new insights into melatonin regulation of cancer growth - Springer Link

38

NEW INSIGHTS INTO MELATONIN REGULATION OF CANCER GROWTH David E. Blask, Leonard A. Sauer, Robert T. Dauchy, Eugene W. Holowachuk, and Mary S. Ruhoff Bassett Research Institute Mary Imogene Bassett Hospital Cooperstown, New York, 13326

INTRODUCTION Numerous studies have confirmed the ability of melatonin to inhibit cancer growth at both physiological and pharmacological concentrations both in vivo and in vitro in experimental model systems utilizing either spontaneous, carcinogen-induced or transplantable murine neoplasms or murine and human cancer cell lines. These model systems have provided important information supporting an antineoplastic role for melatonin at all stages of the tumorigenic process including initiation, promotion, progression and metastasis (1). However, very little is known about the mechanisms by which either the endogenous physiological melatonin signal or the administration of pharmacological doses of melatonin inhibit the various stages of carcinogenesis. Over the past several years, research in our laboratory has focused on the hypothesis that melatonin is an oncostatic neurohormone that inhibits tumor growth promotion. Melatonin inhibits the mitogenic action of a number of growth factors including estradiol (E2), prolactin (PRL) and epidermal growth factor (EGF), on cancer cell proliferation in vitro (1). Additionally our research has addressed the important relationship between dietary fat and cancer, particularly the role of the uptake and metabolism of linoleic acid, an essential fatty acid important in the promotion of tumor growth (2–5) (see below). By combining these two hypotheses, we have formulated a novel hypothesis that melatonin inhibits tumor growth by inhibiting the tumor uptake and metabolism of linoleic acid (6–8). In this article, we will discuss the current evidence from our laboratory that has emerged from the use of the “tissue-isolated’’ tumor preparation, hepatoma 7288CTC, that supports this unique postulate. Melatonin after Four Decades, edited by James Olcese. Kluwer Academic / Plenum Publishers, New York, 2000.

337

338

D. E. Blask et al.

TISSUE-ISOLATED HEPATOMA 7288CTC MODEL SYSTEM In this model system, a 3-mm cube of Morris rat hepatoma is attached to and grown on the end of a vascular stalk composed of the truncated superficial epigastric artery and vein in the inguinal region of adult male Buffalo rats. The tumor implant and adjacent pedicle are then enclosed in a sterile parafilm envelope containing penicillin. The arterial supply to and venous drainage from the tumor is thus exclusively via the superficial epigastric vessels. nmor attachment to other host tissues or vasculature is blocked by the parafilm envelope. Following replacement of the tumor implant into the inguinal fossa and closure of the skin, the tumor is allowed to grow in a “tissueisolated” manner. This arrangement allows the cannulation of the epigastric vessels for the direct perfusion of the tumor itself with physiological and/or pharmacological agents and the measurement of arteriovenous differences across the tumor of various biochemical factors and products important in tumor growth and metabolism. All this is accomplished while the tumor is maintained in a physiological state as reflected by the monitoring of blood gases, blood flow, pH, glucose uptake and lactate release (9). We have taken advantage of such a model system to investigate for the first time, using a totally integrative approach, the mechanisms by which melatonin inhibits tumor growth in vivo from the systemic to the biochemical and molecular levels within a chronobiological context.

