`THE JOURNAL OF' IMMUNOLOGY
`Copyright«> 1982 by Tile American Association of Immunologists
`
`Voj. 1 28, No 3, March 1982
`Printed VI U S A.
`
`STUDIES ON THE ABILITY OF MONOCLONAL ANTIBODIES TO SELECTIVELY MEDIATE
`COMPLEMENT-DEPENDENT CYTOTOXICITY OF HUMAN MYELOGENOUS LEUKEMIA
`BLAST CELLS1
`
`EDWARD 0. BALL,2 JAMES M. KADUSHIN, BERNICE SCHACTER, AND MICHAEL W. FANGER
`Dartmouth Medical School. Hanover. NH 03755 and the Department of Pathology, University Hospitals
`From the Department of Microbiology,
`
`
`
`of Cleveland, Case Western Reserve University, Cleveland. OH 44 1 06
`
`finding was that the antigens determined by these antibodies
`were found to some extent on all cells tested, both normal and
`leukemic. Even so, we have determined that some of these
`
`
`antibodies can permit complem€nt-mediated lysis of leukemic
`
`myeloblasts but not normal cells. This paper describes in detail
`
`the patterns of cytotoxicity of 20 different monoclonal anti
`bodies to cells from patients with leukemia, normal individuals,
`and human leukemia cell lines. Several of these monoclonal
`antibodies
`
`appear exclusively cytotoxic to myeloid leukemia
`cells, whereas others mediate lysis of myeloid leukemia cells
`
`
`as well as subpopulations of normal cells. The potential clinical
`
`
`utility of this panel of monoclonal antibodies is considered.
`
`MATERIALS ANO METHOOS
`
`A panel of monoclonal antibodies that bind to leukemic
`blast cells from patients with acute myelocytic leukemla
`and chronic myelocytic leukemia in blast crisis was stud
`ied for their ability to mediate complement-dependent
`lysis of a variety of cell populations from patients with
`leukemia, normal blood cells, and human leukemia cell
`llnes. Several of these monoclonal antibodies were selec
`tively cytotoxic to myeloid leukemla cells (AML-1-99,
`AML-1-211, AML-2-30, CML-75, CML-115, and CML-150).
`Although they were all capable of binding to normal cell
`populations, none of these hybridomas were cytotoxic to
`normal cells. Three of these antibodies (AML-1-211, CML-
`75, and CML-150) were cytotoxic to some leukemia cell
`samples only after dilution of the hybridoma supernatant,
`i.e., they showed a prozone. Binding of these three anti
`bodies, as well as another, AML-1-201, as determined in
`a radiolmmunoassay, also showed a prozone. Other mon
`oclonal antibodies are described (AML-2-23 and AML-2-
`9) that mediate complement-dependent cytotoxicity to
`myeloid leukemia cells as well as selected normal cell
`types (monocytes and lymphocytes, respectively). The
`potential clinical utility of these monoclonal antibodies is
`considered in the context of recently encountered prob
`lems in the use of monoclonal antibodies to mediate
`leukemia cell lysis in vivo.
`
`Details of the production of these hybridomas and partial
`Hybridomas.
`characterizations of the monoclonal antibodies are reported elsewhere. 3
`Briefly, monoclonal antibodies were developed by immunizing BALB/c mice
`with human myelogenous leukemia cells and fusing spleen lymphocytes
`with cells from the P3-XAg63 murine myeloma cell line by using polyethylene
`glycol (m.w. 1000, J. T. Baker Chemical Co., Phillipsburg, NJ) according to
`the method of Kohler and Milstein (2). A panel of 20 different monoclonal
`antibodies were included in this study. All were lgM immunoglobullns except
`AML-2-23, which is an lgG of the .,., subclass. The hybridomas were grown
`in Dulbecco's minimal essential medium (DMEMl supplemented with 20%
`fetal calf serum CFCS). 2 mM L-glutamine, 5 mM HEPes• buffer. and
`gentamicln, 50 µg/ml. All medium components were obtained from K. C.
`Biologicals, Kansas City. MO, except genta micin (Schering Corp., Kenil
`worth, NJ).
