| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Experimental Therapeutics, Preclinical Pharmacology |
Division of Medical Oncology [C. P. S., E. G. E. d. V.], and Division of Clinical Immunology [B-J. K., L. F. M. H. d. L.], Department of Internal Medicine, University Hospital Groningen, 9700 RB Groningen, the Netherlands
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSComposition of PBSCs and...Hematopoietic Colony FormationDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSComposition of PBSCs and...Hematopoietic Colony FormationDISCUSSIONREFERENCES |
|---|
To obtain a more specific way to eliminate tumor cells from PBSCs,
treatment with monoclonal antibodies has been studied. The use of
antibodies was found to be effective and feasible in purging tumor
cells from PBSCs in hematological cancer patient studies
(7, 8, 9, 10, 11, 12)
, although the binding of a monoclonal antibody
alone does not induce tumor cell lysis. In the systemic
treatment of solid tumors, antibody-based treatment has been shown to
be beneficial in a setting of minimal residual disease
(13)
. Compared with previous disappointing antitumor
effects of immunotherapy in patients with high tumor loads
(14, 15, 16)
, adjuvant administration of monoclonal antibodies
was found to induce a survival benefit in colorectal carcinoma patients
(17)
. Immunotherapy also gained new interest because a
clinically beneficial effect was seen in disseminated breast cancer
patients treated with the humanized anti-HER2 antibody Herceptin
(18)
. However, only a minority of patients are eligible
for this type of treatment because HER2/neu expression in breast cancer
is 
25–30%. Elimination of breast cancer cells from bone marrow
after antigen-binding by means of immunobeads and immunotoxins was
shown to be effective in vitro (19
, 20)
.
To increase cytotoxicity, the use of cytokines also has been studied. IL-2 incubation of PBSCs induced tumor cell killing in vitro up to 50% (21) , and it did not negatively affect stem cell engraftment in breast cancer patients (22) . An additional effect of anti-CD3 antibody OKT3, in conjunction with IL-2, on tumor cell killing was seen in the bone marrow of hematological patients (23) . Also in the hematological setting, Kaneko et al. (24) described activation of peripheral blood mononuclear cells with IL-2 and OKT3, combined with bispecific antibodies, for ex vivo purging of leukemic cells from bone marrow. Adding bispecific antibodies clearly increased cytolysis in this study. A bispecific antibody combines affinity to both target and cytotoxic effector cells, thus allowing more efficient cell lysis than with a monoclonal antibody alone (25) .
In view of the above, it seems reasonable to further evaluate the combination of activation of T cells present in breast cancer patient PBSC harvests and a bispecific antibody for purging of carcinoma cells from PBSCs, which to our knowledge has not been described previously. In our study, we used the bispecific monoclonal antibody BIS-1, which is directed against the pancarcinoma-associated membrane antigen EGP-2 and the CD3 complex present on T cells. EGP-2, also called epithelial cellular adhesion molecule, is a 40-kDa membrane-bound glycoprotein, strongly expressed by most carcinomas and universally expressed in breast cancer specimens (reviewed in Ref. 26 ). As such, EGP-2 is a commonly used target antigen in many carcinoma-directed immunotherapeutic approaches (17 , 25 , 26) . The bispecific antibody BIS-1 creates functional cross-linking of the activated T cells and EGP-2-positive tumor cells, allowing the delivery of a tumor cell-specific lethal hit, and this T cell retargeting with BIS-1 induces specific epithelial tumor cell killing in vitro and in vivo (14 , 27) . The goal of this study was to examine in vitro activation of T cells present in PBSC harvests obtained from breast cancer patients for generation of cytotoxic effector cells and to study purging of epithelial tumor cells from PBSCs by BIS-1 retargeting of activated PBSCs.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSComposition of PBSCs and...Hematopoietic Colony FormationDISCUSSIONREFERENCES |
|---|
Activation.
PBSCs were incubated for 0, 24, and 72 h in the above-described
culture medium containing one of the following additives:
(a) PBS solution [0.14 M NaCl, 2.7
mM KCl, 6.4 mM
Na2HPO4 · 2
H2O, 1.5 mM
KH2PO4, (pH 7.4)];
(b) 100 units/ml IL-2 (aldesleukin; Chiron, Amsterdam, the
Netherlands); (c) 5% (v/v) anti-CD3, (tissue culture
supernatant containing 10 µg/ml OKT3); or (d) 100 units/ml
IL-2 and 5% (v/v) OKT3. Prior to further use, cells were washed in the
culture medium without activating additives.
