- Email: [email protected]
Culture supernatants of lymphokine-activated killer (LAK) cells contain a high-molecular-weight cytotoxic lymphokine
Immunology Letters, 23 (1989/1990) 2 7 - 3 0 Elsevier IMLET 01303
Culture supernatants of lymphokine-activated killer (LAK) cells contain a high-molecular-weight cytotoxic lymphokine Sarit Larisch-Bloch, Barbara Sterman, David Weiss and Dov Sulitzeanu The Lautenberg Center f o r General and TUmor Immunology, The Hebrew University Medical School, Jerusalem, Israel (Received 8 June 1989; revision received and accepted 25 July 1989)
1. Summary
Supernatants of human lymphokine-activated killer (LAK) cells grown in vitro were tested for cytotoxic activity against several mouse and human neoplastic cell lines. All LAK preparations tested (14/14) exhibited cytotoxic activity (40-90°7o killing of the target cells). Sephacryl S-300 Gel filtration experiments indicated that the biological activity of the LAK supernatant is associated with molecular moities ranging from 800 kDa or more, to less than 10 kDa. The finding of strong cytotoxic activity in LAK supernatants against several tumor lines points to the possibility of employing soluble products of these cells, rather than the living cells themselves, for therapeutic purposes. 2. Introduction
Lymphokine-activated killer (LAK) cells have been shown to exert strong anti-tumor activity in adoptive immunotherapy experiments in laboratory animals, and to bring about tumor regression in some cancer patients [1]. The mechanism by which LAK cells function therapeutically remains uncertain. It has been reported that only a very small proportion of the injected LAK cells in fact reach tumor sites [2, 3]. It seems possible, accordingly, that LAK cells act largely indirectly, by releasing in
Key words: LAK cell; Cytotoxic lymphokine Correspondence to: D. Sulitzeanu, Lautenberg Center for General and Tumor Immunology, P.O.B. 1172, Jerusalem 91010, Israel.
vivo soluble growth-inhibiting and/or cytotoxic factors; findings recently reported by Hersh et al. in a preliminary communication [4] support this assumption. We have undertaken a study designed to determine the presence and nature of tumor cytotoxic factors released by cultured LAK cells into the medium. We describe herewith the results of a first series of experiments which demonstrate the in vitro cytotoxicity of human LAK supernatants for a variety of human and mouse tumor cell lines. Fractionation of the supernatants by gel filtration and ultracentrifugation has shown that the cytotoxic activity is associated with molecules over a wide range of molecular weight, including a high-molecularweight substance of more than 900 kDa. Cytotoxic molecules of this size have not previously been described. 3. Materials and Methods 3.1. Cell lines
The following cell lines were used as targets in the cytotoxicity texts: HeLa (human cervical carcinoma), MDA 231 (human breast carcinoma), Raji and DG-75 (EBV-positive and negative Burkitt's lymphoma, respectively), A9 (murine fibrosarcoma), M109 (murine pulmonary carcinoma), P815 (murine mastocytoma). Cells were cultured in RPMI or Dulbecco's modified Eagle's medium (DMEM) with 10o7o fetal calf serum (FCS), streptomycin sulfate (100/~g/ml and penicillin (100 U/ml). Cells were examined to ascertain lack of mycoplasma contamination. Infected
0165-2478 / 89 / $ 3.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
27
cells were freed of mycoplasma by treatment with BM-cycline (Boehringer, Mannheim), according to the manufacturer's instructions.
used to evaluate the significance of the data.
4.