`Cells. Leukemic cells were obtained from patients with AML. a cute
`lymphocytic leukemia (ALL), chronic lymphocylic leukemia (CLL). and CML
`Although monoclonal antibodies that recognize tumor-spe
`
`In blast crisis. and were separated from blood by Ficoll-HypaQue gradient
`cific antigens would be extremely useful in the diagnosis and
`centrifugation (3). Normal blood cells were obtained from the mononuclear
`treatment of human malignant diseases, it is as yet unclear
`cell populations of normal laboratory personnel. Monocytes were separated
`from the mononuclear cell fraction by adherence to pla stic. T and B
`
`whether truly tumor-specific antigens exist on human tumor
`lymphocyte separations were accomplished by filtering lymphocyte prepa
`cells (1 ). Nonetheless, it seems likely that quantitative differ
`
`rations through nylon wool columns as described by Oaynilovs et al. (4). T
`ences in antigen expression will be found on tumor cells
`cells were collected in the initial filtrate while the adherent B cells were
`
`compared to normal cells, and such differences may be useful
`detached by vigorous agitation and rewashing of the column with medium.
`
`in demonstrating subgroups of selected tumor types or in
`The human leukemic cell lines CCRF-CEM, KG-1 a. HL-60. and U937 were
`
`also studied. CCRF-CEM, a lymphoblastoid
`cell line, derived from a patient
`mediating lysis of tumor vs normal cells.
`with T-ALL (obtained from the American Type Culture Collectlon). was
`maintained in culture in RPMI 1 640 containing 20% FCS (5). The KG-1 a
`In a previous paper,3 we reported the preparation and prop.
`
`
`
`erties of a panel of monoclonal antibodies that bind to myelo
`cell line, a subline of the KG-1 cell line derived from a patient with AML (6).
`wa s grown in Alpha medium (Flow Laboratories, Inc., Rockville, MO) con
`
`
`blasts from patients with acute myelocytic leukemia (AML)4 or
`taining 20% FCS and g entamicin . This line was a gift from Or. P. Koeffler,
`
`
`chronic myelocytic leukemia CCML) in blast crisis. An important
`Division of Hematology-Oncology, UCLA. Los Angeles. CA. The HL-60 cell
`line was derived from a patient with acute promyelocytlc leukemla (7). This
`line was a gilt of Dr. Robert Gallo, Labora tory of Tumor Immunology,
`National Institutes of Health and was maintained in RPMI supplemented with
`1 0% FCS. The U937 cell line was derived from a patient with hi stlocyti c
`lymphoma and has some functional and morphologic chara cteristics of
`macrophages (8). This line was a gilt of Dr. Paul Guyre, Department of
`Physiology, Dartmouth Medical School, and was maintained in RPMI 1640
`supplemented with 10% FCS.
`Binding. Supernatants from hybridomas were serially diluted In DMEM
`containing 20% FCS and were added to wells coated with glutaraldehyde
`fixed leukemic cells, normal lymphocytes, or normal monocytes and Incu
`bated for 2 hr a1 37°C according to the method of Kennetrt (9). After
`washng ott unboun d antibody with phosphate-buffered saline CPBS) and
`
`'This work was supported by Grants CA31918 and Al 19053 awarded by the
`National Cancer Institute and the Institute of Allergy and Infectious Oiseases.
`OHHS. respectivel y.
`•Address correspondence to: Edward 0. Ball. M.O .. Department of Mlcrobl·
`ology, Dartmouth Medical School, Hanover. NH 03755.
`3 Ball, E. D .. Fanger. M. w.: Monoclonal antibodies reactive with human
`myeloid leukemia cells. Clln. Exp. lmmunol. In press.
`• Abbrevlalions used in this paper: AML. acute myelocytlc leukemia; CML,
`chronic myelocytic leukemia: ALL. a cute lymphocytlc leukemia: CLL, chronic
`leukemia; CML·BC, chroni c rnyelocytlc leukemia In blast crisis;
`lymphocytic
`HEPES. N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid.
`
`1476
`
`1 of 6
`
`BI Exhibit 1005
`
`
`
`1982]
`
`MONOCLONAL ANTIBODIES CYTOTOXIC TO MYELOID LEUKEMIA CELLS
`
`1477
`
`1201-anti-mouse ic-antibody was added and incubated for 2 hr at 37°C. After
`washing with PBS the wells were counted in a gamma counter (Beckman
`4000).
`Cytotoxicity. Cytotoxicity was tested by dye exclusion in microtiter trays
`according to the method of Amos (10). Cells from patients with leukemia
`and mormal peripheral blood were suspended in Verona! Buffer (Oxoid, K.
`C. Biologicals. St. Louis. MO) or calcium and magnesium free Hanks'
`balanced salt solution (M. A. Bioproducts. Walkersville. MD). Hybridoma
`supernatants were diluted in hybridoma culture medium and incubated with
`cells for 30 min at 22°C before a wash step and the addition of rabbit serum
`(diluted 1 :4.2 in Verona! buffer) as a source of complement (Low-Tox H,
`Cedarlane Laboratories. Ltd., Hornby, Ontario. Canada). Incubation with
`complement proceeded for 1 hr at 22°C. Trypan blue exclusion was used
`to estimate cytolysis. Controls included anti·.B• microglobulin and anti-la
`antisera (Accurate Chemical and Scientific Corp., Westbury, NY). Negative
`controls included human AB serum. the myeloma parent supernatant. and
`heat-inactivated complement. Each experiment was done at least in dupli
`cate and in many cases. in quadruplicate. The data reported are the means
`of replicate assays in which deviation from the mean was fess then 10%.
