Leukemia Research 25 (2001) 1115–1125
`
`www.elsevier.com/locate/leukres
`
`Direct effect of bispecific anti-CD33×anti-CD64 antibody on
`proliferation and signaling in myeloid cells
`
`Larisa Balaian a, Edward D. Ball b,*
`a Department of Medicine and Cancer Center, Uni6ersity of California, San Diego School of Medicine, La Jolla, CA, USA
`b Blood and Marrow Transplantation Di6ision, Uni6ersity of California, 9500 Gilman Dri6e, San Diego, La Jolla, CA 92093-0960, USA
`
`Received 22 January 2001; accepted 27 April 2001
`
`Abstract
`
`Bispecific anti-CD33×anti-CD64 antibody (BsAb) directly inhibited proliferation and colony formation of human acute
`myeloid leukemia cell lines, without affecting the function of normal monocytes. Addition of BsAb to normal monocytes induced
`tyrosine phosphorylation of Cbl and Vav, association of these molecules with CD33, and downstream signaling. In leukemia cells
`that were insensitive to BsAb treatment, Vav and Cbl were constitutively phosphorylated and, therefore, constitutively associated
`with CD33. Direct growth inhibition is an additional mechanism by which BsAb may be useful in the therapy of AML. © 2001
`Elsevier Science Ltd. All rights reserved.
`
`Keywords: Myeloid leukemia; Fc receptors; Monocytes; Bispecific antibodies
`
`1. Introduction
`
`Leukemia cells from patients with acute myeloid
`leukemia (AML) commonly express certain myeloid-
`specific antigens such as CD33 and CD64 [1–3]. Con-
`siderable attention has been focused on targeting
`these antigens with monoclonal antibodies (mAb) for
`therapeutic effects [4]. Bispecific antibodies (BsAb),
`targeting leukemia-associated antigens simultaneously
`with targeting activating antigens on cytotoxic effec-
`tor cells, offer a novel and powerful approach to
`anti-leukemia therapy [5]. We have previously re-
`ported the production of BsAb CD33×anti-CD64
`[6,7] and have demonstrated its in-vivo and in-vitro
`activities in mediating lysis of AML cells by cytotoxic
`myeloid effector cells [6,8,9]. However, the direct ef-
`fect of the BsAb on AML cells was not previously
`examined. Since both CD33 and CD64 are present on
`AML cells, the effects induced by this BsAb after
`binding to either CD33 alone, CD64 alone, or both
`antigens, could be determined by both the molecular
`
`* Corresponding author. Tel.: +1-858-657-7053; fax: +1-858-657-
`6837.
`E-mail address: tball@ucsd.edu (E.D. Ball).
`
`signaling cascades of either one or both of the anti-
`gens.
`CD64 is a member of the Fcg receptor family of
`cell-surface proteins that play an important role in
`both host defense and autoimmune disorders. Fcg re-
`ceptor signaling can lead to downstream events such
`as phagocytosis, ADCC, enhancement of antigen pre-
`sentation and the release of
`intracellular cytokines
`and reactive oxygen intermediates [10]. Ligation of
`monocyte-associated CD64 results in rapid tyrosine
`phosphorylation
`on
`several
`signal
`transduction
`molecules. These molecules include the immunorecep-
`tor tyrosine-based activation motif (ITAM) of the g
`chain [11,12], and several non-receptor PTKs: the src-
`family members Hck, Lyn [13] and Fgr [14]; Syk [14–
`16]; the protooncogenes c-Cbl
`[17,18] and Vav [14];
`and phospholipase C (PLC) [19]. However, the leuko-
`cyte immunoglobulin-like receptors (LIR) -1 and -2,
`which are also expressed on monocytes, are bound to
`the tyrosine phosphatase SHP-1. Co-ligation of either
`LIR with CD64 inhibits tyrosine phosphorylation of
`the associated Fc receptor g chain and Syk molecules,
`as well as intracellular calcium mobilization [20]
`CD33 is a cell-surface antigen specifically expressed
`on myeloid cells including myeloid leukemia cells. It
`
`0145-2126/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
`PII: S0145-2126(01)00084-4
`
`1 of 11
`
`BI Exhibit 1015
`
`

