Cell-Surface Antigens From Human Breast Tumor Cells 1 ,2 Michael P. Lerner, 3 J Hili Anglin, 4 and Robert E. Nordquist 5 , 6 , 7
ated with this human breast tumor cell line, BOT-2, and the reactions of patients' sera against these cells and their extracted antigens.
ABSTRACT-Biochemical and immunologie studies on breast cancer with the use of cells from a human ductal cell carcinoma, BOT-2, were initiated. Antigens were extracted from the cells by mild sonication and purified by gel filtration chromatography. Only one of the three peaks from gel filtration chromatography reacted with antiserum prepared against whole BOT-2 cells. Analysis by polyacrylamide gel electrophoresis of the BOT-2 cell extract revealed many protein bands, whereas analysis of the antibody-reactive peak after gel filtration chromatography revealed fewer protein bands. Immunologie tests to identify human serum antibodies agalnst BOT-2 cells or cell extracts were performed by fixed cell immunofluorescence, Iiving cell membrane immunofluorescence, and indirect hemagglutination. Depending on the test, the sera from women with diagnosed, untreated mammary cancer were positive in 45-80% of the cases, whereas the sera from women without apparent breast diseases (controls) were positive in only 5-10% of the cases. The results suggested that the antigens from the BOT-2 cells will be useful in understanding the processes involved in human mammary neoplasia. - J Natl Cancer Inst 60: 39-44, 1978.
MATERIALS AND METHODS
Breast cancer is the greatest cause of death, due to cancer, among American women. Although surgical and chemotherapeutic methods are available to treat women with breast cancer, the mortality rate is still distressingly high. Part of the reason for the high mortality rate is the lack of information about specific biochemical and immunologie events that occur during the neoplastic process. However, tumor-specific antigens and antibodies associated with mammary cancer have been demonstrated by lymphocyte cytotoxicity and blocking (1, 2), lymphocyte blastogenesis (3), leukocyte migration inhibition (4, 5), delayed hypersensitivity reactions (6, 7), and circulating tumor-specific antibodies (8-11). Antigens from neoplastic mammary tissue have been extracted, purified, and used in a skin testing procedure (7). All the above data, plus information from studies on other tumor systems, indicate that tumor cells have distinct antigens which may help the tumor cell metastasize or avoid immunologie destruction. Despite recent re ports concerning antigens and antibodies associated with human breast cancer, knowledge about the nature of the antigens is lacking. Part of this problem can be attributed to difficulties associated with antigen extraction from tissue and to the small amounts of antigens recovered from whole tumor tissue. The use of cultured tumor cells to isolate tumor antigens provides a greater amount of uniform antigens than is available from whole tumor tissue. Nordquist et al. (12) recently described the establishment and long-term cultivation of a human ductal cell carcinoma. Here we re port our results from studies of the antigens associVOL. 60, NO. 1, JANUARY 1978
ABBREVIATIONS USED: FCS = fetal calf serum; PBS = phosphatebuffered saline; PGE = polyacrylamide gel electrophoresis; SDS = sodium dodecyl sulfate; FCF = fixed cell immunofluorescence; FITC = fluorescein isothiocyanate; LCMF = living cell membrane immunofluorescence; HA = hemagglutination. Received March 15, 1977; accepted August 1,1977. Supported by grant BC-230 from the American Cancer Society and a Maizie Wilkonson Award (M.P.L.) and in part by Public Health Service grants GRS 5-S01-RR05411 from the Division of Research Resources, National Institutes of Health, and a grant from the Oklahoma Heart Association. 3 Department of Microbiology and Immunology, University of Oklahoma Health Seiences Center, Oklahoma City, Okla. 73190. 4 Department of Biochemistry and Molecular Biology, University of Oklahoma Health Seiences Center. 5 Cancer Research Program , Oklahoma Medical Research Foundation, Oklahoma City, Okla. 73104. 6 Present address: Department of Anatomical Sciences, University of Oklahoma Health Seiences Center. 7 We thank Richard Strecker, Peggy Riggs, Peggy Munson, and Lucy Durbin for excellent technical assistance. 1
2
39
J
NATL CANCER INST
Downloaded from http://jnci.oxfordjournals.org/ at Dalhousie University on December 11, 2012