`
`RESULTS
`
`Sixteen of the 20 monoclonal antibodies studied gave maxi
`mal binding to both leukemic and normal cells in undiluted or
`weakly diluted supernatant. The titer. defined as the dilution
`that gave 50% of maximal binding, ranged from 1 :2 to 1: 1024
`for this group of monoclonal antibodies. The binding of several
`representative monoclonal antibodies to AML cells is seen in
`Figure 1A. Either a plateau or linear decline in binding was
`seen with decreasing concentration of antibody.
`In contrast. four of the 20 monoclonal antibodies studied,
`AML-1-201, AML-1-211, CML-75. and CML-150, showed an
`increase in binding to AML cells as the hybridoma supernatant
`was diluted. with maximal binding observed between 1 :8 and
`1 /32 dilution (Fig. 18).
`Binding of the majority of these monoclonal antibodies to
`morrnal lymphocytes paralleled that to AML cells (Fig. 2). A
`notable exception was AML-2-23, which showed binding to
`lymphocytes only slightly above background while binding
`significantly to AML cells over a broad range of dilutions. The
`binding of monoclonal antibodies AML-2-23, AML-1-211. AML-
`2-9, AML-1-99, and CML-75 to rnonocytes is shown in Fig. 3.
`With the exception of monoclonal antibody AML-2-23, the
`binding patterns of these monoclonal antibodies to monocytes
`was similar to their binding to lymphocytes and AML cells.
`AML-2-23 showed similar binding to both monocytes and AML
`cells.
`Several monoclonal antibodies selectively mediated comple
`ment-dependent lysis of leukemic myetoblasts (AML-1-211,
`AML-1-99. CML-75, CML-115. and CML-150). Three of these
`clones. AML-1-211. CML-75, and CML-150, mediated lysis of
`some leukemia cell samples only upon dilution of the super
`natant, whereas other samples were lysed by both diluted and
`neat supernatant. Monoclonal antibody AML-1-211 permitted
`lysis of four different AML cell samples only when diluted
`between 1 :8 and 1: 128 (see Figs. 4 and 5 for representative
`cytotoxicity curves). Monoclonal antibody CML-75 permitted
`lysis of one of these AML cell samples at dilutions between 1:
`16 and 1 :51 2 (Figs. 4 and 5). The dilutions of these two
`monoclonal antibodies
`that mediated cytotoxicity corre
`sponded to those that yielded maximal binding. CML-150,
`another monoclonal antibody that had a prozone in binding to
`AM L cells, was able to fyse one sample of leukemic myeloblasts
`in undiluted supernatant. However, in contrast to monoclonal
`antibodies AML-1-211 and CML-75, CML-150 was still not
`cytotoxic to the AML samples shown in Figures 4 and 5.
`Subsequently, other AML samples were lysed by hybridomas
`CML-75, AML-1-211, and CML-150 only at dilutions of super
`natant between 1: 1 0 and 1: 100. ft is notable that each of these
`
`1:4
`
`1:256
`1:64
`1: 16
`Dilution of Anti body
`
`1:1024
`
`t•
`
`1:256
`1:&4
`1:16
`Dilution of Antibody
`
`1:102•
`
`Figure 1. Binding of serially diluted hybridoma supernatant to 1 o• AML cells.
`A. Hybridomas represented are AML-2-1 7 (0). AML-2-23 (6), AML-1-22 (0), and
`AML-1-104 <•>. B. Hybridomas represented are AML-1-201 (6). AML-1-211
`<•>. CML-75 (0), and CML-150 (0). The binding of the P3-X63Ag8 parent
`myeloma supernatant (.A.) is also represented In both t A and 18. Supernatants
`were diluted with DMEM containing 20% calf serum and incubated for 2 hr at
`37°C on glutaraldehyde-tixed leukemia cells. Bound monoclonal antibody was
`detected with a rabbit-anti-mouse kappa chain antibody labeled with '251. Each
`sample was counted for 2 min in a gamma counter.
`
`three monoclonal antibodies (AML-1-211, CML-75. and CML-
`1 50) mediated lysis of some leukemic cells only upon dilution
`while other samples were lysed by neat supernatant. Moreover.