`

`2.2. Materials
`
`All chemicals were purchased from Sigma Chemical
`Co. (St. Louis, MO).
`
`2.3. Cells and cells lines
`
`Human AML cell lines HL-60 and U-937 were ob-
`tained from ATCC. Cell-line DER was created earlier
`in our laboratory. The NB4 cell line was a gift from
`Dr. Lanotte (Hospital St. Louis of Paris, France). All
`leukemia cells were cultured under standard condi-
`tions in RPMI-1640 containing 10% FCS.
`Monocytes were isolated via adherence from pe-
`ripheral blood mononuclear cells collected from nor-
`mal donors as described elsewhere [24].
`
`L. Balaian, E.D. Ball /Leukemia Research 25 (2001) 1115–1125
`1116
`is the smallest member of the siglec (s6 ialic acid-bind-
`ing Ig-related lectins)
`family, and is a 67 kDa
`transmembrane glycoprotein expressed by the earliest
`myeloid progenitors. It continues to be present during
`myelomonocytic differentiation on monocytes and
`neutrophils. However,
`the expression of CD33 on
`neutrophils is much lower than on monocytes. It has
`been shown recently that
`the cytoplasmic tail of
`CD33 contains two tyrosine-based motifs, both poten-
`tial ITIMs. MAb directed against CD33 are used in
`the diagnosis of leukemia and for therapeutic target-
`ing and purging in AML. However, despite its clinical
`importance, little is known of its role in myeloid cells,
`except that it may be involved in sialic acid-depen-
`dent cell interactions. While the expression of CD33
`on myeloid cells was first described in the early 1980s
`[1], only very recently have data regarding the possi-
`ble function(s) of CD33 been reported [21–23]. Tay-
`lor
`demonstrated
`that CD33
`becomes
`tyrosine
`phosphorylated in myeloid cells after both pervana-
`date treatment and CD33 receptor cross-linking, re-
`sulting in recruitment of tyrosine phosphatases SHP-1
`and SHP-2 [22]. Phosphorylation of CD33 was spe-
`cifically inhibited by an Src family tyrosine kinase
`inhibitor. Moreover,
`the first cytoplasmic tyrosine
`residue (LXY340XXL) of CD33 is dominant in SHP-
`1/SHP-2 binding, and mutation of this same tyrosine
`enhances CD33-mediated adhesion. Co-ligation of
`CD33 and CD64 on monocytes resulted in decreased
`phosphorylation and duration of phosphorylation of
`various proteins, suggesting that recruitment of SHP-
`1 to CD33 resulted in a general decrease in kinase
`activity [21]. This raises the possibility that, similar to
`CD22 in B cells and p75/AIMR1 in NK cells, CD33
`might act as an inhibitory receptor in myeloid cell
`signaling leading to inhibition of cell growth and
`apoptosis.
`Here, we report our findings on the direct effects of
`anti-CD33×anti-CD64 BsAb on cells from AML cell
`lines and normal monocytes from healthy donors.
`
`2.4. Proliferation assay
`
`AML cells were cultured in triplicate wells in 96-
`well round-bottomed plates, with or without various
`amounts of BsAb in a final volume of 200 ml of
`RPMI-10 at 37 °C in humidified 5% CO2 in air, at
`105 cells per well. The cultures were pulsed with 1
`mCi [3H] thymidine per well during the last 3 h of the
`3 h culture. The amount of 3H-thymidine incorpo-
`rated into acid-precipitable DNA was assessed via liq-
`uid scintillation counting.
`
`2.5. Colony formation assay
`
`AML cells were cultured in 0.3% agarose as de-
`scribed elsewhere [25]. Briefly, 104 cells were placed in
`35 mm tissue culture dishes containing tissue culture
`media RPMI-1640 supplemented with 10% FBS and
`agarose at final concentration of 0.3%. After a 10–14
`day incubation in the presence of various concentra-
`tion of antibodies at 37 °C, the number of colonies
`was estimated under the light microscope. Clusters
`containing less than 40 cells were excluded from anal-
`ysis.
`
`2. Materials and methods
`
`2.1. Antibodies
`
`The mAb anti-CD33, anti-CD64 and anti-CD33×
`anti-CD64 BsAb were obtained from Medarex, Inc.
`(Annandale, NJ). The anti-Syk, anti-Vav, anti-Cbl,
`anti-CD33 rabbit or goat polyclonal antibodies were
`purchased from Santa Cruz Biotechnology Inc. (Santa
`Cruz, CA). A horse-radish-peroxidase-conjugated anti-
`phosphotyrosine mAb, 4G10, was obtained from Up-
`state Biotechnology Inc.
`(Lake Placid, NY). The
`fluorescein-labeled anti-CD14, anti-CD33 and anti-
`CD64 mAb was obtained from B&D (San Jose, CA).
`
`2.6. FACS analysis for CD64 and CD33
`
`Normal or leukemic cells were washed and then
`suspended in staining media (SM), containing RPMI-
`1640, 3% FCS, 0.01% NaN3 and 1 mg/ml of propid-
`ium iodide
`(Calbiochem, La
`Jolla, CA),
`plus
`saturating amounts of FITC-conjugated anti-CD33,
`anti-CD64 mAbs, or an isotype-matched control mAb
`of irrelevant specificity. After 30 min at 4 °C, the
`cells were washed in SM twice and then analyzed on
`a FACScan (Becton Dickinson). Dead cells and de-
`bris were excluded from analysis by characteristic for-
`ward and side scatter profiles and propidium iodide
`staining.
`
`2 of 11
`
`BI Exhibit 1015
`
`

