Multiple Tumoricidal Effector Mechanisms Induced by Adriamycin1
7
[CANCER RESEARCH 44, 2561-2566, June 1984] Multiple Tumoricidal Effector Mechanisms Induced by Adriamycin1 Daniel Salazar and Stefan A. Cohen2 Department of Medicine, State University of New York at Buffalo, School of Medicine, The Buffalo General Hospital, Buffalo, New York 14203 ABSTRACT The antitumor cytotoxic mechanisms of Adriamycin-elicited peritoneal exúdate cells were investigated. Peritoneal exúdate cells from mice collected 1 day after an i.p. injection of Adriamycin (10 mg/kg) displayed enhanced cytotoxicity against P815 (natu ral killer-insensitive, macrophage-sensitive) but not YAC-1 (nat ural killer-sensitive) tumor cell lines. These cells contained a sufficient concentration of the drug to be cytotoxic for P815 tumor cells in 18-hr chromium release assays. Freeze-thaw lysates of these peritoneal exúdate cells were found to be as cytotoxic to P815 as their corresponding whole cells. The lytic activity of these lysates was removed by centrifugation at 100,000 x g, indicating the insolubility of the effector moiety. These cells were also shown to produce significant amounts of Superoxide anión and H2O2 in response to phorbol myristate acetate. A catalase-inhibitable augmentation of the cytotoxicity of these cells against P815 was observed when phorbol myris tate acetate was added to the assay. Neutrophils and not macrophages were likely responsible for this effect. Peritoneal lymphocytes from mice given injections of Adriamycin 5 to 7 days previously were cytotoxic to YAC-1 tumor cells in 4-hr assays. Finally, peritoneal macrophages harvested 5 to 7 days after Adriamycin administration were cytotoxic to P815 in the absence of detectable Adriamycin. The addition of phorbol my ristate acetate inhibited the lysis of P815 by these cells. INTRODUCTION NK3 lymphocytes and tumoricidal macrophages, which can recognize and destroy tumor cells without prior exposure to tumor-associated antigens, constitute an important primary de fense against neoplastic disease (4, 11, 12). The anthracycline antibiotic ADM, which has direct antitumor effects (5), has re cently been shown to have significant influence on these non specific cytotoxic effector cells (8). The ADM-induced antitumor response in vivo has been shown to be radioresistant (21), inhibited by antimacrophage agents such as silica and carra- geenan (16), and detectable after the drug is eliminated (22). Administration i.p. of ADM elicits, in the peritoneal exúdate of mice, both NK cell activity (24) and macrophages (29) capable of inhibiting the growth of tumor cells in vitro. However, Haskill (10) has presented evidence that ADM-elicited macrophages may owe their antitumor cytostatic properties to in vivo uptake of the drug and subsequent transfer to tumor cells. These studies raised 2 questions: (a) can ADM activate these cytotoxic effector cells in a manner similar to bacteria or bacterial products such as lipopolysaccharide, Bacillus Calmette-Guerin, 1Supported by NIH Grant CA 28835-04. 2 To whom requests for reprints should be addressed, at The Buffalo General Hospital, 100 High Street, Buffalo, NY 14203. $The abbreviations used are: NK, natural killer; ADM, Adriamycin; PE, peritoneal exúdate; PEC, peritoneal exúdate cells; ADM 1d PEC, PEC from mice given ADM 1 day prior to sacrifice; PMA, phorbol myristate acetate; E/T, effector/target ratio; FBS, fetal bovine serum; HBSS, Hanks' balanced salt solution; LU, lytic units. Received December 16,1982; accepted March 6,1984. and Comyebacterium parvum; or (o) are the cytotoxic effects a result of drug storage by macrophages and subsequent delivery to tumor. The present study demonstrates that ADM-elicited PEC can lyse tumor cells through both mechanisms. MATERIALS AND METHODS Animals. C3Heb/FeJ mice, 6 to 10 weeks old, were obtained from The Jackson Laboratory (Bar Harbor, ME). Reagents. Catalase, Superoxide dismutase, ferricytochrome c (type VI), horseradish peroxidase, and H2O2 were purchased from Sigma Chemical Co. (St. Louis, MO). PMA was purchased from Consolidated Midland Corporation (Brewster, NY), it was stored as a 2 RIM solution in dimethyl sulfoxide (J. T. Baker Co., Philadelphia, PA) at -70° in the dark. ADM was obtained from Adria Laboratories (Columbus, OH). ADM was dissolved in pyrogen-free sterile H2O and injected i.p. into mice immedi ately as a single dose of 10 mg/kg in all experiments. C. parvum was purchased from Wellcome Research (Beckenham, England), and the suspension was injected i.p. into mice as a single dose of 700 tig in 0.1 ml in all experiments. Preparation of Resident or Elicited Peritoneal Exúdate Cells. Mice were sacrificed by cervical dislocation and the peritoneal cavity lavaged with 6 ml of cold HBSS without Ca2+ and Mg2+ (pH 7.2) using a syringe with an 18-gauge needle. PEC (pooled from 5 to 10 mice) were washed twice with HBSS, counted, and assessed for viability by trypan blue exclusion. In some experiments, PEC were aliquoted into 96-well tissue culture plates (Costar, Cambridge, MA) in 0.2 ml/well and allowed to adhere for 2 hr. The medium and nonadherent cells were aspirated, and the monolayers were washed twice with RPM11640. Differential counts of stained monolayers of the cells showed they were >90% macro phages. Preparation of PEC Nonadherent to Nylon Wool. PEC (4 x 107) were passed though nylon wool columns according to published procedures (13). Briefly, the PEC were allowed to adhere to the column for 60 min at 37°C,and the nonadherent PEC were displaced by dropwise addition of 40 ml of warm medium. Colls were washed in medium and adjusted to the appropriate concentration. Histology of PEC. Coded samples were given to the Department of Hematology, Buffalo General Hospital, and slides of PEC preparations were made by cytocentrifugation and stained with Wright-Giemsa. The differential counts on samples were performed by counting at least 200 cells. Tumor Target Cells. The murine tumors YAC-1 (a lymphoma of A/Sn origin) and P815 (a mastocytoma of DBA/2 origin) were maintained as suspension cultures in RPMI 1640 medium with gentamicin (50 pg/m\) and 5% FBS. Radiolabeled tumor target cells were prepared by incubat ing 107 tumor cells in 