`a prozone was seen with some leukemic cells with only one or
`two of these monoclonal antibodies while another antibody
`permitted lysis in neat supernatant, yet the pattern was oppo
`site for other cell samples. A summary of the collected cytotox
`icity data for these monoclonal antibodies is shown in Table I.
`Neither AML-1-211, AML-1-99, CML-75, CML-115, or
`CML 1 50 mediated lysis of any normal T and B lymphocytes or
`monocytes studied at any dilution of antibody (Figs. 6 and 7).
`Monoclonal antibodies AML-2-9 and AML-2-23 were cyto
`toxic to a number of leukemia cell samples as well as selected
`normal cell populations. AML-2-23 mediated lysis of eight of
`13 myeloblast samples as well as all normal monocytes studied
`(Fig_ 6), but was not cytotoxic to any lymphocytic leukemia cell
`samples or normal lymphocytes (Fig. 7). In contrast, AML-2-9
`mediated lysis of four of 12 myeloid leukemias as well as some
`lymphocytic leukemias and normal lymphocyte samples.
`One monoclonal antibody, AML-1-201, mediated lysis of
`nearly every cell population studied. both normal and leukemic.
`The exception was the cell line Oaudi, which does not express
`class I HLA antigens or ,82-microglobulin on the cell surface
`(11 • 1 2). This antibody appeared specific to ,82-microglobulin
`as determined by an enzyme-linked immunosorbent assay with
`purified human ,82-microglobulin (a gift of Dr. George Bernier,
`Department of Medicine, Dartmouth Medical School). Finally,
`a number of monoclonal antibodies did not lyse any cell pop
`ulation studied despite the demonstration that significant bind
`ing occurred and the fact they were lgM immunoglobulins.
`Several monoclonal antibodies (AML-1-2 2, AML-1-99, AML-
`1-116, AML-1-201. AML-2-9, and AML-2-23) mediated lysis
`
`2 of 6
`
`BI Exhibit 1005
`
`
`
`1478
`
`16
`
`.... 12
`Q
`..
`�8
`8
`
`12
`
`.., Q 8
`>< ..,;
`c:
`g 4
`u
`
`c
`
`1:4
`
`BALL, KADUSHIN, SCHACTER. AND FANGER
`
`{VOL. 128
`
`I: 16
`1:64
`1:256
`01 lution of An Ii body
`
`I:�
`
`><
`..,;
`c:
`
`g 4 u
`
`12
`
`r:>
`Qs
`><
`..,; c
`g 4
`u
`
`1:4
`
`1:16
`1:64 1:25 6
`Dilution of Antibody
`
`1:1024
`
`Figvre 2. Binding of serially diluted hy-
`bridoma supernatant to 1 o• normal lympho-
`cytes and 1 o• AML cells. Figure 2A shows lhe
`binding of AML-1-211 to lymphocytes (0) and
`AML cells <•>. The P3-X63Ag8 myeloma cell
`line supernatant is also represented (A) in
`Figures 2A. 8. C. D, and £. Figure 28 shows
`the binding of AML-2-23 to lymphocyles (0)
`and AML cells<•>. Figure 2C shOws the bind·
`ing of AML-1-201 to lymphocytes (0) and AML
`cells<•>. Figure 2D shows the binding of CML·
`150 to lymphocytes (0) and AML cells <•>.
`Figure 2E shows the binding of CML-75 10
`lymphocytes (0) and AML cells <•>. Samples
`were counted for 2 min.
`
`I: 16
`1:64 1:256 1:1024
`Dilution of Antibody
`
`0
`
`1:4
`
`1:16
`1:64 1 : 2 56
`Dilution of Antibody
`
`1:1024
`
`16
`
`IZ
`
`.,
`'Q
`
`E
`
`1:4
`
`1:256
`1:64
`I: 16
`Oilolion of Antibody
`
`I; I024
`
`of one or more of the CCRF-CEM, KG-1 a, HL-60, or U937 cell
`lines (Table I). Only AML-1-99 and AML-1-201 were capable
`of lysing all cell lines while AML-2-9 and AML-1-22 permitted
`lysis of CCRF-CEM and AML-1-11 6 permitted lysis of KG-1 a
`cells. Both AML-2-9 and AML-2-23 permitted lysis of the HL-
`60 cell line.