Results
3.2. Preparation of L A K supernatants LAK supernatants were prepared by growing human peripheral blood lymphocytes (PBL) from healthy donors in RPMI + 10% FCS, with 1000 U/ml recombinant IL-2, (Cetus Corp., Emeryville) as described [1]. FCS was omitted in some experiments, to obtain LAK supernatants in serum free medium (LAK-SFM). 3.3. Cytotoxicity assay The cytotoxic activity of the LAK cell supernatants was tested by means o f a terminal 51Cr labeling assay [6, 7] as follows: cells were added to 96-well microtiter plates (15 000 cells in 0.1 ml medium). After 24 h in culture, the medium was removed and fresh medium added, followed by various dilutions of the LAK supernatant (0.1 ml) in triplicates. Eighteen wells (6 groups of 3 wells each) which served as controls were filled with medium rather than LAK supernatant. After incubation for 72 h, the medium was removed and the ceils washed. 30/~1 of sodium 51chromate (0.1 mCi/ml; Nuclear Research Center, Negev, Israel) was added to each well and the plates were incubated at 37°C for 90 min. Only the living cells incorporate 51Cr. The cells were washed again and dissolved in 0.2 ml 1 M NaOH, then transferred into tubes for counting in a gamma counter. Percent cytotoxicity was calculated according to the formula: 1
-
mean cpm in test sample mean cpm in controls
A total of 8 LAK supernatants prepared in medium with FCS, and 6 supernatants prepared in parallel in SFM, were tested. All of these showed significant cytotoxicity in repeated experiments (at least two cell lines tested with each preparation). Fig. 1 shows the results of several experiments carried out with LAK supernatants (tested on A9 and HeLa cells, 3 experiments with each cell line) or with LAK-SFM (4 experiments with the HeLa cells). The LAK-SFM preparation retained its cytotoxicity after lyophilization, as seen in the two experiments performed with P815 cells. Very good cytotoxicity was obtained in all cases, ranging from 40-90°7o.
4.1. Gel filtration analysis Gel filtration analysis was performed 5 times on 4 different LAK supernatants, with identical patterns of protein distribution (Fig. 2). The first highmolecular-weight fraction (F1) contained molecules larger than 900 kDa. When a 970-kDa IgM marker and a 2000-kDa dextran blue marker were run on the same Sephacryl S-300 column, F1 was seen to be located between the IgM and dextran blue peaks. As seen in Fig. 3, F1 displayed clear cytotoxicity (15-45o70), although the effect was somewhat less -e'c
A9
-ec
100 8O
x
100
3.4. Fractionation of L A K supernatant
>, c, o
~.,
6O 40
<.9
LAK supernatants pretested for cytotoxic activity were separated by gel filtration on a large (110 × 3 cm) Sephacryl S-300 column. Fractions of 2 - 3 ml were collected, dialyzed against culture medium (RPMI) and tested for cytotoxic reactivity. 3.5. Statistical analysis The non-parametric Mann-Whitney U test was 28
2O
LAK supernatont with FCS
LAK supernatant without FCS
Fig. 1. Cytotoxic activity of LAK supernatants against 4 tumor cell lines. LAK preparations were tested at 1/4 dilution. Data shown are from repeat experiments with LAK supernatants prepared with FCS (HeLa, A9), with SFM (HeLa) or with lyophilized, reconstituted SFM.
280 O.D 0.36
CPM
0.30
o---.~
supernatont IgG- 1125
~l L , '
3600
o.24
;tv ; ,L jl?, p
o.1a
I,
,, f,
3000
~,"', :
0.12 O.lO;
2400
laOO
Jl
0.07 0.04
.!
~: ;', I:
12oo
',
,
600
i 0.01
0 FR1
Fig. 2. Analysis of LAK supernatants on Sephacryl S-300. The gel separation was performed 5 times on 4 different LAK supernatants. Protein concentrations were determined by spectrophotometry at 280 nm. The protein pattern shown is from one representative run. All LAK preparations yielded identical patterns of protein distribution.
A9
Hela
DG75
P815
I00 80 x
2 2
60 40
o~
}'t
20 0
A
[3
Fig. 3. Cytotoxic activity of LAK supernatant Fraction 1, recovered by gel filtration on Sephacryl S-300. (A) Fraction 1 of LAK supernatant. (B) Fraction 1 of lyophilized, reconstituted LAK supernatant.
than that obtained with the original, nonfractionated supernatants. Statistical analysis confirmed that all cytotoxicity data were highly significant ( P < 0.05-0.005).
