|
|
||||||||
Department of Human Genetics, University of Michigan, AnnArbor,Michigan 481090618
| Abstract |
|---|
|
|
|---|
Key Words: copper angiogenesis cancer tetrathiomolybdate zinc
| The War Against Copper: Is It Almost Over or Is It Just Beginning? |
|---|
|
|
|---|
Zinc is almost ideal for the maintenance therapy of Wilson's disease, but is too slow acting to be optimal for patients as they initially present with acute copper toxicity, particularly those patients presenting with neurological symptoms. Penicillamine is disastrous in a large proportion of these patients because it makes about one-half of the patients worse, probably by mobilizing hepatic copper and by temporarily elevating brain copper further. Many patients, about 25% of those treated, never recover to their pre-penicillamine baseline, and are often left with serious, permanent, penicillamine-induced disability (8). For these patients we have developed tetrathiomolybdate (TM), a compound with a long and interesting history, beginning with molybdenum-induced copper deficiency in ruminants in Australia and New Zealand (911). It was eventually discovered that the sulfur-rich rumen was converting the molybdenum to thiomolybdates, and thiomolybdates given to rats proved to be potent anticopper agents, whereas molybdenum itself given to nonruminants was ineffective (1215). Of the thiomolybdates, the tetra-substituted compound TM is the most potent. We have used TM for the initial treatment of 63 neurologically presenting Wilson's disease patients with excellent results (1618). TM acts in the gut by forming a tripartite complex with food copper and food protein, preventing copper absorption. TM absorbed into the blood forms a tripartite complex with albumin and blood copper, rendering such bound copper nontoxic. Neurological function is well preserved during an initial treatment period lasting 8 weekslong enough to control copper toxicity. Then the patients are treated with zinc maintenance therapy and most slowly recover much of their neurological function over 1 or 2 years. As with zinc, TM is essentially nontoxic in these patients, with the only risk appearing to be overtreatment and the depletion of the bone marrow of copper, producing a reversible anemia (16).
With our therapeutic cupboard now well stocked with two new, effective, and nontoxic anticopper therapies, each designed to complement the other and allow a continuum of therapy in Wilson's disease, one might assume that the medical war against copper has been won, or is at least almost over. However, it turns out that there is yet another ``front'' to this warthat of reducing copper levels to inhibit angiogenesis as a therapy for cancer and other diseases of neovascularization. That battle is just beginning. It appears that the lessons we have learned from treating Wilson's disease with anticopper drugs will be very useful in this new war.
| Copper and Angiogenesis: Basic Studies |
|---|
|
|
|---|
The two types of experiments show different things. The first shows that copper and copper-containing molecules are angiogenic in this model. The second, however, is more interesting in the context of our topic. It shows that mild copper deficiency reduces angiogenesis produced by a known angiogenic stimulant.
| Tumor Growth and Angiogenesis |
|---|
|
|
|---|
In the normal human, angiogenesis is necessary and critically important during early embryonic and fetal growth and development. A certain amount of angiogenesis is required during childhood growth and development, although to less and less an extent as the individual matures. In adults, the only obvious requirements for angiogenesis are for wound healing and in the uterus during the menstrual cycle.
The absolute dependence of tumor growth on angiogenesis and the relative lack of a need for angiogenesis in adults has suggested that the need for angiogenesis might be an Achilles' heel of cancer, including advanced and metastatic cancer (2230). This theory proposes that if therapies could be found that inhibit angiogenesis, they might be effective against further cancer growth while being nontoxic to normal tissue.
Angiogenesis in normal cells and tissues involves a complex interaction and balance between stimulators and inhibitors. Brem (29) lists 35 endogenous stimulators and 18 inhibitors. These would be available to the normal organism at various times and places according to the types of cells and tissues, developmental states, and the environmental situation. Most of the angiogenic promoters also appear to be available for recruitment by various types of tumors. Tumors can not only synthesize their own angiogenic promoters within their own cells, but they can also recruit host cells, which then become involved in the invasive process. These host cells can synthesize still other angiogenic promoters, thus assisting in the tumor growth and invasion process.