Cell culture and antigen extraction. -BOT-2 cells were grown in Eagle's minimum essential medium containing 10% heat-inactivated FCS. The morphology and growth properties of this cell line have been described (12). The growth medium from 10 confluent 75-cm 2 flasks was decanted, and the cells were resuspended in 3 ml of PBS (0.15 M NaCI, 0.01 M Na+, K+ P0 4 , pH 7.1). The cells were washed by low-speed centrifugation and resuspended in 5 ml of PBS. Antigens were extracted from the cells by sonication in a tank sonifier (40 mA) for 60 minutes at 4° C. The sonicate was centrifuged at 1,500Xg for 10 minutes to pellet the cells, and the supernatant was centrifuged at 140,000Xg for 60 minutes to remove large aggregates. The resulting supernatant was used to study the properties of the BOT-2 cell antigens. Antigen purification and analysis.-The supernatant was concentrated to about 5 mg protein/rnl by Sephadex G200 hydration. This sampIe was then applied to Sephadex G-50, G-I00, G-150, and G-200 columns equilibrated at 4° C with PBS. Fractions were collected, and the absorbency at 280 nm was determined in a Gilford spectrophotometer (Gilford Instrument Laboratories , Inc., Oberlin, Ohio). Contiguous fractions of absorbing material were pooled, dialyzed against several changes of water, and lyophilized. The lyophilized sampIes were dissolved in PBS and analyzed by PGE according to
40
LERNER, ANGLIN , AND NORDQUIST
J
NATL CANCER INST
ing positive HA was 1:8. For control experiments, FCS or HeLa cell extracts were coupled to tanned erythrocytes. For all experiments, proteins were estimated by the absorbeney ratio at 260 and 280 nm (17) . RESULTS Antigen Extraction and Purification
We were interested in learning whether the BOT-2 ceIls synthesized surface antigens useful for the study of many aspects of human mammary cancer. Three different methods were used to obtain antigen material from BOT-2 ceIls: 1) extraction of antigens from eells with the use of 3 M KCI, 2) extraction with I-butanol, and 3) extraction by sonication in PBS. We monitored the extraction procedures by protein determination and by immunoelectrophoresis with the use of antiserum prepared against living BOT-2 eells. We found that extraction of antigens from the cells with 3 M KCI or 1butanol yielded large amounts of cellular protein containing little antibody-reactive material. Extraction of antigens from cells by mild sonication at 4° C in PBS yielded less protein and more total antibody-reactive material than did the other methods. We found that, after sonication, BOT-2 cells, replated in growth medium, were still viable. In fact, enumeration studies showed that almost all the eells after sonication were viable and could be replated to produce more breast tumor cell antigens. We used gel filtration chromatography to purify the antibody-reactive material from BOT-2 cells. Antigens were extraeted by sonication and then eoncentrated as described in "Materials and Methods." Appropriate amounts of antigen were chromatographed with PBS on Sephadex G-50 medium at 4° C. The results are shown in text-figure lA. Extracts from sonically vibrated BOT-2 cells separated on Sephadex G-50 yielded two peaks of material that absorbed light at 280 nm. The first peak, in the excluded volume, determined with dextran blue , contained material that reacted with antiserum against living BOT-2 cells. The second peak, in the included volume , contained about one-half the extracted protein but did not contain any antibodyreactive material. When the antigen material was chromatographed on Sephadex G-I00, the absorbency and antibody-reaeting profiles were similar to the profiles obtained with the use of Sephadex G-50. T 0 obtain a better separation of BOT-2 cell antigen, extracts from sonically vibrated, centrifuged cells were separated on Sephadex G-200. The profiles (text-fig. IB) indicated that three peaks of material absorbed light at 280 nm. The first peak, in the excluded volume, contained about one-third the extracted proteins but only trace amounts of material reacting with antiserum against BOT-2 cells. The last peak, weIl within the included volume, did not contain any antibody-reactive material. The second pcak, eluted after the excluded volume, contained about one-third the protein and almost all the material that reacted with antiserum against BOT-2 cells. Finally, BOT-2 cell antigens were chromato-
VOL. 60, NO. 1 JANUARY 1978
Downloaded from http://jnci.oxfordjournals.org/ at Dalhousie University on December 11, 2012
published procedures (13, 14). Gels were 6x70 mm and contained 7.5% acrylamide and a 3% acrylamide stacker. Gel buffer (pH 8.3) consisted of 0.025 M tris0.192 M glycine or 0.1% SDS in tris-glycine. Sampies for SDS-PGE were heated at 100° C for 2 minutes. Sampies in 10% glycerol were applied to the gels, which were electrophoresed at 3 mA/gel until the tracking dye was at the bottom of the gel. Gels were stained for 2 hours at 37° C in 0.3% Coomassie brilliant blue in methanol:acetic acid:water (50:7:43) and destained by diffusion in methanol:acetic acid:water (10:7:83). To loeate protein bands, destained gels were scanned in a Gilford spectrophotometer equipped with a 20-em linear transport. Antiserum.