0.1 ml with 0.1 mCi of [51Cr]sodium enrómate (Amersham, Arlington Heights, IL) for 45 min at 37°.The cell suspension was washed with HBSS 3 times and diluted to 4 x 105 cells/ml in RPMI 1640 containing gentamicin and 10% FBS. Cytotoxicity Assay. PEC at various concentrations in 0.2 ml of RPMI 1640 medium supplemented with 10% FBS, gentamicin (50 jig/ml), and 25 mM A/-2-hydroxyethylpiperazine-A/1-2-ethane-sulfonic acid were incu bated with 2x10* 51Cr-labeled target cells for 4 or 18 hr at 37°in flat- bottomed 96-well plastic microtiter plates in a humidified atmosphere containing 5% CO2. The microplates were then centrifuged, and 0.1 ml of supernatant was counted for radioactivity for 1 min in a Packard j- Counter. Spontaneous release was determined by incubating 51Cr-la- beled target cells alone, and total labeling was determined by counting JUNE 1984 2561 Research. on November 26, 2014. © 1984 American Association for Cancer cancerres.aacrjournals.org Downloaded from
Transcript of Multiple Tumoricidal Effector Mechanisms Induced by Adriamycin1
[CANCER RESEARCH 44, 2561-2566, June 1984]
Multiple Tumoricidal Effector Mechanisms Induced by Adriamycin1
Daniel Salazar and Stefan A. Cohen2
Department of Medicine, State University of New York at Buffalo, School of Medicine, The Buffalo General Hospital, Buffalo, New York 14203
ABSTRACT
The antitumor cytotoxic mechanisms of Adriamycin-elicited
peritoneal exúdate cells were investigated. Peritoneal exúdatecells from mice collected 1 day after an i.p. injection of Adriamycin(10 mg/kg) displayed enhanced cytotoxicity against P815 (natural killer-insensitive, macrophage-sensitive) but not YAC-1 (natural killer-sensitive) tumor cell lines. These cells contained a
sufficient concentration of the drug to be cytotoxic for P815tumor cells in 18-hr chromium release assays. Freeze-thaw
lysates of these peritoneal exúdate cells were found to be ascytotoxic to P815 as their corresponding whole cells. The lyticactivity of these lysates was removed by centrifugation at100,000 x g, indicating the insolubility of the effector moiety.These cells were also shown to produce significant amounts ofSuperoxide anión and H2O2 in response to phorbol myristateacetate. A catalase-inhibitable augmentation of the cytotoxicity
of these cells against P815 was observed when phorbol myristate acetate was added to the assay. Neutrophils and notmacrophages were likely responsible for this effect. Peritoneallymphocytes from mice given injections of Adriamycin 5 to 7days previously were cytotoxic to YAC-1 tumor cells in 4-hr
assays. Finally, peritoneal macrophages harvested 5 to 7 daysafter Adriamycin administration were cytotoxic to P815 in theabsence of detectable Adriamycin. The addition of phorbol myristate acetate inhibited the lysis of P815 by these cells.
INTRODUCTION
NK3 lymphocytes and tumoricidal macrophages, which can
recognize and destroy tumor cells without prior exposure totumor-associated antigens, constitute an important primary de
fense against neoplastic disease (4, 11, 12). The anthracyclineantibiotic ADM, which has direct antitumor effects (5), has recently been shown to have significant influence on these nonspecific cytotoxic effector cells (8). The ADM-induced antitumor
response in vivo has been shown to be radioresistant (21),inhibited by antimacrophage agents such as silica and carra-
geenan (16), and detectable after the drug is eliminated (22).Administration i.p. of ADM elicits, in the peritoneal exúdate ofmice, both NK cell activity (24) and macrophages (29) capable ofinhibiting the growth of tumor cells in vitro. However, Haskill (10)has presented evidence that ADM-elicited macrophages may
owe their antitumor cytostatic properties to in vivo uptake of thedrug and subsequent transfer to tumor cells.
These studies raised 2 questions: (a) can ADM activate thesecytotoxic effector cells in a manner similar to bacteria or bacterialproducts such as lipopolysaccharide, Bacillus Calmette-Guerin,
1Supported by NIH Grant CA 28835-04.2To whom requests for reprints should be addressed, at The Buffalo General
Hospital, 100 High Street, Buffalo, NY 14203.$The abbreviations used are: NK, natural killer; ADM, Adriamycin; PE, peritoneal
exúdate; PEC, peritoneal exúdate cells; ADM 1d PEC, PEC from mice given ADM1 day prior to sacrifice; PMA, phorbol myristate acetate; E/T, effector/target ratio;FBS, fetal bovine serum; HBSS, Hanks' balanced salt solution; LU, lytic units.
Received December 16,1982; accepted March 6,1984.
and Comyebacterium parvum; or (o) are the cytotoxic effects aresult of drug storage by macrophages and subsequent deliveryto tumor. The present study demonstrates that ADM-elicited
PEC can lyse tumor cells through both mechanisms.
MATERIALS AND METHODS
Animals. C3Heb/FeJ mice, 6 to 10 weeks old, were obtained fromThe Jackson Laboratory (Bar Harbor, ME).
Reagents. Catalase, Superoxide dismutase, ferricytochrome c (typeVI), horseradish peroxidase, and H2O2 were purchased from SigmaChemical Co. (St. Louis, MO). PMA was purchased from ConsolidatedMidland Corporation (Brewster, NY), it was stored as a 2 RIM solution indimethyl sulfoxide (J. T. Baker Co., Philadelphia, PA) at -70° in the dark.
ADM was obtained from Adria Laboratories (Columbus, OH). ADM wasdissolved in pyrogen-free sterile H2O and injected i.p. into mice immedi
ately as a single dose of 10 mg/kg in all experiments. C. parvum waspurchased from Wellcome Research (Beckenham, England), and thesuspension was injected i.p. into mice as a single dose of 700 tig in 0.1ml in all experiments.
Preparation of Resident or Elicited Peritoneal Exúdate Cells. Micewere sacrificed by cervical dislocation and the peritoneal cavity lavagedwith 6 ml of cold HBSS without Ca2+ and Mg2+ (pH 7.2) using a syringe
with an 18-gauge needle. PEC (pooled from 5 to 10 mice) were washed
twice with HBSS, counted, and assessed for viability by trypan blueexclusion. In some experiments, PEC were aliquoted into 96-well tissue
culture plates (Costar, Cambridge, MA) in 0.2 ml/well and allowed toadhere for 2 hr. The medium and nonadherent cells were aspirated, andthe monolayers were washed twice with RPM11640. Differential countsof stained monolayers of the cells showed they were >90% macrophages.