`

`L. Balaian, E.D. Ball /Leukemia Research 25 (2001) 1115–1125
`
`1117
`
`2.7. Phagocytosis assay
`
`Monocytes were cultured 5 days in the presence of 50
`ng/ml of GM-CSF (Immunex Inc., Seattle, WA). After
`two washes with PBS, cells were stained with FITC-la-
`beled anti-CD14 (green fluorescence) mAb for 30 min
`at room temperature then washed twice with PBS and
`transferred to a polypropylene 96-well plate at a con-
`centration of 105/well. E. coli particles (Molecular
`Probes Inc., Eugene, OR) pre-stained with tetramethyl-
`rhodamine (red fluorescence) were added to monocytes
`at a 10:1 ratio. Various concentrations of BsAb or
`control antibody were added in a 10 ml volume. After
`incubation at 37 °C for 1 or 3 h, each well was flushed
`with medium and the cell suspension transferred to a
`fresh tube. After fixation with 2% paraformaldehyde,
`the cells were deposited on slides by cytocentrifugation
`and then analyzed by fluorescent microscopy. The per-
`centage of double-stained cells was calculated.
`
`2.8. Cell acti6ation and immunoprecipitation
`
`Monocytes from normal donors or AML cell lines
`(5–6× 106) were activated by 10 mg/ml BsAb anti-
`CD33×anti-CD64 or unconjugated anti-CD33 and
`anti-CD64 mAb for 20 min at room temperature, fol-
`lowed by the addition of polyclonal anti-mouse IgG at
`20 mg/ml for different time points (from 1 min to 1 h).
`The reaction was stopped by adding ice-cold PBS. After
`three washes in ice-cold PBS, cells were lysed in lysis
`buffer, containing 1% (v/v) Triton X-100, 0.15 M NaCl,
`50 mM Tris–HCl [pH=7.2], 0,1% SDS, 1 mM Na-or-
`thovanadate, 1 mM phenylmethylsulfonyl fluoride, 1%
`[v/v] EDTA, 1% [v/v] Aprotinin, and 0.03 mM Leu-
`peptin. After 30 min on ice, the nuclear debris was
`removed by centrifugation for 15 min at 13,000×g.
`Lysates were equalized with respect to the amount of
`protein, as assessed by O.D. at 280 nM. Specific im-
`munoprecipitation was performed for 2 h toovernight
`in the presence of 30% (vol./vol.) ‘Protein A/G’, conju-
`gated with agarose (Santa Cruz Biotech., CA). Im-
`munoprecipitates were washed three times in lysis
`buffer and then suspended in an equal volume of
`Laemli sample buffer for SDS-PAGE.
`
`2.9. SDS-PAGE and Western blotting
`
`idase-conjugated matching secondary antibodies. The
`filters then were washed in PBS-T, incubated with the
`enhanced
`chemoluminescence
`detection
`reagents
`(Pierce, Rockford, IL), and exposed to X-ray film (Fuji
`Film, Fischer, Tustin, CA).
`
`3. Results
`
`3.1. Inhibition of AML cell proliferation and colony
`formation upon BsAb ligation
`
`We examined AML cells for their response to recep-
`ligation measuring 3H-thymidine incorporation.
`tor
`HL-60, U-937, NB4 and DER cells were incubated with
`the anti-CD33×anti-CD64 BsAb, or unconjugated sin-
`gle mAb anti-CD33 or anti-CD64 or both mAB. We
`found that BsAb significantly (PB0.001) inhibited pro-
`liferation of HL-60 and U-937 cells in a dose-dependant
`manner, but failed to decrease proliferation of DER
`cells (Fig. 1). In NB4 cells we also observed significant
`(about 60%) inhibition of proliferation initiated by
`BsAb (data not shown). Interestingly, the level of inhi-
`bition was maximal at BsAb concentrations of 0.01–0.1
`mg/ml. These data suggest that the effect was not caused
`by antibody toxicity. Meanwhile,
`incubating HL-60
`cells with the unconjugated anti-CD33 or anti-CD64
`mAb did not decrease proliferation of leukemia cells. In
`contrast, U-937 cells were responsive to the combina-
`tion of both unconjugated mAb. However, DER cells
`were not responsive to either BsAb or unconjugated
`mAb.
`The ability of BSAB to inhibit the leukemia cells
`growth was confirmed by a similar dose-dependent
`reduction of colony formation in both HL-60 (PB
`0.001) and U-937 (PB0.001) cells (Fig. 2), but not in
`DER cells (data not shown). However, we found that
`incubation of leukemia cells with both unconjugated
`mAb did not affect colony formation. Importantly, in
`both proliferation and colony formation assays, the
`maximum BSAB activity was observed at a concentra-
`tion of 0.01–0.1 mg/ml and therefore was probably not
`caused by antibody toxicity.
`
`3.2. Effect of IFN-k on sensiti6ity of AML cells to
`antibody treatment
`
`Total cell lysates or immune-precipitates were added
`to separate wells (8 mg/well) of SDS-PAGE (7.5–10%
`acrylamide) gel, electrophoretically size-separated under
`reducing conditions, and then transferred onto nitrocel-
`lulose for immunoblotting. The filters were first incu-
`bated for 1 h in 5% non-fat dry milk in PBS-T (PBS
`plus 0.01% Tween 20), and then incubated with the
`primary antibody for 2 h. After washing in PBS-T, the
`filters were incubated for 1 h inhorse-radish perox-
`
`We investigated the levels of CD33 and CD64 expres-
`sion on three leukemia cell lines: HL-60, U-937 and
`DER. CD33 was expressed at 95, 97 and 96%, respec-
`tively (data not shown). CD64 was expressed at 79, 95
`and 30%, respectively (data not shown). Since our
`experiments did not detect differences between AML
`cells in the level of CD33 expression, but indicated that
`the level of CD64 expression on DER cells was signifi-
`cantly lower, we addressed the question of whether the
`
`3 of 11
`
`BI Exhibit 1015
`
`