THE ROLE OF LINOLEIC ACID IN TUMOR METABOLISM AND GROWTH The Morris rat hepatoma 7288CTC is among a variety of transplantable tumors that are characterized by a unique growth requirement for linoleic acid, an essential polyunsaturated fatty acid ( 6,C18:2). Linoleic acid is the major polyunsaturated fatty acid consumed in the human diet and numerous investigations have shown that high fat diets, particularly those containing linoleic acid as the major fatty acid, increase the growth rates of transplantable tumors in murine species (10). More specifically, Sauer and Dauchy (2–5) have shown that the increased blood concentrations of linoleate following its dietary intake results in an increased arterial supply to and uptake of this fatty acid by tissue-isolated hepatoma 7288CTC. They further demonstrated that the increased tumor uptake of linoleate directly stimulates the growth of this tumor and inferred from these results that either the fatty acid itself or a metabolite initiated or enhanced specific tumor growth processes. These results implied to these investigators that linoleate is more than merely an energy source for tumor growth and may function as a specific tumor growth signal transduction molecule (11). Support for this hypothesis comes from evidence showing that linoleic acid is oxidized intracellularly to 13-hydroxy-9, 11-octadecadienoic acid (13-HODE) by an n-6 lipoxygenase which is most likely part of the 15-lipoxygenase family of enzymes. The activity of this lipoxygenase is regulated by the tyrosine kinase moiety of the EGF receptor such that binding of EGF to its cognate receptor stimulates lipoxygenase activity and the metabolism of linoleic acid to 13-HODE. Following its formation, 13-HODE apparently enhances the EGF-induced autophosphorylation of tyrosine kinase of the EGF receptor as well as the tyrosine phosphorylation of key downstream signal transduction proteins such as GTPase activating protein and

New Insights into Melatonin Regulation of Cancer Growth

339

mitogen-activated protein kinase (MAPK). Recent evidence indicates that 13-HODE up-regulates EGF-dependent tyrosine phosphorylation, and thus EGF-induced mitogenesis, by inhibiting the dephosphorylation of the EGF receptor presumably by altering the interaction of tyrosine phosphatases with the EGF receptor. Therefore, linoleic appears to play a critical role in transducing the EGF-induced mitogenic signal from the cell surface to the nucleus of cancer cells (11,12). In tissue-isolated hepatoma 7288CTC, the rate of 13-HODE released by this tumor is directly dependent upon the rate of tumor linoleic acid uptake. Furthermore, the lipoxygenase inhibitor, nordihydroguairetic acid (NDGA), suppressed 13-HODE production and tumor growth without affecting linoleic acid uptake. Moreover, the addition of 13-HODE to linoleic acid deficient donor blood perfusing tumors of linoleic acid-deficient recipient animals resulted in a dose-dependent increase in [3H]thymidine incorporation into DNA (13). Taken together, these results indicate that 13HODE is the mitogenic signal responsible for linoleate-dependent growth in hepatoma 78288CTC in vivo.

EFFECTS OF MELATONIN ON THE GROWTH AND LINOLEIC ACID METABOLISM OF HEPATOMA 7288CTC We initially determined the sensitivity of hepatoma 7288CTC to the oncostatic effects of pharmacological doses of melatonin (50 to 200 µg) injected S.C. into tumorbearing rats, maintained on a 12L:12D light:dark cycle, every afternoon one to two hours prior to lights off (PM). Injections with melatonin or vehicle began one week prior to tumor implantation and continued until the end of the experiment three to four weeks later. Melatonin treatment was effective in delaying the appearance and suppressing the growth of hepatoma 7288CTC at all doses tested. Furthermore, the tumor uptake, content and release of linoleic acid, total fatty acids and 13-HODE, respectively, was suppressed in animals receiving PM melatonin therapy in a dosedependent manner. Interestingly, in a study in which melatonin (200 µg to 1mg) was injected in the morning (AM) two to three hours following lights on, there was no effect on tumor growth or linoleic acid uptake and metabolism (6–8; unpublished results). However, unlike animals maintained in diurnal lighting, tumor growth and linoleic acid uptake and metabolism in rats maintained under constant light conditions were equally inhibited regardless of whether melatonin was injected during either the subjective AM or PM (unpublished results). We concluded from these results that melatonin’s inhibitory effect on the growth of and linoleic acid uptake and metabolism by hepatoma 7288CTC was both dose- and circadian-time dependent. It is unclear at this point what mechanism is responsible for the apparent diurnal rhythm of tumor sensitivity to melatonin in rats maintained on diurnal lighting. However, the elimination of this sensitivity rhythm of hepatoma 7288CTC to melatonin in constant light strongly suggests that the normal endogenous melatonin rhythm itself may drive this tumor rhythm under diurnal lighting conditions. We next turned our attention to the issue of whether the physiological nocturnal melatonin signal itself exerted an oncostatic effect on tumor growth and metabolism. Previous studies in vivo had demonstrated that pinealectomy or constant light exposure increased the incidence of carcinogen-induced mammary cancer in rats while in vitro studies have shown that physiological concentrations of melatonin inhibit cancer cell proliferation (1). We approached this question in vivo by extinguishing

