JOURNAL
OF SURGICAL
RESEARCH
50,410-415
(19%)
RESIDENT RESEARCH AWARD Natural Killer Cell Stimulatory Factor (NKSF) Augments Natural Cell and Antibody-Dependent Tumoricidal Response against Colon Carcinoma Cell Lines MICHAEL D. LIEBERMAN, Division
Presented
M.D.,RoBERTK.
of Surgical
at the Annual
SIGAL, M.D., NOELN.
Oncology and Harrison School of Medicine, Meeting
of the Association
for Academic
INTRODUCTION
Texas,
of Pennsylvania
November
14-17,
1990
AND METHODS
Cell lines. Colon carcinoma cell lines WC 1, SW 1116, and 498 LI were obtained from Dr. Meenhard Herlyn from the Wistar Institute (Philadelphia, PA). 498 LI 410
Inc. reserved.
Houston,
MATERIAL
Colorectal carcinoma is diagnosed in approximately 150,000 patients each year in the United States and accounts for 75,000 deaths annually [l]. Alternative and $1.50 0 1991 by Academic Press, of reproduction in any form
Surgery,
ANDJOHN M. DALY, M.D.
improved forms of adjuvant therapy to treat colorectal cancer are necessary. Biologic response modifiers have been utilized to enhance anti-tumor effector mechanisms in the treatment of colon cancer. Initial trials with nonspecific immune stimulators or interferon-a were unsuccessful in patients with colorectal cancer [2, 31. However, interleukin-2 (IL-2) administration in patients with metastatic colon carcinoma has resulted in tumor regression [4]. IL-2 enhances the cytolytic activity of natural killer (NK) cells in vitro and in Go, enabling them to lyse tumor cell targets that are resistant to unstimulated NK cells [5,6]. In one trial, IL-2 and/or lymphokine-activated killer cell therapy in patients with advanced colorectal cancer achieved a 17% response rate [4]. IL-2-related systemic toxicity severely limited the efficacy of this cytokine. The search for a more potent cytokine which augments host cytotoxic function with less toxicity may improve therapeutic efficacy. Isolation of a novel cytokine, natural killer cell stimulatory factor (NKSF), from an Epstein-Barr virustransformed B-lymphocyte cell line, may potentially benefit patients with malignancy. NKSF in low concentrations induces interferon-y production from lymphocytes, enhances NK cell lysis of lymphoma cell line targets, and serves as a comitogen with IL-2 to stimulate lymphocyte proliferation [7]. Thus, the biologic function of NKSF may stimulate natural immune effector mechanisms toward colorectal cancer cells. This study investigated the ability of NKSF to activate NK cell cytotoxicity and lymphocyte-mediated antibody-dependent cellular cytotoxicity of colon carcinoma cell lines.
Inc.
0022-4804/91 Copyright All rights
II,FRCS,
Department of Surgical Research, University Phikzdelphia, Pennsylvania 19104
The therapy of colorectal cancer may be improved by biologic response modifiers that enhance natural killer (NK) cell and antibody-dependent tumoricidal mechanisms. This study examined the effect of a recently discovered cytokine purified from the supernatant of an Ebstein-Barr virus-transformed B-lymphoblastoid cell line (RPMI-8866), natural killer cell stimulatory factor (NKSF), on NK and antibody-dependent cellular cytotoxicity (ADCC) of human colon adenocarcinoma cell lines. Human peripheral blood lymphocytes were cultured for 24 hr in the presence or absence of NKSF (3.6 p&f) or interleukin-2 (1 n2M). The cultured lymphocytes were analyzed for lytic potential toward chromium-51-labeled colon carcinoma targets SW 1116, 498 LI, and WC 1. ADCC was measured by incubating chromium-51-labeled SW 1116 or WC 1 targets with the monoclonal antibody C017-lA, an IgG2a antibody reactive with gastrointestinal cancer-associated cell antigen, or control mouse IgG prior to testing NKSFtreated or control PBL effecters in a 6-hr cytotoxicity assay. NKSF significantly enhanced NK cytolysis of colon carcinoma and NK-resistant lymphoma cell lines, and on a molar basis was approximately 300 times more potent than interleukin-2 in generating NK cytotoxicity. Furthermore, NKSF significantly augmented lymphocyte-mediated ADCC against colon carcinoma targets, and the combination of NKSF with the antibody C017-1A had an additive effect on lymphocyte tumoricidial capacity. Thus, NKSF may have a potential role in the treatment of colon cancer. o 1991 Academic Press,