`
`DISCUSSION
`
`We presented data previously that described the binding of
`a panel of monoclonal antibodies to leukemic and normal cell
`populations. None of these hybridomas were specific for any
`leukemia cell type, yet significant quantitative differences in
`antigen expression on leukemic cells compared to normal cells
`were shown with several of these antibodies. In spite of a lack
`of absolute binding specificity, however, several antibodies are
`capable of selective complement-dependent cytotoxicity of
`leukemic, and, in some cases, myeloid leukemia cells. The
`present report documents these observations and demon
`strates that some monoclonal antibodies can. under appropri
`ate conditions, express specific cytotoxic activities that were
`not evident in the initial screenings.
`Monoclonal antibodies AML-1-99. AML-1-211, AML-2-30,
`CML-75, CML-115, and CML-150 all demonstrated cytotoxic
`ity to myeloid leukemias exclusively while sparing normal cell
`populations. AML-1-211 mediated lysis of seven of 13 myeloid
`leukemia samples. This antibody showed a prozone effect with
`
`1:256 1:1024
`1:4 1:16 1:64
`Oilulioo of Antibody
`Figvre 3. Binding of serially diluted hybridoma supernatant to 1 o• normal
`monocytes. Hybridomas represenled are AML-1-99 a:J>. AML-t-211 <•>. AML-2·
`9 (&-&). AML-2-23 (6.). and CML-75 (0). The P3-X63Ag8 myeloma cell line
`supernatant is also represented (A). Each sample was counted tor 2 min.
`
`tour of these AML samples that coincided with the binding
`behavior of the hybridoma supernatant. Lysis of normal lym
`phocytes or monocytes was not observed with this antibody at
`any dilution. Monoclonal antibody CML-75 was able to cause
`complement-dependent lysis of five of 1 3 myeloid leukemia
`samples but was not cytotoxic to normal lymphocytes and
`monocytes despite the demonstration of large amounts of CML-
`75 antigen on these cells. Antibodies AML-1-99, CML-115,
`and CML-150 were not cytotoxic to normal cells yet AML-1-99
`
`3 of 6
`
`BI Exhibit 1005
`
`
`
`1982]
`
`MONOCLONAL ANTIBODIES CYTOTOXIC TO MYELOID LEUKEMIA CELLS
`
`1479
`
`100
`90
`80
`10
`
`�
`"' 60
`·;;;
`,... ...J �o
`"ii u
`40
`30
`20
`10
`
`1:512
`
`t2048
`
`1:2
`
`1:8
`
`1:126
`1:32
`of Anlibody
`Dilution
`Figure 4. �pendent cytotoxJcity mediated by monoclonal antibodies to
`cells from a patient with AML Equal volumes of hybrldoma supernatant and cells
`at 2 x 10•1m1 were incubated for 30 min at 22•c In mlcrotiter wells. Rabbit C
`was added and incubation continued for 60 min. Cytolysis was estimated by
`Trypan blue exclusion. Hybridomas represented are AML-1-211 (9). AML-2-23
`(6). and CML-75 (0). The P3-X63Ag8 myetoma supernatant is also represented
`(A). All assays were done in duplicate.
`
`divalent or trivalent binding. More easily dissociable complexes
`
`
`would be formed that would result in some loss of antibody by
`washing during the assay. As less antibody is presented to the
`
`cell surface more cross-linking of receptors might occur. lead
`ing to a more stable complex. Although this behavior was noted
`for four of the 20 antibodies studied, all of which were lgM
`
`immunoglobulins, 15 other lgM monoclonal antibodies showed
`either a plateau or a nearly linear decrease in binding upon
`
`
`
`dilution. Differences in antigen density or distribution and/or
`
`that some of the monoclonal antibodies were of a different lgM
`
`subclass may also contribute to these results. The monoclonal
`antibodies that showed a prozone were among those with the
`
`highest binding to a variety of cells under standardized condi
`tions, whereas none of the monoclonal antibodies recognizing
`less densely expressed epitopes had a prozone in the binding
`
`
`assay. The prozone effect noted in the cytotoxicity assay may
`
`have the same explanation because a wash step is included in
`the assay.
`These studies illustrate that for selected monoclonal anti
`
`bodies, binding and/or effector functions can either be under
`
`estimated or undetectable if screenings are limited to undiluted
`
`
`for both the hybridoma supernatants. This has implications
`initial selection of hybridomas and attempts to demonstrate
`
`
`specificity of hybridomas as it is clear that cytotoxicity can be
`demonstrated for some monoclonal antibodies only at critical
`
`
`antibody concentrations. Furthermore, this behavior needs to
`be considered in trials of serotherapy of human leukemia.