`

`1118
`
`L. Balaian, E.D. Ball /Leukemia Research 25 (2001) 1115–1125
`
`increased level of CD64 on DER cells would alter the
`cells’ response to antibody ligation. After treatment of
`U-937 (responsive) and DER (unresponsive) cells with
`200 mg/ml of INF-g at 37 °C for 24 h, the level of
`CD64 expression was 98 and 64%, respectively. We
`performed proliferation assays in the presence or ab-
`sence of 0.1 mg/ml of various antibodies. We found that
`IFN-g was able to increase significantly (PB0.001) the
`sensitivity of DER cells (which were initially unrespon-
`sive)
`to antibody ligation (Fig. 3). U-937 cells re-
`sponded to BsAb and the combination of anti-CD33
`and anti-CD64 mAb even without INF-g treatment.
`
`However, the effects of BsAb and single mAb were
`increased after treatment of the cells with IFN-g.
`
`3.3. Stimulation of normal monocyte phagocytosis by
`BsAb
`
`Since CD33 and CD64 antigens are expressed by
`both AML cells and normal monocytes, we examined
`whether or not BsAb ligation could affect the phago-
`cytic function of monocytes. We incubated monocytes
`from five normal donors in the presence of 50 ng/ml of
`GM-CSF for 5 days and then analyzed their ability to
`
`Fig. 1. Effect of BsAb on proliferation of leukemia cells. Leukemia cells were cultured for 4 h intriplicate for each condition in tissue culture
`media supplemented with 0.5% FBS in the presence of various concentrations of antibodies. 3H-thymidine was added at the initiation of culture.
`The basal level of cell proliferation was considered as 100%, and the percentage of basal stimulation was calculated for each condition. Each
`histogram represents results of five independent experiments. Error bars indicate the standard error about the mean (S.E.M.).
`
`Fig. 2. Effect of BsAb on leukemia cells colony formation. Leukemia cells were cultured in 0.3% agarose in triplicate for each condition in the
`presence of various concentrations of antibodies. After 14 days of culture, the number of colonies was enumerated in each dish. The basal level
`of colony formation was considered as 100%, and the percentage of stimulation was calculated for each condition. Each histogram represents
`results of four independent experiments. Error bars indicate S.E.M.
`
`4 of 11
`
`BI Exhibit 1015
`
`

