- Email: [email protected]
An interaction between type 1 and type 2 programmed cell death and clonogenic survival
Medical Hypotheses (2003) 61(5–6), 583–585 ª 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0306-9877(03)00236-6
An interaction between type 1 and type 2 programmed cell death and clonogenic survival K. M. Anderson,1,2 P. Bonomi,3 Y. Hu,1 J. E. Harris1 1
Department of Medicine, The University of Illinois and the West Side V A Hospital, Chicago, USA; 2Department of Biochemistry, The University of Illinois and the West Side V A Hospital, Chicago, USA; 3Department of Pathology, The University of Illinois and the West Side V A Hospital, Chicago, USA
Summary We suggest two additional reasons why current, non-surgical therapies for most solid, epithelialderived cancers can lack effectiveness. Studies with panc-1 human pancreatic cancer cells cultured with actinomycin D and/or MK 886 indicate firstly, that type 2 (intrinsic, autophagic, mitochondrial-dependent, MK 886induced) programmed cell death is less effective than the type 1 (apoptotic, extrinsic, ligand-dependent, actinomycin D-induced) form in reducing the number of residual clonogenic cells, and secondly, that activation of cellular suicide during their combined culture results in a greater number of residual clonogenic cells compared with either agent alone. Hypothesis. Based on results from the culture of panc-1 cells with MK 886 and/or actinomycin D, we suggest that in this system, and possibly in others: (a) type 2 programmed cell death is a less effective inhibitor of residual cells with clonogenic potential, and (b) activation together of both forms of PCD increases the number of residual clonogenic cells. ª 2003 Elsevier Ltd. All rights reserved.
INTRODUCTION Hematopoietic cancer cells generally respond to various therapies with a type 1 ‘apoptotic’ form of programmed cell death (PCD), while solid, epithelial-derived cancers seem more often to undergo a type 2 response (1,2). Whether all cells retain an ability to express either form of cellular suicide, as defined above, or as seems to be the case, have evolved preferentially to express one rather than the other is not certain. This of course depends upon which endogenous signaling pathways are present and whether they can be accessed by the agent of interest.
When cultured with actinomycin D, an inhibitor of RNA synthesis, panc-1 human pancreatic cancer cells express a type 1 response (3) and a type 2 response after challenge with MK 886, a 5-lipoxygenase inhibitor that induces additional biochemical functions (4–7). These cells provide an experimental system with which interaction between the 2 pathways can be studied. Whether their mutual activation results in synergism, simple additivity or interference and the extent to which either form reduces the number of residual clonogenic cells are important questions bearing on the lack of successful response, especially by solid cancers, to most available non-surgical therapies.
Received 22 April 2002 Accepted 12 July 2002
OBSERVATIONS AND DISCUSSION Correspondence to: Anderson K.M. MD PhD, Rush Medical College, 1753 W Harrison, Chicago, Il 60612 USA. Phone: +1-312-942-6423; Fax: +1-312042-3192; E-mail: [email protected]
There is preliminary evidence consistent with several conclusions (Fig. 1 is one of three representative examples leading to the same conclusions).
583
584 Anderson et al.
maining after culture with MK 886 was greater; (c) when the 2 agents were cultured together rather than separately, the number of residual clonogenic cells was increased, compared with the most effective inhibitor, actinomycin D, and (d) this last result implies an interaction between these 2 forms of cellular suicide, consistent with an interference from the type 2 PCD in the reduction in residual clonogenic cells by the more effective type 1 process. This inference is supported by results of caspase activation studied with flow cytometry in which actinomycin D-activation of caspase activity, absent in MK-treated cells, was reduced in cells cultured with both agents (work in progress). In various clinical chemotherapy, radiation and biological response-modifier protocols, both forms of programmed cell death may have been inadvertently activated. To the extent that in at least some cancers, type 1 PCD may reduce the numbers of clonogenic cells to a greater extent than a type 2 response, and that activation of both forms can mutually interfere with this reduction, failure of response becomes more understandable. If a type 2 response is more commonly evoked by therapy in solid epithelial cancers, and is itself less effective in inducing PCD and reducing the number of potentially clonogenic cells, this could contribute to the differences often observed between the responses of solid cancers and of hematopoietic cancers with their more frequent and possibly more effective type 1 response to therapy. To the extent that these observations and their interpretation are correct and applicable to any other types of cancer cells treated with agents that induce both forms of PCD, they provide further reasons why the response to chemicals or radiation by solid cancers is usually less than that of hematopoietic malignancies. CRITIQUE
Fig. 1 Representative clonogenic assay of panc-1 cells cultured for 18 days with 100 ng/ml of actinomycin D, 40 lm MK 886 or their combination. Cells were first cultured for 48 h with these agents, washed twice with RPMI, detached with trypsin-EDTA, and 1000 viable cells cultured in triplicate for each category in 25 cm2 flasks with 5 ml RPMI containing 10% fetal calf serum, 50 U/ml penicillin, 50 lgm/ml of streptomycin and 25 mM HEPES. Media was changed every 3 days and after 18 days, detached cells were gently washed away, retained cells stained with crystal violet in methanol and colonies of 50 or more cells scored as positive. C, control; A, actinomycin D; M, MK 886 and A + M, their combination. The number of identified clones in the MK 886-treated sample was greater than in the actinnomycin D-treated one; to be presented in greater detail elsewhere.
