (0).
`
`
`
`
`labcled antimousc antibody. The results from three experiments
`Each poinl represents the mean or three individual analyses and the SD
`
`
`arc shown in Fig. 3. The percentage of positive cells did not
`
`
`is less than Ill% for positivity and 20% for MFI. lsotypic control or
`
`change significantly during the first 4-h period. After 24-h
`
`mouse lgG was included in all or the analyses. Both the percentage or
`
`
`
`
`incubation, the majority of immunoconjugate binding was lost.
`positive cells and the MFI increase with the amount or immunoconju­
`Binding of the Lys3-BN x 22 lmmunoconjugate to
`gatc added. Compared with the immunoconjugatc made between
`Lys·'-BN and mAb 22. the immunoconjugate made between Lys·'-BN
`
`
`Monocytes and Neutrophils. We also tested the binding of
`and F(ah'h has a similar percentage or positivity. but consistently lower
`
`the Lys"-BN x 22 immunoconjugate to peripheral monocytes
`
`MFI.
`
`
`
`before and after incubation with 200 units/ml rlFN-'Y for 18 h.
`
`
`
`This is shown in Table I. rlFN-'Y dramatically increased the
`
`
`
`expression of FcyRI on human monocytes as defined by the
`
`of MFI from <30 to > 120. In contrast, there was no
`increase
`<he binding of the immunoconjugate to SCCL cells could be
`
`
`
`
`
`change in the expression of FqRI on human peripheral lym­
`
`
`
`
`detected by the addition of FITC-labeled antimouse an<ibody.
`
`phocytes. The conjugation of Lys"-BN to mAb 22 did not
`
`
`
`The resuhs of flow cytometric analysis of the binding of two
`
`
`
`interfere with its binding to FqRI. Neutrophils without rlFN-'Y
`
`
`immunoconjugates to four SCCL cell lines are illustrated in
`
`
`
`
`treatment did not express cell surface FcyRl or bind to the
`
`
`
`Fig. 2. The binding was directly proportional to the amount of
`
`
`
`immunoconjugate. After incubation with rlFN-'Y for 18 h,
`
`
`immunoconjugate used to stain the cells. This was manifested
`Lys"-BN X 22 immunocon­
`
`>90% of neutrophils bound to the
`
`
`
`both by an increase in the absolute percentage of cells stained
`jugate with a median MFI of 50.
`
`
`
`positively and by an augmentation of the MFI of the entire cell
`Lys3-BN x 22 lmmunoconjugate Redirected Tumor
`was increased
`
`population. As the amount of immunoconjugate
`
`Cell lysis. The ability of the Lys"-BN x 22 immunoconju­
`
`
`from 5-50 µg/ml, the percentage of positive cells generally
`
`
`
`gatc to elicit monocyte-mediated and neutrophil-mediated tumor
`
`
`
`increased from 50-85%, and the MFI generally increased from
`of � 1 Cr release
`cell lysis was tested in a series
`
`assays. The
`
`
`
`< 100 to > 200. In general, the immunoconjugate prepared be­
`
`
`
`optimal concentration of the immunoconjugate-induced cyto-
`
`
`had a higher tween the whole antibody of mAb 22 and Lys·'-BN
`
`0 10 20 30 •o 50
`
`10 20 30 •o 50
`
`SHP-77 coll• �
`
`ii: ::I
`
`100
`
`! 80
`!: 80
`� '°
`.. • 20 ...
`
`SHP-77 cell•
`
`300
`
`200 �
`
`100
`0 0--
`
`Downloaded from
`
`clincancerres.aacrjournals.org
`
`on August 31, 2015. © 1995 American Association for Cancer
`Research.
`
`4 of 11
`
`BI Exhibit 1016
`
`

`

`Clinical Cancer Research 429
`
`OMS273
`•
`II H345
`0 H69
`8 SHP-77
`
`120
`
`�
`..... 100
`Cl
`c
`.,, 80
`c
`:0
`-0
`• Cl •
`c •
`c.>
`... • 11.
`
`40
`
`20
`
`60
`
`0
`
`24h
`1h
`0
`4h
`
`Fig. J Stability of the immunoconjugate to four SCCL cell lines was analyzed by now cytome1ry. The percentage of binding is defined as
`binding
`
`100% at the initiation of incubation al 37°C (lime 0). Resuhs from three individual are presented as mean ::!: SD. There is no significant
`experiments
`change in immunoconjugale binding to SCCL cells in the first 4 h.
