---start---- bbd 9.14.98 prendergast - rational approaches to attacking cancer Q: the MYC oncogene is a potent cell growth activator that is tightly controlled in normal cells, but frequently deregulated in cancer cells. in the deregulated form found in cancer cells, MYC induces malignant cell growth. You have discovered a novel MYC binding protein, DONALD, and suspect it might be important for MYC activity in tumor cells. What questions must be addressed experimentally for you to determine if MYC-DONALD interaction is a target for drug discovery? Focus on the questions to be addressed and suggest general experiments to do so, not actual technical details. Note: Dr P is a molecular biologist, therefore a reductionist :) he will try to relate his talk back to biological problems, however Cancer is very complicated. Today we'll discuss some genes that control it and ways that have been approached to attack those genes as a way to attack the tumor. we'll discuss a class of nontoxic compounds to attack tumors specifically pathway to cancer: cell + environmental insult --> mutation in cell growth regulatory genes --> benign growth (mole, etc), hyperplasia ---> malignant conversion (set of steps we do not understand) ---> invasion ---> metastasis via blood, lymphatics, etc. cancer is mainly a disease of gene mutations. cell focus assay for oncogenes - people looked at DNA in tumor cells- first human tumor cell lines were cultured in the 50s and 60s. Rodent cells and other animal cells have been easier to culture. So anyway, they isolated some DNA - can precipitate DNA with calcium phosphate, and other cells will take up that precipitate and express those genes. After two weeks - normal rodent cells exposed to tumor cell DNA precipitate will overgrow and change shape. those clones were then examined for specific human genes - and they found the Ras oncogene this way using a "rodent assay" Ras is mutated in cancer a lot and is very powerful. one of the first oncogenes found. a normal cell with Ras introduced to it will become a tumor cell -> gene dependent malignant transformation. Human cancers - many have mutations in a small set of genes, including Ras. This suggests we may be able to attack cancer genetically. WHat is Ras? a small GTPase. it binds and hydrolizes GTP. when bound to GTP it is active and sends a signal. When bound to GDP it is off. Ras*GTP is the form that gets stuck on in cancer cells. this is due to a mutation in Ras. larger world view - normally, Ras is on inner leaflet of plasma membrane- growth factor receptor normally turns on Ras, which then turns on genes in nucleus and so forth. Ras signal transduction is very complex. Rhetorical point- it links many growth factor receptors - soluble and solid state growth factors- and then Ras activates a number of cascades - lipid and protein kinases, cytoskeletal signals, etc. when stuck on, generates a lot of traffic. how can we go after Ras? In 1990 they found that for Ras to function it must be posttranslationally modified. Ras as originally produced is inactive and can't cause transformation. needs farnesyl transfase (FTase) to modify it. This FTase is a classical catalytic enzyme. we know how to inhibit enzymes like that. so, we thought if you kill FTase, we'd kill Ras. FTase causes isoprenylation (like phosphorylation, but different) of Ras. the farnesyl groups are intermediates on a pathway leading to cholesterol something or other. there are other chemicals which shunt the intermediates from cholesterol pathway into another pathway. Ras has a CAAX box at the end of it, and FTase adds a fatty group onto that. to make it short, compounds that competitively inhibit binding of this protein were developed. they made a drug that looked like it, that bound CAAX. they inhibit hte enzyme by competing off Ras. If you look in cell,s you see you do block Ras modification using this method. if you titer in the drug, adding more and more of it, gradually you see Ras shifts into unmodified form. so we kill Ras prenylation and this kills Ras function. then you put these cells into culture and you can see that the cells do not become transformed. when you transform cells they usually change shape. when you add drug, these cells become normal. A Ras independent tumor isn't affected at all by the drug. cell biology version of tumor assay - transformed cells can grow without anchorage and will colonize soft agar. other oncogenes - src (needs ras) and raf (ras independent). so adding drug to src and ras tumor cells stops the cancer; using drug in raf cells doesn't bother it at all. what happens in animal with tumor? MMTV - mouse mammary tumor virus - with Ras downstream of it- mammary tumors develop. If you treat with the FTase inhibitor, tumors go away. these animals have no toxicity from treatment either. these drugs are now in phase I trials, starting phase II. definitely nontoxic. no maximum tolerated endpoint. FTIs identify a fundamental difference in the physiology of normal and transformed cells. what is the nature of this difference?? FTIs - questions: what is the mechanism behind the anti-transforming effects of FTIs? These drugs will target any Ras, not just mutant Ras - so why is it tumor specific? How can FTIs be nontoxic yet cause tumor regression? Why do tumor cells persist in oncomice after treatment? Tumor comes back when drug is withdrawn. What is the basis for FTI resistance? a research problem starts with interesting biology; that's why this is so interesting. so what do we do? Ras isn't the main target for these drugs, is the thing. FTIs are neither cytostatic nor cytotoxic - they don't stop growth or kill cells of any kind. they slow them down but do not stop them. reversion phenomenon - very fast. over in 20-24 hrs. dramatic shift in shape for a cell. Ras is a long lived protein and it takes awhile to replenish it. WE're now blocking replenishing