-To prepare antiserum against living BOT-2 ceIls, 106 washed cells were mixed with an equal volume of Freund's complete adjuvant and injected into the flanks of New Zealand White rabbits. Rabbits received three weekly injections and were bled 4 weeks after the first injection. The antiserum was chromatographed on DEAE-Sephadex A-50 in PBS to isolate the y-globulin fraction, which was then fully absorbed with cross-linked FCS, human tissue acetone extracts, and packed HeLa cells (15). The antiserum reacted only against BOT-2 cells and breast tumor tissue. No fluorescence was observed when the antiserum was tested against HeLa ceIls, human melanoma ceIls, alveolar carcinoma ceIls, or normal ductal cells within a breast tissue specimen. Immunofluorescence techniques.-For FCF, washed BOT-2 ceIls were fixed in acetone, dried, and rehydrated with PBS. The eeIls, plus a drop of patients' sera, were in cubated for 60 minutes at 23° C and then rinsed with three changes of PBS. The ceIls were then coupled with a 1:10 dilution of FITC-tagged goat anti-human y-globulins for 60 minutes, washed in PBS, and mounted in buffered glycerin. For LCMF, washed BOT-2 ceIls were incubated with an equal volume of patients' sera for 60 minutes at 23° C. The ceIls were then washed , coupled to FITC, and mounted as described for FCF. Sampies were studied on an Olympus UV microscope with the use of FITC interference filter and a Y-52 barrier filter. Less than 5% of serum samples from patients with malignant melanoma or alveolar cell carcinoma reacted with BOT-2 ceIls. This value of reactivity was equal to that of the reactivity of normal control patients' sera. Indirect HA.-For HA studies, BOT-2 cell antigen, after sonication and high-speed centrifugation, was coupled to tanned sheep or human group 0 erythrocytes. Washed erythrocytes (2.5% suspension) were mixed with an equal volume of 0.0294 mM tannic acid and incubated for 10 minutes at 37° C (16). The cells were washed with PBS (pH 6.4), incubated for 10 minutes at 23° C with 1 mg of BOT-2 antigen, and washed in PBS containing 1% normal rabbit serum (16). Serial twofold dilutions of patients' sera (undiluted to diluted 1:256) were added to an equal volume of 0.5% sensitized erythrocytes, and the mixture was incubated at 23° C for 2 hours. HA was determined by visual inspection. The average maximum dilution of patients' sera show-
41
BREAST TUMOR CELL SURFACE ANTIGENS
.2
.1
E
c
B
0
co .2 N
-
E
IV
c:::
>.
u
o
c
IV .&:J
~
LO LO
.1
III
.&:J
+-'
.
c:::
~
.c
o
.05
C
Cf)
.c -c
C1>
>
10
30
50
Fraction Number TEXT-FIGURE I.-Gel filtration chromatography of BOT-2 cell antigens. A) Sephadex G-50 medium (35X 1.3 cm) with 5 mg of antigen; I-mI fractions were collected. B) Sephadex G-200 (43x 1.5 cm) with 8 mg of antigen; 3-ml fractions were collected. C) Sephadex G-I50 (49x2.7 cm) with 15 mg of antigen; 3-ml fractions were collected. The ! indicates the excluded volume as determined with dextran blue. The + indicates fractions reacting with rabbit antiserum against living BOT-2 cells as determined by a standard Ouchterlony precipitin test.
graphed on Sephadex G-150 (text-fig. lC). The results were similar to those seen with the separation obtained with Sephadex G-200. The excluded peak from Sephadex G-150 contained only trace amounts of antibodyreactive material, whereas the second, included peak had almost all the antibody-reactive material. The results from these experiments showed that chromatography of BOT-2 cell extracts on Sephadex G-150 or G200 yielded an included peak of proteins that contained almost all the material that reacted with antiserum against living BOT-2 cells but substantially less protein than the original antigen extract. Protein fractions extracted from BOT-2 cells and purified by chromatography were analyzed by PGE to determine the number of proteins in the antigen preparation before and after Sephadex purification. We used nondenaturing PGE to analyze antigens extracted from BOT-2 and HeLa cells (text-fig. 2). Extracts from HeLa cells (text-fig. 2A) had two intenselv staining bands of material and several lighter staining bands. Extracts from BOT-2 cells (text-fig. 2B) had one intensely staining band and severallighter staining bands. Comparison of the PGE profiles from BOT-2 and HeLa cells clearly showed that these profiles differed. A PGE profile of the antibody-reactive peak from BOT-2 cells after chromatography on Sephadex G-200 is shown in text-figure 2C. This peak contained an intensely stainVOL. 60, NO. 1, JANUARY 1978
+:i
~
Q)
0::
Migration Distance .. TEXT-FIGURE 2.-PGE profiles of antigens from BOT-2 and HeLa cells. A) HeLa cells; B) BOT-2 cells; C) Sephadex G-200 antibodyreactive peak from BOT-2 cells; D) Sephadex G-200 excluded peak from BOT-2 cells. Gels were 7.5% acrylamide and were run in trisglycine buffer, pH 8.3. Procedures for antigen extraction, chromatography, and PGE are described in "Materials and Methods" and in text-fig. 1. Approximately 50 p,g of protein was applied to each gel.