Flow Cytometry.
After PBSC activation as described above, phenotyping of T cells was
assessed using: FITC- or phycoerythrin-labeled anti-CD4, anti-CD8,
anti-CD25, anti-CD69, and anti-HLA DR monoclonal antibodies (Immuno
Quality Products, Groningen, the Netherlands). PBSCs were incubated for
30 min at 4°C (5 µl of antibody for 1 x
106 cells in 100 µl of PBS containing 1% HPS),
washed once in PBS, and resuspended in 150 µl of PBS. Samples were
analyzed on a Coulter Elite Cytometer (Coulter Electronics, Hialeah,
FL) using an argon laser (488 nm) for FITC and phycoerythrin
excitation.
Target Cell Lines.
GLC1 (EGP-2-negative parental cell line) and GLC1M13 (EGP-2-positive
subclone) are small-cell lung cancer-derived cell lines
(29)
. These cell lines were cultured according to routine
procedures in RPMI 1640-based medium supplemented with 14%
heat-inactivated FCS, 2 mM glutamine, 60 µg/ml
gentamicin, 0.05 mM ß-mercaptoethanol, and 1
mM sodium pyruvate at 37°C in a humidified atmosphere
containing 5% CO2. The EGP-2-positive (GLC1M13)
and EGP-2 negative (GLC1) cell models of similar origin were used in
the 51Cr-release assay. For morphological
reasons, the EGP-2-positive breast cancer-derived cell line MCF-7 was
used in the log-depletion assay. MCF-7 was cultured according to
routine procedures in RPMI-based medium supplemented with 10% FCS at
37°C in a humidified atmosphere containing 5%
CO2.
BIS-1.
The BIS-1-producing quadroma was made in our department by fusion of
the hybridomas RIV-9 and MOC-31, producing anti-CD3 (IgG3) and
anti-EGP-2 (IgG1) antibodies, respectively, according to De Lau
et al. (30)
. Preparation and purification were
performed as described previously (14)
. Briefly, BIS-1 was
produced on large scale by means of a hollow fiber culture system
(Endotronics, Minneapolis, MN). Purification of the hybrid antibodies
(IgG3/IgG1) from parental-type antibodies, also produced by the
quadroma, was performed by protein A column chromatography. BIS-1
F(ab')2 was then produced by means of digestion
by pepsin followed by G-150 Sephadex gel filtration, and was added to a
0.9% sodium chloride solution to obtain a final concentration of 0.2
mg/ml.
51Cr-Release Assay.
51Cr-release assays were performed according to
standard procedures to assess BIS-1-redirected T-cell cytotoxicity
(14)
. All determinations were executed in triplicate.
Target cells (5 x 106; tumor cells GLC1 or
GLC1M13) were suspended in 100 µl of culture medium containing 3.7
MBq of [51Cr]sodium chromate (Amersham
Pharmacia Biotech Benelux, Roosendaal, the Netherlands) and incubated
for 1 h at 37°C in a humidified 5% CO2
atmosphere. Unbound [51Cr]sodium chromate was
removed by washing the tumor cells three times with medium.
Subsequently, aliquots of 100 µl of medium containing 0, 2.5 x
103, 2.5 x 104, or
2.5 x 105 PBSCs (effector cells) after the
above-mentioned 24 or 72 h with PBS, IL-2, OKT3, or IL-2/OKT3 were
added into a 96-well round-bottomed microtiter plate (Greiner no.
650180; Greiner, Alphen aan de Rijn, the Netherlands). To each well, 50
µl of medium containing 2.5 x 103
51Cr-labeled target tumor cells were also added,
resulting in E:T ratios of 0, 1, 10, and 100 in a final volume of 200
µl/well. Finally, 50 µl of medium containing 0.4 µg/ml BIS-1
F(ab')2 (final concentration during the assay,
0.1 µg/ml) or 50 µl of medium without BIS-1 were pipetted. The
microtiter plates were centrifuged at 46 x g for 2 min
and incubated at 37°C in 5% CO2. After a 4-h
incubation, the plates were centrifuged at 182 x g for
5 min, and 100-µl samples taken from the supernatant were counted in
a gamma counter (Wizard; EG&G/Wallac). Cell lysis was calculated from
the percentage of 51Cr released, according to the
formula: [(experimental release - spontaneous release)/(maximal
release - spontaneous release)] x 100%. Maximal release was
determined from a sample to which 100 µl of 2% Triton X-100 solution
were added instead of effector cells. Spontaneous release was
determined from a sample to which 100 µl of medium were added instead
of effector cells.