ability. You'd think it would take a long time to see drug effect, b/c it takes time to lose the store of long-lived Ras present when you start giving drug. you're trying to get rid of all the Ras in the membrane, should take a while. effect shows up too soon.by day 1 we see an actual effect - but we do not see loss of Ras until after day 2. also, if you add drug, cells revert to benign phenotype, and while Ras recovers after day 7, it takes about 2 weeks for phenotype to revert. also, if you transform cells with a different version of Ras, as long as Ras gets to membrane it transforms cells...but [missed his point here] you can separate biology from effects on Ras. so, inhibition of Ras farnylation is NOT critical to revert the transformed phenotype. there is no correlation in human tumor cell lines b/w Ras dependence and susceptibility to tumor treatment with FTI. cells transformed by farnesyl independent forms of Ras--oops, missed that. ihibiting Ras prenylation happens but isn't really useful... so we are seeing all this stuff but we do not know what is going on since data do not fit original hypothesis.... actin- cell structure - a filament of G actin subunits that makes F actin filaments. several types. actin was the DNA of the 40s. something we used to think was crap taking up space with no information content. Actin shape is important. shape affects function. actin sets up platform for other proteins that bind....integrins mediate contact with extracellular environment. Drugs seem to target cell shape is regulated in part by actin. it wouldn't be so surprising to find biochemical basis of disease having to do with protein involved in shape. Ras transformed fibroblast - actin stained in green - cortical actin is ruffled...we don't know what it means but it is transformation. we also see small cells, sometimes mistaken as cancer cells.... normal cells - also respond to FTIs if you look closely enough. no growth effect but there is a shape effect. the cells get thicker, most stress fibers, they get larger. [argh. i am falling asleep] Ras superfamily - 100 diff GTPases regulating diff phenomena. some upregulate growth, some suppress it, activate actin, intracellular membrane trafficking, etc. these are all prenylated but not farnesylated. some are farnesylated though. RhoB - farnesylated rhoB protein is implicated in stress fiber regulation. FTIs block ras transforamation by ihibiting rhoB dependent signal transduction. predictions of this model: fti's inhibit RhoB farnesylation nd cell localizaition. RhoB has short halflife. dominant inhibitors of RhoB block transformation by ras but not by raf. exctopic expression of farnesyl independent foms...[went too fast...] pulse chase experiment- finds out how long it stays so we foundn that RhoB turns over prettyquick. so biology and kinetics fit pretty good. can we genetically mimic effects of the drug? Transfer oncogenes into normal cells and make clones that transform... RhoB mutants bind to GTP exchange factors and get stuck in on form. RhoB inhibitor prevents transformation....but with Raf you can cause it again. Farnesyl independent Rho - N myristylated RhoB - what happens if you put that ito a cell and a drug blocks it? it goes to internal vesicles, when you add drugs it fans out a litle. biochemically - resistant to FTIs. [argh! i have no idea what is going on!] soft agar assay - control cells lose ability to grow, cell with Myr-rhoB grow. RhoB is short lived, rapidly depleted by FTI tx. dominant inhibitors of RhoB... why does the regression and reapparance occur, with the mammary tumor mice? how come we can treat but not cure (tumor comes back when drug is withdrawn...) can we do anything to fix this? apoptosis - cell suicide that maintains tissue homeostasis - culling process for cells. apoptosis is noninflammatory, normal homeostatic process. not like necrosis. necrosis can occur with certain chemotherapy agents or with radiation. but with this FTI drug, we see apoptosis, not necrosis. Rho regulates integrin function - these are surface proteins that talk to Ras, bind to ECM. Rho proteins promote cell adhesion, ability to attach, actin network, etc. adehsion, angiogenesis, apoptosis... varying adhesion conditions profoundly changes biology of tumor cells. loss of growth occurs when cells are not able to adhere. a normal cell has a physiological adhesion - tumor cell doesn't - tumor cell can grow w/o adhesion. if you lose that ability cell is susceptible to death. FTIs induce apoptosis in Ras transformed cells denied substratum attachment. features of FTI induced apoptosis - p53 independent. kinetics are the same as for reversion, susceptibility abolished by farnesyl independent forms of RhoB malignant cells lacking a context for attachemnet may respond to tx with FTI by undergoing apoptosis - but malignant cells in priveleged locations where attachment is possible,may respond to tx by assuming a benign phenotype that allows them to persist in the organism - without being killed. like musical chairs. if cells can "find a chair" (attachement) they can persist. Rho is an important target of FTIs. deep insight here is that tumor cells and cancer are reallya bout aberrancy of cell adhesion, subverted by oncogenes. normal cells aren't set up to persist outside of normal context. FTIs affect shape and attachment properties, same as oncogenes do. so instead of blocking Ras itself, we go after Rho, which affects shape. FTIs block Ras transformation by inhibiting RhoB dependent signal transduction. hopefully you have some conception of what biochemical approaches to disease can do that biological approaches can't do... proof of principle: i have an idea - if i do this, something will happen. show that any way you do it, the effect will occur. knock out a gene, cause the effect. then if you make a drug to get rid of protein made by gene, maybe it will work. ---end----