Preparation of PEC Nonadherent to Nylon Wool. PEC (4 x 107) were
passed though nylon wool columns according to published procedures(13). Briefly, the PEC were allowed to adhere to the column for 60 minat 37°C,and the nonadherent PEC were displaced by dropwise addition
of 40 ml of warm medium. Colls were washed in medium and adjustedto the appropriate concentration.
Histology of PEC. Coded samples were given to the Department ofHematology, Buffalo General Hospital, and slides of PEC preparationswere made by cytocentrifugation and stained with Wright-Giemsa. The
differential counts on samples were performed by counting at least 200cells.
Tumor Target Cells. The murine tumors YAC-1 (a lymphoma of A/Sn
origin) and P815 (a mastocytoma of DBA/2 origin) were maintained assuspension cultures in RPMI 1640 medium with gentamicin (50 pg/m\)and 5% FBS. Radiolabeled tumor target cells were prepared by incubating 107 tumor cells in 0.1 ml with 0.1 mCi of [51Cr]sodium enrómate(Amersham, Arlington Heights, IL) for 45 min at 37°.The cell suspensionwas washed with HBSS 3 times and diluted to 4 x 105 cells/ml in RPMI
1640 containing gentamicin and 10% FBS.Cytotoxicity Assay. PEC at various concentrations in 0.2 ml of RPMI
1640 medium supplemented with 10% FBS, gentamicin (50 jig/ml), and25 mM A/-2-hydroxyethylpiperazine-A/1-2-ethane-sulfonic acid were incubated with 2x10* 51Cr-labeled target cells for 4 or 18 hr at 37°in flat-
bottomed 96-well plastic microtiter plates in a humidified atmosphere
containing 5% CO2. The microplates were then centrifuged, and 0.1 mlof supernatant was counted for radioactivity for 1 min in a Packard j-Counter. Spontaneous release was determined by incubating 51Cr-la-
beled target cells alone, and total labeling was determined by counting
JUNE 1984 2561
Research. on November 26, 2014. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from
D. Salazar and S. A. Cohen
a 0.1-ml aliquot after resuspension of the target cells. The following
formula was used to compute percentage of cytotoxicity:
Test cpm - spontaneous cpm% of cytotoxicity = _ t , =- x 100
Total cpm - spontaneous cpm
Spontaneous release by target cells alone did not exceed 10% for 4-hr assays or 25% for 18-hr assays. The results were also expressed as
the LU per PE with 1 LU being defined as the number of cells requiredto achieve 15% cytotoxicity of target cells (14). All E/T ratios were carriedout in triplicate.
Measurement of ADM Concentration in PEC by Spectrofluorome-try. ADM was assayed by the procedure of Schwartz (27); 1 x 106
resident or ADM-elicited PEC washed with 0.9% (w/v) NaCI solutionwere resuspended in 0.5 ml of 1% (w/v) sodium lauryl sulfate, n-Butyl
alcohol (1.5 ml) was added, and the mixture sonicated for 1 min at 0.7output using a Somfier Cell Disruptor (Model W 140 D; Heat SystemsUltrasonics, Plainview, NY) fitted with a microtip. The solution was thencentrifuged at 800 x g for 10 min, and the fluorescence intensity of thebutyl alcohol layer measured at 468-nm excitation and 588-nm emissionwith an Aminco Model SPF-500 spectrofluorometer. ADM concentration
was determined by extrapolation from a standard curve of fluorescenceintensities of known amounts of ADM added to resident PEC prior tosonication. Standard curves produced had a correlation coefficient>0.995, and the experiments presented in the results were repeated atleast 3 times with similar results.
Measurement of Superoxide Anión Production. The assay for su-
peroxide anión was slightly modified from the procedure of Pick andKeisari (20); 1x10« PEC in 1.3 ml of reagent solution (140 mw NaCI, 10
mw K2HPO4, 5.5 m« dextrose, and 100 mw ferricytochrome c) wereincubated at 37° in microcentrifuge tubes for 15 min; 13 /il of 4 /IM
solution of PMA in reagent solution were added, and the tubes incubatedfor 0 (control) or 60 min (assay). Additional assays were performed withSuperoxide dismutase (50 units/ml) added to reagent solution for demonstration of specificity. After incubation, the tubes were centrifuged at11,500 x g for 1 min in a Fisher microcentrifuge (Model 235), and thesupernatant was decanted and placed on ice. The ASSOof the assaysupernatants with the control supernatant as reference was determinedwith a Gary Vanan Model 118 dual-beam spectrophotometer. The follow
ing formula was used to determine the amount of ferricytochrome creduced:
ASSO= 2.1 x 104 mmol/cm
Since the reduction was inhibited by Superoxide dismutase, it represented the amount of Superoxide aniónproduced by the cells in responseto the PMA. The experiment presented in "Results" was repeated 3
times with similar results.Measurement of Hydrogen Peroxide Production. The assay for H2O2
production was slightly modified from the procedure of Pick and Keisari
(20); 1 x 106 PEC were incubated in 1.3 ml of reagent solution (without
ferricytochrome c), 0.28 rriM phenol red, and horseradish peroxidase(type II) (8.5 units/ml) for 0 (control) or 60 min (assay) with PMA. Afterincubation and centrifugaron (11,500 x g; 1 min), the supematants wereadded to test tubes containing 13 n\ of 1 M NaOH and gently shaken.A«ioof assay supernatant was determined with control supernatant asreference. H2O2concentration was determined by linear regression froma standard curve obtained by incubating known amounts of H202 in theabove procedure. Standard curves produced always had a correlationcoefficient of >0.995, and the experiment presented in the results wasrepeated 3 times with similar results.
RESULTS
Composition of PEC Suspensions after Various i.p. Treatments. Mice given ADM or C. parvum i.p. at 1 day before sacrificeyielded approximately 3 times as many PEC as H2O-injected
mice or untreated mice (Table 1). These ADM 1d PEC containeda higher percentage of neutrophils, a slightly higher percentageof macrophages, and a lower percentage of lymphocytes compared to H2O 1d PEC or resident PEC. No mast cells wereobserved in these ADM-induced PEC. There were approximately
5 to 7 times more PEC obtained from mice given ADM at 5 and7 days or C. parvum at 7 days previously as compared withcontrol mice. Mast cells were also absent from these PECsuspensions, while the percentage of lymphocytes and neutrophils returned to levels similar to H20-induced PEC. ADM 5d
PEC and ADM 7d PEC did not contain significantly differentpercentages of macrophages than H2O 5d PEC.