Hematopoietic Colony Formation.
Toxicity of prior T-cell activation and subsequent BIS-1 treatment on
hematopoietic stem cell recovery was studied with the CFU-GM assay
(31)
. Briefly, hematopoietic colony formation was assessed
in 1 ml of DMEM including 1.1% methyl cellulose, 20% FCS, 1%
deionized BSA, 0.001% 
-thioglycerol, and 10 ng/ml IL-3 and
granulocyte macrophage-colony stimulating factor. PBSCs (2 x 105 cells, after the prior activations as
mentioned above) were plated after a 4-h incubation with or without
BIS-1 (0.1 µg in 200 µl DMEM), with or without GLC1M13 at an E:T
ratio of 100:1. Cells were plated in 35-mm dishes and cultured for 14
days at 37°C. Hematopoietic colonies containing 
40 cells were
counted under an inverse microscope.
Log-Depletion Assay.
To assess the log-depletion of tumor cells by activated PBSC, MCF-7
tumor cells were added to PBSCs after 72 h prior activation with
OKT3 under the conditions mentioned above. MCF-7 tumor (target) cells
were added to (effector) PBSCs in an E:T ratio of 1 x
104:1 in a total volume of 6 ml of RPMI 1640
(supplemented with HPS, gentamicin, and glutamine, as mentioned above),
in the presence or absence of BIS1 (final concentration during the
assay, 0.1 µg/ml). As a control, MCF-7 tumor cells were also added to
6 ml of RPMI 1640 without PBSCs. After a 4-h incubation at 37°C in
5% CO, sedimentation of cells onto slides was performed. Cells were
stained with monoclonal antibody MOC31, directed against EGP-2, using
indirect immunoperoxidase staining with horseradish
peroxidase-conjugated rabbit antimouse as a second antibody and
3-amino-g-ethylcarbazol as a substrate. Slides were routinely
counterstained with H&E.
Statistics.
Cytotoxic cell lysis, hematopoietic colony formation, and leukocyte
phenotype were analyzed by means of Student’s t test.
P < 0.05 was considered significant.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSComposition of PBSCs and...Hematopoietic Colony FormationDISCUSSIONREFERENCES |
|---|
. Tumor cell lysis was increased by the
addition of BIS-1, after all prior activations, compared with cell
lysis without BIS-1. The maximal effect of BIS-1 was seen after 72 h of prior PBSC activation with OKT3 (P < 0.0005
compared with lysis without BIS-1). Tumor cell lysis in the
presence of BIS-1 was not significantly different after prior PBSC
activation with IL-2/OKT3 compared with OKT3 alone. The addition of
BIS-1 did not increase lysis of control GLC1 (the EGP-2-negative
counterpart of GLC1M13 and therefore incapable of binding BIS-1)
compared with tumor cell lysis without BIS-1 (not shown).
|
Tumor cell lysis of GLC1M13 in the presence of BIS-1 was increased nearly 100% after 72 h of PBSC activation compared with 24 h of activation with OKT3 and IL-2/OKT3 (P < 0.005 and P < 0.025, respectively).
In Fig. 2
, the effect of increasing E:T
ratios is shown. After all PBSC activations, increasing E:T ratio
coincided with increased BIS-1 redirected cytotoxicity (maximal GLC1M13
lysis after OKT3 activation, P < 0.0005).
|
). In the absence of BIS-1, only >1 log
depletion of MCF-7 tumor cells (mean, 6% MCF-7 cells remaining) was
observed with OKT3-activated PBSCs. The sensitivity of tumor cell
detection in these experiments was 1 MCF-7 tumor cell in the total
number (e.g., 6 x 107) of PBSCs
screened, in line with results obtained by Ross et al.
(32)
and Brügger et al.