As a result, tumor angiogenesis is a very complex subject. The review by Beckner (30) lists nine positive tumor angiogenic factors (or promoters) made by inflammatory cells, 13 made by endothelial cells, 15 made by fibroblasts and/or occurring in the extracellular matrix, and 20 made by tumor cells. Of course, there is some overlap of factors appearing in more than one of these lists, but nonetheless, the total number of angiogenic promoters possibly involved in tumorogenesis is daunting. Further, there is a complex array of interactions among the various promoters, as illustrated in a table provided by Beckner (30). For example, vascular endothelial growth factor (VEGF) is shown as interacting with 16 other angiogenic promoters.
One question that can be asked is: Why all this redundancy in angiogenic promoters? One possible explanation is that normal growth and tissue repair, both dependent upon angiogenesis, has to occur in a large array of environments such as embryonic, fetal, and beyond life stages, in a variety of tissues, in a variety of metabolic and oxygenation states, and in a variety of tissue injury conditions (septic, hypoxic, etc.). Thus, diversity in protecting angiogenesis and tissue growth may be generally critical to development and survival.
Irrespective of the evolutionary reasons for angiogenic promoter redundancy, it makes a diverse array of factors potentially available to tumors to facilitate their growth, invasion, and metastatic processes. This diversity, in turn, potentially complicates the development of effective antiangiogenic therapies aimed at stopping tumor growth. Some key questions with respect to antiangiogenic therapy are asked in Table I
.
|
Unfortunately, however, the answer to Question 2 also appears to be yes, at least partially. While some positive factors act primarily on VEGF, it appears that a fairly large number can act independently to stimulate angiogenesis in tumors (30). This probably means that a therapy aimed at inhibiting only VEGF will not be successful, or will be only partially successful. Some tumor types may have redundant systems, and other tumor types may develop resistance to VEGF inhibition by recruiting alternate angiogenic pathways.
So far the answer to Question 3 also seems to be yes, with a wide variety of angiogenic promoters isolated from various types of tumors (29, 30). Again, this complicates the therapeutic task, suggesting that different inhibitors may be required for different tumors.
It would, of course, be wonderful if the answer to Question 4 were to be yes because it would mean that a therapy designer would have the potential to shut down tumor angiogenesis with a single approach, much as would be the case if the answer to Question 1 were yes. So, is there a common regulatory element? We believe there is. It appears that a requirement for relatively high levels of copper for activity is a property shared by a large number of angiogenic promoters, which may allow a reduction in the body's copper levels as a therapeutic tool to shut down angiogenesis in many cancers. We will return to this exciting topic after briefly reviewing other antiangiogenic strategies tried so far.
| Antiangiogenic Anticancer Approaches (Noncopper Related) |
|---|
|
|
|---|
Because of the key role of VEGF, several agents are aimed at interfering with its action (29). There is a VEGF antibody being studied by Genetech, and another agent is under study by Sugen and another by Novartis that blocks VEGF receptor signaling. Thalidomide is a drug used many years ago in Europe for sedation that turned out to have teratogenic side effects. It is being studied by Celgene because it blocks the activity of VEGF and possibly other angiogenic promoters. Interferon-
is undergoing trial as an inhibitor of VEGF production.
Two protein agents that are natural substances in the body and are angiogenic inhibitors have been studied by O'Reilly et al. (32, 33). One is angiostatin (32), a 38-kDa protein that inhibits angiogenesis in a variety of ways, including inhibition of endothelial cell proliferation. Angiostatin is an internal fragment of the plasminogen molecule and has demonstrated potent antitumor effects in animal models.
The other molecule, endostatin (33), has received a great deal of public attention. It is a 20-kDa C-terminal fragment of collagen XVIII. It is an antiangiogenic molecule that inhibits endothelial cell proliferation and has also proven to be a potent antitumor agent in animal models. Both angiostatin and endostatin are protein molecules and therefore cannot be given orally.