`
`90
`80
`70
`�
`.'!! 60
`"' � 50
`� 40
`30
`
`20-----
`
`1:51 2
`
`1:2048
`
`Cell Ltne•
`KH-18 CORF HL-60 U937
`++ ++ ++ ++
`++ ++ ++ ++
`
`++ ++
`++
`
`Monoclonal
`Anllbody
`
`AML-1-99
`AML-1-201
`AML-1-211
`AML-2·9
`AML-2·23
`AML-2·30
`CML-75
`CML-1 t5
`CML-150
`
`90
`ao
`� 70
`160
`- &O
`� �o
`
`30
`20
`
`1:2
`
`1:8
`
`1:542
`
`1:2048
`
`1:2
`
`1:3
`
`1:32
`1:128
`of Antibody
`Dilution
`Figure 5. C-depenClent cytotoxlclty mediated by monoclonal antibodies to
`cells from a patient with AML (a different patient from that shown in Fig. 6).
`Hybrklomas represented aro AML-1-211 c•>. AML-2-23 (6), and CML-75 (0).
`The P3-X63Ag8 myeloma supernatant Is also represented (&). All assays were
`Clone In duplicate.
`
`TABLE I
`Cytotoxic1ty of leukemls cells and cell lines med/Bled by monoc/onsl sntibodies
`in the presence of complement
`Laul<emia Cell Type•
`AML CML-BC ALL
`CLL
`1/7
`0/4
`0/3
`0/4
`8/8
`3/3
`4/4
`3/4
`6/8
`1/5
`0/3 0/4
`2/4
`2/8
`1/3
`1/4
`5/8
`0/4
`0/3
`3/5
`0/4
`1/5
`0/2
`0/4
`0/3
`1/5
`0/4
`4/8
`2/8
`0/4
`0/3
`0/3
`2/8
`1/4
`0/3
`0/3
`• Cells were obtained from patients with a variety of leukemias. The numerator
`refers lo lhe number of different individuals whose cells showed positive reactions
`and the denominator refers to the number of individuals whose cells were tested.
`Positive Is defined as >50% lysis of leukemia cells by hybfldoma supernatant.
`and CML-115 each permitted lysis of one of the AML samples
`• + + refers to >50% lysis ot the cell line mediated by monoclonal antibody.
`studied and CML-150 permitted lysis of three AML samples.
`No score indicates that no lysis was observed.
`
`Monoclonal antibodies AML-2-9 and AML-2-23 were cyto
`toxic to some myeloid leukemias as well as some normal cell
`types. AML-2-9 permitted lysis of some samples of normal
`lymphocytes but not of normal monocytes. AML-2-23 was
`
`highly active in killing monocytes and six of eight AML and
`three of five CML-blast crisis cell samples but was not cytotoxic
`
`to lymphoid cells. The specificity of this antibody is similar to
`et at.
`that of monoclonal antibodies reported by linker-Israeli
`(13) and Ugolini et al. (14) that also were capable of mediating
`lysis of monocytes (13) and some myelomonocytic leukemia
`cells. Todd et al. (15) reported the production of a monoclonal
`antibody, Mo2, that binds to normal monocytes and some
`
`leukemic myeloblasts. This antibody does not bind to the HL-
`60 cell line, which suggests that AML-2-23 and Mo2 bind to
`
`
`different antigenic determinants because AML-2-23 both binds
`1:121
`l:Jl
`
`to and permits complement-dependent lysis of this cell line.
`Dilution d Antibody
`
`The explanation for the prozone effect in the binding of some
`Figure 6. C-dependent cytotoxicity mediated by monoclonal antibodies to
`
`of these lgM monoclonal antibodies is not clear. One possibility
`normal human rnonocytes. Hybridomas represented are AML-1-201 <D>. AML-1-
`is that when excess antibody is present, binding may be
`211 <•>. AML-2-23 (6), and CML-75 (0). The P3-X63Ag8 myeloma supernatant
`Is also represented (A). All assays were done in duplicate.
`
`
`relatively more monovalent and thus inherently less stable than
`
`4 of 6
`
`BI Exhibit 1005
`
`
`
`1480
`
`BALL, KADUSHIN, SCHACTER. AND FANGER
`
`(VOL. 128
`
`100-----.-----..
`90
`80
`70 �
`. .ff 60
`� ...J 50
`� 40
`30
`20
`10
`
`1:2
`
`1:128
`1:8
`1:32
`Dilution of Antibody
`Figure 7. C-<tependent cytotoxicity mediated by monoclonal antibodies to
`normal human lymphocytes. Identical results were obtained with unfractionated
`peripheral blood lymphocytes. T lymphocytes, and B lymphocytes. Hybridomas
`represented are AML-1-201 �. AML-1-211 <•>. AML-2-23 (6), and CML-75
`(0). The P3-X63Ag8 myeloma supernatant is also represented (&). All assays
`were done in duplicate.