ing band and several lighter staining bands. The position of the intensely staining band from Sephadex G200 had the same mobility as the intensely staining band observed in the nonchromatographed BOT-2 cell extracts (text-fig. 2B). HeLa cell extracts did not have an intensely staining band in the same position as observed in BOT-2 cells. A profile of the excluded peak obtained after chromatography of BOT-2 cell extracts on Sephadex G-200 is shown in text-figure 2D. As expected, this sample contained some protein bands found in the original BOT-2 cell extracts. However, this fraction did contain a lightly staining band migrating at the same position as the intensely staining, antibody-reactive band shown in text-figure 2C. The excluded fraction contained a small amount of antibodyJ
NATL CANCER INST
Downloaded from http://jnci.oxfordjournals.org/ at Dalhousie University on December 11, 2012
o
42
LERNER, ANGLIN , AND NORDQUIST
reactive material as determined by immunoprecipitation. We next analyzed the crude and purified BOT-2 cell extracts with the use of SDS-PGE. A profile of the original extract, after sonication and high-speed centrifugation, is shown in text-figure 3A. The original extract
A
B
0
LO LO +-'
«J ~
U
C «J
.J:) ~
0
(J)
.J:)
ce Cl)
>
TABLE
~
«J
1.-Immunologic tests for antibodies against BOT-2 cell antigens a
Q) 0::
Results of tests with the use of: FCF
0
Diseases
None-control Fibroadenoma Cystic mastitis Mammary cancer
Migration Distance.. TEXT-FIGURE 3.-SDS-PGE profiles of antigens from BOT-2 cells. A) BOT-2 cells; B) Sephadex G-150 excluded peak from BOT-2 cells; C) Sephadex G-150 antibody-reactive peak from BOT-2 cells; D) Sephadex G-200 antibody-reactive peak from BOT-2 cells. Gels were 0.1% SDS-7.5% acrylamide and were run in 0.1% SDS, trisglycine buffer, pH 8.3. Procedures for antigen extraction, chrornatography, and PGE are described in "Materials and Methods" and in text-fig. 1. Approximately 50 /Lg of protein was applied to each gel.
contained many proteins and several intensely staining bands. After chromatography of the BOT-2 cell extract on Sephadex G-150, the excluded peak was found to contain many stained bands (text-fig. 3B). Analysis of the proteinsin the antibody-reactive peak obtained with J NATL CANCER INST
a
LCMF
NO.of patients
Percent positive
21 30 36 68
5 90 56 80
NO.of pa-
tients 22 27 32 61
HA
Percent positive
No.of patients
Percent positive
5 4 41 45
29 29 38 59
10 21 16 46
Sera from women with diagnosed, untreated breast diseases.
age of positive responses in the three tests that used sera from patients with no known breast diseases (controls) ranged from 5 to 10%. For patients with benign breast tumors (fibroadenoma), the positive tests ranged from 4% (LCMF) to 21% (HA), whereas the FCF test indicated 90% positive responses. This high percentage of positive reactions may be indicative of organ-specific reactions without a clear differentiation of tumor specificity. For patients with nonneoplastic breast diseases (cystic mastitis), the HA test indicated only 16% positive, the FCF test 56% positive, and the LCMF test 41% positive. However, in patients with diagnosed mammary carcinoma, at least 45% of all three tests were positive. VOL. 60, NO. 1 JANUARY 1978
Downloaded from http://jnci.oxfordjournals.org/ at Dalhousie University on December 11, 2012
E c
Sephadex G-150 revealed fewer protein bands (text-fig. 3C). A similar profile was obtained after PGE analysis of the material in the antibody-reactive peak after Sephadex G-200 chromatography (text-fig. 3D). These results indicate that there were fewer proteins in the Sephadex-purified, antibody-reactive peaks than in the original BOT-2 ceIl extracts. Concurrent with our biochemical studies on the breast tumor cell antigens, we have started clinical studies to determine whether patients with mammary cancer have circulating antibodies against BOT-2 cell antigens. From a local hospital, we obtained coded sarnples of sera from normal patients and patients with diagnosed fibroadenoma, diagnosed cystic mastitis, and diagnosed but untreated breasr cancer. We used three tests to identify antibodies in patients' sera that reacted against BOT-2 cells or the sonicaIly vibrated, centrifuged extracts from BOT-2 cells. In the FCF test, patients' sera and fluorescent-tagged goat anti-human y-globulins (FITC) were added to acetone-fixed BOT-2 ceIls. For living cell membrane immunofluorescence, patients' sera and FITC were added to washed, living BOT-2 ceIls. The presence (or absence) of fluorescence, which indicated patients' antibodies against BOT-2 cells , was noted after sample observation with a fluorescent microscope. For HA, sonicaIly vibrated, centrifuged extracts from BOT-2 ceIls were attached to tanned erythrocytes . Serum samples from patients were added, and the presence or absence of HA was determined. The results of these experiments are shown in table 1. The percent-
43
BREAST TUMOR CELL SURFACE ANTIGENS
DISCUSSION
VOL. 60, NO. 1, JANUARY 1978
REFERENCES (l) SINKOVICS JG, REEVES wj, CABINESS JR: Cell- and antibodymediated immune reactions of patients to cultured cells of breast carcinoma. J Natl Cancer Inst 48: 1145-1149, 1972
J
NATL CANCER INST
Downloaded from http://jnci.oxfordjournals.org/ at Dalhousie University on December 11, 2012
Currently, little is known about the properties of the mammary cancer cell and its interrelationship with the host during the neoplastic process. There are, however, several re ports dealing with antigens extracted from whole breast tumor tissue. A human breast tumor antigen was isolated after an antigen-antibody complex of tissue extracts and immunoglobulins was separated by Sephadex G-200 chromatography (10). The antigen appeared to be in the included volume after Sephadex chromatography. In another study, an extract from a breast carcinoma, obtained from a liver metastases, was separated by gel filtration chromatograp~y (11). The isolated antigen material, which reacted with sera from breast cancer patients, also appeared in the included volume after Sephadex G-200 gel filtration. An
ated with this human breast tumor cell line, BOT-2, and the reactions of patients' sera against these cells and their extracted antigens.