From the examples cited here, it should be clear that the approaches being taken by investigators are quite diverse. While some encouraging clinical results are beginning to emerge, it is too early to say which of these approaches, if any, will yield useful anticancer therapies.
| Copper Control as an Antiangiogenic Anticancer Therapy: Animal Studies |
|---|
|
|
|---|
Brem and coworkers (34, 35) were the first to examine the effect of copper limitation on tumor growth in animals. They made rabbits and rats partially copper deficient by using penicillamine and a low copper diet. They then implanted brain tumors into the brains of these and control animals. The results were very impressive in one aspect and very disappointing in another.
First, as to the positive results, the brain tumors in the copper-deficient animals were markedly smaller than in the control animals (34). The surface of the brains over the implants in the control animals showed a plethora of new vascularization, whereas this was not present in the copper-deficient animals. Cut sections through the tumors and surrounding normal brain tissue revealed marked invasion of blood vessels from the tumor into the normal brain tissue in control animals, and this was completely missing in copper-deficient animals (35).
The disappointing results were a lack of survival advantage in the copper-deficient animals (34). The treated animals died at approximately the same time after tumor transplantation as the control animals. It appears that cerebral edema and brain stem herniation, the likely cause of death in most of the animals, was approximately equal. The bottom line was that while the proof of the principle that relative copper deficiency inhibited tumor angiogenesis and reduced tumor growth was supported, the therapy seemed to have no overall benefit. Another negative aspect of the study was that metastases to lung from tumor implanted in the thigh was not affected by a copper-deficient state (34).
Reading this literature and having developed an extremely potent and safe anticopper therapy in TM (as discussed in the first section of this paper), I decided to reexamine the question of anticopper antiangiogenesis therapy in cancer. I carried out a mouse sarcoma tumor model study using MCA205 cells injected subcutaneously, and I obtained positive results with TM treatment with a significantly reduced rate of tumor growth (Brewer GJ, unpublished data). With TM therapy, a tripartite complex of albumin, copper, and TM is formed in the blood, which clears slowly and makes serum copper an inaccurate reflector of body copper status. I, therefore, used plasma ceruloplasmin as a surrogate marker of copper status and it worked well. Ceruloplasmin is a copper-containing plasma protein synthesized by the liver and released into the blood according to copper availability in the liver (36).
However, this tumor mouse model was not ideal, and I joined forces with an oncologist, Dr. Sofia Merajver. Dr. Merajver designed an elegant study using the HER2-neu transgenic mouse model in which control mice develop mammary cancer during the first year of life. Treated animals were given TM daily by gastric gavage. Over a period of 270 days, 18 of 22 control mice developed obvious mammary cancer, while all 15 of the TM-treated mice were free of visible tumor, although small clusters of malignant cells were present on microscopic examination (37). Release of a few of the treated mice from TM therapy caused their tumors to grow into detectable masses. TM treatment of a few of the control mice allowed tumor growth stabilization in mice whose tumors could be followed humanely for a long enough period of observation.
These results were very exciting to us because they suggested not only tumor growth stabilization with TM therapy, but also potential prophylaxis in patients with high genetic risks of developing cancer.
| Copper Control as an Antiangiogenic Anticancer Therapy: Clinical Studies |
|---|
|
|
|---|
To be evaluable for efficacy, patients have to stay on therapy long enough to have their ceruloplasmin levels in the target range of 5 to 15 mg/dl (normal is 2035 mg/dl) for at least 60 days. The reason for this is that stability is not expected until significant copper depletion occurs, including copper depletion from the tumor, and there is good evidence that tumors sequester extra copper (3942). If the observation period is shorter than 60 days, there probably has not been adequate opportunity for disease stabilization to occur. In the original publication (38), six of the 18 patients were evaluable for efficacy. Five achieved disease stabilization, including one with regression of lung metastases, and the sixth had stable disease except for progression at one site. The length of disease stability, or freedom from progression (FFP) ranged from 120 to 413 days. For comparison, FFP in the usual phase 1 study averages about 60 days. The only side effect of TM was rare overtreatment resulting in anemia, which required temporary cessation of drug (38).