`
`1:5 12
`
`I:�
`
`Optimal conditions for leukemia cell lysis may vary with the
`monoclonal antibody used and the leukemia cell population.
`There are now several human leukemia cell lines available
`for in vitro studies of leukemia cell biology (16). It is of interest
`that few of the monoclonal antibodies from the panel reported
`here mediated tysis of the myeloid leukemia cell tines, KG-1 a
`or Hl-60. Only one of the hybridomas that showed selective
`complement-dependent cytotoxicity for fresh leukemia sam
`ples from patients, AML-1-99, was cytotoxic to both of these
`cell lines despite evidence that significant binding occurred
`with all hybridomas but AML-2-23. Antibody AML-1-99, how
`ever, mediated lysis of the KG-1a, HL-60, U937, and CCRF
`CEM cell lines, suggesting that the AML-1-99 antigen is ex
`pressed similarly on these cell lines. Because these cell lines
`are of diverse lineage, their susceptibility to lysis by AML-1-99
`may possibly be due to binding to an antigen expressed on
`rapidly proliferating cells, a "division antigen." Monoclonal
`antibody AML-2-23, which mediated tysis of 8 of 13 myeloid
`leukemia cell populations, also mediated lysis of the HL-60 cell
`line but not of another myeloid cell line, KG-1 a. It is of interest
`that AML-2-23 did not permit lysis of the U937 cell line. These
`cells seem to require activation by medium conditioned by
`mixed lymphocyte cultures to assume some of the functional
`characteristics of monocytes (17). The failure of AML-2-23 to
`bind to and permit complement-dependent lysis of this cell line
`in its unstimulated state also shows that at least one normal
`surface marker of normal monocytes is absent. Studies with
`stimulated U937 cells to determine if the AML-2-23 antigen is
`expressed on stimulated cells are planned.
`In contrast, the expression of the AML-2-23 antigen on cells
`of the HL-60 cell line shows that these cells express an antigen
`characteristic of mature monocytes. We have also found that
`cells of the HL-60 cell tine express an antigen characteristic of
`potymorphonuclear leukocytes that is recognized by a mono
`clonal antibody, PMN-29, developed in our laboratory (unpub
`lished observation). The expression of two different antigens
`on these cells in the native state that correlate with specific
`states of differentiation Is interesting in light of the demon
`strated ability of these cells to differentiate into either mature
`granulocytes or monocytes with appropriate chemical inducers
`(18). The present report suggests that certain markers of
`differentiated cells are present before induction of differentia
`tion with chemical mediators. Studies to determine the quanti-
`
`tative changes. if any, of these specific markers on cells
`stimulated to differentiate are in progress.
`Although the monoclonal antibodies reported here did not
`detect leukemia-specific antigens, there are several reasons to
`consider their utility in leukemia treatment. Some experimental
`evidence exists that monoclonal antibodies directed toward
`normal cell surface antigens can be used to effectively treat
`animal tumors that bear that antigen without necessarily caus
`ing untoward effects. The studies of Bernstein et al. (19) using
`a monoclonal anti-Thy-1 . 1 antibody to treat a Thy-1 . 1-positive
`murine leukemia in vivo have shown efficacy in the elimination
`of tumor metastases while demonstrating no serious toxicity as
`a result of normal T lymphocyte binding and lysis. The use of
`monoclonal antibodies directed toward antigens that are ex
`pressed more, quantitatively, on leukemic cells may be equally
`efficacious and yet not seriously deleterious to normal host
`cells. It has also been shown that under appropriate conditions
`subpopulations of tumor cells can develop that do not express
`certain surface antigens (20). Thus, monoclonal antibody ther
`apy directed toward a given single antigen could result in
`selection of a subpopulation of tumor cells. Finally, modulation
`of cell surface antigens can occur as a result of monoclonal
`antibody binding (18).
`It seems likely that successful treatment of human tumors
`may require the use. in combination, of panels of monoclonal
`antibodies such as AML-1-99. AML-1-211, AML-2-9, AML-2-
`23, AML-2-30, CML-75, CML-115, and CML-150. In this man
`ner, the problem already encountered in t1 ials of serotherapy
`in humans, that of antigen modulation (21-23), may be circum
`vented, as multiple determinants are used as targets for anti
`body-mediated leukemia cell lysis. Studies utilizing the panel
`of monoclonal antibodies described in this report are being
`conducted to evaluate tumor cell escape from antibody-medi
`ated lysis and its prevention.