ABSTRACT-Biochemical and immunologie studies on breast cancer with the use of cells from a human ductal cell carcinoma, BOT-2, were initiated. Antigens were extracted from the cells by mild sonication and purified by gel filtration chromatography. Only one of the three peaks from gel filtration chromatography reacted with antiserum prepared against whole BOT-2 cells. Analysis by polyacrylamide gel electrophoresis of the BOT-2 cell extract revealed many protein bands, whereas analysis of the antibody-reactive peak after gel filtration chromatography revealed fewer protein bands. Immunologie tests to identify human serum antibodies agalnst BOT-2 cells or cell extracts were performed by fixed cell immunofluorescence, Iiving cell membrane immunofluorescence, and indirect hemagglutination. Depending on the test, the sera from women with diagnosed, untreated mammary cancer were positive in 45-80% of the cases, whereas the sera from women without apparent breast diseases (controls) were positive in only 5-10% of the cases. The results suggested that the antigens from the BOT-2 cells will be useful in understanding the processes involved in human mammary neoplasia. - J Natl Cancer Inst 60: 39-44, 1978.
MATERIALS AND METHODS
Breast cancer is the greatest cause of death, due to cancer, among American women. Although surgical and chemotherapeutic methods are available to treat women with breast cancer, the mortality rate is still distressingly high. Part of the reason for the high mortality rate is the lack of information about specific biochemical and immunologie events that occur during the neoplastic process. However, tumor-specific antigens and antibodies associated with mammary cancer have been demonstrated by lymphocyte cytotoxicity and blocking (1, 2), lymphocyte blastogenesis (3), leukocyte migration inhibition (4, 5), delayed hypersensitivity reactions (6, 7), and circulating tumor-specific antibodies (8-11). Antigens from neoplastic mammary tissue have been extracted, purified, and used in a skin testing procedure (7). All the above data, plus information from studies on other tumor systems, indicate that tumor cells have distinct antigens which may help the tumor cell metastasize or avoid immunologie destruction. Despite recent re ports concerning antigens and antibodies associated with human breast cancer, knowledge about the nature of the antigens is lacking. Part of this problem can be attributed to difficulties associated with antigen extraction from tissue and to the small amounts of antigens recovered from whole tumor tissue. The use of cultured tumor cells to isolate tumor antigens provides a greater amount of uniform antigens than is available from whole tumor tissue. Nordquist et al. (12) recently described the establishment and long-term cultivation of a human ductal cell carcinoma. Here we re port our results from studies of the antigens associVOL. 60, NO. 1, JANUARY 1978
ABBREVIATIONS USED: FCS = fetal calf serum; PBS = phosphatebuffered saline; PGE = polyacrylamide gel electrophoresis; SDS = sodium dodecyl sulfate; FCF = fixed cell immunofluorescence; FITC = fluorescein isothiocyanate; LCMF = living cell membrane immunofluorescence; HA = hemagglutination. Received March 15, 1977; accepted August 1,1977. Supported by grant BC-230 from the American Cancer Society and a Maizie Wilkonson Award (M.P.L.) and in part by Public Health Service grants GRS 5-S01-RR05411 from the Division of Research Resources, National Institutes of Health, and a grant from the Oklahoma Heart Association. 3 Department of Microbiology and Immunology, University of Oklahoma Health Seiences Center, Oklahoma City, Okla. 73190. 4 Department of Biochemistry and Molecular Biology, University of Oklahoma Health Seiences Center. 5 Cancer Research Program , Oklahoma Medical Research Foundation, Oklahoma City, Okla. 73104. 6 Present address: Department of Anatomical Sciences, University of Oklahoma Health Seiences Center. 7 We thank Richard Strecker, Peggy Riggs, Peggy Munson, and Lucy Durbin for excellent technical assistance. 1
2
39
J
NATL CANCER INST
Downloaded from http://jnci.oxfordjournals.org/ at Dalhousie University on December 11, 2012
Cell culture and antigen extraction. -BOT-2 cells were grown in Eagle's minimum essential medium containing 10% heat-inactivated FCS. The morphology and growth properties of this cell line have been described (12). The growth medium from 10 confluent 75-cm 2 flasks was decanted, and the cells were resuspended in 3 ml of PBS (0.15 M NaCI, 0.01 M Na+, K+ P0 4 , pH 7.1). The cells were washed by low-speed centrifugation and resuspended in 5 ml of PBS. Antigens were extracted from the cells by sonication in a tank sonifier (40 mA) for 60 minutes at 4° C. The sonicate was centrifuged at 1,500Xg for 10 minutes to pellet the cells, and the supernatant was centrifuged at 140,000Xg for 60 minutes to remove large aggregates. The resulting supernatant was used to study the properties of the BOT-2 cell antigens. Antigen purification and analysis.-The supernatant was concentrated to about 5 mg protein/rnl by Sephadex G200 hydration. This sampIe was then applied to Sephadex G-50, G-I00, G-150, and G-200 columns equilibrated at 4° C with PBS. Fractions were collected, and the absorbency at 280 nm was determined in a Gilford spectrophotometer (Gilford Instrument Laboratories , Inc., Oberlin, Ohio). Contiguous fractions of absorbing material were pooled, dialyzed against several changes of water, and lyophilized. The lyophilized sampIes were dissolved in PBS and analyzed by PGE according to
40
LERNER, ANGLIN , AND NORDQUIST
J
NATL CANCER INST
ing positive HA was 1:8. For control experiments, FCS or HeLa cell extracts were coupled to tanned erythrocytes. For all experiments, proteins were estimated by the absorbeney ratio at 260 and 280 nm (17) . RESULTS Antigen Extraction and Purification