`
`Table I Binding of the immunoconjugale 10 monocyies before and after rlFN--y incubation"
`Before rlFN--y incubation
`
`After rlFN--y incubation
`
`% Positive :t SD
`MFI :!: SD
`MFI :t SD
`% Positive :t SD
`52.0 :t 26.0 (2)
`210.1 ::!: 46.7 (2)
`mAb 22
`85.2 ;!: 16.7
`83.5 :t 2.2
`124.6 ::!: 25.5 (2)
`70.7 :t 15.6
`84.7 :t 17.2
`29.9 :t 14.6 (2)
`F(ab')2
`188.0 ::!: 85.I (2)
`86.7 :t 3.4
`92.J :t 7.60
`mAb 22 x Lys3-BN
`25.7 ;!: 0.10 (2)
`85.8 ::!: 1 6.3
`72.3 ::!: 10.7
`I 19.6 :!: 20.9 (2)
`26.7 ;!: 8.57 (2)
`F(ab'h X Lys3-BN
`"Flow cy1ometric analysis of the binding of mAb 22, F(ab')2 of mAb 22, and two immunoconjuga1es lo human monocytes before and after 18-h
`rIFN--y incubation. MFI increased significantly after rIFN--y incubation.
`
`150
`
`?:
`:? 100
`;:; •
`0 50
`.,,.
`
`-
`Elopo-11
`--<>- Elopo"'-1
`12
`
`The results of monocyte-mediated cytotoxicity of SCCL
`
`
`
`
`cells are presented in Fig. 5. Since the effector
`cells came from
`
`
`
`
`different donors in each experiment, the cytotoxic potency var­
`
`
`
`ied greatly among donors. For each SCCL cell line, >80% of
`
`
`cells could be lysed by monocytes from some donors. Tumor
`
`
`cell lysis was primarily dependent on E:T ratio in each individ­
`at an E:T
`
`
`
`ual donor. The greatest lysis was consistently achieved
`
`
`ratio of 100:1. Pilot experiments using 4-h assays showed mod­
`4 h to
`
`
`
`erate cytotoxicity. Increasing the incubation time from
`
`
`
`
`18 h resulted in a higher percentage of tumor cell lysis. We also
`
`observed that monocytes from the same donor could have dif­
`
`
`
`
`
`
`ferent cytotoxic potency against different SCCL cell lines. One
`
`
`SCCL cell line could be belier killed by monocytes from one
`
`
`
`donor, but resistent to monocytes from another donor.
`Monocyte-mediated
`
`
`cytotoxicity was studied under differ­
`toxicity was determined by a dose-response assay using acti­
`
`
`
`
`
`
`ent assay conditions in each experiment. The result of a typical
`
`vated monocytes as effector cells and SHP-77 cells as target
`experiment
`
`is shown in Fig. 6. Incubating target cells with
`
`
`cells. Results of two such assays are presented in Fig. 4. Max­
`
`
`
`
`
`activated monocytes in the absence of immunoconjugate re­
`
`
`sulted in 15-60% of tumor cell lysis, depending on the donor.
`imum tumor cell lysis was observed over a wide range of
`mAb 22, did not further
`
`
`
`
`concentrations. Since the flow cytometry analysis indicated that
`
`
`The addition of parental antibody,
`
`
`
`
`increase the lysis. In the presence of immunoconjugate, tumor
`
`
`the maximum binding of the immunoconjugate to SCCL cells
`
`occurred at concentrations above 25 µg/ml, we used 25 µg/ml
`
`
`cell lysis was increased to 50-95%. This increase is significant
`
`
`
`for each individual donor (P < 0.05) and is consistent in each
`
`
`immunoconjugate in most cytotoxicity assays.
`
`.I
`
`100
`
`1000
`
`10
`10000 100000
`lmmunocon)ugate concentration (ng/ml)
`Fig. 4 Effect of different concentrations of immunoconjugate Lys·'-BN
`x 22 on cyto1oxicity of �'Cr-labeled SHP-77 cells by human monocytes
`al an E:T ratio of 100:1. Rcsuhs of two experiments arc shown. Tumor
`cell lysis by monocytes achieved with 25 µ.g/ml immunoconjugale was
`defined as I 00%.
`
`Downloaded from
`
`clincancerres.aacrjournals.org
`
`on August 31, 2015. © 1995 American Association for Cancer
`Research.