5. Discussion
Adoptively transferred lymphokine-activated killer (LAK) cells are endowed with appreciable antitumor activity, as demonstrated in animal immunotherapy experiments and by the tumor regres-
sion in some cancer patients so treated [1]. The mechanism of LAK anti-tumor activity in vivo is as yet unclear. Recent studies have indicated that only a very small proportion of the injected LAK cells reach tumor sites. It seems reasonable to assume, accordingly, that the systemic effects exerted by these cells are mediated largely by soluble cytotoxic substances they release. To test this possibility, we examined the supernatants of cultured LAK cells for cytotoxic activity in vitro against several neoplastic cell lines of human and murine origin. The results showed conclusively that all LAK preparations tested were cytotoxic. To date, only one preliminary report has appeared describing the release by cultured LAK cells of a cytotoxic/cytostatic factor [4], although secretion of a variety of cytotoxic molecules by activated cytotoxic T-cells has been reported extensively [8, 9]. To identify and characterize the soluble LAK cytotoxic factor(s), we fractionated the LAK supernatants on a Sephacryl S-300 gel filtration column. All fractions tested exhibited cytotoxic activity against the target tumor cell lines. Since the known lymphocyte-produced cytotoxins have molecular weights lower than 140000, with the exception of the 200-kDa lymphotoxin described by Granger et al. [5], we focused our attention on the high-molecular-weight fraction, F1, which contains molecules larger than 900 kDa. This fraction proved cytotoxic against all target cells tested in numerous experiments. Cytotoxic preparations were tested routinely at several dilutions and, in most cases, the effects were dose-dependent. However, irregular dose-response curves were obtained occasionally, suggesting the co-presence of substances possessing growthstimulating or cytotoxicity-inhibitory activity, or both. Because of this, only results obtained at one dilution (generally 1/4) are shown in this report. The extensive clinical trials conducted with intact LAK cells have yielded only very limited therapeutic gains. Tumor regression has been documented in only a small proporation of advanced cancer patients, and is often restricted in extent and duration. Serious side-effects of treatment with LAK cells and IL-2 are common, and some patients have succumbed to these. There are, moreover, formidable logistic difficulties involved in obtaining the large numbers of LAK cells required for treatment, and 29
the costs o f such treatment are correspondingly large. I f it can be shown that the a n t i - t u m o r activity o f L A K ceils indeed resides with the soluble products they release, it might be h o p e d that these substances will afford a logistically and economically more feasible therapeutic modality, one that can be better standardized, and that might prove less toxic and more efficacious.
Acknowledgements We gratefully acknowledge the support o f The Society o f Research Associates o f the Lautenberg Center, The C o n c e r n Foundation, Los Angeles, and the Miriam Jaffe Research Grant f r o m the Israel Cancer Association. We also t h a n k Ms. Judith Blakemore, Cetus C o r p o r a t i o n , Emeryville, USA,
30
for a generous gift o f recombinant IL-2, and Dr. N. Grover for help with the statistical analysis.
References [1] Rosenberg, S. A. (1988) Immunol. Today 9, 58-62. [2] Maghazachi, A. A., Herberman, R. B., Vujanovic, N. L. and Hiserodt, J. C. (1988) Cell Immunol. 115, 179-194. [3] Herberman, R. B. (1989) J. Clin. Oncol. 7, 1-4. [4] Hersh, E. M., Scuderi, P., Salmon, S., Chong, A. and Grimes, W. (1988) Proc. Am. Assoc. Cancer Res. 29, 1588. [5] Granger, G. A., Yamamoto, R. S., Fair, D. S. and Hiserodt, J. C. (1978) Cell Immunol. 38, 388-402. [6] Neville, M. E. (1987) J. Immunol. Methods 99, 77-82. [7] More, R., Yron, I., Ben Sasson, S. and Weiss, D. W. (1975)Cell Immunol. 15, 382-391. [8] Ware, C.F. and Green, L.M. (1987) Lymphokines 14, 307- 324. [9] Ding-E Young, J. and Liu, C. C. (1988) Immunol. Today 9, 140-144.