We were interested in learning whether the BOT-2 ceIls synthesized surface antigens useful for the study of many aspects of human mammary cancer. Three different methods were used to obtain antigen material from BOT-2 ceIls: 1) extraction of antigens from eells with the use of 3 M KCI, 2) extraction with I-butanol, and 3) extraction by sonication in PBS. We monitored the extraction procedures by protein determination and by immunoelectrophoresis with the use of antiserum prepared against living BOT-2 eells. We found that extraction of antigens from the cells with 3 M KCI or 1butanol yielded large amounts of cellular protein containing little antibody-reactive material. Extraction of antigens from cells by mild sonication at 4° C in PBS yielded less protein and more total antibody-reactive material than did the other methods. We found that, after sonication, BOT-2 cells, replated in growth medium, were still viable. In fact, enumeration studies showed that almost all the eells after sonication were viable and could be replated to produce more breast tumor cell antigens. We used gel filtration chromatography to purify the antibody-reactive material from BOT-2 cells. Antigens were extraeted by sonication and then eoncentrated as described in "Materials and Methods." Appropriate amounts of antigen were chromatographed with PBS on Sephadex G-50 medium at 4° C. The results are shown in text-figure lA. Extracts from sonically vibrated BOT-2 cells separated on Sephadex G-50 yielded two peaks of material that absorbed light at 280 nm. The first peak, in the excluded volume, determined with dextran blue , contained material that reacted with antiserum against living BOT-2 cells. The second peak, in the included volume , contained about one-half the extracted protein but did not contain any antibodyreactive material. When the antigen material was chromatographed on Sephadex G-I00, the absorbency and antibody-reaeting profiles were similar to the profiles obtained with the use of Sephadex G-50. T 0 obtain a better separation of BOT-2 cell antigen, extracts from sonically vibrated, centrifuged cells were separated on Sephadex G-200. The profiles (text-fig. IB) indicated that three peaks of material absorbed light at 280 nm. The first peak, in the excluded volume, contained about one-third the extracted proteins but only trace amounts of material reacting with antiserum against BOT-2 cells. The last peak, weIl within the included volume, did not contain any antibody-reactive material. The second pcak, eluted after the excluded volume, contained about one-third the protein and almost all the material that reacted with antiserum against BOT-2 cells. Finally, BOT-2 cell antigens were chromato-
VOL. 60, NO. 1 JANUARY 1978
Downloaded from http://jnci.oxfordjournals.org/ at Dalhousie University on December 11, 2012
published procedures (13, 14). Gels were 6x70 mm and contained 7.5% acrylamide and a 3% acrylamide stacker. Gel buffer (pH 8.3) consisted of 0.025 M tris0.192 M glycine or 0.1% SDS in tris-glycine. Sampies for SDS-PGE were heated at 100° C for 2 minutes. Sampies in 10% glycerol were applied to the gels, which were electrophoresed at 3 mA/gel until the tracking dye was at the bottom of the gel. Gels were stained for 2 hours at 37° C in 0.3% Coomassie brilliant blue in methanol:acetic acid:water (50:7:43) and destained by diffusion in methanol:acetic acid:water (10:7:83). To loeate protein bands, destained gels were scanned in a Gilford spectrophotometer equipped with a 20-em linear transport. Antiserum.-To prepare antiserum against living BOT-2 ceIls, 106 washed cells were mixed with an equal volume of Freund's complete adjuvant and injected into the flanks of New Zealand White rabbits. Rabbits received three weekly injections and were bled 4 weeks after the first injection. The antiserum was chromatographed on DEAE-Sephadex A-50 in PBS to isolate the y-globulin fraction, which was then fully absorbed with cross-linked FCS, human tissue acetone extracts, and packed HeLa cells (15). The antiserum reacted only against BOT-2 cells and breast tumor tissue. No fluorescence was observed when the antiserum was tested against HeLa ceIls, human melanoma ceIls, alveolar carcinoma ceIls, or normal ductal cells within a breast tissue specimen. Immunofluorescence techniques.-For FCF, washed BOT-2 ceIls were fixed in acetone, dried, and rehydrated with PBS. The eeIls, plus a drop of patients' sera, were in cubated for 60 minutes at 23° C and then rinsed with three changes of PBS. The ceIls were then coupled with a 1:10 dilution of FITC-tagged goat anti-human y-globulins for 60 minutes, washed in PBS, and mounted in buffered glycerin. For LCMF, washed BOT-2 ceIls were incubated with an equal volume of patients' sera for 60 minutes at 23° C. The ceIls were then washed , coupled to FITC, and mounted as described for FCF. Sampies were studied on an Olympus UV microscope with the use of FITC interference filter and a Y-52 barrier filter. Less than 5% of serum samples from patients with malignant melanoma or alveolar cell carcinoma reacted with BOT-2 ceIls. This value of reactivity was equal to that of the reactivity of normal control patients' sera. Indirect HA.-For HA studies, BOT-2 cell antigen, after sonication and high-speed centrifugation, was coupled to tanned sheep or human group 0 erythrocytes. Washed erythrocytes (2.5% suspension) were mixed with an equal volume of 0.0294 mM tannic acid and incubated for 10 minutes at 37° C (16). The cells were washed with PBS (pH 6.4), incubated for 10 minutes at 23° C with 1 mg of BOT-2 antigen, and washed in PBS containing 1% normal rabbit serum (16). Serial twofold dilutions of patients' sera (undiluted to diluted 1:256) were added to an equal volume of 0.5% sensitized erythrocytes, and the mixture was incubated at 23° C for 2 hours. HA was determined by visual inspection. The average maximum dilution of patients' sera show-
41
BREAST TUMOR CELL SURFACE ANTIGENS
.2
.1
E
c
B
0
co .2 N
-
E
IV
c:::
>.