340

D. E. Blask et al.

the nocturnal melatonin peak by either pinealectomy or exposure of tumor-bearing animals to either constant bright light or a 12L:12D lighting regimen in which the dark phase was “contaminated” by low intensity light (0.2 lux) (14). In each of these scenarios, the onset of palpable tumors was substantially advanced and the tumor growth rate was accelerated by two-fold over that of 12L:12D intact or shampinealectomized controls. Additionally, the tumor uptake of linoleic acid and its conversion to 13-HODE were also markedly elevated in these animals providing strong evidence that the physiological nocturnal melatonin signal is critical for restraining the growth of these tumors, presumably by inhibiting linoleic acid uptake and metabolism to 13-HODE. We reasoned that if the circadian melatonin signal itself is a critical inhibitory regulator of tumor growth and metabolism as well as a driving force for a diurnal rhythm of tumor sensitivity to exogenous melatonin, then hepatoma 7288CTC should evince a circadian rhythm of linoleic acid uptake and metabolism that is the “mirror-image” of the melatonin rhythm. Arteriovenous difference measurements made across tumors at six different circadian time points showed that linoleic acid uptake and oxidation to 13-HODE were highest during the light phase when circulating melatonin levels were low and lowest when melatonin levels were at their peak. This circadian rhythm of tumor metabolism was completely eliminated by pinealectomy (8; unpublished results). These results make a convincing argument for a melatonin-driven circadian rhythm of linoleic acid uptake and metabolism.

DIRECT EFFECTS OF MELATONIN ON TUMOR LINOLEIC ACID UPTAKE AND METABOLISM AND THE SIGNAL TRANSDUCTON MECHANISMS INVOLVED In the tumor growth experiments cited above, it was not completely clear whether the inhibitory action of melatonin on tumor growth was the result or cause of the inhibition of linoleic acid uptake and its conversion to 13-HODE. To more precisely address the question of whether melatonin directly inhibits tumor fatty acid uptake and metabolism, we perfused tissue-isolated hepatoma 7288CTC in situ with melatonin. Following a 60 minute perfusion with donor whole blood from 48-hr fasted rats with elevated circulating fatty acid levels, melatonin was added to the whole blood perfusate at near physiological nocturnal peak levels. Approximately 40 minutes following the addition of melatonin to the perfusate, linoleic acid uptake decreased by nearly 70% from the steady-state control perfusion levels. Over this same time-course, tumor 13HODE production declined to undetectable levels in response to tumor perfusion with melatonin. The suppressive effects of melatonin on linoleic acid uptake and metabolism to 13-HODE were reversible following the removal of melatonin from the perfusate. Neither N-acetylserotonin nor 6-hydroxymelatonin had any effect on tumor linoleate uptake and metabolism in this perfusion system (6–8; unpublised results). These findings indicate that melatonin’s inhibition of linoleic acid uptake and metabolism is a rapid and specific effect exerted directly on the tumor itself and is likely the cause of tumor growth inhibition rather than its result. To further clarify that melatonin’s inhibition of tumor growth was ultimately due to its ability to halt the production of the mitogenic signaling molecule 13-HODE, we examined the incorporation of [3H]-thymidine into DNA in hepatoma 7288CTC in


