The discovery of angiogenesis inhibitors: A new class of drugs: Transcript: Part 2
Judah Folkman:
Oncogenes/ Suppressor Genes
|
|
Oncogenes/ Suppressor Genes There is increasing evidence that the angiogenic switch is governed in part by oncogenes and tumor suppressor genes. To examine this aspect of tumor angiogenesis, what other evidence do we have, besides breast biopsies, that certain human tumors may be non-angiogenic for prolonged periods of time before switching to the angiogenic phenotype?
|
16. Dormant Primary Tumors |
This slide illustrates some of that evidence. For many years pathologists have published case reports of autopsies of people who died of trauma, but who never had cancer in their lifetime. Microscopic in situ cancers were found in many different organs. Then, in 1993 Professors Black and Welch at Dartmouth published in the New England Journal[of Medicine] a review of many of these papers. In autopsies of women from 40 to 50 years of age, 39% had small in situ carcinomas in their breasts (called prevalence), but only 1 out of 100 would ever be expected to have that diagnosis in their lifetime (called incidence). In men over 60... 46% had microscopic in situ carcinomas of the prostate &but only 1 out of 100 would be expected to have that diagnosis in their lifetime. But the most surprising finding was that more than 98% of the autopsies showed carcinomas in situ of the thyroid, but only 1 out of 1,000 would be expected to have that diagnosis made in their lifetime. To make a long story short, I have been collaborating with Professor Lief Andersson of the pathology department at the Helsinki Central Hospital in Finland, who has sent me a whole set of slides of thyroid glands taken from autopsies of people who died of trauma. He has stained them so that the blood vessels are highlighted. In our laboratory we can measure the size of the in situ carcinomas with an accuracy to 1 micron. They range in size from 250 to 500 microns diameter. There are also larger carcinomas in the thyroid ranging up to >4000 microns in diameter. So far in unpublished work I have found that the carcinomas below 500 microns are not neovascularized, but that above approximately 1000 microns ( 1 millimeter), carcinomas are intensely neovascularized. Thus, in this group of specimens it is possible to see that the angiogenic switch occurs at a particular size range. How does the tumor switch on angiogenesis? If that mechanism were understood, it might be possible to turn angiogenesis activity off in tumor cells. Currently angiogenesis inhibitors inhibit endothelial cells from responding to an angiogenic stimulus from the tumor.
|
|
This slide shows a recent paper in the February 2001 issue of the Journal of the National Cancer Institute this year by Robert D'Amato in our department and his student Antonio Fernandez. It has been known for years that when an oncogene is abnormally expressed by a tumor cell that cell survival increases and the cell is more refractory to apoptosis, especially by cytotoxic chemotherapy. The oncogene bcl-2 decreases tumor cell apoptosis in vitro. D'Amato discovered that this same gene also upregulates, a potent angiogenic stimulator--vascular endothelial growth factor (VEGF) and this turns on angiogenesis.
|
<
18. Increased angiogenesis in prostate carcinomas transfected with Bcl-2 |
This slide shows the increased microvessel density when bcl-2 is overexpressed in two types of indolent human prostate cancer growing in a mouse. These tumors were already angiogenic at a low level, but their angiogenesis has been increased further by transfection with the bcl-2 oncogene. |
19. Chemotherapy-Mitomycin C
20. Antiangiogenic therapy-TNP-470 |
This slide shows that when D'Amato treated these tumor-bearing mice with the cytotoxic drug, mitomycin, the tumors lacking bcl-2 were inhibited or eradicated. However, the tumors expressing bcl-2 escaped mitomycin and grew like the untreated controls. The oncogene had made the tumor cells so refractory to apoptosis that the cytotoxic chemotherapy, which targets these cells, could not inhibit their growth. In contrast, when D'Amato used an angiogenesis inhibitor TNP-470, tumor growth was almost completely inhibited. This angiogenesis inhibitor targets endothelial cells recruited to the tumor and when they undergo apoptosis, the tumor cells supported by them die, regardless of expression of the bcl-2 oncogene. In other words, the administration of the angiogenesis inhibitor overcame all of the anti-apoptotic advantage conferred on the tumor by the oncogene. (TNP-470 means Takeda Neoplastic Product and is a synthetic analog of fumagillin discovered by Donald Ingber in our laboratory in 1985 and now produced by Takeda as an angiogenesis inhibitor for Phase II clinical trials in the U.S. It is a very specific endothelial cell inhibitor, but does not inhibit tumor cells at the doses given.)