We conclude that (a) Actinomycin D-induced type 1 PCD reduced the number of residual panc-1 clonogenic cells to a greater extent than MK 886-induced type 2 PCD; (b) the maximum diameter of larger colonies reMedical Hypotheses (2003) 61(5–6), 583–585
It is understood that other explanations could be invoked to account for these results. However, the questions raised merit examination. While the reduction in cell numbers under the conditions described was comparable, it is not established that the subset of potentially clonogenic cells was ab initio equally at risk from either form of PCD. Any role for the length of the cell cycle or for the numbers or distribution of cells in cycle has not been examined. The origin of accompanying necrotic cells seen by electron microscopy, to be reported elsewhere, was unclear. They could have resulted from cells first entrained by either form of PCD that subsequently become necrotic due to the lack of scavanger cells in this in vitro system and/or because cells directly entered that state. An activity associated with MK 886 function such as increased oxidative stress may have directly affected the number of residual cloª 2003 Elsevier Ltd. All rights reserved.
An interaction between type 1 and type 2 programmed cell death and clonogenic survival
nogenic cells, unrelated to the induction of type 2 PCD. Certainly other cell types exposed to different agents able to elicit both forms of cellular suicide may not exhibit comparable changes leading to the conclusions advanced. Inhibition of RNA synthesis by actinomycin D and induction of oxidative stress, inhibition of 5-lipoxygenase activity and possible induction of NF-kB by MK 886 may be responsible for some of the results observed, due to their diverse effects on target cell biochemistry (5–7). The questions addressed may not arise with the use of closely targetted agents, e.g., tyrosine kinase inhibitors (8), etc. ACKNOWLEDGEMENTS We thank the Weinberg Family Trust, the Wadsworth Memorial Trust and the Seidel Family Trust for their support.
REFERENCES 1. Bortner C. D., Oldenburg N. B. E., Cidlowski J. A. The role of DNA fragmentation in cancer. Trends Cell Biol 1995; 5: 21–26.
ª 2003 Elsevier Ltd. All rights reserved.
585
2. Kroemer G., Reed J. C. Mitochondrial control of cell death. Nat Med 2000; 6: 513–519. 3. Kleeff J., Kornmann M., Sawhney H., Kore M. Actinomyhcin D induces apoptosis and inhibits growth of pancreatic cancer cells. Int J Can 2000; 86: 399–407. 4. Anderson K. M., Seed T. et al. Five-lipoxygenase inhibitors reduce panc-1 survival: The mode of cell death and synergism of MK 886 with gamma linolenic acid. Antican Res 1998; 18: 791–800. 5. Datta K., Biswal S. S., Keher J. P. The five-lipoxygenase activating protein (FLAP), MK 886, induces apoptosis independently of FLAP. Biochem J 1999; 340: 371–375. 6. Anderson K. M., Alrefai W. A., Anderson C. A. et al. Eicosanoids and other Bioactive Lipids in Cancer, Inflammation, and Radiation Injury, Vol. 5. In: Honn et al. (eds). Kluwer Academic, 2002: 451–456. 7. Avis I., Hong S. H., Martinez A. et al. Five-lipoxygenase inhibitors can mediate apoptosis in human breast cancer cell lines through complex eicosanoid interactions. FASEB J 2001; 15: 2005–2007. 8. Shawver L. K., Slamon D., Ullrich A. Smart drugs: Tyrosine kinase inhibitors in cancer therapy. Cancer cell 2002; 1: 117–123.