This phase 1 study has been ongoing and a long-term follow-up is being reported (43 [Merajver, et al., unpublished]). A total of 40 patients have been entered by now (including the original 18), and 18 of these are evaluable by our criteria. All 18 have achieved disease stabilization for longer than 2 months, with the average being 9.5 months, two as long as 30 months, with six patients still on study. Eleven different types of malignancies are represented by these 18 patients.
It must be admitted that this clinical trial is uncontrolled, and final conclusions cannot be drawn until large numbers of patients with specific cancers are studied under controlled conditions. However, my tentative interpretation of these results is that they show reasonable evidence of efficacy. While it is true that occasional patients with certain types of advanced metastatic cancer will have periods of disease stability, this type of spontaneous stability for a lengthy period is rare with many of the cancers in our study. And lengthy periods of disease stabilization in our study are relatively common, not just the exception. For example, if one selects 6 months of disease stabilization as somewhat unusual spontaneous occurrences with the malignancies under study, nine of our 18 evaluable patients achieved this period of stabilization or better, with two who have not yet achieved it still on study. Clearly, there is very strong, but anecdotal, evidence of efficacy in two patients, one with metastatic breast cancer and one with metastatic chondrosarcoma, who have had stable disease while on TM therapy for over 2.5 years (43).
An important question is: Will TM (and its associated clinical copper deficiency) be effective therapy in all metastatic solid tumors? We have evidence that it is effective against many in both the clinical and preclinical studies, as shown in Table II
, but will there be exceptions? Part of the answer will probably come from whether some tumors use angiogenic promoters that do not depend upon copper, or whether they use promoters that have a high enough affinity for copper to obtain copper under TM treatment conditions. If VEGF is as critical to tumor angiogenesis as now seems likely, and if VEGF is truly copper-dependent for activity (see next section), these situations may not be common. Another question is whether bone metastases can obtain nutrients from bone marrow and not depend upon neovascularization.
|
Of course, it will be important to determine whether antiangiogenesis therapy with TM interacts favorably with other modalities of cancer treatment. We have a positive mouse model study of lung cancer in which radiation therapy and TM therapy combined to have a greater effect than either alone (44). So far, we only have anecdotal (but encouraging) clinical evidence that interactions with other anticancer modalities will be positive. Additional studies to answer this question have been initiated.
Finally, if TM does prove safe, would it prove effective as a preventive in patients at high risk for getting cancer, such as those with cancer predisposing genotypes, as suggested by our findings with the HER2-neu transgenic mouse study (37)? And will it be useful in other noncancer diseases of neovascularization such as diabetic retinopathy, rheumatoid arthritis, and psoriasis?
| Copper Control as an Antiangiogenic Anticancer Therapy: Possible Mechanisms of Action |
|---|
|
|
|---|
|
Fibroblastic growth factors (FGF), acidic and basic (synonymous with FGF1 and FGF2) are also angiogenic promoters and also bind copper with a high affinity (51, 52). Angiotropin, a monocyte-derived angiogenic factor, contains copper within the molecule (53). The binding of angiogenin, a potent angiogenic factor, to endothelial cells appears to be modulated by copper (54), although this has been challenged (55). SPARC (secreted protein, acidic and rich in cysteine) is a source of copper-binding peptides that stimulate angiogenesis (56). Ceruloplasmin, heparin, and the tripeptide Gly-His-Lys (also referred to as liver growth factor) are angiogenic when complexed to copper (57).
The above citations list nine angiogenic substances that are reported to bind or contain copper, and they support a preliminary hypothesis (Hypothesis 2 of Table III
) that copper is critically involved in the action of one or more of them, and this provides a mechanism for the therapeutic efficacy of a lowered copper level. Admittedly, final proof of the hypothesis that lowered copper levels reduce angiogenesis through reducing the angiogenic stimulating activity of one or more of the listed factors is not in, but for the sake of further discussion I will make the assumption that this hypothesis will be proven correct.
Hard upon its heels, then, must come Hypothesis 3 of Table III
, which suggests that one or more key angiogenic promoters is sensitive to activation by copper in the ``mid-range'' of body copper status. That is, well before cellular enzymes requiring copper are affected, one or more of these factors becomes less active. In this sense, the ability of the liver to put copper into ceruloplasmin and secrete it into the blood seems to be a reflection of the copper availability required for the hypothesized copper-dependent angiogenic promoter(s).