`

`L. Balaian, E.D. Ball /Leukemia Research 25 (2001) 1115–1125
`
`1119
`
`Fig. 3. Effect of INF-g on sensitivity of AML cells to antibody treatment. U-937 and DER cells were cultured in the presence or absence of 200
`u/ml of INF-g for 24 h under standard conditions. After two washings proliferation assays were performed in the presence of various antibodies
`as described in Fig. 1. Each bar represents the percentage of inhibition, calculated as the difference between the basal level of 3H uptake (media
`alone) and the results of proliferation for each condition. Error bars indicate S.E.M.
`
`integrate pre-stained E. coli particles in the presence of
`BSAB, unconjugated anti-CD33 and anti-CD64 anti-
`bodies or the combination of both mAb. We found
`that, in contrast to anti-CD33 mAb, anti-CD64 mAb as
`well as BsAb significantly increased (PB0.001) the
`phagocytic activity of monocytes (Fig. 4). This effect
`was present when we treated cells with a combination
`of two unconjugated antibodies, but the level of phago-
`cytosis was lower. Time-course experiments revealed
`that in all five samples tested, the level of BsAb-medi-
`ated phagocytosis after 3 h oftreatment was higher
`compared to that at 1 h. Meanwhile, at both time
`points, the effect was dose-dependent.
`
`3.4. Induction of tyrosine phosphorylation by BsAb
`ligation
`
`In order to determine the mechanisms of the func-
`tional diversity between normal monocytes and AML
`cells, we analyzed the signaling cascade leading to
`leukemia
`cell
`growth
`inhibition.
`Since
`tyrosine
`phoshorylation is an early and obligatory event in cell
`activation, we examined protein phosphorylation in-
`duced by BsAb ligation by immunoblot analysis using
`an antibody to phosphotyrosine. After the addition of
`BsAb to normal monocytes or AML cells, in normal
`monocytes, we observed rapid phosphorylation of
`protein(s) that were of approximately 100–120 kDa
`(data not shown). In contrast, similar treatment of
`AML cells did not induce phosphorylation of addi-
`tional proteins. However, the majority of proteins in
`AML cells were constitutively phosphorylated (data not
`shown). Time-course experiments showed that induced
`protein tyrosine phosphorylation could be detected as
`early as 1 min after BsAb ligation. Protein phosphory-
`lation reached a peak level at approximately 2 min, but
`still could be detected for at least 1 h after stimulation
`
`(data not shown). This indicates that BsAb ligation
`induces rapid tyrosine phosphorylation of cytosolic
`proteins independent of other stimulatory signals and
`serves as a stimulatory factor that is sufficient to induce
`activation of phagocytosis in normal monocytes. How-
`
`Fig. 4. Effect of BsAb on phagocytosis of normal monocytes. Normal
`monocytes were stained with CD14-FITC and mixed with pre-stained
`tetramethylrhodamine E. coli particles in the presence of various
`concentrations of antibodies. The numbers of double stained (phago-
`cytic) cells were estimated for each condition under fluorescent micro-
`scope. Each bar indicates the percentage of phagocytic cells for each
`condition. Error bars indicate S.E.M. Bars marked with an asterisk
`have mean values that are significantly higher than that of the control
`condition (PB0.005, Student’s t-test).
`
`5 of 11
`
`BI Exhibit 1015
`
`