ADM-induced Anti-YAC-1 Activity. Table 2 summarizes theresults of experiments measuring the cytotoxic effect of ADM-induced PEC assayed against the NK-sensitive YAC-1 tumor cellline (23). ADM 1d PEC displayed anti YAC-1 (4 hr) activity
equivalent to H20 1d PEC. In contrast, ADM 5d PEC hadsignificantly augmented anti-YAC-1 activity at all E/T ratios
tested compared to all other suspensions tested. When thenumber of LU/PE was determined, mice treated 5 days previously with ADM displayed an 8- to 16-fold increase compared toH2O-treated mice.
ADM-induced Anti-P815 Activity. To determine whetherADM-elicited PEC can lyse a macrophage-sensitive, NK-resistanttumor cell (23), 51Cr-labeled P815 mastocytoma cells were used
as target cells. No cytotoxicity against P815 tumor cells wasdetected in 4 hr with any of these PEC (data not shown). Givingmice Injections of H2O at 1 or 5 days prior to sacrifice had nosignificant effect on the ability of PEC to lyse P815 in 18 hr
Table 1Compositionof PECsuspensionsinduced by ADM
No. of ex-% of peritonealcells
Treatment ofmice*NoneHzO,
1dayHzO,5daysADM,1dayC.
parvum, 1dayADM,5daysADM,7daysC.
parvum, 7 days10-«2.0
±0.2°(16)"1.7
±0.2(12)1.8±0.3(5)5.7±0.4(20)5.7
(1)9.7±0.9(20)10.8±1.6(3)13.1±0.5 (3)periments"
Macrophages2556151238
±3.043±5.839±5.855±1.61056
±8.14252
±12.2Lymphocytes50
±2.039±3.145±4.813±1.61532
±5.45326
±7.1Mast
cells10
±2.012+3.111
±1.802000Neutrophils2±
1.09±5.85
±2.732±2.47311
±2.2520
±5.0
PEC were harvested 1, 5, or 7 days after i.p. injection into C3HeB/FeJ mice of H20 (0.1 ml), ADM (10 mg/kg), or C.parvum (0.7 mg/mouse)."The compositions of the PEC suspensions were determined by differential counting in the indicated number of
individualexperiments.cMean ±S.E."Numbers in parentheses, number of determinations.
2562 CANCER RESEARCH VOL. 44
Research. on November 26, 2014. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from
compared to resident PEC (Table 3). However, ADM 1d PECdisplayed significantly greater cytotoxicity in 18 hr than didresident or H2O 1d PEC at E/T ratios of 25/1 and 12.5/1. Thesemice also had a 4-fold increase in the number of LU/PE compared
to control mice. ADM 5d PEC or ADM 7d PEC displayed significantly greater cytotoxicity than resident or H2O-induced PEC
Table 2Natura/ cytotoxicity of resident or ADM-elicited PECagainst YAC-1tumor cells
after 4 hr
Treatment of% cytotoxicity
miceNone
H20, 1 dayH20, 5 daysADM, 1 dayADM, 5 days50/1"17.0
±4.3°(7)"
13.1 ±8.8 (3)14.9 ±6.7 (2)16.9 ±2.4 (6)39.9 ±3.8 (12)25/17.3
±2.8 (11)19.4 ±6.7 (4)8.7 ±2.9 (2)
12.2 ±1.9 (9)32.2 ±4.4 (12)12.5/11.2
±0.5(5)10.4 ±3.5 (4)2.9 ±0.2 (2)6.3 ±1.4(6)
15.0 ±2.6(7)LU/PE"1.3
2.10.92.9
15.8"E/T ratio."Number of PEC recovered from mouse (as determined in Table 1) divided by
the number of PEC required to achieve 15% cytotoxicity (1 LU).cMean ±S.E." Numbers in parentheses, number of experiments.
TablesNatural cytotoxicity of resident or ADM-elicited PECagainst P815 tumor cells after
18 hr
Treatment ofmiceNone
H2O,1 dayH20, 5 daysADM, 1 dayADM, 5 daysADM, 7 daysC. parvum, 7
days%
cytotoxicity25/1a10.5±
2.7°(10)"
12.7 ± 2.7 (9)10.6± 4.4 (5)22.4 ± 3.0 (17)21.1 ± 3.3 (10)28.2 + 11.5 (3)35.4 ± 2.9 (3)12.5/10.3
±0.2 (6)5.5 ±2.4 (8)4.9 ±2.1 (6)9.4 ±2.6 (13)"
22.9 ±3.1 (9)36.0 ±8.8 (3)38.7 ±3.3 (3)6.25/1NDe
0.5 ±0.3 (7)2.7 ±1.6(6)0.6 ±0.6 (7)
15.4 + 3.8(5)28.6 + 5.3 (3)37.6 ±1.6(3)LU/PE"2.1
1.61.3
9.126.3
72.7
" E/T ratio."Number of PEC per mouse (as determined in Table 1) divided by the number
of PEC required to achieve 15% cytotoxicity (1 LU).°Mean±S.E."Numbers in parentheses, number of experiments.8ND, not determined.'p < 0.01 (compared to resident or H20 1d PEC by Wikxwon matched-pairs
signed-rankstest).9p < 0.05.
Adriamycin-induced Tumoricldal Activity
against P815 at all E/T ratios assayed. These mice had 13 and30 times more LU/PE than their respective H2O-treated mice. C.parvum 7d PEC also displayed significantly greater anti-P815
activity than controls at all E/T ratios tested.ADM PEC Lysate Cytotoxicity. To ascertain if there was any
direct involvement of ADM in the cytotoxicity against YAC-1 orP815 tumor cells, PEC from untreated or ADM-treated micewere disrupted by 3 cycles of freezing and thawing. These PEC-lysates caused no chromium release from YAC-1 tumor cells in4 hr (Table 4, Experiment A). If ADM was added directly to 51CrYAC-1 cells at 10~5 to 10~9 M, no cytotoxicity was observed
(data not shown). Freeze-thaw lysates obtained from ADM 1d,
5d, 7d PEC were also tested against P815 tumor cells. ADM 1dPEC-lysates were found to be as effective as whole PEC in
killing P815 tumor cells (Table 4, Experiments A and B). Incontrast, lysates of resident, ADM 5d PEC, or ADM 7d PECcaused no chromium release from P815 cells. Furthermore, theability of ADM 1d PEC-lysates to kill P815 cells was lost whenthese lysates were centrifugea at 100,000 x g for 1 hr at 4°in
order to remove organelles and membrane fragments (Table 4,Experiment B).