u
o
c
IV .&:J
~
LO LO
.1
III
.&:J
+-'
.
c:::
~
.c
o
.05
C
Cf)
.c -c
C1>
>
10
30
50
Fraction Number TEXT-FIGURE I.-Gel filtration chromatography of BOT-2 cell antigens. A) Sephadex G-50 medium (35X 1.3 cm) with 5 mg of antigen; I-mI fractions were collected. B) Sephadex G-200 (43x 1.5 cm) with 8 mg of antigen; 3-ml fractions were collected. C) Sephadex G-I50 (49x2.7 cm) with 15 mg of antigen; 3-ml fractions were collected. The ! indicates the excluded volume as determined with dextran blue. The + indicates fractions reacting with rabbit antiserum against living BOT-2 cells as determined by a standard Ouchterlony precipitin test.
graphed on Sephadex G-150 (text-fig. lC). The results were similar to those seen with the separation obtained with Sephadex G-200. The excluded peak from Sephadex G-150 contained only trace amounts of antibodyreactive material, whereas the second, included peak had almost all the antibody-reactive material. The results from these experiments showed that chromatography of BOT-2 cell extracts on Sephadex G-150 or G200 yielded an included peak of proteins that contained almost all the material that reacted with antiserum against living BOT-2 cells but substantially less protein than the original antigen extract. Protein fractions extracted from BOT-2 cells and purified by chromatography were analyzed by PGE to determine the number of proteins in the antigen preparation before and after Sephadex purification. We used nondenaturing PGE to analyze antigens extracted from BOT-2 and HeLa cells (text-fig. 2). Extracts from HeLa cells (text-fig. 2A) had two intenselv staining bands of material and several lighter staining bands. Extracts from BOT-2 cells (text-fig. 2B) had one intensely staining band and severallighter staining bands. Comparison of the PGE profiles from BOT-2 and HeLa cells clearly showed that these profiles differed. A PGE profile of the antibody-reactive peak from BOT-2 cells after chromatography on Sephadex G-200 is shown in text-figure 2C. This peak contained an intensely stainVOL. 60, NO. 1, JANUARY 1978
+:i
~
Q)
0::
Migration Distance .. TEXT-FIGURE 2.-PGE profiles of antigens from BOT-2 and HeLa cells. A) HeLa cells; B) BOT-2 cells; C) Sephadex G-200 antibodyreactive peak from BOT-2 cells; D) Sephadex G-200 excluded peak from BOT-2 cells. Gels were 7.5% acrylamide and were run in trisglycine buffer, pH 8.3. Procedures for antigen extraction, chromatography, and PGE are described in "Materials and Methods" and in text-fig. 1. Approximately 50 p,g of protein was applied to each gel.