`

`1120
`
`L. Balaian, E.D. Ball /Leukemia Research 25 (2001) 1115–1125
`
`Fig. 5. Effect of BsAb on Cbl phosphorylation. Normal monocytes (6A) or leukemia cell line DER cells (6B) were treated for 20 min with 10
`mg/ml of BsAb, anti-CD33 mAb, anti-CD64 mAb or a combination of both single antibodies followed by the addition of 20 mg/ml of polyclonal
`mouse IgG for 2 min. Cell lysates were prepared and then used to generate immune precipitates with normal rabbit IgG or anti-Cbl antibody.
`Top row: immunoblot with anti-phosphotyrosine antibody; bottom row: immunoblot with anti-Cbl antibody.
`
`ever, BsAb ligation did not induce phosphorylation in
`AML cells and appears to inhibit proliferation and
`colony formation due to detected differences between
`normal monocytes and AML cells in their basal level
`of tyrosine phosphorylation.
`
`3.5. Identification of proteins that are
`tyrosine-phosphorylated following BsAb ligation
`
`Having established that BsAb ligation induced
`phosphorylation of 120 kDa cytosolic protein(s), we
`analyzed whether Cbl and Vav are involved in its
`signaling cascade. We performed immunoprecipitation
`of these proteins in resting and antibody-treated cell
`
`immunoblot
`followed by phosphotyrosine
`lysates
`analysis. We found that the addition of BsAb, as well
`as unconjugated anti-CD33 and anti-CD64 mAb or
`both mAb, did induce rapid tyrosine phosphorylation
`of Cbl and Vav in monocytes of normal donors
`(Figs. 5A and 6). In contrast, Cbl and Vav were con-
`stitutively phosphorylated in resting leukemic cells
`prior to antigen ligation (Figs. 5B and 6). Further-
`more, the addition of BsAb or unconjugated anti-
`CD33 or anti-CD64 mAb to AML cells did not
`increase the level of tyrosine phosphorylation of Cbl
`or Vav. These data demonstrate the difference be-
`tween normal monocytes and AML cells in BsAb sig-
`naling.
`
`6 of 11
`
`BI Exhibit 1015
`
`

`

`L. Balaian, E.D. Ball /Leukemia Research 25 (2001) 1115–1125
`
`1121
`
`3.6. Identification of proteins that are tyrosine-phosphory-
`lated following per6anadate treatment
`
`BsAb ligation leads to activation of both the CD33 and
`the CD64 signaling pathways. The effects of CD64
`ligation have been well described, but CD33 signaling
`pathways are less well studied. Therefore, we examined
`the proteins that are involved in CD33 signaling. Normal
`monocytes and AML cells were treated with pervana-
`date, which is known to induce phosphorylation by
`blocking the activity of tyrosine phosphatases, and then
`performed co-immunoprecipitation of CD33 with several
`protein kinases and their substrates.
`Co-immunoprecipitation of CD33 and Syk revealed
`no differences in the phosphorylation pattern between
`normal monocytes and AML cells (Fig. 7). In both cases,
`Syk formed a complex with CD33 only after it became
`phosphorylatred upon pervanadate treatment.
`Similar experiments with proto-oncogenes Vav and
`
`Cbl revealed not only differences between normal mono-
`cytes and AML cells but also the heterogeneicity of AML
`cells. In two AML cells lines, HL-60 (Figs. 8 and 9) and
`U-937 (data not shown), as well as in normal monocytes
`(data not shown), Vav and Cbl formed complexes with
`CD33 only when they were phosphorylated after per-
`vanadate treatment. However, in DER cells (Figs. 8 and
`9), these proteins were constitutively phosphorylated
`and, therefore, associated with CD33. These data demon-
`strate that: (a) CD33 uses Syk, Vav and Cbl for its
`signaling and (b) CD33 associates with their phosphory-
`lated forms. In some AML cells, Cbl and Vav are
`constitutively phoshorylated, a situation that may alter
`CD33 signaling.
`
`4. Discussion
`
`An understanding of steps involved in the anti-CD-
`
`Fig. 6. Effect of BsAb on Vav phosphorylation. Normal monocytes or leukemia cell line DER cells were incubated with BsAb (+ lines) as
`described above or left untreated (− lines). Cell lysates were prepared and then used to generate immune precipitates with normal rabbit IgG or
`anti-Vav antibody. Top row: immunoblot with anti-phosphotyrosine antibody; bottom row: immunoblot with anti-Vav antibody.
`
`Fig. 7. Co-immunoprecipitation of CD33 and Syk in pervanadate-treated cells. Normal monocytes and leukemia cell line HL-60 were incubated
`with 1 mM pervanadate for 10 min at 37 °C (+ lines) or left untreated (− lines). Cell lysates were prepared and then used to generate immune
`precipitates with rabbit IgG or anti-Syk antibody. Top row (A): immunoblot with anti-CD33 antibody. Medium row (B): immunoblot with
`anti-phosphotyrosine antibody. Bottom row (C): immunoblot with anti-Syk antibody.
`
`7 of 11
`
`BI Exhibit 1015
`
`