These experiments suggested that ADM sequestered by ADM1d PEC could account for the cytotoxicity against P815 cells.Therefore, the ADM concentration in these PEC was determinedas well as the ability of ADM to lyse P815 cells. As shown inChart 1, ADM was directly cytotoxic to P815 cells after 18 hrwhen added at 1 x 10~7 M or greater. It can also be seen that
ADM 1d PEC-lysates contained a concentration of drug more
than sufficient to be cytotoxic to P815 cells, while ADM 5d PECcontained a lower concentration of drug. No ADM was detectedin lysates from PEC harvested 7 days after drug administration(data not shown).
Characterization of Peritoneal Effector Cells from ADM-treated Mice. In order to characterize the effector cell(s) in PECpopulations active against YAC-1 and P815 tumor cells, ADM-or C. pa/vum-elicited PECs were fractionated either by nylon
wool columns or by adherence to plastic prior to determinationof both cytotoxicity and cell composition. Nylon wool nonadher-
ent ADM 1d PEC contained significant numbers of macrophages
Table 4Natural cytotoxicity of resident and ADM-elicited PECor their lysates against YAC-1or P815 tumor cells
% cytotoxicity
Treatment ofmiceExperiment
ANoneNoneADM,
1dayADM.1dayADM,5daysADM.5daysExperiment
BNoneNoneADM,
1dayADM,1dayADM,1dayADM,7daysADM,7 daysTreatment
ofPECNoneFT0NoneFTNoneFTNoneFTNoneFTFT/100.000
xg"NoneFT25/1"3.4
±0.5"0.0
±0.018.5±0.92.1
±0.825.9±0.70.0±0.0\
' â„¢/12.5/12.1
±1.20.0±0.012.2
±0.41.9±0.420.4±1.20.0
±0.025/15.2
±0.80.4±0.116.9
±0.716.2±1.115.9
±2.02.1±0.213.2
±1.40.0±0.030.0±0.423.0±1.54.5±0.319.9
±0.40.0±0.012.5/10.0
±0.00.0±0.011.9±
1.411.8±2.019.3
±1.40.0+0.01.6
±0.80.0±0.021.6+2.117.0
±2.12.3±1.025.2
±2.91.4±0.8
"E/T ratio." Mean + S.E.CFT,freezing and thawing; the PEC were subjected to 3 cycles of freezing and thawing, and lysates were
incubated with 5'Cr-labeledtumor cells.dThe PEC lysates were centrifuged at 100,000 x g for 1 hr at 4°,and the resulting supematants were
incubated with 5'Cr-!abeledtumor cells.
JUNE 1984 2563
Research. on November 26, 2014. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from
D. Salazar and S. A. Cohen
(Table 5, Experiment A), whereas nonadherent ADM 5d PEC,ADM 7d PEC, or C. para/m 7d PEC contained greater than 90%lymphocytes (Table 5, Experiments A and B). All of these nonadherent PEC fractions displayed significantly diminished activityagainst P815 compared to unfractionated PEC. In contrast,nonadherent ADM 5d PEC displayed approximately twice thecytotoxicity of corresponding unfractionated PEC against YAC-
1 tumor cells (Table 5, Experiment B). PEC adherent to plasticfrom mice treated with ADM 1 and 7 days or C. parvum 7 dayspreviously contained greater than 85% macrophages (Table 5,Experiment A). These fractions displayed very similar cytotoxicactivity compared to corresponding unfractionated cells.
PMA Augmentation of Natural Cytotoxicity of ADM-elicited
PEC. Nathan ef al. (18, 19) have shown a close correlationbetween PMA-triggered PEC cytotoxicity against certain tumorcells and the release of H;.o by these PMA-stimulated cells. The
70
60>
5 50x 40oo 30
> 20
•\!10 -
uL iiL10
-10 io • io ' io 'ADM CONCENTRATION <M)
10~6 10~5
Chart 1. Cytotoxicity of various concentrations of ADM against 51Cr-P815eelsafter 18 hr (O). The anti-P8l5 cytotoxicity and ADM concentration of lysates(corresponding to an E/T ratio of 25/1) from PEC harvested 1 day P) and 5 days(D)after i.p. ADM administration.ADM 7d PEClysatesdisplayedneithercytotoxicitynor detectable ADM (data not shown).
experiment presented in Table 6 demonstrates that ADM 1dPEC released significant amounts of Superoxide aniónand H202in response to 40 HM PMA. Smaller amounts of both oxygenproducts were also detected with ADM 5d PEC. Furthermore,ADM 1d PEC or ADM 5d PEC did not produce significantamounts of Superoxide aniónor H2O2 in the presence of super-
oxide dismutase or catatase, respectively (Table 6); or in theabsence of PMA (data not shown). Experiments were also conducted to assess whether ADM may be released into the assaybuffer by ADM 1d PEC in response to PMA and thus interferewith the detection of these oxygen products. ADM 1d PEC didnot release any substances into the assay buffer with detectablefluorescence at 468 nm excitation and 588 nm emission (fluorescence maximum for ADM) or with absórbanos at 550 or 610 nm(absorbance maximum for ferricytochrome c and phenol red,respectively) within 1 hr after the addition of PMA (data notshown).
Cytotoxicity assays of ADM- or C. parvum-induced PEC
against P815 were conducted for 4 and 18 hr in the presenceand absence of 40 nM PMA. Resident or ADM id PEC did notcause chromium release from P815 with PMA in 4 hr (data notshown). In contrast, PMA was capable of significantly augment-
Table6PMA-inducedreleaseof Superoxideaniónand hydrogen peroxide by resident and
ADU-elicitedPEC
nM<V/10*ceHs/hr nMhfeOz/IO'cells/hr
Treatment ofmiceNone
ADM, 1 dayADM, 5 daysPMA"7.5
±0.188.1 ±0.851.8 ±0.6PMA
+SOD*-"0.0
±0.00.0 ±0.00.0 ±0.0PMA"2.6
±0.129.1 ±0.94.2 ±0.1PMA
+catatase60.2
±0.13.2 ±0.20.3 ±0.1
" PEC (10*)were incubated with 40 nm PMA for 1 hr at 37°,and the amount of
Superoxideaniónwas determined by reduction of ferricytochrome C. There wasno detectable Superoxideaniónrelease before adding PMA.