In regard to copper concentrations, many investigators using in vitro experiments appear to be unaware of the very low concentrations of ionic or free copper that exists in vivo, even extracellularly. While serum copper in the human is about 100 µg/dl, or about 1.5 µM, 90% of that copper is in ceruloplasmin and is therefore not readily available to act on angiogenic promoters. Even the remaining 10% is bound to protein and other molecules. Intracellularly, ionic copper is kept so low (in yeast, less than one atom per cell!) that some processes use copper chaperones to move copper from one molecule to another (58, 59). If some of the angiogenic promoters require copper chaperones, it may be the relative affinity of the chaperone for copper that will determine the activation of the angiogenic promoter.
| Why Copper? Is Copper a Basic Regulator of Growth Control? |
|---|
|
|
|---|
I hypothesize that the key is variability in the environmental availability of copper relative to other cations, which given that the organism has an absolute requirement for copper for survival, determines whether a primitive organism should also grow or remain in a relatively quiescent state.
To sustain this hypothesis, there must be variability in the distribution, or at least the bioavailability, of copper around the world. Is there such evidence? First, there is some evidence that copper content of soils in various parts of the world is quite variable (6063) and the same is not true for other essential trace metals except, occasionally, zinc (61). There is even better evidence for geographic variation in bioavailability of copper (6469). One excellent kind of evidence is that grazing animals such as sheep, which obtain their minerals from grass that obtains its minerals from soil (in the days before mineral supplementation), were not infrequently affected by copper deficiency diseases (6569). This indicates that their particular pastures were deficient in copper. To my knowledge, this kind of deficiency in grazing animals has not been generally described for other metals such as zinc and iron.
So if copper is variably available in the environment, did primitive organisms require copper, and if so, did they develop strategies to cope with this variability? In answer to the first question, yes, they require copper. Yeast, for example, has multiple vital copper-dependent enzymes (70). The essentiality of copper for yeast physiology is in keeping with an elaborate system in yeast to try to ensure adequate cellular levels of copper. In a low copper environment, there is derepression of a series of genes whose products enhance copper uptake (7075). Two genes provide high affinity plasma membrane copper permeases (71, 72). Another gene codes for a metalloreductase that elaborates a reductant into the extracellular environment to mobilize copper (73). The copper regulatory enzymes and systems in yeast seem to be more complex and numerous than for other cations, with the possible exception of iron, and homologues of these same yeast genes seem to usually be present in mammals (58, 59).
Thus, yeast has evolved as if it has had to deal with low copper availability and it has the tools to acquire copper in a low copper environment. Now, the really important question for us is: If copper availability is too low, does it affect the growth (division) of yeast? Certainly in yeast, copper is required for growth, and yeast appear to be more sensitive to relative copper depletion than relative depletion of other cations in terms of growth. Yeast shift from aerobic metabolism to fermentation in the absence of copper, which is less able to sustain high rates of cell division. So the answer to our question (Does copper availability in yeast affect growth?) appears to be ``yes.''
Some recent findings in the organism Podospora anserina add further support for a role for copper as a primitive regulator of growth (7678). P. anserina is a filamentous fungus with a limited lifespan. Mutation in the gene grisea results in reduced growth, along with other phenotypic changes. Grisea codes for a transcription factor, and this transcription factor requires copper for activity. In other words, adequate levels of copper, acting through this transcription factor, are required to maintain a normal growth rate.
Thus, it seems feasible that adequate availability of copper has been a primitive signal that growth should occur. If we assume that certain copper enzymes such as cytochrome oxidase were an absolute requirement for health and growth, relative copper availability as a growth signal would make sense, particularly if copper availability was somewhat variable. Perhaps other essential metals were much less variable in the environment of early organisms.