Culture supernatants of lymphokine-activated killer (LAK) cells contain a high-molecular-weight cytotoxic lymphokine Sarit Larisch-Bloch, Barbara Sterman, David Weiss and Dov Sulitzeanu The Lautenberg Center f o r General and TUmor Immunology, The Hebrew University Medical School, Jerusalem, Israel (Received 8 June 1989; revision received and accepted 25 July 1989)
1. Summary
Supernatants of human lymphokine-activated killer (LAK) cells grown in vitro were tested for cytotoxic activity against several mouse and human neoplastic cell lines. All LAK preparations tested (14/14) exhibited cytotoxic activity (40-90°7o killing of the target cells). Sephacryl S-300 Gel filtration experiments indicated that the biological activity of the LAK supernatant is associated with molecular moities ranging from 800 kDa or more, to less than 10 kDa. The finding of strong cytotoxic activity in LAK supernatants against several tumor lines points to the possibility of employing soluble products of these cells, rather than the living cells themselves, for therapeutic purposes. 2. Introduction
Lymphokine-activated killer (LAK) cells have been shown to exert strong anti-tumor activity in adoptive immunotherapy experiments in laboratory animals, and to bring about tumor regression in some cancer patients [1]. The mechanism by which LAK cells function therapeutically remains uncertain. It has been reported that only a very small proportion of the injected LAK cells in fact reach tumor sites [2, 3]. It seems possible, accordingly, that LAK cells act largely indirectly, by releasing in
Key words: LAK cell; Cytotoxic lymphokine Correspondence to: D. Sulitzeanu, Lautenberg Center for General and Tumor Immunology, P.O.B. 1172, Jerusalem 91010, Israel.
vivo soluble growth-inhibiting and/or cytotoxic factors; findings recently reported by Hersh et al. in a preliminary communication [4] support this assumption. We have undertaken a study designed to determine the presence and nature of tumor cytotoxic factors released by cultured LAK cells into the medium. We describe herewith the results of a first series of experiments which demonstrate the in vitro cytotoxicity of human LAK supernatants for a variety of human and mouse tumor cell lines. Fractionation of the supernatants by gel filtration and ultracentrifugation has shown that the cytotoxic activity is associated with molecules over a wide range of molecular weight, including a high-molecularweight substance of more than 900 kDa. Cytotoxic molecules of this size have not previously been described. 3. Materials and Methods 3.1. Cell lines
The following cell lines were used as targets in the cytotoxicity texts: HeLa (human cervical carcinoma), MDA 231 (human breast carcinoma), Raji and DG-75 (EBV-positive and negative Burkitt's lymphoma, respectively), A9 (murine fibrosarcoma), M109 (murine pulmonary carcinoma), P815 (murine mastocytoma). Cells were cultured in RPMI or Dulbecco's modified Eagle's medium (DMEM) with 10o7o fetal calf serum (FCS), streptomycin sulfate (100/~g/ml and penicillin (100 U/ml). Cells were examined to ascertain lack of mycoplasma contamination. Infected
0165-2478 / 89 / $ 3.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
27
cells were freed of mycoplasma by treatment with BM-cycline (Boehringer, Mannheim), according to the manufacturer's instructions.
used to evaluate the significance of the data.
4.