WILLIAMS,
Killer
LIEBERMAN
ET
AL.:
NATURAL
KILLER
was established by serial subcapsular splenic injection of SW 1116 cells and subsequent harvesting of hepatic nodules in the nude mouse. Colorectal cell lines were maintained in W487 medium at 37”C, pH 7.4, in 5% CO,. Daudi, derived from a human Burkitt lymphoma, was maintained in continuous culture. Medium. Colon carcinoma cell lines were maintained in medium W487 which consisted of 4 parts MCDB 202 (GIBCO, Grand Island, NY), 1 part L-15 (Flow Laboratories, McLean, VA), epidermal growth factor (5 rig/ml, Collaborative Research, Bedford, MA), transferrin (10 pg/ml), insulin (5 pg/ml), hydrocortisone ( lop6 M), ethanolamine ( 10e4 M), phosphoethanolamine (lop4 M, all Sigma, St Louis, MO), and 2% fetal calf serum (Flow Laboratories, Inc., Rockville, MD). Cytotoxicity assays were performed in RPM1 1640 (GIBCO) supplemented with 10% heat-inactivated fetal calf serum, 100 units/ml penicillin, 100 @g/ml streptomycin, 50 pg/ml gentimicin, 1% glutamine, 1% pyruvate/insulin, and 1% nonessential amino acids (all Sigma). Daudi was maintained in this medium. Reagents. NKSF was produced by Dr. Bite Perussia, Wistar Institute, (Philadelphia, PA). Briefly, cells from the Ebstein-Barr virus-transformed B-cell line RPMI8866 were cultured for 48 hr with 1O-7 M phorbol-12,13dibutyrate (PDBu). The conditioned medium was sequentially purified by filtration through a QAE-Zetaprep cartridge, affinity chromatography on a lentil lectin-Sepharose column, hydroxylapatite column chromatography of the activity recovered in the bound fraction from the lentil lectin column, affinity chromatography on heparin-Sepharose, and anion exchange chromatography on a Mono-Q resin column [7]. Recombinant human IL-2, specific activity 15 X lo6 units/mg protein, was provided by Hoffman-LaRoche (Nutley, NJ). Monoclonal antibody (MAb) CO 17-IA was provided by Dr. Meenhard Herlyn, W6/32 was obtained from Dr. Bite Perussia, and anti-CD16 and anti-HLA-Dr was purchased from Becton-Dickinson (Sunnyvale, CA). Peripheral blood lymphocytes (PBL). Peripheral blood mononuclear cells from healthy volunteers were isolated from heparinized venous blood by Ficoll-Hypaque density gradient centrifugation. Lymphocytes were then purified by plastic adherence and passaged through nylon wool columns to deplete B-lymphocytes and monocytes. Cytotoxicity assays (NK and ADCC). Labeling of targets was performed by incubation of 2 X lo6 viable tumor cells (Daudi, WC 1, SW 1116, and 498 LI) with 125 ~1 of Na5Cr0 (2 mCi/ml, New England Nuclear, Boston, MA) for 90 min at 37°C followed by washing the cells three times in complete medium. In the ADCC assay chromium (Cr)-51-labeled tumor targets were then sensitized with saturating amounts of the MAb C017-1A or a control of mouse IgG for 90 min at 22°C on a shaker. The antibody-labeled targets were then washed, centri-
CELL
STIMULATORY
411
FACTOR
fuged (400g X 6 min), and resuspended in complete medium. The cytotoxicity assays were performed in U-bottom microwells by adding serial dilutions of effector cells, each dilution being plated in triplicate, to 1 X lo4 51Cr-labeled target cells in a total volume of 200 ~1 complete medium. The plates were centrifuged (1OOg X 5 min) at the beginning of the assay, incubated at 37°C in 5% CO, for 6 hr, and centrifuged again (400g X 10 min). A lOO-~1 aliquot of supernatant was removed for gamma counting (Wallac Clinigamma, Gaithersburg, MD). Percentage specific lysis was computed by the formula (E - S)/(M - S) x 100, where E is the 51Cr release in the presence of effecters, S is the spontaneous 51Cr release in the presence of medium, and M is the maximum release obtained by lysing targets with 1% Triton X-100. Percentage ADCC was calculated by subtracting percentage lysis with MAb CO 17-1A from percentage lysis with control IgG. Fluorescence-activated cell sorter (FACS) analysis. Fluorescein isothiocynate-conjugated (FITC)-labeled anti-CD16 and anti-HLA-Dr was