6 SOD, Superoxidedismutase.' PEC were assayed with Superoxidedismutase (50 units/ml)." PEC (10") were incubated with 40 MMPMA for 1 hr at 37°,and the amount of
H20r produced was determined by horseradish peroxidase-catalyzed oxidation ofphenol red at Agioâ„¢by reference with incubated sample of cells to which PMAwas added directly before centrifugatton. There was no detectable HjO2 releasedwithout PMA." PEC were assayed with catatase(2000 units).
TablesEffect of nylon wool separation and plastic adherence on the natural cytotoxicity of ADM-elicited PECagainst
YAC-1orP815
% OfPECTreatment
ofmiceExpenment
AH.,0.1dayADM,1dayHjO,
7daysADM,7daysC.
parvum,7daysExperimentsHaO,
5daysADM,5 daysCell
fractionWholeWholeNW-NACPL-ADWholeWholeNW-NAPL-ADWholeNW-NAPL-ADWholeWholeNW-NAMacrophage4855668746536894069246613Lymphocyte3414303374393234864302996%
cytotoxicityYAC-1(4hr)
P815(18hr)25/1"
12.5/1 25/11
3.6 ±0.7o35.2
±2.17.5±0.927.1
±1.610.5±0.346.8
±0.23.6±1.538.2
±1.345.2+1.310.5±0.538.7
±1.30.0
±0.4 1.3 ±1.2 4.8±1.222.6±0.5 11.8 ±0.9 14.9 ±0.543.7±0.8 25.6 ±0.5 9.8 ±1.312.5/18.5
±0.513.1±2.52.8
±1.215.3±3.57.3
±1.535.1±2.12.5
±1.032.0±1.439.9
•2.32.3±0.736.3±0.60.0
±1.110.8+1.41.6±0.8
' E/T rat».6 Mean ±S.E.c NW-NA, PEC nonadherent to nylon wool column; PL-AD, PEC adherent to plastic.
2564 CANCER RESEARCH VOL. 44
Research. on November 26, 2014. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from
Adriamycin-induced Tumoriddal Activity
Table?Lysis of P8T5 tumor cells by resident or ADM-eliciteo PEC triggered with PMA
% cytotoxicity (P815,18 hr)
Treatment of mice Treatment of PEC 25/1" 25/1 + PMA6
ExperimentAHzO,1dayADM,1dayADM,1dayADM,1dayExperiment
BH20,1dayC.
parvum. 1dayExperimentCNoneADM,
1dayADM,1dayADM,5daysADM,5daysExperiment
DH20,1dayADM,1dayADM,1dayADM,7daysADM,7daysC.
parvum, 7daysC.parvum, 7 daysNoneNoneCátalas»CytC8NoneNoneNoneNoneCatataseNoneCatalaseNoneNoneCatalaseNoneCatalaseNoneCatalase15.3
±1.1"25.5
±2.523.3±0.922.0
±1.320.1
±0.618.2±1.022.5
±1.434.0+1.229.6±3.533.5±2.331
.2±2.87.3
±1.513.1±2.510.7±0.646.8±0.250.7±0.845.2+1.346.3±0.65.9
±0.874.3±0.94.1±2.050.2
±1.59.7
±0.663.5±1.2Ntf55.5
±0.924.1±2.39.2±0.610.2
±1.30.0
±0.032.4±2.91.6
±0.824.7±0.623.8±0.830.1±0.627.5±1.7
"E/T ratio."PMA (40 nM) was added to the cytotoxicity assay.cMean±S.E.''Catalase (2000 units/ml) was added to the cytotoxicity assay.''CytC, ferricytochrome c (260 /<M)added to the cytotoxicity assay.'NO,notdetermined.
¡ngthe cytotoxicity of ADM 1d PECor C.parvum 1d PECagainstP815 tumor cells in 18 hr (Table 7, Experiments A and B). ThePMA augmentation of ADM 1d PEC against P815 was totallyinhibited by the addition of catatase (2000 units/ml) and partiallyinhibited by the addition of 260 MMof ferricytochrome c to theassay. Neither compound affected the cytotoxicity of ADM 1dPECwithout PMA nor changed the spontaneous releaseof P815alone. This PMA-induced augmentation of anti-P815 cytotoxicitywas not observed with ADM 5d PEC, ADM 7d PEC, or C.parvum 7d PEC; in fact, the addition of PMA suppressed thecytotoxicity of these cells (Table 7, Experiments C and D).
DISCUSSION
Injection i.p. of bacteria (e.g., C. parvum. Bacillus Calmette-Guerin)or cell-wall components of bacteria (e.g., lipopolysaccha-ride) has been shown to elicit in the peritoneum both tumoricidalmacrophages (1, 12, 25) and NK lymphocytes (30, 32). Thepresent study examined the ability of the antitumor antibioticADM to induce these nonspecific cytotoxic effector cells afteri.p. administration. The increase in the number of PEC demonstrates that ADM provokes a local inflammatory response in theperitoneal cavity. The initial response to ADM (1 day after administration) was characterized by an increased percentage of bothneutrophils and macrophages and the disappearance of mastcells from the PEC (Table 1). Neutrophils have been reported tobe the first cells which invade the site of inflammation in responseto stimuli such as thioglycollate (18) or C. parvum (Table 1) Sincesome of cells, histologically identified as macrophages in theADM 1d PEC were nonadherent to nylon wool (Table 5), ADMmay cause either an influx of immature macrophages into theperitoneum or may affect the surface properties of these cells.Previously, Cohen ef al. (6) have reported that ADM causes aninflux of nonadherent, nonphagocytic macrophages into thespleen, while Facchinetti ef al. (9) have shown that ADM causes
peritoneal macrophages to detach from coverslips on which theywere plated.
Martin et al. (17) have studied the effects of ADM treatmenton the composition of rat PEC. They found that ADM depletedmast cells from the PEC population. This effect, however, doesnot appear to be unique to ADM, since C. para/m-elicited PECwere also depleted of mast cells (Table 1). In contrast to ourstudies, Martin ef al. (17) observed decreased yields of rat PECwith time after i.p. injection. This may indicate a differencebetween the rat and the mouse in their response to ADM.