ing band and several lighter staining bands. The position of the intensely staining band from Sephadex G200 had the same mobility as the intensely staining band observed in the nonchromatographed BOT-2 cell extracts (text-fig. 2B). HeLa cell extracts did not have an intensely staining band in the same position as observed in BOT-2 cells. A profile of the excluded peak obtained after chromatography of BOT-2 cell extracts on Sephadex G-200 is shown in text-figure 2D. As expected, this sample contained some protein bands found in the original BOT-2 cell extracts. However, this fraction did contain a lightly staining band migrating at the same position as the intensely staining, antibody-reactive band shown in text-figure 2C. The excluded fraction contained a small amount of antibodyJ
NATL CANCER INST
Downloaded from http://jnci.oxfordjournals.org/ at Dalhousie University on December 11, 2012
o
42
LERNER, ANGLIN , AND NORDQUIST
reactive material as determined by immunoprecipitation. We next analyzed the crude and purified BOT-2 cell extracts with the use of SDS-PGE. A profile of the original extract, after sonication and high-speed centrifugation, is shown in text-figure 3A. The original extract
A
B
0
LO LO +-'
«J ~
U
C «J
.J:) ~
0
(J)
.J:)
ce Cl)
>
TABLE
~
«J
1.-Immunologic tests for antibodies against BOT-2 cell antigens a
Q) 0::
Results of tests with the use of: FCF
0
Diseases
None-control Fibroadenoma Cystic mastitis Mammary cancer
Migration Distance.. TEXT-FIGURE 3.-SDS-PGE profiles of antigens from BOT-2 cells. A) BOT-2 cells; B) Sephadex G-150 excluded peak from BOT-2 cells; C) Sephadex G-150 antibody-reactive peak from BOT-2 cells; D) Sephadex G-200 antibody-reactive peak from BOT-2 cells. Gels were 0.1% SDS-7.5% acrylamide and were run in 0.1% SDS, trisglycine buffer, pH 8.3. Procedures for antigen extraction, chrornatography, and PGE are described in "Materials and Methods" and in text-fig. 1. Approximately 50 /Lg of protein was applied to each gel.
contained many proteins and several intensely staining bands. After chromatography of the BOT-2 cell extract on Sephadex G-150, the excluded peak was found to contain many stained bands (text-fig. 3B). Analysis of the proteinsin the antibody-reactive peak obtained with J NATL CANCER INST
a
LCMF
NO.of patients
Percent positive
21 30 36 68
5 90 56 80
NO.of pa-
tients 22 27 32 61
HA
Percent positive
No.of patients
Percent positive
5 4 41 45
29 29 38 59
10 21 16 46
Sera from women with diagnosed, untreated breast diseases.
age of positive responses in the three tests that used sera from patients with no known breast diseases (controls) ranged from 5 to 10%. For patients with benign breast tumors (fibroadenoma), the positive tests ranged from 4% (LCMF) to 21% (HA), whereas the FCF test indicated 90% positive responses. This high percentage of positive reactions may be indicative of organ-specific reactions without a clear differentiation of tumor specificity. For patients with nonneoplastic breast diseases (cystic mastitis), the HA test indicated only 16% positive, the FCF test 56% positive, and the LCMF test 41% positive. However, in patients with diagnosed mammary carcinoma, at least 45% of all three tests were positive. VOL. 60, NO. 1 JANUARY 1978
Downloaded from http://jnci.oxfordjournals.org/ at Dalhousie University on December 11, 2012
E c
Sephadex G-150 revealed fewer protein bands (text-fig. 3C). A similar profile was obtained after PGE analysis of the material in the antibody-reactive peak after Sephadex G-200 chromatography (text-fig. 3D). These results indicate that there were fewer proteins in the Sephadex-purified, antibody-reactive peaks than in the original BOT-2 ceIl extracts. Concurrent with our biochemical studies on the breast tumor cell antigens, we have started clinical studies to determine whether patients with mammary cancer have circulating antibodies against BOT-2 cell antigens. From a local hospital, we obtained coded sarnples of sera from normal patients and patients with diagnosed fibroadenoma, diagnosed cystic mastitis, and diagnosed but untreated breasr cancer. We used three tests to identify antibodies in patients' sera that reacted against BOT-2 cells or the sonicaIly vibrated, centrifuged extracts from BOT-2 cells. In the FCF test, patients' sera and fluorescent-tagged goat anti-human y-globulins (FITC) were added to acetone-fixed BOT-2 ceIls. For living cell membrane immunofluorescence, patients' sera and FITC were added to washed, living BOT-2 ceIls. The presence (or absence) of fluorescence, which indicated patients' antibodies against BOT-2 cells , was noted after sample observation with a fluorescent microscope. For HA, sonicaIly vibrated, centrifuged extracts from BOT-2 ceIls were attached to tanned erythrocytes . Serum samples from patients were added, and the presence or absence of HA was determined. The results of these experiments are shown in table 1. The percent-
43
BREAST TUMOR CELL SURFACE ANTIGENS
DISCUSSION
VOL. 60, NO. 1, JANUARY 1978
REFERENCES (l) SINKOVICS JG, REEVES wj, CABINESS JR: Cell- and antibodymediated immune reactions of patients to cultured cells of breast carcinoma. J Natl Cancer Inst 48: 1145-1149, 1972
J
NATL CANCER INST
Downloaded from http://jnci.oxfordjournals.org/ at Dalhousie University on December 11, 2012
Currently, little is known about the properties of the mammary cancer cell and its interrelationship with the host during the neoplastic process. There are, however, several re ports dealing with antigens extracted from whole breast tumor tissue. A human breast tumor antigen was isolated after an antigen-antibody complex of tissue extracts and immunoglobulins was separated by Sephadex G-200 chromatography (10). The antigen appeared to be in the included volume after Sephadex chromatography. In another study, an extract from a breast carcinoma, obtained from a liver metastases, was separated by gel filtration chromatograp~y (11). The isolated antigen material, which reacted with sera from breast cancer patients, also appeared in the included volume after Sephadex G-200 gel filtration. An