`

`1122
`
`L. Balaian, E.D. Ball /Leukemia Research 25 (2001) 1115–1125
`
`Fig. 8. Co-immunoprecipitation of CD33 and Vav in pervanadate-treated cells. Two leukemia cell lines DER and HL-60 were treated with
`pervanadate as described in Fig. 8. Cell lysates were prepared and then used to generate immune precipitates with rabbit IgG or anti-Vav
`antibody. Top row (A): immunoblot with anti-CD33 antibody. Medium row (B): immunoblot with anti-phosphotyrosine antibody. Bottom row
`(C): immunoblot with anti-Vav antibody.
`
`Fig. 9. Co-immunoprecipitation of CD33 and Cbl in pervanadate-treated cells. Two leukemia cell lines DER and HL-60 were treated with
`pervanadate as described in Fig. 8. Cell lysates were prepared and then used to generate immune precipitates with rabbit IgG or anti-Cbl antibody.
`Top row (A):
`immunoblot with anti-CD33 mAb. Medium row (B):
`immunoblot with anti-phosphotyrosine antibody. Bottom row (C):
`immunoblot with anti-Cbl antibody.
`
`33×anti-CD64 BsAb-induced signaling cascade in
`myeloid cells is of potential clinical significance. This
`antibody has been prepared using a chemical linkage
`technique [6,7] and was active in vitro in mediating ADCC
`by monocytes [6,8,9]. In a mouse model in vivo, anti-
`CD33×anti-CD64 BsAb demonstrated the ability to
`increase the anti-tumor effects of INFg-stimulated mono-
`cytes (unpublished results). Currently, this antibody is
`under study in a Phase 1 clinical trial for patients with
`relapsed AML. Therefore, we decided to investigate the
`direct effect of the BsAb on leukemia cells to further define
`its mechanism(s) of action. We found that BsAb was able
`to decrease proliferation and colony formation of some
`AML cells. BsAb-induced inhibition of colony formation
`demonstrated that the effect of BsAb on leukemia cells
`was most likely a direct effect rather than the alternative
`possibility that reciprocal
`lysis occurred from one
`leukemia cell expressing CD64 killing an adjacent
`leukemic cell expressing CD33 through BsAb bridging.
`
`Recently, it was demonstrated that anti-CD33 antibody
`diminished the level of leukemia cell proliferation [26] and
`prevented generation of dendritic cells from both mono-
`cytes and CD34+ myeloid precursors [27]. This function
`was ascribed to the inhibitory nature of the CD33
`molecule. In our experiments, however, single anti-CD33
`or anti-CD64 mAb failed to demonstrate inhibitory
`activity in both proliferation and colony formation assays.
`After 3 h of culture, we detected an inhibitory effect only
`in the proliferative response of U-937 cells when we treated
`these cells with a combination of both single anti-CD33
`and CD64 antibodies. These discrepancies could be
`explained by the variety of reasons: source of anti-CD33
`antibody and/or time of culture. We indicated that a
`longer (48 h) treatment of AML cells with anti-CD33 mAb
`resulted in inhibition of AML cell proliferation (data not
`shown).
`Since normal monocytes express on their surface both
`CD33 and CD64 molecules [1,5], and we found that BsAb
`
`8 of 11
`
`BI Exhibit 1015
`
`