`Several of these monoclonal antibodies have immediate di
`agnostic value based on their ability to selectively lyse leukemic
`myeloblasts but not leukemic lymphoblasts. This panel of mon
`oclonal antibodies may also be useful in defining subsets of
`myetoid leukemia cells based on variable binding to and tysis
`of cells from individual patients. Such data can be correlated
`with cellular morphology and histochemical staining. It may be
`possible to show characteristic "fingerprints" of certain sub
`types of leukemia defined by binding to panels of monoclonal
`antibodies that could aid in both diagnosis and possibly in
`determining optimal therapy.
`
`Acknowledgments. We wish to thank Ors. Hillard Lazarus
`and Roger Herzig for providing some of the leukemia cells used
`in these studies.
`
`REFERENCES
`1. Old, L. J. 1981. Cancer Immunology. The search for specificity. G. H. A.
`Clowes Memorial Lecture. Cancer Res. 41 :361.
`2. Kohler. G .. and C. Milstein. 1975. Continuous cultures of fused cells se
`creting antibody of predefined specificity. Nature 256:495.
`3. Boyum. A. 1976. Isolation of lymphocytes. granulocytes. and macrophages.
`Scand. J. lmmunot. 5:9.
`4. Daynilovs. J., G. Ayoub, and P. I. Terasakr. 1980. B-lymphocyte isolation by
`thrombin-nyton. Histocompatibllity testing rePOrt of the 8th International
`Histocompatibility Workshop, 287.
`5. Foley, G. E .. H. Lazarus. S. Farber, B. G. Uzman. B. A. Boone, and R. E.
`McCarthy. 1965. Continuous culture of human lymphoblasts from peripheral
`blood of a child with acute leukemia. Cancer 18:522.
`6. Koeffler. H. P .. R. Billing, A. J. Lusis, R. Sparkes. and D. w. Golde. 1980.
`An undifferentiated variant derived from the human acute myelogenous
`leukemia cell line (KG-1). Blood 56:265.
`7. Colllus. S. J .. R. C. Gallo, and R. E. Gallagher. 1977. Continuous growth
`and differentiation of human myeloid leukaemlc cells in suspension culture.
`Nature 270:347.
`
`5 of 6
`
`BI Exhibit 1005
`
`
`
`1982]
`
`MONOCLONAL ANTIBODIES CYTOTOXIC TO MYELOID LEUKEMIA CELLS
`
`1481
`
`8. Sundstrom, C .. and K. Nilsson. t 976. Establishment and characterization of
`a human histiocytic lymphoma cell line (U-937). Int. J. Cancer 17:565.
`9. Kennett, R. H. 1960. Enzyme-linked antibody assay with cells attached to
`polyvinyl chloride plates. in Monoclonal Antibodies. Edited by Kennett. R.
`H .. T. J. McKearn. and K. B. Bechtol. Plenum Press. New York. P. 376.
`10. Amos. 0. B. Cytotoxicity testing. rn Manual of Tissue Typing Techniques,
`Department of Health. Education, and Welfare publication (National Institutes
`of Health). Government Printing Office. Washington. O.C. 80-545. P. 42.
`11. Fellous. M., F. Martchelewicz. M. Kamorn, and S. Oavsset. 1975. The use
`of a lymphoid cell line to define new B-lymphocyte specificities probably
`controlled by the MHC Region. Histocompatibility Testing. Edited by Kiss
`meyer-Nielson. Munksgaard, Copenhagen. P. 708.
`12. Evrir., P. E .. and K. Nilsson. 1974. /J2 microglobulin production in vitro by
`human hematopoietic. mesenchymal, and epithelial cells. J. lmmunol. 112:
`137.
`13. Linker-Israeli, M .. R. J. Bltling, K. A. Foon. J. H. Fitchen, and P. I. Terasaki.
`1961 . Monoclonal antibodies reactive with acute myelogenous leukemia
`cells. Fed. Proc. 40: 1118 CAbstr .)
`14. Ugolini. v .. G. Nuney, G. R. Smith, P. Stasney, and J. D. Capra. 1960. Initial
`characterization of monoclonal antibodies against human monocytes. Proc.
`Nall. Acad. Sci. 77:6764.
`15. Todd. R. F. Ill, L. M. Nadler, and S. F. Schlossman. 1981. Antigens on
`human monocytes identified by monoclonal antibodies. J. lmmunol. 126:
`1435.
`
`16. Koeffler, H.P .. and 0. W. Golde. 1980. Human myeloid leukemia cell lines:
`a review. Blood 56:344.
`17. Larrick. J. W .. D. G. Fischer, S. J. Anderson. and H. S. Koren. 1980.
`Characterization of a human macrophage-like cell line stimulated in vitro: a
`model of macrophage functions. J. lmmunol. 125:6.
`18. Fontana, J. A .. 0. A. Colbe