(33)
.
|
|
Composition of PBSCs and Activation Markers on T Lymphocytes |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSComposition of PBSCs and...Hematopoietic Colony FormationDISCUSSIONREFERENCES |
|---|
.
|
.
|
|
Hematopoietic Colony Formation |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSComposition of PBSCs and...Hematopoietic Colony FormationDISCUSSIONREFERENCES |
|---|
. No
effect of 24 h of prior activation was seen when compared with the
PBS control. The PBS control was not different after 24 or 72 h
(mean, 70 versus 67 CFU-GM; n = 3; not
significant). Also after 72 h, no effect of prior PBSC activation
with IL-2 or OKT3 was seen. However, after 72 h of IL-2/OKT3
activation, CFU-GM numbers were decreased (mean, 39; n = 3; P < 0.0005). No negative effect of BIS-1 alone,
or BIS-1 and GLC1M13 tumor cells, on CFU-GM numbers was observed after
any of the prior PBSC activations (data not shown).
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSComposition of PBSCs and...Hematopoietic Colony FormationDISCUSSIONREFERENCES |
|---|
). Activation of PBSCs was a prerequisite for effective
BIS-1-mediated cell lysis, which is compatible with studies showing
that T cells need prior activation to gain cytolytic potential [see
Kroesen et al. (25
for review]. When we
activated PBSCs with OKT3 and then subsequently with BIS-1, within the
4 h of the assay tumor cell depletion of >3 logs was
observed. It can be argued that an even higher purging efficiency can be expected with this format in the clinical setting for a number of reasons. PBSC harvests, without further purification except for erythrocyte lysis, were used for T-cell activation and tumor cell killing. Selection by means of a density gradient was considered to be less desirable in view of the actual clinical situation. Thus, the "effector cells" consisted of only a minority of CD8+ T cells. E:T ratios in the clinical setting (i.e., the ratio of potential cytotoxic effector cells to tumor cells in tumor-contaminated PBSCs) are likely to be >1 x 102:1 (as used in the 51Cr-release assay in this study) or 1 x 104:1 (log-depletion assay). This is generally the case because highly sensitive methods, including immunocytochemistry (32 , 33) and reverse transcriptase-PCR (34) , are required to detect single tumor cells in 1 x 106–1 x 107 PBSCs. Tumor cell lysis clearly increased with increasing E:T ratios, and therefore a high purging efficiency may be expected in the clinical setting.
The CFU-GM assay, which has predictive value for hematological recovery
after stem cell transplantation (35)
, was used as a
functional evaluation of hematopoietic colony formation after the
purging procedure in this study. CFU-GM numbers were not affected by
PBSC treatment with OKT3, BIS-1, or even by tumor cell killing during
the course of the cytotoxicity assay. In a number of studies, the use
of antibodies for purging purposes was also found not to affect
hematopoietic colony formation in vitro (20)
or
engraftment in patients (7, 8, 9
, 11)
. PBSC treatment with
OKT3 was found to suppress hematopoietic colony formation in
hematological malignancies (36)
, whereas normal control
bone marrow was not affected (37)
. Furthermore, no adverse
effect on hemopoiesis was seen in vivo when patients were
treated i.v. with low-dose OKT3, as is used for induction of antitumor
immunomodulation (38)
. The fact that we did not see a
negative effect of prior activation with OKT3 alone on PBSCs of breast
cancer patients is consistent with these findings. In vitro
IL-2 incubation of breast cancer patient-derived PBSCs did not
negatively affect hematopoietic colony formation in three studies
(21
, 22
, 39)
. Our data confirm and extend these findings.
Despite this, the combination of OKT3 and IL-2 stimulation appeared to
have a clear negative effect on hematopoietic colony formation in our
study (Fig. 5)
. In hematological malignancies, it was suggested that
activated T cells could suppress hematopoietic colony formation
(36)
. This might possibly explain our findings because the
degree of T-cell activation after prior treatment with IL-2/OKT3 was
indeed higher compared with the other treatments (e.g., 73%
of CD8+ T cells were also positive for HLA DR after 72 h of
IL-2/OKT3 activation compared with 53% after OKT3 activation) in this
study.