Medical Hypotheses (2003) 61(5–6), 583–585
An interaction between type 1 and type 2 programmed cell death and clonogenic survival K. M. Anderson,1,2 P. Bonomi,3 Y. Hu,1 J. E. Harris1 1
Department of Medicine, The University of Illinois and the West Side V A Hospital, Chicago, USA; 2Department of Biochemistry, The University of Illinois and the West Side V A Hospital, Chicago, USA; 3Department of Pathology, The University of Illinois and the West Side V A Hospital, Chicago, USA
Summary We suggest two additional reasons why current, non-surgical therapies for most solid, epithelialderived cancers can lack effectiveness. Studies with panc-1 human pancreatic cancer cells cultured with actinomycin D and/or MK 886 indicate firstly, that type 2 (intrinsic, autophagic, mitochondrial-dependent, MK 886induced) programmed cell death is less effective than the type 1 (apoptotic, extrinsic, ligand-dependent, actinomycin D-induced) form in reducing the number of residual clonogenic cells, and secondly, that activation of cellular suicide during their combined culture results in a greater number of residual clonogenic cells compared with either agent alone. Hypothesis. Based on results from the culture of panc-1 cells with MK 886 and/or actinomycin D, we suggest that in this system, and possibly in others: (a) type 2 programmed cell death is a less effective inhibitor of residual cells with clonogenic potential, and (b) activation together of both forms of PCD increases the number of residual clonogenic cells. ª 2003 Elsevier Ltd. All rights reserved.
INTRODUCTION Hematopoietic cancer cells generally respond to various therapies with a type 1 ‘apoptotic’ form of programmed cell death (PCD), while solid, epithelial-derived cancers seem more often to undergo a type 2 response (1,2). Whether all cells retain an ability to express either form of cellular suicide, as defined above, or as seems to be the case, have evolved preferentially to express one rather than the other is not certain. This of course depends upon which endogenous signaling pathways are present and whether they can be accessed by the agent of interest.
When cultured with actinomycin D, an inhibitor of RNA synthesis, panc-1 human pancreatic cancer cells express a type 1 response (3) and a type 2 response after challenge with MK 886, a 5-lipoxygenase inhibitor that induces additional biochemical functions (4–7). These cells provide an experimental system with which interaction between the 2 pathways can be studied. Whether their mutual activation results in synergism, simple additivity or interference and the extent to which either form reduces the number of residual clonogenic cells are important questions bearing on the lack of successful response, especially by solid cancers, to most available non-surgical therapies.
Received 22 April 2002 Accepted 12 July 2002
OBSERVATIONS AND DISCUSSION Correspondence to: Anderson K.M. MD PhD, Rush Medical College, 1753 W Harrison, Chicago, Il 60612 USA. Phone: +1-312-942-6423; Fax: +1-312042-3192; E-mail: [email protected]
There is preliminary evidence consistent with several conclusions (Fig. 1 is one of three representative examples leading to the same conclusions).
583
584 Anderson et al.
maining after culture with MK 886 was greater; (c) when the 2 agents were cultured together rather than separately, the number of residual clonogenic cells was increased, compared with the most effective inhibitor, actinomycin D, and (d) this last result implies an interaction between these 2 forms of cellular suicide, consistent with an interference from the type 2 PCD in the reduction in residual clonogenic cells by the more effective type 1 process. This inference is supported by results of caspase activation studied with flow cytometry in which actinomycin D-activation of caspase activity, absent in MK-treated cells, was reduced in cells cultured with both agents (work in progress). In various clinical chemotherapy, radiation and biological response-modifier protocols, both forms of programmed cell death may have been inadvertently activated. To the extent that in at least some cancers, type 1 PCD may reduce the numbers of clonogenic cells to a greater extent than a type 2 response, and that activation of both forms can mutually interfere with this reduction, failure of response becomes more understandable. If a type 2 response is more commonly evoked by therapy in solid epithelial cancers, and is itself less effective in inducing PCD and reducing the number of potentially clonogenic cells, this could contribute to the differences often observed between the responses of solid cancers and of hematopoietic cancers with their more frequent and possibly more effective type 1 response to therapy. To the extent that these observations and their interpretation are correct and applicable to any other types of cancer cells treated with agents that induce both forms of PCD, they provide further reasons why the response to chemicals or radiation by solid cancers is usually less than that of hematopoietic malignancies. CRITIQUE
Fig. 1 Representative clonogenic assay of panc-1 cells cultured for 18 days with 100 ng/ml of actinomycin D, 40 lm MK 886 or their combination. Cells were first cultured for 48 h with these agents, washed twice with RPMI, detached with trypsin-EDTA, and 1000 viable cells cultured in triplicate for each category in 25 cm2 flasks with 5 ml RPMI containing 10% fetal calf serum, 50 U/ml penicillin, 50 lgm/ml of streptomycin and 25 mM HEPES. Media was changed every 3 days and after 18 days, detached cells were gently washed away, retained cells stained with crystal violet in methanol and colonies of 50 or more cells scored as positive. C, control; A, actinomycin D; M, MK 886 and A + M, their combination. The number of identified clones in the MK 886-treated sample was greater than in the actinnomycin D-treated one; to be presented in greater detail elsewhere.