|
21. Postive Regulators of Angiogenesis |
This slide shows a list of angiogenic proteins, which are commonly expressed in human tumors. It important to understand that a human breast cancer may begin by expressing mainly VEGF. In fact, approximately 60% of breast cancers at the time of first diagnosis express VEGF, as shown by Relf, Harris and Bicknell in Oxford. But, subsequently the metastases may express up to 6 different angiogenic proteins, including VEGF. |
|
In this review published by Kerbel and Rak, at least 15 oncogenes are now known to upregulate expression of pro-angiogenic proteins, and in some cases downregulate expression of angiogenic inhibitors. For example, the ras oncogene upregulates VEGF and downregulates expression of thrombospondin, an angiogenesis inhibitor. V-src does the same. Her2 upregulates VEGF expression. The anti-cancer drug, Herceptin, among its other activities, blocks angiogenesis induced by VEGF. The EGF receptor tyrosine kinase upregulates VEGF, bFGF and interleukin -8, three strong angiogenic proteins.
|
23. Oncogene H-ras |
Jack Arbiser in our lab showed four years ago that when endothelial cells are transfected by the large T antigen of SV40 virusanother oncogene--they become immortal in vitro but are poorly tumorigenic in vivo. There is "no take" when millions of these cells are injected into mice. But when you look under the skin you see a pinhead size little white tumor, similar to in situ tumors. When these tumor cells were subsequently transfected with an additional oncogene H-ras, they became highly angiogenic angiosarcomas, which grew rapidly and killed the animals. The interpretation of this experiment is that the second oncogene increased the angiogenic activity of the tumor cells so that they became lethal to the host.
|
24. Role of oncogenic ras in tumor maintenance |
In another paper, Ronald DePinho and his colleagues at the Dana Farber Cancer Institute, reported that the H-ras oncogene contributes to the angiogenic phenotype because withdrawal of the ras oncogene-in a doxycycline-inducible H-ras mouse melanoma model--led to apoptosis of vascular endothelial cells in the tumor bed and subsequent tumor inhibition. In this same issue of Nature, Bob Weinberg at MIT showed that transfection of the large T antigen of SV40 virus oncogene plus telomerase immortalized cells in vitro, but there was "no tumor take" in vivo until the cells had been trasnfected by H-ras.
|
25. Squamous Cell Carcinoma transfected with Thrombospondin ½ |
And this concept is driven home by another experiment by Michael Detmar at the Massachusetts General Hospital. He used squamous cell tumors that were already highly angiogenic, and he transfected these tumor cells with an angiogenesis inhibitor, thrombospondin. It is a normal protein in our blood at low levels. He collected those tumor cells, which produced high levels of thrombospondin as well as those cells that produced low levels of thrombospondin and implanted these cells into immunodeficient mice. The high thrombospondin producers yielded the smallest tumors, which were almost invisible. The low thrombospondin-producing cells resulted in large highly vascularized tumors. The stunning part of the paper was that the proliferation rate of the tumor cells was similar in all of them. Therefore, the tumor size was governed by tumor cell apoptosis. Inhibition of tumor angiogenesis by high production of thrombospondin led to high tumor cell apoptosis and inhibition of tumor growth. Low production of thrombospondin resulted in high angiogenesis and low tumor cell apoptosis.