used for staining PBL and tumor cell lines, respectively. In all analyses 1 X lo6 splenocytes were incubated for 45 min at 4°C with 5 ~1 of primary antibody diluted in 45 ~1 of phosphate-buffered saline containing 2% FCS and 0.01% sodium azide (staining buffer). Cells were then washed three times in staining buffer. A similar procedure was used to apply FITC-labeled goat anti-mouse Ab for secondary labeling of W6/32. Analysis of 5 X lo4 cells/sample was performed on a FACS microfluorimeter (Ortho Cytofluorograf, Ortho Diagnostic Instruments, MA). Statistics. The data were assessed by ANOVA with the Scheffee multiple comparison procedure. The level of significance was determined at P < 0.05. Data are expressed as means * standard deviation. RESULTS
Pathologic
Features
of Colon Carcinoma
Cell Lines
The clinical and pathologic features of the colon carcinoma cell lines are provided in Table 1. Previous studies have demonstrated an inverse relationship between the relative state of differentiation of a tumor target and susceptibility to NK cell lysis [B]. Thus, cytospin analysis was performed on the colon cancer cell lines to define the state of differentiation based on cellular and nuclear morphology, mucin production, and pattern of gland formation. 498 LI represented the least differentiated adenocarcinoma, WC 1 more differentiated than 498 LI, and SW 1116 the most differentiated tumor studied. Interestingly, 498 LI was derived from the SW 1116 cell line which was repeatedly passaged through subcapsular splenic injections and subsequent harvesting of hepatic nodules in nude mice. This process selected for a more undifferentiated population of colon carcinoma cells. In this study there was no definitive relationship between
412
JOURNAL
OF
SURGICAL
RESEARCH:
VOL.
TABLE Pathologic Cell line
Features
Site
WC1 SW 1116 498 LI
Colon Colon Colon
+ xenogeneic
stage
NK lytic potential and state of tumor target cell differentiation (Figs. lB-1D). Because of the possible relationship between major histocompatibility complex (MHC) antigen expression and NK cell cytotoxicity [9], the tumor cell lines were analyzed for MHC class I and class II antigen expression (Table 2). The colon carcinoma cell lines demonstrated significant differences in the expression of MHC class I antigens. W6/32, a monomorphic IgG antibody, was used to stain for MHC class I antigen on the tumor cells. All tumor cells stained for MHC class I antigen and the mean channel fluorescence for WC 1, SW 1116, and 498 LI was 248 +- 51,477 191, and 944 f 67, respectively (P < .05). The colon carcinoma cell lines did not express class II HLA-Dr molecules. Previous studies have demonstrated CO 17-1A expression on the cell surface of the colorectal cell lines [lo].
1991
Cell Lines
C A A
passage
4, APRIL
1
of Colorectal
Duke’s
50, NO.
Differentiation
Mucin
Glands
Undifferentiated Moderate Undifferentiated
Yes Yes No
No Yes No
Effect of NKSF on NK Cell Cytotoxicity Colon carcinoma cell lines were utilized to measure basal- and NKSF-activated NK cell cytotoxicity. PBL were isolated from healthy adults (n = 5) and cultured for 24 hr in the presence or absence of NKSF (1 p/ml = 3.6 PM) or IL-2 (50 p/ml = 1 ti), a cytokine known to activate NK cell cytotoxicity. The cultured lymphocytes were then measured for cytotoxicity toward WC 1, SW 1116, and 498 LI colon carcinoma cell lines (Figs. lA-1D). Basal NK cytotoxicity toward the tumor cell lines was variable. WC 1 and SW 1116 were resistant to basal NK lysis, while 498 LI was relatively sensitive to unstimulated NK lysis. For example, at an effector to target ratio (E:T) of 50:1, mean NK specific lysis was 6 + 4% for WC 1,16 f 4% for SW 1116, and51 -+ 6% for498LI (PcO.05
A100 60 ml3
503
l&i:1
253
625:1
m
!al
12!3l
6251
125:l
6251
E:T RATIO
E:T RATIO
ff ------_____
-N
OF SURGICAL
RESEARCH
50,410-415
(19%)
RESIDENT RESEARCH AWARD Natural Killer Cell Stimulatory Factor (NKSF) Augments Natural Cell and Antibody-Dependent Tumoricidal Response against Colon Carcinoma Cell Lines MICHAEL D. LIEBERMAN, Division
Presented
M.D.,RoBERTK.
of Surgical
at the Annual
SIGAL, M.D., NOELN.