341

response to perfusion with melatonin either alone or in combination with 13-HODE. The incorporation of [3H]-thymidine by hepatoma 7288CTC was reduced by approximately 43% in response to perfusion with melatonin (1 nM) for two hours; linoleic acid uptake was reduced by 80% and 13-HODE production was completely negated. When melatonin was co-perfused with 13-HODE (12 µg/ml) the tumors took up 50% of the 13-HODE resulting in inhibition of [3H]-thymidine incorporation by melatonin though even linoleic acid uptake was still completely blocked (unpublished results). These data strongly support the hypothesis that melatonin inhibition of the production of the mitogenic metabolite 13-HODE, via the inhibition of linoleic acid uptake, is responsible for the inhibition of tumor growth. The rapid time-course and the circadian nature of melatonin inhibition of tumor uptake and metabolism of linoleic acid suggested that high affinity melatonin receptors (i.e., mt1 and/or MT2) (15) mediated the first step of a signal transduction cascade that ultimately culminated in the inhibition of tumor growth. Moreover, since tumor linoleate uptake and metabolism to 13-HODE could be rapidly and reversibly regulated by melatonin, this suggested that a recently cloned fatty acid transport protein (FATP) (16) may be involved via a functional link to melatonin receptors. Indeed, Northern blot analysis revealed the overexpression of mRNA transcripts for FATP while RT-PCR demonstrated the presence of both mt1 and MT2 receptor mRNAs in hepatoma 7288CTC (8; unpublished results). This provided us with compelling evidence that the molecular substrates were present in hepatoma 7288CTC for facilitated linoleate transport and conversion to 13-HODE and their rapid inhibition by melatonin. Interestingly, neither daily afternoon melatonin injections nor constant light exposure altered the expression of either FATP or melatonin receptor mRNAs. Functional melatonin receptors negatively coupled to adenylate cyclase via a pertussis toxin (PTX)-sensitive G protein such that melatonin suppresses cAMP have been demonstrated in a number of tissues (15). It stood to reason that if high affinity melatonin receptors mediated melatonin’s blockade of linoleate uptake and metabolism to 13-HODE via a suppression of cAMP, then such an effect should be reversible with either PTX, forskolin or cAMP itself. High constitutive levels of cAMP have been observed in several types of malignancies including liver and mammary adenocarcinoma (17,18). In our tumor perfusion system, we found that PTX, forskolin and 8bromo-CAMP completely reversed the melatonin-induced inhibition of linoleate uptake and metabolism to 13-HODE. Additionally, NF-023, an inhibitor of inhibitory G proteins, also reversed the suppressive action of melatonin on linoleic acid uptake and metabolism (8; unpublished results). These results argued strongly for the hypothesis that melatonin inhibits tumor growth by suppressing linloeate uptake and conversion to 13-HODE via a suppression of cAMP through FTX-sensitive melatonin receptors that may be functionally linked to FATP.

CONCLUSIONS The tissue-isolated rat hepatoma model has provided us with an unprecedented opportunity to address the role of melatonin in the regulation of cancer progression from the organismal to the molecular level in the context of circadian biology. Our results have confirmed that the oncostatic effects of physiological and pharmacologia1 concentrations of melatonin are indeed circadian time-dependent. Furthermore,

342

D. E. Blask et al.

circadian time-dependent inhibition of tumor growth is mediated via melatonin’s ability to inhibit the tumor uptake of dietary linoleic acid and its conversion to 13HODE, an amplifying signal for EGF-induced mitogenesis (11,12). Such a mechanism may explain our earlier reported results showing the melatonin inhibits EGF-induced mitogenesis in human breast cancer cells in culture (19). Moreover, a signal transduction cascade involving melatonin receptor-induced suppression of cAMP may mediate melatonin’s ability to suppress the function of FATP and prohibit the entry of linoleate into tumor cells thereby obstructing the transduction of the EGF-induced mitogenic signal from the tumor cell surface to the nucleus.