|
26. Oncogene/Potential Oncogene Regulation of Angiogenesis |
This slide shows that when pharmaceutical companies screen for a small molecule that can target a product caused to be overexpressed by an oncogene, they usually look for inhibition of tumor cell proliferation in vitro. However, because of the data I just showed, these same small molecules, by inhibiting oncogene products, may also inhibit angiogenic proteins. For example, farnesyl transferase inhibitors block ras and thus downregulate the angiogenic protein, VEGF. Herceptin blocks HER-2 and thus downregulates VEGF. The C225 antibody to the EGF receptor tyrosine kinase or the oral inhibitor, Irressa, down-regulates VEGF, bFGF, and IL-8. |
27. Types of Angiogenesis Inhibitors |
This slide makes the distinction between direct and indirect angiogenesis inhibitors. Most of the small molecules, which are employed to inhibit oncogene products, are considered as indirect angiogenesis inhibitors, for example when they turn off VEGF. Such indirect angiogenesis inhibitors, by targeting a tumor cell product, are susceptible to the development of a form of drug resistance if a mutant tumor cell arises, which makes a different angiogenic factor. In contrast, a direct angiogenesis inhibitor specifically or selectively targets endothelial cells. Examples are angiostatin, endostatin, thrombospondin, tumstatin, and others. Drug resistance rarely, if ever, develops to these inhibitors because endothelial cells, like bone marrow cells rarely mutate. There are more than a dozen of each type of angiogenesis inhibitor. They can of course be used in combination.
|
28. Mechanisms of the Angiogenic Switch |
This slide shows that the switch to the angiogenic phenotype is regulated not only at the genetic level, but at epigenetic levels as well. For example, in the tumor cell, angiogenesis is genetically controlled by oncogenes and tumor suppressor genes as we have already discussed. But the host's genetic background also plays a role in determining the angiogenic response. For example, Robert D'Amato in our department is an ophthalmologist. Ophthalmologists know that approximately 11 million patients in the U.S. have macular degeneration, a disease in which new blood vessels grow abnormally in the back of the eye (choroid), and which may lead to blindness, but that the disease is rare or nonexistent in African Americans. This observation led D'Amato to demonstrate that mice with different genetic backgrounds may have either a low angiogenic or a high angiogenic response to a standard angiogenic stimulus (such as a VEGF or bFGF pellet implanted in the cornea). This was published in FASEB [Federation of American Societies for Experimental Biology Journal] last year.
|
29. Id1 and Id3 required for neurogenesis, angiogenesis |
Another piece of evidence that the host's genetic background can regulate angiogenesis is this report by Robert Ben Ezra and his colleagues at Sloan Kettering, in collaboration with Richard Hynes at MIT. When two developmental genes Id1 and Id3 are deleted in mice, the embryos die with pulmonary hemorrhages. However, if one allele (pair) of Id1 is deleted and both alleles of Id3 are deleted, relatively normal appearing mice are born, except that when the mice are implanted with tumor cells, there is no angiogenic response from the host or it is extremely weak. As a result tumors do not grow, nor do they metastasize. Ben Ezra's next experiment was to do a bone marrow transplant from mice carrying normal Id1 and Id3 genes. It has been known for the past 5 years since the original reports of Jeffrey Isner in Boston and Shaheen Rafii in New York, that endothelial progenitor cells from the bone marrow are found in the circulation at low levels. These cells can home to the neovascular bed of tumor, where they take up residence among the endothelial cells recruited by the tumor from the local neighborhood. However, the bone marrow--derived endothelial cells found in a tumor bed do not make up more than approximately 1% of the microvascular endothelial cells in the tumor bed. In contrast, in Ben Ezra's Id1, Id3 deficient mice, where there were virtually no blood vessels at the site of tumor cell implantation, the bone marrow derived endothelial cells supplied almost all of the endothelial cells to form new tumor vessels. Large lethal tumors then grew at these sites. In one of the control experiments, the angiogenic protein VEGF was implanted in the opposite flank in a collagen gel, called Matrigel. The bone marrow--derived circulating endothelial cells also homed to the implanted gel, demonstrating that they are attracted by the VEGF.
|
< previous section | next section >
More: The discovery of angiogenesis inhibitors: A new class of drugs
|