If a copper requirement for growth was fixed in primitive organisms in early evolution, copper modulation of growth could very well have remained in higher organisms, even if availability of copper became less variable. Moving up to mammals, copper deficiency is a potent teratogen (79, 80), probably more so than deficiency of other metals, and development of a normal mammalian fetus requires normal growth, which requires normal angiogenesis. Although the data are piecemeal at present, it seems possible that throughout evolution copper has played a fundamental role in regulating growth, and in higher organisms, perhaps part of that regulation has been retained by copper regulation of growth factors and angiogenesis.
| My Current Interpretation of the Meaning of These Various Studies |
|---|
|
|
|---|
Second, I believe there will be two separate and distinct levels of copper effects in mammals, holdovers from the early evolution discussed above. The higher level allows angiogenic promotion and growth, as well as cellular health. The lower level does not allow angiogenic promotion and growth, but does allow cellular health. This is in parallel with simple organisms: ample copper equals growth and health and low copper allows the cells to live, but not grow. Also in parallel is the fetus: ample copper equals growth and health and low copper allows the fetus to live, but not to develop properly.
Third, I believe that it is highly likely that lowering copper to levels in between the high and low levels described above, the therapeutic ``window'' discussed in the paper, will offer a safe and effective way to stabilize the growth of a wide variety of advanced and metastatic cancers. Given the evidence that some of the same angiogenic and growth factors are elevated in lymphomas and leukemias, this therapeutic approach may extend beyond solid tumors.
The correctness of my interpretations will determine the eventual usefulness of this approach to cancer therapy and other diseases of neovascularization. Based on encouraging results so far, it may be reasonable to conclude that the war on copper is just beginning.
| Footnotes |
|---|
1 To whom requests for reprints should be addressed at Department of Human Genetics, 4909 Buhl Building, University Of Michigan, Ann Arbor, MI 481090618. E-mail: brewergj{at}umich.edu ![]()
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. Du, G. Filiz, A. Caragounis, P. J. Crouch, and A. R. White Clioquinol Promotes Cancer Cell Toxicity through Tumor Necrosis Factor {alpha} Release from Macrophages J. Pharmacol. Exp. Ther., January 1, 2008; 324(1): 360 - 367. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Cuzzocrea, T. Genovese, M. Failla, G. Vecchio, M. Fruciano, E. Mazzon, R. Di Paola, C. Muia, C. La Rosa, N. Crimi, et al. Protective effect of orally administered carnosine on bleomycin-induced lung injury Am J Physiol Lung Cell Mol Physiol, May 1, 2007; 292(5): L1095 - L1104. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Hassouneh, M. Islam, T. Nagel, Q. Pan, S. D. Merajver, and T. N. Teknos Tetrathiomolybdate promotes tumor necrosis and prevents distant metastases by suppressing angiogenesis in head and neck cancer Mol. Cancer Ther., March 1, 2007; 6(3): 1039 - 1045. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Chen, Q. C. Cui, H. Yang, R. A. Barrea, F. H. Sarkar, S. Sheng, B. Yan, G. P. V. Reddy, and Q. P. Dou Clioquinol, a Therapeutic Agent for Alzheimer's Disease, Has Proteasome-Inhibitory, Androgen Receptor-Suppressing, Apoptosis-Inducing, and Antitumor Activities in Human Prostate Cancer Cells and Xenografts Cancer Res., February 15, 2007; 67(4): 1636 - 1644. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. F. McCarty and K. I. Block Preadministration of High-Dose Salicylates, Suppressors of NF-{kappa}B Activation, May Increase the Chemosensitivity of Many Cancers: An Example of Proapoptotic Signal Modulation Therapy. Integr Cancer Ther, September 1, 2006; 5(3): 252 - 268. [Abstract] [PDF] |
||||
![]() |
M. K. Khan, F. Mamou, M. J. Schipper, K. S. May, A. Kwitny, A. Warnat, B. Bolton, B. M. Nair, M. S. T. Kariapper, M. Miller, et al. Combination tetrathiomolybdate and radiation therapy in a mouse model of head and neck squamous cell carcinoma. Arch Otolaryngol Head Neck Surg, March 1, 2006; 132(3): 333 - 338. [Abstract] [Full Text] [PDF] |
||||