PEC from mice treated 5 days (but not 1 day) previously withADM demonstrated significantly increased killing of YAC-1 targetcells (Table 2). This lysis occurred within 4 hr and was enrichedby passage of the PEC over nylon wool columns. Since YAC-1tumor cells were insensitive to ADM as well as to ADM-PEClysates in the 4-hr assay, this seems to rule out the possibilitythat ADM retention by PEC and transfer to tumor could beresponsible for lysis of this tumor. These experiments supportthe conclusion of Santoni ef a/. (24) that ADM can elicit NK cellactivity in the peritoneal cavity. The mechanism for this augmentation may simply be that ADM administration causes infiltrationof NK cells into the peritoneal cavity. This laboratory has demonstrated that ADM 1d PEC produced a soluble, nondialyzablemediator(s) capable of augmenting cultured splenic NK activityas well as inducing the proliferation of cultured murine thymo-cytes (7). Therefore, it is possible that resident peritoneal NKcells become activated by this mechanism.
ADM-elicited PEC also demonstrated increased cytotoxicityagainst P815 tumor cells which are NK-resistant but macrophagesensitive (23). Whole and adherent PEC harvested 1 day afterADM treatment displayed a statistically significant cytotoxicityagainst this tumor in 18 hr (Tables 3 and 5). Furthermore, lysatesof these PEC were as effective as their corresponding wholecells in lysing P815. It was also determined that ADM 1d PECcontain concentrations of ADM sufficient to cause lysis of P815tumor cells. These experiments are consistent with the hypothesis of Haskill (10) that ADM stored by peritoneal macrophagesmay be transferred to tumor cells and result in their lysis. C.parvum 1d PEC, unlike ADM 1d PEC, did not display increasedP815 activity compared to control activity (Table 7). Additionally,it was found that the cytotoxicity of ADM 1d PEC lysates waslost after centrifugation (100,000 x g), suggesting that ADMretained by macrophages was not free in the cytoplasm. In thisregard, several groups have reported that ADM is stored bymacrophages in vacuoles or cytoplasmic inclusions (9, 10,17).
PEC from mice treated 5 or 7 days previously with ADMdisplayed significant cytotoxicity against P815. In contrast toADM "id PEC lysates, no P815 lysis was observed with lysates
from ADM 5d PEC or ADM 7d PEC. This cytotoxicity wasmediated by cells adherent to plastic and diminished by nylonwool passage, indicating that macrophages were the effectorcells. Since the cytotoxicity of ADM PEC increased with timeafter administration and the concentration of ADM within thePEC decreased, it seems likely that ADM administration to micecan induce tumoricidal macrophages in the peritoneal cavity.However, some caution is needed in this interpretation. ADM 5dPEC contained detectable quantities of the drug and, thus, theconcentration of ADM in the macrophage-tumor cell microenvi-ronment may be sufficient to kill tumor. Additionally, it cannot beentirely ruled out that ADM is metabolized to nonfluorescentspecies with antitumor activity (28).
JUNE 1984 2565
Research. on November 26, 2014. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from
D. Saiazar and S. A. Cohen
Nathan et al. (18,19) have presented evidence for the involvement of H202 in the lysis of certain tumors (P815, P388) byneutrophils or macrophages from the peritoneum when triggeredby PMA. Our results show that ADM 1d PEC produce significantamounts of Superoxide anión and H2O2 in response to PMA.Reduced levels of these oxygen metabolites were detected withADM 5d PEC. When PMA was added to ADM 1d PEC or C.parvum 1d PEC, significantly augmented P815 lysis was obtained. In the case of ADM 1d PEC, this increase was partiallyinhibited by the addition of ferricytochrome-C (which oxidizes
Superoxide anión and thus diminishes its dismutation to H2O2)and completely inhibited by catatase (which eliminates H2O2).Furthermore, plastic-adherent ADM 1d PEC retained anti-P815
activity but lost their ability to lyse P815 in response to PMA(data not shown) indicating that adherent macrophages were notresponsible for PMA-augmented lysis. Since ADM 1d PEC pro
duced significantly greater amounts of H2O2 and contained significantly greater percentages of neutrophils but not macrophages than ADM 5d PEC or ADM 7d PEC, it is likely that PMAaugmentation of cytotoxicity in this system was due to neutrophils and not macrophages. Furthermore, PMA suppressed ADM5 PEC, ADM 7d PEC, and C. parvum 70 PEC cytotoxicity againstP815 cells. Keller (15) also found that PMA suppressed killing byC. pa/vu/n-induced rat adherent PEC in a 36-hr cytotoxicity
assay. Since P815 is less susceptible to H2O2 than P388 (18),the concentration of H2O2 released from ADM 5d PEC by PMAmay be too low to kill the P815 cells. Catatase did not reversethis PMA-induced suppression. Prostaglandin E2,which has been
shown to inhibit macrophage cytotoxicity (26) and be releasedby macrophages in response to PMA (3), may be responsible forthis suppression.
In light of the findings obtained in the present study as well asother studies, it seems pertinent to regard the antitumor actionsof ADM as consisting of several components. ADM may actdirectly on tumor cells either through intercalation with DNA (5),the intracellular production of reactive oxygen metabolites (2), orthrough interaction at the cell surface in an as yet undefinedmanner (31). The drug may also be delivered to tumor afterstorage by macrophages (10). ADM induces neutrophils in theperitoneum which have the potential to release H2O2, a potentmediator of cytolysis (18). Finally, ADM may stimulate the host's
natural defense system consisting of tumoricidal macrophagesand NK cells. The relative role of each of these possible mechanisms in ADM therapeutic effect is unknown and remains forfuture investigation.
ACKNOWLEDGMENTS
We thank JanuszWicher. Nick C. Gagliardi,and DebbieWohlert for their valuabletechnicalassistance.We also thank Or. J. Waltersand Or. M J. Ehrkefor theircriticalreviewof the manuscript.
REFERENCES
1. Alexander, P., and Evans, R. Endotoxin and double standard RNA rendermacrophagescytotoxic. Nat. New BioL 232:76-78,1971.
2. Bachur, N. R., Gordon, S. L., and Gee, M. V. A general mechanism formicrosomai activation of quinone anticancer agents to free radicals. CancerRes., 38: 1745-1750,1978.
3. Bonney, R. J., Wightman, P. D., Dahlgren, M. E., Davtes,P., KueM,F. A., andHumes, J. L. Effect of RNA and protein synthesis inhibitors on the release ofinflammatory mediators by macrophages responding to phorbol myristateacetate. Btochim. Btophys. Acta, 663: 410-417.1980.
4. Burnet, F. M. The concept of immunological surveillance Prog. Exp. TumorRes., T3:162-185,1970.
5. Carter, S. K. Adriamycin—areview. J. Nati Cancer Inst. 55: 1265-1267,1975.
6. Cohen, S. A., Ehrke, M. J., Ryoyama, K., and Mihich. E. Augmentation of thephagocyte activity of murine spleen cell populations induced by Adriamycin.Immunopharmacology.5: 75-83,1982.