`

`L. Balaian, E.D. Ball /Leukemia Research 25 (2001) 1115–1125
`
`1123
`
`inhibited proliferation of leukemia cells, we examined
`whether it could alter the phagocytic activity of normal
`monocytes. We demonstrated that BsAb increased the
`phagocytic activity of normal monocytes against E. coli
`particles. Earlier in our laboratory, Zhong et al. (unpub-
`lished results) observed a similar effect of BsAb on
`phagocytosis of normal monocytes against tumor cells.
`Since we found that single anti-CD64 mAb was also able
`to increase phagocytosis, we concluded that this effect
`might be ascribed to the activating function of the CD64
`molecule. Our data are in concordance with earlier
`published results [10,28].
`We detected differences between various leukemia
`cells in their levels of surface CD64 expression while
`levels of CD33 expression were less variable. Impor-
`tantly, we noted that a low level of CD64 molecule
`expression was associated with the inability of leukemia
`cells to respond to antibody stimulation in DER cells.
`However, after INF-g mediated upregulation of CD64,
`these cells became responsive to BsAb and single mAb
`treatment. Taken together, these data prompted us to
`investigate BsAb-mediated signaling in myeloid cells.
`Both CD33 and CD64 molecules have been noted to
`initiate the downstream signaling cascade in myeloid
`cells resulting in activation of various classes of protein
`tyrosine kinases (PTK) and protein tyrosine phos-
`phatases (PTP). Monocyte CD64 ligation triggers cellu-
`lar events that are crucial for a variety of immune
`responses. These include phagocytosis, production of
`cytokines, release of agents that damage microorgan-
`isams or infected cells, and changes in expression of cell
`surface proteins
`involved in cell-to-cell
`interaction
`[29,30]. Signaling events triggered by CD64 cross-linking
`result in phosphorylation of tyrosine on several impor-
`tant signal transduction molecules. Tyrosine kinases of
`the Src-family [13,14] and Syk [14–16,31,32] become
`activated and associate with ITAM. Targets of these
`activated PTKs include the Fc-g R itself [11,12], enzymes
`that generate second messengers (PLC-g and PI3-kinase)
`[19,33],
`and regulators of Ras, Vav
`and Cbl
`[14,17,18,33–35].
`Little is known about CD33 signaling in myeloid cells
`except that the cytoplasmic tail of CD33 contains two
`ITIMs, and after activation, CD33 becomes tyrosine-
`phosphorylated and recruits tyrosine phosphatases SHP-
`1 and SHP-2 [21–23]. However, the balance between
`CD33 and CD64 molecules on the cell surface and/or
`their possible interaction needs to be investigated in
`more detail.
`We detected differences between normal monocytes
`and AML cells in their phosphorylation pattern and
`response to BsAb treatment. In normal cells, BsAb
`ligation induced rapid phosphorylation of protooncoge-
`nes Vav and Cbl. In contrast, in some AML cells, these
`proteins were constitutively phosphorylated, and BsAb
`cross-linking did not change their levels. Moreover, the
`
`majority of proteins in untreated AML cells were consti-
`tutively phosphorylated. This finding is important since
`PTKs regulate a variety of cellular functions, including
`cell proliferation and differentiation [36]. Enhanced or
`dysregulated PTK activity has been implicated in many
`hematopoietic and non-hematopoietic cancers [37,38].
`Excessive PTK activity can result in hematopoietic cell
`transformation; perturbation of either PTKs or PTPs
`may result in uncontrolled cell growth. Myeloid cells are
`responsive to cytokines that induce tyrosine phosphory-
`lation and can become ligand-independent when endoge-
`nous PTKs become dysregulated. Specific PTPs, through
`mutation or altered expression, may enhance PTK activ-
`ity and also cause myeloid ligand independence [39].
`In order to understand how important phosphoryla-
`tion of Vav and Cbl are in CD33 signaling in general,
`we treated normal myeloid and leukemia cells with
`pervanadate. This results in blocking of PTPs and
`subsequent phosphorylation of PTKs and their sub-
`strates,
`leading to CD33 downstream signaling. We
`examined Vav, Cbl and Syk phosphorylation since we
`found involvement of the first two in BsAb-mediated
`signaling. Additionally, Syk is a very important PTK
`often associated with Vav and Cbl.
`Our co-immunoprecepitation studies demonstrated
`that in normal monocytes, CD33 signals using Syk, Vav
`and Cbl, and associates only with their phosphorylated
`forms. In AML cells, the phosphorylation patterns are
`different from normal cells and, importantly, differ from
`one leukemia cell line to another. Although we did not
`detect any differences in the phosphorylation pattern of
`Syk, we noted major differences in the tyrosine phospho-
`rylation of Vav and Cbl between normal and some AML
`cells. Unlike in normal monocytes, these proteins were
`phosphorylated in untreated DER cells and, therefore,
`are in complex with CD33. Moreover, the level of
`tyrosine phosphorylation in these cells was not changed
`by pervanadate treatment. Constitutively phosphory-
`lated forms of Cbl and Vav may preclude leukemic cells
`from responding to CD33 signaling or alter the response.
`This notion is supported by the finding that AML cells
`are heterogeneous in their response to BsAb signaling. In
`contrast to HL-60 and U-937 cells, in DER leukemic
`cells, BsAb could not induce inhibition of proliferation
`and colony formation, unless they were treated with
`INF-g.
`Our data suggest that in normal monocytes, CD33
`recruits Syk, Cbl and Vav for its signaling. In leukemia
`cells, which,
`in contrast to normal monocytes, have
`constitutively phoshorylated forms of Cbl and Vav,
`downstream events following CD33 ligation might be
`altered. This results in an anergic, impaired response of
`AML cells to CD33 signaling. However, we cannot
`exclude the possibility that sensitivity of leukemia cells
`to BsAb treatment is associated with the level of CD64
`expression. Similar impairment due to constitutive phos-
`phorylation of certain proteins may occur in CD64
`
`9 of 11
`
`BI Exhib