REFERENCES 1. Blask, D.E. Melatonin in oncology. in Melatonin—Biosynthesis, Physiological Effects, and Clinical Implications (eds Yu H.S. and Reiter, R.J.) (CRC Press, Boca Raton), 447–475 (1993). 2. Sauer, L.A. and Dauchy, R.T. Identification of linoleic and arachidonic acids as the factors in hyperlipemic blood that increase [3H]thymidine incorporation in hepatoma 7288CTC perfused in situ. Cancer Res. 48:3106–3111 (1988). 3. Sauer, L.A. and Dauchy, R.T. Uptake of plasma lipids by tissue-isolated hepatoma 7288CTC and 7777 in vivo. Br. J. Cancer 66:290–296 (1992). 4. Sauer, L.A. and Dauchy, R.T. The effect of omega-6 and omega-3 fatty acids on [3H]thymidine Incorporation in hepatoma 7288CTC perfused in situ. Cancer Res. 66297-303 (1992). 5. Sauer, L.A., Dauchy, R.T., and Blask, D.E. Dietary linoleic acid intake controls the arterial blood plasma concentration and the rates of growth and linoleic acid uptake and metabolism in hepatoma 7288CTC in Buffalo rats. J. Nutrition 127:1412–1421 (1997). 6. Blask, D.E., Sauer, L.A., and Dauchy, R.T. Melatonin suppression of tumor growth in vivo: a novel mechanism involving inhibition of fatty acid uptake and metabolism. in Therapeutic Potential of Melatonin, Frontiers of Hormone Research, vol. 23, (eds Maestroni, G.J.M., Conti, A., and Reiter, R.J.) (Basel, Karger) 107–114 (1997). 7. Blask, D.E., Sauer, L.A., and Dauchy, R.T. Melatonin regulation of tumor growth and the role of fatty acid uptake and metabolism. Neuroendocrinol. Lett. 18:59–62 (1997). 8. Blask, D.E., Sauer, L.A., Dauchy, R.T., Holowachuk, E.W., and Ruhoff, M.S. Circadian rhythm in experimental tumor carcinogenesis and progression and the role of melatonin. in Biological Clocks— Mechanisms and Applications, (ed. Touitou, Y.) (Amsterdam, Elsevier) 469–474 (1998). 9. Dauchy, R.T. and Sauer, L.A. Preparation of “tumor-isolated” rat tumors for perfusion: a new surgical technique that perserves continuous blood flow. Lab. Animal Sci. 36:678–681 (1986). 10. Klurfield. D.M. Fat effects on experimental tumorigenesis. J. Nutr. Biochem. 6:201–205 (1995). 11. Eling, T.E. and Glasgow, W. Cellular proliferation and lipid metabolism: importance of lipoxygenases in regulating epidermal growth factor-dependent mitogenesis. Cancer Metast. Rev. 13:397–410 (1994). 12. Glasgow,W.E., Hui, R., Everhart, A.L., Jayawickreme, S.P., Angerman-Stewart, J., Han, B.B., and Eling, T.F.The linoleic acid metabolite (13 S)-hydroxyoctadecadienoic acid, aug-ments the epidermal growth factor receptor signaling pathway by attenuation of receptor dephosphorylation. J. Biol. Chem. 272:119269–19276 (1997). 13. Sauer, L.A., Dauchy, R.T., Blask, D.E., Armstrong, B.J., and Scalici, S. 13-Hydroxyoctadecadienoic acid is the mitogenic signal for linoleic acid -dependent growth in rat hepatoma 7288CTC in vivo. Cancer Res.59:4688–4692(1999). 14. Dauchy, R.T., Sauer, L.A., Blask, D.E., and Vaughan, G.M. Light contamination during the dark phase in “photoperiodically controlled” animal rooms: effect on tumor growth and metabolism in rats. Lab. Animal Sci. 47:511–518 (1997). 15. Reppert, S.M. and Weaver, D.R. Melatonin madness. Cell 83:1059–1062 (1995). 16. Schaffer, J.E. and Lodish, H.F. Expression cloning and characterization of a novel adipocyte long chain fatty acid transport protein. Cell 79:427–436 (1994). 17. DeRubertis, F.R. and Craven, PA. Sequential alterations in the hepatic content and metabolism of cyclicAMP and cyclic GMP induced by DL-ethionine: evidence for malignant transformation of liver with a sustained increase in cyclic AMI? Metabolism 25:1611–1625 (1976).


343

18. Cho-Chung, Y.S. Role of cyclic AMP receptor proteins in growth, differentiation, and suppression of malignancy: new approaches to therapy. Cancer Res. 50:7093–7100 (1990). 19. Cos, S. and Blask, D.E. Melatonin modulates growth factor activity in MCF-7 human breast cancer cells. J. Pineal Res. 17:25–32 (1994).