Results
3.2. Preparation of L A K supernatants LAK supernatants were prepared by growing human peripheral blood lymphocytes (PBL) from healthy donors in RPMI + 10% FCS, with 1000 U/ml recombinant IL-2, (Cetus Corp., Emeryville) as described [1]. FCS was omitted in some experiments, to obtain LAK supernatants in serum free medium (LAK-SFM). 3.3. Cytotoxicity assay The cytotoxic activity of the LAK cell supernatants was tested by means o f a terminal 51Cr labeling assay [6, 7] as follows: cells were added to 96-well microtiter plates (15 000 cells in 0.1 ml medium). After 24 h in culture, the medium was removed and fresh medium added, followed by various dilutions of the LAK supernatant (0.1 ml) in triplicates. Eighteen wells (6 groups of 3 wells each) which served as controls were filled with medium rather than LAK supernatant. After incubation for 72 h, the medium was removed and the ceils washed. 30/~1 of sodium 51chromate (0.1 mCi/ml; Nuclear Research Center, Negev, Israel) was added to each well and the plates were incubated at 37°C for 90 min. Only the living cells incorporate 51Cr. The cells were washed again and dissolved in 0.2 ml 1 M NaOH, then transferred into tubes for counting in a gamma counter. Percent cytotoxicity was calculated according to the formula: 1
-
mean cpm in test sample mean cpm in controls
A total of 8 LAK supernatants prepared in medium with FCS, and 6 supernatants prepared in parallel in SFM, were tested. All of these showed significant cytotoxicity in repeated experiments (at least two cell lines tested with each preparation). Fig. 1 shows the results of several experiments carried out with LAK supernatants (tested on A9 and HeLa cells, 3 experiments with each cell line) or with LAK-SFM (4 experiments with the HeLa cells). The LAK-SFM preparation retained its cytotoxicity after lyophilization, as seen in the two experiments performed with P815 cells. Very good cytotoxicity was obtained in all cases, ranging from 40-90°7o.
4.1. Gel filtration analysis Gel filtration analysis was performed 5 times on 4 different LAK supernatants, with identical patterns of protein distribution (Fig. 2). The first highmolecular-weight fraction (F1) contained molecules larger than 900 kDa. When a 970-kDa IgM marker and a 2000-kDa dextran blue marker were run on the same Sephacryl S-300 column, F1 was seen to be located between the IgM and dextran blue peaks. As seen in Fig. 3, F1 displayed clear cytotoxicity (15-45o70), although the effect was somewhat less -e'c
A9
-ec
100 8O
x
100
3.4. Fractionation of L A K supernatant
>, c, o
~.,
6O 40
<.9
LAK supernatants pretested for cytotoxic activity were separated by gel filtration on a large (110 × 3 cm) Sephacryl S-300 column. Fractions of 2 - 3 ml were collected, dialyzed against culture medium (RPMI) and tested for cytotoxic reactivity. 3.5. Statistical analysis The non-parametric Mann-Whitney U test was 28
2O
LAK supernatont with FCS
LAK supernatant without FCS
Fig. 1. Cytotoxic activity of LAK supernatants against 4 tumor cell lines. LAK preparations were tested at 1/4 dilution. Data shown are from repeat experiments with LAK supernatants prepared with FCS (HeLa, A9), with SFM (HeLa) or with lyophilized, reconstituted SFM.
280 O.D 0.36
CPM
0.30
o---.~
supernatont IgG- 1125
~l L , '
3600
o.24
;tv ; ,L jl?, p
o.1a
I,
,, f,
3000
~,"', :
0.12 O.lO;
2400
laOO
Jl
0.07 0.04
.!
~: ;', I:
12oo
',
,
600
i 0.01
0 FR1
Fig. 2. Analysis of LAK supernatants on Sephacryl S-300. The gel separation was performed 5 times on 4 different LAK supernatants. Protein concentrations were determined by spectrophotometry at 280 nm. The protein pattern shown is from one representative run. All LAK preparations yielded identical patterns of protein distribution.
A9
Hela
DG75
P815
I00 80 x
2 2
60 40
o~
}'t
20 0
A
[3
Fig. 3. Cytotoxic activity of LAK supernatant Fraction 1, recovered by gel filtration on Sephacryl S-300. (A) Fraction 1 of LAK supernatant. (B) Fraction 1 of lyophilized, reconstituted LAK supernatant.
than that obtained with the original, nonfractionated supernatants. Statistical analysis confirmed that all cytotoxicity data were highly significant ( P < 0.05-0.005).