`
`5 of 11
`
`BI Exhibit 1016
`
`

`

`430 lmmunoconjugate of Lys-'-BN and mAb 22 Can Induce Lysis of SCCL
`
`DMS273 cells
`
`H69 cells
`
`100
`
`80
`
`l
`•
`•
`J!: 60
`
`40
`
`•
`..
`0
`E 20
`:>
`...
`
`0
`
`-- Donor 1'
`- Donor2
`-- Donor 3#
`
`12:1 25 :1 50:1 100:1
`
`too
`
`l
`80
`•
`•
`J!: 60
`•
`..
`..
`0
`E 20
`:>
`...
`
`40
`
`0
`
`-- Donor 1'
`- Oonor2
`-- Donor 3#
`- Donor 4#
`
`12:1 25:1 50:1 1 OO:t
`
`H345 cells
`
`SHP-77 cells
`
`l
`•
`
`100
`
`80
`
`60
`
`40
`
`'i
`..
`0
`E 20
`:>
`...
`
`-- Donor 1
`- Oonor2
`-- Oonor3#
`
`• � .?:'
`
`T
`r
`A--··-·y
`
`./
`
`-- Donor 1·
`--- Donor2'
`-- Donor 3•
`--- Donor4'
`--·--!::---
`Donor S•
`
`100
`
`l
`!! 80
`.
`.?:' 60
`•
`.. 40
`.. 0
`E 20
`:>
`...
`
`0
`
`12: 1 25:1 50:1 100:1
`
`12:1 25:1 50:1 100:1
`Ellector
`to tuget cell
`ratio
`Fig. S Cytotoxicity of �'Cr-labeled SCCL cells by human monocytes. Assays were performed al E:T ratios ranging from 12: I 10 100: I. Monocytes
`were separated from differenl donors in each individual assay. A 4-h assay was performed in pilot experiments. Since the tumor cell lysis was greatly
`
`enhanced by the monocy1es of 1he same donor after 18-h incubation. subsequent assays were done using an 18-h incubation. There is no difference
`in cytotoxici1y between the immunoconjugate made from Lys'-BN and mAb 22 or its F(ab'h fragmen1s. •. 4-h assay; It, immunoconjugate made
`between L-BN and F(ab'h of mAb 22.
`
`individual assay. The immunoconjugate-induced SCCL cell ly­
`
`
`bated with target cells. Again, the immunoconjugate-induced
`(P < 0.05) blocked by adding exces­
`sis could be significantly
`
`
`tumor cell lysis could be significantly blocked by adding excess
`
`sive amounts of either unconjugated mAb 22 or Lys3-BN in
`
`amounts of either mAb 22 or Lys3-BN in each individual assay
`each assay. The presence of 1 mg/ml human IgG did not
`(data not shown).
`
`
`interfere with immunoconjugate-induced tumor cell lysis (data
`
`not shown). We also compared the potency of tumor cell lysis
`DISCUSSION
`
`
`between rlFN--y stimulated and nonstimulated monocytes from
`Recently, a variety of bispecific antibodies and immuno­
`
`
`
`the same donor. The result is shown in Fig. 7. Monocytes after
`in vitro as
`
`toxins have been shown to confer antitumor effects
`
`
`
`rlFN--y stimulation had a significantly higher cytotoxic potency
`
`
`well as in vivo (26-29). The major advantage of these bispecific
`
`
`
`(P < 0.05) in different E:T ratios. There was no difference in the
`
`molecules lies in their ability to target therapy selectively to
`
`
`observed cytotoxicity from immunoconjugate made with whole
`
`
`tumor cells while simultaneously triggering cytotoxicity by en­
`
`mAb 22 versus its F(ab'h fragments.
`
`
`dogenous effector cells. In this report we have demonstrated that
`
`The results of neutrophil-mediated SCCL cell lysis are
`immunoconjugates
`made of Lys3-BN and mAb 22, or F(ab'h
`shown in Fig. 8. Cytotoxic potency varied greatly among do­
`
`
`
`fragments of mAb22, can effectively induce lysis of SCCL cells
`
`nors. More than 75% of SHP-77 cells were lysed by neutrophils
`
`in the presence of rIFN--y activated monocytes and neutrophils
`
`
`from two different donors. The percentage of tumor cell lysis
`in vitro. The Lys3-BN x 22 immunoconjugate
`binds to more
`
`
`was greatly increased by neutrophils from one donor when the
`
`than 85% of SCCL cells at concentrations above 25 µg/ml,
`from 4 h to 18 h. It
`
`incubation time of the assay was prolonged
`
`while cytotoxicity is noted at 1000-fold lower concentrations.