Oncology and Harrison School of Medicine, Meeting
of the Association
for Academic
INTRODUCTION
Texas,
of Pennsylvania
November
14-17,
1990
AND METHODS
Cell lines. Colon carcinoma cell lines WC 1, SW 1116, and 498 LI were obtained from Dr. Meenhard Herlyn from the Wistar Institute (Philadelphia, PA). 498 LI 410
Inc. reserved.
Houston,
MATERIAL
Colorectal carcinoma is diagnosed in approximately 150,000 patients each year in the United States and accounts for 75,000 deaths annually [l]. Alternative and $1.50 0 1991 by Academic Press, of reproduction in any form
Surgery,
ANDJOHN M. DALY, M.D.
improved forms of adjuvant therapy to treat colorectal cancer are necessary. Biologic response modifiers have been utilized to enhance anti-tumor effector mechanisms in the treatment of colon cancer. Initial trials with nonspecific immune stimulators or interferon-a were unsuccessful in patients with colorectal cancer [2, 31. However, interleukin-2 (IL-2) administration in patients with metastatic colon carcinoma has resulted in tumor regression [4]. IL-2 enhances the cytolytic activity of natural killer (NK) cells in vitro and in Go, enabling them to lyse tumor cell targets that are resistant to unstimulated NK cells [5,6]. In one trial, IL-2 and/or lymphokine-activated killer cell therapy in patients with advanced colorectal cancer achieved a 17% response rate [4]. IL-2-related systemic toxicity severely limited the efficacy of this cytokine. The search for a more potent cytokine which augments host cytotoxic function with less toxicity may improve therapeutic efficacy. Isolation of a novel cytokine, natural killer cell stimulatory factor (NKSF), from an Epstein-Barr virustransformed B-lymphocyte cell line, may potentially benefit patients with malignancy. NKSF in low concentrations induces interferon-y production from lymphocytes, enhances NK cell lysis of lymphoma cell line targets, and serves as a comitogen with IL-2 to stimulate lymphocyte proliferation [7]. Thus, the biologic function of NKSF may stimulate natural immune effector mechanisms toward colorectal cancer cells. This study investigated the ability of NKSF to activate NK cell cytotoxicity and lymphocyte-mediated antibody-dependent cellular cytotoxicity of colon carcinoma cell lines.
Inc.
0022-4804/91 Copyright All rights
II,FRCS,
Department of Surgical Research, University Phikzdelphia, Pennsylvania 19104
The therapy of colorectal cancer may be improved by biologic response modifiers that enhance natural killer (NK) cell and antibody-dependent tumoricidal mechanisms. This study examined the effect of a recently discovered cytokine purified from the supernatant of an Ebstein-Barr virus-transformed B-lymphoblastoid cell line (RPMI-8866), natural killer cell stimulatory factor (NKSF), on NK and antibody-dependent cellular cytotoxicity (ADCC) of human colon adenocarcinoma cell lines. Human peripheral blood lymphocytes were cultured for 24 hr in the presence or absence of NKSF (3.6 p&f) or interleukin-2 (1 n2M). The cultured lymphocytes were analyzed for lytic potential toward chromium-51-labeled colon carcinoma targets SW 1116, 498 LI, and WC 1. ADCC was measured by incubating chromium-51-labeled SW 1116 or WC 1 targets with the monoclonal antibody C017-lA, an IgG2a antibody reactive with gastrointestinal cancer-associated cell antigen, or control mouse IgG prior to testing NKSFtreated or control PBL effecters in a 6-hr cytotoxicity assay. NKSF significantly enhanced NK cytolysis of colon carcinoma and NK-resistant lymphoma cell lines, and on a molar basis was approximately 300 times more potent than interleukin-2 in generating NK cytotoxicity. Furthermore, NKSF significantly augmented lymphocyte-mediated ADCC against colon carcinoma targets, and the combination of NKSF with the antibody C017-1A had an additive effect on lymphocyte tumoricidial capacity. Thus, NKSF may have a potential role in the treatment of colon cancer. o 1991 Academic Press,