5. Discussion
Adoptively transferred lymphokine-activated killer (LAK) cells are endowed with appreciable antitumor activity, as demonstrated in animal immunotherapy experiments and by the tumor regres-
sion in some cancer patients so treated [1]. The mechanism of LAK anti-tumor activity in vivo is as yet unclear. Recent studies have indicated that only a very small proportion of the injected LAK cells reach tumor sites. It seems reasonable to assume, accordingly, that the systemic effects exerted by these cells are mediated largely by soluble cytotoxic substances they release. To test this possibility, we examined the supernatants of cultured LAK cells for cytotoxic activity in vitro against several neoplastic cell lines of human and murine origin. The results showed conclusively that all LAK preparations tested were cytotoxic. To date, only one preliminary report has appeared describing the release by cultured LAK cells of a cytotoxic/cytostatic factor [4], although secretion of a variety of cytotoxic molecules by activated cytotoxic T-cells has been reported extensively [8, 9]. To identify and characterize the soluble LAK cytotoxic factor(s), we fractionated the LAK supernatants on a Sephacryl S-300 gel filtration column. All fractions tested exhibited cytotoxic activity against the target tumor cell lines. Since the known lymphocyte-produced cytotoxins have molecular weights lower than 140000, with the exception of the 200-kDa lymphotoxin described by Granger et al. [5], we focused our attention on the high-molecular-weight fraction, F1, which contains molecules larger than 900 kDa. This fraction proved cytotoxic against all target cells tested in numerous experiments. Cytotoxic preparations were tested routinely at several dilutions and, in most cases, the effects were dose-dependent. However, irregular dose-response curves were obtained occasionally, suggesting the co-presence of substances possessing growthstimulating or cytotoxicity-inhibitory activity, or both. Because of this, only results obtained at one dilution (generally 1/4) are shown in this report. The extensive clinical trials conducted with intact LAK cells have yielded only very limited therapeutic gains. Tumor regression has been documented in only a small proporation of advanced cancer patients, and is often restricted in extent and duration. Serious side-effects of treatment with LAK cells and IL-2 are common, and some patients have succumbed to these. There are, moreover, formidable logistic difficulties involved in obtaining the large numbers of LAK cells required for treatment, and 29
the costs o f such treatment are correspondingly large. I f it can be shown that the a n t i - t u m o r activity o f L A K ceils indeed resides with the soluble products they release, it might be h o p e d that these substances will afford a logistically and economically more feasible therapeutic modality, one that can be better standardized, and that might prove less toxic and more efficacious.
Acknowledgements We gratefully acknowledge the support o f The Society o f Research Associates o f the Lautenberg Center, The C o n c e r n Foundation, Los Angeles, and the Miriam Jaffe Research Grant f r o m the Israel Cancer Association. We also t h a n k Ms. Judith Blakemore, Cetus C o r p o r a t i o n , Emeryville, USA,
30
for a generous gift o f recombinant IL-2, and Dr. N. Grover for help with the statistical analysis.
References [1] Rosenberg, S. A. (1988) Immunol. Today 9, 58-62. [2] Maghazachi, A. A., Herberman, R. B., Vujanovic, N. L. and Hiserodt, J. C. (1988) Cell Immunol. 115, 179-194. [3] Herberman, R. B. (1989) J. Clin. Oncol. 7, 1-4. [4] Hersh, E. M., Scuderi, P., Salmon, S., Chong, A. and Grimes, W. (1988) Proc. Am. Assoc. Cancer Res. 29, 1588. [5] Granger, G. A., Yamamoto, R. S., Fair, D. S. and Hiserodt, J. C. (1978) Cell Immunol. 38, 388-402. [6] Neville, M. E. (1987) J. Immunol. Methods 99, 77-82. [7] More, R., Yron, I., Ben Sasson, S. and Weiss, D. W. (1975)Cell Immunol. 15, 382-391. [8] Ware, C.F. and Green, L.M. (1987) Lymphokines 14, 307- 324. [9] Ding-E Young, J. and Liu, C. C. (1988) Immunol. Today 9, 140-144.