7. Cohen, S. A., Saiazar, D., and Wiener,J. Adriamycnvinducedactivation of NKactivity may initially involve LAP production. Cancer immunoi. Immunother.,Õ5:188-193,1983.
8. Ehrke, M. J., Cohen, S. A., and Mihich, E. Selective effects of Adriamycin onmurine host defense systems. Immunoi. Rev., 65: 55-78,1982.
9. Facchinetti, T., Raz, A., and Goldman, R. A differential interaction of daunc-mycin. Adriamycin, and W-trifluoroacetyladriamycm14-valerate with mouseperitoneal macrophages.Cancer Res., 38: 3944-3949,1978.
10. Haskill. J. S. Adriamycm-activatedmacrophages as tumor growth inhibitors.Cancer Res., 41: 3852-3856,1981.
11. Herberman. R. B . and Holden, H. T. Natural eel-mediated immunity. Adv.Cancer Res., 27: 303-377,1978.
12. Hibbs,J. B.. Taintor, R. R., Chapman, H. A., and Weinberg,J. B. Macrophagetumor killing: influence of the local environment. Science (Wash. DC), 797:279-280,1977.
13. Julius, M. H., Simpson, E.. and Herzenberg, L. A. A rapid method for theisolation of functional thymus-derived murine lymphocytes. Eur. J. Immurai.,3: 645-649, 1973.
14. Kay, H. D., Bonnard, G. D., West, W. H., and Herberman, R. H. A functionalcomparison of human Fc-receptor-bearinglymphocytes active in natural cytotoxicity and antibody-dependentcellularcytotoxicity. J. Immunoi., 118: 2058-2066,1977.
15. Keller, R. Suppression of natural antitumour defence mechanisms by phorbolesters. Nature (Lond.), 282:729-731,1979.
16. Mantovani, A., Polentarutti. N., Luini, W., Peri, G., and Spreafico, F. Rote ofhost defense mechanisms in the antitumor activity of Adriamycin and dauno-mycmin mice. J. Nati. Cancer Inst., 63: 61-66,1979.
17. Martin, F.. Caignard,A., Olsson, O.. Jeannm.J. F.. and Ledere, A. Tumoricidaleffect of macrophagesexposed to Adriamycin in vivo or in vitro. Cancer Res.,42: 3851-3857,1982.
18. Nathan, C. F., Brukner. L. H., Silverstem.S. C.. and Conn, Z. A. Extracellularcytolysis by activated macrophages and granulocytes I. Pharmacologietriggering of effector cells and the release of hydrogen peroxide. J. Exp. Med.,749: 84-99,1979.
19. Nathan, C. F., and Root, R. K. Hydrogen peroxide release from mouseperitoneal macrophages. Dependenceon sequential activation and triggering.J. Exp. Med., 746:1648-1662,1977.
20. Pick, E.. and Keisan.Y. Superoxide aniónand hydrogen peroxide productionby chemically elicited peritoneal macrophages—inductionby multiple non-phagocytic stimuli. Cell. Immunoi.,59: 301-318,1981.
21. Riccardi. C., Pucetti, P., Santoni, A., Herberman, R. B., and Bonmassar. E.Adnamycin-mducedantitumor response in lethally irradiated mice. Immunopharmacology, 7: 211-218,1979.
22. Riccardi, C., Santoni, A., Bartozzari, T., Pucetti, P., and Herberman, R. B. Invivo natural reactivity of mice against tumor cells. Inst. J. Cancer, 25: 475-486,1980.
23. Roder, J. C., Lohman-Matthes, M., Domzig, W., Kiessling, R., and Halter, O.A functional comparison of tumor killing by activated macrophagesand naturalkiller cells. Eur. J. Immunoi.,9:283-288,1979.
24. Santoni,A., Riccardi,C., Sorci, U., and Herberman.R. B. Effects of Adriamycinon the activity of mouse natural killer cell. J. Immunoi., 724:2329-2335,1980.
25. Schultz, R. M . and Chingos, M. A. Macrophage activation for nonspecifictumor cytotoxicity. Adv. Pharmacol.Chemother., 77:157-186,1980.
26. Schultz, R. M., Paulidis. N. A., Stylos, W. A., and Chingos, M. A. Regulationof macrophage tumoricidal function: a rote for prostagiandins of the E series.Science(Wash. DC),202: 320-321.1980.
27. Schwartz, H. S. A fluorometric assay for daunomycmand Adriamycin in animaltissues. Biochem. Med., 7: 396-404,1973.
28. Schwartz, H. S., Kanter, P. M., Mannello,A. J., and Gurtoo, H. L. Biotransformations of anthracylines in the smooth endoplasmic retjculum of rat liver. In:F. M. Muggia (ed.), Proceedingsof the InternationalSymposium on Anthracy-clmeAntibiotics in CancerTherapy, pp. 108-123. Hingham,MA: Martin Nijhoff,1982.
29. Stoychkov, J. N., Schultz, R. M., Chingos, M. A., Paulidis, N. A., and Goldin,A. Effects of Adriamycin and cydophosphamide treatment on induction ofmacrophagecytotoxic function in mice. Cancer Res., 39: 3014-3017,1979.
30. Tracey, D. E., Wolfe, S. A., Durdik, J. M., and Henney, C. S. BCG-mducedmurine effector cells I. Cytoiytic activity in peritoneal exúdales:an earlyresponse to BCG. J. Immunoi., 7J9:1145-1151,1977.
31. Tritton, T. R., and Yee, G. The antìcanceragent Adriamycin can be activelycytotoxic without entering cells. Science (Wash. DC), 277:248-250,1982.
32. Wolfe, S. A., Tracey, D. E., and Henney, C. S. BCG-mducedmurine effectorcells. II. Characterization of natural killer cells in peritoneal exúdales. J.Immunoi., 779:1152-1158,1977.
2566 CANCER RESEARCH VOL. 44
Research. on November 26, 2014. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from
1984;44:2561-2566. Cancer Res Daniel Salazar and Stefan A. Cohen AdriamycinMultiple Tumoricidal Effector Mechanisms Induced by
Updated version
http://cancerres.aacrjournals.org/content/44/6/2561
Access the most recent version of this article at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
Subscriptions
Reprints and
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Permissions
To request permission to re-use all or part of this article, contact the AACR Publications
Research. on November 26, 2014. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from