In search of purging methods efficient in both eliminating tumor cells and maintaining sufficient hemopoiesis, nonselective purging methods using chemotherapy failed to prove useful because of the negative effect on hematopoietic colony formation (5) . As an alternative procedure, in vitro stem cell selection through enrichment of CD34-positive cells has been used, but tumor cells may not be completely eliminated this way (6) . Antibody-based purging methods, e.g., with immunotoxins, proved efficient in eliminating tumor cells (3–4 log depletion, compatible with our results) but were shown to have varying effects on hematopoietic stem cells (19 , 20) . To find a universally expressed epitope in solid tumors is considered difficult, at least when compared with the situation in hematological malignancies (3) . However, antibody-based therapy using epitopes that are not universally expressed (15 , 16 , 18 , 20) is obviously of little clinical significance. The method presented here may offer a good possibility for antibody-based tumor elimination from hematopoietic stem cell harvests because the EGP-2 transmembrane marker is not shed into the circulation, is frequently present and overexpressed in carcinoma cells, and is absent from bone marrow cells (26) . The use of the patient material (PBSCs) itself to eliminate tumor cells is an additional asset of this method. Furthermore, highly sensitive methods for detection of tumor cells in peripheral blood and PBSCs, i.e., immunocytochemistry and a quantitative reverse transcriptase-PCR based on EGP-2 expression, have been developed in our institute (34) . This may allow us to evaluate our purging efficiency in clinically relevant patient samples, which may otherwise be potentially difficult.
It has been stated that an immunocompetent graft may provide antitumor activity, which may also affect possible residual disease in the patient (3) . Long-term follow-up analyses after CD34+ stem cell selection of PBSC grafts (which do not include immunocompetent natural killer cells or T cells) may shed more light on the impact of the immunocompetence of the graft. At this point, data on small numbers of patients are available after a short follow-up, not yet allowing conclusions on disease-free or overall survival (40 , 41) . If indeed the immunocompetence of the graft should play a role, the purging method with BIS-1 described here is likely of interest because immunocompetent cells remain in the graft. Both OKT3 and BIS-1 are used clinically, and the toxicities of OKT3 and BIS-1 are well known in vitro (25, 26, 27) as well as in vivo (14 , 27 , 38) . Autologous patient serum could replace NCS or HPS in this setting.4 Therefore, we are investigating the possibility of performing a clinical study including the use of OKT3 for T-cell activation and retargeting by BIS-1 for purging epithelial tumor cells from PBSCs.
The results of the present in vitro study indicate that specific purging of epithelial cancer cells by means of bispecific antibody BIS-1 is feasible and effective in vitro.
|
FOOTNOTES |
|---|
1 This work was supported by the Dutch Cancer
Society, Grant 96-1277. 
2 To whom requests for reprints should be
addressed, at Division of Medical Oncology, Department of Internal
Medicine, University Hospital Groningen, P.O. Box 30.001, 9700 RB
Groningen, the Netherlands. Phone: 31-50-3616161; Fax: 31-50-3614862;
E-mail: e.g.e.de.vries{at}int.azg.nl
3 The abbreviations used are: PBSC, peripheral
blood stem cell; IL, interleukin; EGP-2, epithelial glycoprotein-2;
BIS-1, bispecific antibody-1; NCS, newborn calf serum; HPS, human
pooled serum; CFU-GM, granulocyte macrophage colony-forming unit.
4 C. P. Schröder, personal observation.
Received 10/24/99; revised 3/16/00; accepted 3/17/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSComposition of PBSCs and...Hematopoietic Colony FormationDISCUSSIONREFERENCES |
|---|
RIII. J. Hematotherapy, 4: 453-456, 1995.[Medline]
This article has been cited by other articles:
|
|
 |
|
  C.P. Schroder, H. Timmer-Bosscha, J.G. Wijchman, L.F.M.H. de Leij, H. Hollema, M.J. Heineman, and E.G.E. de Vries An in vitro model for purging of tumour cells from ovarian tissue Hum. Reprod., May 1, 2004; 19(5): 1069 - 1075. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. H. N. Wong, J. Chan, J. Wong, and P. K. H. Tam Ubiquitous Aberrant RASSF1A Promoter Methylation in Childhood Neoplasia1 Clin. Cancer Res., February 1, 2004; 10(3): 994 - 1002. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Y. Koh, J. R. Ortaldo, B. R. Blazar, M. Bennett, and W. J. Murphy NK-cell purging of leukemia: superior antitumor effects of NK cells H2 allogeneic to the tumor and augmentation with inhibitory receptor blockade Blood, December 1, 2003; 102(12): 4067 - 4075. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
,
PBS; 
, IL-2; 
, OKT3; 
, IL-2/OKT3). * indicates significant
difference compared with E:T ratios of 1:1 and 10:1.