WILLIAMS,
Killer
LIEBERMAN
ET
AL.:
NATURAL
KILLER
was established by serial subcapsular splenic injection of SW 1116 cells and subsequent harvesting of hepatic nodules in the nude mouse. Colorectal cell lines were maintained in W487 medium at 37”C, pH 7.4, in 5% CO,. Daudi, derived from a human Burkitt lymphoma, was maintained in continuous culture. Medium. Colon carcinoma cell lines were maintained in medium W487 which consisted of 4 parts MCDB 202 (GIBCO, Grand Island, NY), 1 part L-15 (Flow Laboratories, McLean, VA), epidermal growth factor (5 rig/ml, Collaborative Research, Bedford, MA), transferrin (10 pg/ml), insulin (5 pg/ml), hydrocortisone ( lop6 M), ethanolamine ( 10e4 M), phosphoethanolamine (lop4 M, all Sigma, St Louis, MO), and 2% fetal calf serum (Flow Laboratories, Inc., Rockville, MD). Cytotoxicity assays were performed in RPM1 1640 (GIBCO) supplemented with 10% heat-inactivated fetal calf serum, 100 units/ml penicillin, 100 @g/ml streptomycin, 50 pg/ml gentimicin, 1% glutamine, 1% pyruvate/insulin, and 1% nonessential amino acids (all Sigma). Daudi was maintained in this medium. Reagents. NKSF was produced by Dr. Bite Perussia, Wistar Institute, (Philadelphia, PA). Briefly, cells from the Ebstein-Barr virus-transformed B-cell line RPMI8866 were cultured for 48 hr with 1O-7 M phorbol-12,13dibutyrate (PDBu). The conditioned medium was sequentially purified by filtration through a QAE-Zetaprep cartridge, affinity chromatography on a lentil lectin-Sepharose column, hydroxylapatite column chromatography of the activity recovered in the bound fraction from the lentil lectin column, affinity chromatography on heparin-Sepharose, and anion exchange chromatography on a Mono-Q resin column [7]. Recombinant human IL-2, specific activity 15 X lo6 units/mg protein, was provided by Hoffman-LaRoche (Nutley, NJ). Monoclonal antibody (MAb) CO 17-IA was provided by Dr. Meenhard Herlyn, W6/32 was obtained from Dr. Bite Perussia, and anti-CD16 and anti-HLA-Dr was purchased from Becton-Dickinson (Sunnyvale, CA). Peripheral blood lymphocytes (PBL). Peripheral blood mononuclear cells from healthy volunteers were isolated from heparinized venous blood by Ficoll-Hypaque density gradient centrifugation. Lymphocytes were then purified by plastic adherence and passaged through nylon wool columns to deplete B-lymphocytes and monocytes. Cytotoxicity assays (NK and ADCC). Labeling of targets was performed by incubation of 2 X lo6 viable tumor cells (Daudi, WC 1, SW 1116, and 498 LI) with 125 ~1 of Na5Cr0 (2 mCi/ml, New England Nuclear, Boston, MA) for 90 min at 37°C followed by washing the cells three times in complete medium. In the ADCC assay chromium (Cr)-51-labeled tumor targets were then sensitized with saturating amounts of the MAb C017-1A or a control of mouse IgG for 90 min at 22°C on a shaker. The antibody-labeled targets were then washed, centri-
CELL
STIMULATORY
411
FACTOR
fuged (400g X 6 min), and resuspended in complete medium. The cytotoxicity assays were performed in U-bottom microwells by adding serial dilutions of effector cells, each dilution being plated in triplicate, to 1 X lo4 51Cr-labeled target cells in a total volume of 200 ~1 complete medium. The plates were centrifuged (1OOg X 5 min) at the beginning of the assay, incubated at 37°C in 5% CO, for 6 hr, and centrifuged again (400g X 10 min). A lOO-~1 aliquot of supernatant was removed for gamma counting (Wallac Clinigamma, Gaithersburg, MD). Percentage specific lysis was computed by the formula (E - S)/(M - S) x 100, where E is the 51Cr release in the presence of effecters, S is the spontaneous 51Cr release in the presence of medium, and M is the maximum release obtained by lysing targets with 1% Triton X-100. Percentage ADCC was calculated by subtracting percentage lysis with MAb CO 17-1A from percentage lysis with control IgG. Fluorescence-activated cell sorter (FACS) analysis. Fluorescein isothiocynate-conjugated (FITC)-labeled anti-CD16 and anti-HLA-Dr was used for staining PBL and tumor cell lines, respectively. In all analyses 1 X lo6 splenocytes were incubated for 45 min at 4°C with 5 ~1 of primary antibody diluted in 45 ~1 of phosphate-buffered saline containing 2% FCS and 0.01% sodium azide (staining buffer). Cells were then washed three times in staining buffer. A similar procedure was used to apply FITC-labeled goat anti-mouse Ab for secondary labeling of W6/32. Analysis of 5 X lo4 cells/sample was performed on a FACS microfluorimeter (Ortho Cytofluorograf, Ortho Diagnostic Instruments, MA). Statistics. The data were assessed by ANOVA with the Scheffee multiple comparison procedure. The level of significance was determined at P < 0.05. Data are expressed as means * standard deviation. RESULTS
Pathologic
Features
of Colon Carcinoma
Cell Lines
The clinical and pathologic features of the colon carcinoma cell lines are provided in Table 1. Previous studies have demonstrated an inverse relationship between the relative state of differentiation of a tumor target and susceptibility to NK cell lysis [B]. Thus, cytospin analysis was performed on the colon cancer cell lines to define the state of differentiation based on cellular and nuclear morphology, mucin production, and pattern of gland formation. 498 LI represented the least differentiated adenocarcinoma, WC 1 more differentiated than 498 LI, and SW 1116 the most differentiated tumor studied. Interestingly, 498 LI was derived from the SW 1116 cell line which was repeatedly passaged through subcapsular splenic injections and subsequent harvesting of hepatic nodules in nude mice. This process selected for a more undifferentiated population of colon carcinoma cells. In this study there was no definitive relationship between
412
JOURNAL
OF
SURGICAL
RESEARCH:
VOL.
TABLE Pathologic Cell line
Features
Site
WC1 SW 1116 498 LI
Colon Colon Colon
+ xenogeneic
stage
NK lytic potential and state of tumor target cell differentiation (Figs. lB-1D). Because of the possible relationship between major histocompatibility complex (MHC) antigen expression and NK cell cytotoxicity [9], the tumor cell lines were analyzed for MHC class I and class II antigen expression (Table 2). The colon carcinoma cell lines demonstrated significant differences in the expression of MHC class I antigens. W6/32, a monomorphic IgG antibody, was used to stain for MHC class I antigen on the tumor cells. All tumor cells stained for MHC class I antigen and the mean channel fluorescence for WC 1, SW 1116, and 498 LI was 248 +- 51,477 191, and 944 f 67, respectively (P < .05). The colon carcinoma cell lines did not express class II HLA-Dr molecules. Previous studies have demonstrated CO 17-1A expression on the cell surface of the colorectal cell lines [lo].
1991
Cell Lines
C A A
passage
4, APRIL
1
of Colorectal
Duke’s
50, NO.
Differentiation
Mucin
Glands
Undifferentiated Moderate Undifferentiated
Yes Yes No
No Yes No
Effect of NKSF on NK Cell Cytotoxicity Colon carcinoma cell lines were utilized to measure basal- and NKSF-activated NK cell cytotoxicity. PBL were isolated from healthy adults (n = 5) and cultured for 24 hr in the presence or absence of NKSF (1 p/ml = 3.6 PM) or IL-2 (50 p/ml = 1 ti), a cytokine known to activate NK cell cytotoxicity. The cultured lymphocytes were then measured for cytotoxicity toward WC 1, SW 1116, and 498 LI colon carcinoma cell lines (Figs. lA-1D). Basal NK cytotoxicity toward the tumor cell lines was variable. WC 1 and SW 1116 were resistant to basal NK lysis, while 498 LI was relatively sensitive to unstimulated NK lysis. For example, at an effector to target ratio (E:T) of 50:1, mean NK specific lysis was 6 + 4% for WC 1,16 f 4% for SW 1116, and51 -+ 6% for498LI (PcO.05
A100 60 ml3
503
l&i:1
253
625:1
m
!al
12!3l
6251
125:l
6251
E:T RATIO
E:T RATIO
ff ------_____
-N