We conclude that (a) Actinomycin D-induced type 1 PCD reduced the number of residual panc-1 clonogenic cells to a greater extent than MK 886-induced type 2 PCD; (b) the maximum diameter of larger colonies reMedical Hypotheses (2003) 61(5–6), 583–585
It is understood that other explanations could be invoked to account for these results. However, the questions raised merit examination. While the reduction in cell numbers under the conditions described was comparable, it is not established that the subset of potentially clonogenic cells was ab initio equally at risk from either form of PCD. Any role for the length of the cell cycle or for the numbers or distribution of cells in cycle has not been examined. The origin of accompanying necrotic cells seen by electron microscopy, to be reported elsewhere, was unclear. They could have resulted from cells first entrained by either form of PCD that subsequently become necrotic due to the lack of scavanger cells in this in vitro system and/or because cells directly entered that state. An activity associated with MK 886 function such as increased oxidative stress may have directly affected the number of residual cloª 2003 Elsevier Ltd. All rights reserved.
An interaction between type 1 and type 2 programmed cell death and clonogenic survival
nogenic cells, unrelated to the induction of type 2 PCD. Certainly other cell types exposed to different agents able to elicit both forms of cellular suicide may not exhibit comparable changes leading to the conclusions advanced. Inhibition of RNA synthesis by actinomycin D and induction of oxidative stress, inhibition of 5-lipoxygenase activity and possible induction of NF-kB by MK 886 may be responsible for some of the results observed, due to their diverse effects on target cell biochemistry (5–7). The questions addressed may not arise with the use of closely targetted agents, e.g., tyrosine kinase inhibitors (8), etc. ACKNOWLEDGEMENTS We thank the Weinberg Family Trust, the Wadsworth Memorial Trust and the Seidel Family Trust for their support.
REFERENCES 1. Bortner C. D., Oldenburg N. B. E., Cidlowski J. A. The role of DNA fragmentation in cancer. Trends Cell Biol 1995; 5: 21–26.
ª 2003 Elsevier Ltd. All rights reserved.
585
2. Kroemer G., Reed J. C. Mitochondrial control of cell death. Nat Med 2000; 6: 513–519. 3. Kleeff J., Kornmann M., Sawhney H., Kore M. Actinomyhcin D induces apoptosis and inhibits growth of pancreatic cancer cells. Int J Can 2000; 86: 399–407. 4. Anderson K. M., Seed T. et al. Five-lipoxygenase inhibitors reduce panc-1 survival: The mode of cell death and synergism of MK 886 with gamma linolenic acid. Antican Res 1998; 18: 791–800. 5. Datta K., Biswal S. S., Keher J. P. The five-lipoxygenase activating protein (FLAP), MK 886, induces apoptosis independently of FLAP. Biochem J 1999; 340: 371–375. 6. Anderson K. M., Alrefai W. A., Anderson C. A. et al. Eicosanoids and other Bioactive Lipids in Cancer, Inflammation, and Radiation Injury, Vol. 5. In: Honn et al. (eds). Kluwer Academic, 2002: 451–456. 7. Avis I., Hong S. H., Martinez A. et al. Five-lipoxygenase inhibitors can mediate apoptosis in human breast cancer cell lines through complex eicosanoid interactions. FASEB J 2001; 15: 2005–2007. 8. Shawver L. K., Slamon D., Ullrich A. Smart drugs: Tyrosine kinase inhibitors in cancer therapy. Cancer cell 2002; 1: 117–123.
Medical Hypotheses (2003) 61(5–6), 583–585