`
`
`appears that H69 cells are more resistent to cytotoxicity from
`
`
`This suggests that tumor cell lysis does not require saturation of
`
`
`neutrophils. The percentage of cell Jysis ranged from 17-45% at
`
`an E:T ratio of 100: I. Since neutrophils do not express Fc-yRI on
`
`
`BN/GRP receptors. The cytotoxicity observed at low immuno­
`
`
`
`
`conjugate concentrations is particularly important since the BN/
`
`their cell surface without rIFN--y stimulation, nonstimulated
`
`
`neutrophils had much less cytotoxicity when they were incu-
`
`GRP receptors are occupied to a certain extent by endogenous
`
`Downloaded from
`
`clincancerres.aacrjournals.org
`
`on August 31, 2015. © 1995 American Association for Cancer
`Research.
`
`6 of 11
`
`BI Exhibit 1016
`
`

`

`
`
`
`
`431 Clinical Cancer Research
`
`DMS27l c•ll•
`
`HH ceJI•
`
`!100
`. .. :;. ..
`
`= ..
`
`0
`
`,.
`� ..
`
`z ...
`
`. II)
`
`80
`
`v ..
`•
`E ,.
`�
`
`H34S ct Ill
`
`SHP·77 c•U•
`
`_100
`!
`. II)
`
`� ..
`.
`v ..
`•
`� ,.
`..
`
`GI C2
`
`CJ C4
`
`cs
`
`GI G2 T Q
`
`C4 cs
`
`Fig. 6 Comparison of cytotoxic effects by monocytes under different conditions. They were incorporated
`
`
`
`
`
`
`in each cytotoxicity a5say with 25 µglml
`
`
`
`
`One typical experiment is presented here. Cl. incubation of target cells with immunoconjugate. without
`and an E:T ratio of IOO:l.
`immunoconjugate
`
`
`
`
`C2. incubation of target cells with rlFN-y activated monocytes. without
`itself has no cytotoxic effects.
`monocytes. The immunoconjugate
`cells with rlFN-y activated
`
`
`immunoconjugatc. T, incubation of target
`
`
`monocytes in the presence of the immunoconjugate. The cytotoxicity of
`
`
`
`
`monocytcs was significantly higher (P < 0.05) in the presence of immunoconjugate. CJ, incubation of target cells with rlFN-y activated monocytes
`
`and mAb 22, but no immunoc onjugatc. C4. incubation of target cells with rlFN-y activated monocytcs.
`and I 0-fold excess of mAb
`immunoconjugatc.
`
`of target cells of excessive amounts of mAb 22. C5, incubation tly blocked (P < 0.05) by the addition
`
`
`with rlFN--r
`
`22. The cytotoxicity was significan
`
`(P < 0.05) by the addition of
`
`excess of Lys�-BN. The cytotoxicity was signifi cantly blocked
`
`
`
`activated monocytes. immunoconjugate. and 10-fold
`
`excessive amounts of L-BN.
`
`BN/GRP, and antibody delivery to tumor sites in vivo may be
`
`22 had a similar binding profile to both target and effector cells.
`
`
`
`
`limited. Compared with the approach of using BN/GRP antag­
`
`and equal capacity to induce ADCC against SCCL cells. Since
`
`
`
`
`onists, the immunotherapy described in this article offers several
`
`the F(ab'h fragments arc smaller than the whole molecule of
`mAb, they should more readily reach tumor sites i11 1-fro.
`
`
`advantages. First, some antagonists against BN/GRP receptors
`Fi­
`
`
`have been shown to partially inhibit the growth of SCCL cells in
`
`nally, it has been shown that multiple growth factors are in­
`
`vitro and less effectively in vfro (30). They could be quickly
`
`volved in the growth stimulation of SCCL cells (16). The
`
`degraded by serum proteases before they can reach tumor sites
`
`blockage of one pathway, such as BN/GRP receptors by their
`
`antagonists, is not likely to inhibit the growth of SCCL cells i11
`
`
`(16). The conjugation of these small peptides to a large immu­
`
`
`
`vivo given the redundancy of growth factor receptors. However,
`
`
`noglobulin should greatly retard their degradation. Although the
`Lys3-BN used in our