Colon Cancer Resources

today we're going to talk about signal transduction, especially we're going to be focusing on reactive oxygen species and reactive nitrogen species, the take home point is low dosees of these chemicals will stimulate proliferation of cancer cells, high doses will cause apoptosis

of the cancer cell. so we mentioned before about nitric oxide synthase using arginine as substrate to make nitrous oxide and citrulline, and elevated calcium is one of the things that activates the nitrous oxide synthase. the key second messenger is

cyclic gmp, cyclic amp goes up within minutes after you add a stimulus. cyclic gmp takes half an hour to an hour in cells to go up. and in particular, we're going to focus on soluble g uanylyl psych lace, the final product is cyclic gmp.

so this is sort of a cartoon illustrating plasma membrane and calcium enters from the extracellular side to the intracellular side of the cell. and then this in turn leads to activation of the nitric oxide synthase, when the nitric oxide goes up it can interact with

irons illustrated by these stars that are in the soluble guanylate cyclase, it goes up and then it's degraded by the phospho-diesoteric, a transient signal. in the cell, there's a receptor for the atrial natriueritic

peptide, this binds with high affinity and activates a kinase and in turn it has a gunalylate cyclase, it's used to prevent daniel to the heart. so the anp receptor is big, a little over a thousand amino acids, and on the outside we

have about 450 amino acids binding the anp, crosses the memorial brain, there's a kinase component that can phosphorylate proteins and a gunalylate psych late. it's bound to the membrane, this has 700 amino acids, this is 130 amino acids.

the cyclic amp can go up through the soluble guanalyl cyclase. this is what stimulates the enzyme activity. there's a hemebinding component, that's where the iron is in the heme, there's a dimerization component, and then the c terminal a catalytic component.

when the iron is nitrousylated, enzymatic catalysic. pkg causes phosphorylation of protein leading to vasodilation. there's also a cyclic nucleotide beta channel which translates signals to nerve impulses and

then final target is the phosphodiesterase. and we note that one of the drugs that's used for vasodilation is viagra, and basically what it does is that inhibits phosphodiesterase so that the cyclic gmp stays elevated for longer period of

time. in the lab we can use several chemicals to deliver the nitric oxide, such as sper/no, and when add this agent cyclic gmp goes up, protein extracellular regulated kinase, erk, gets activated. cyclic gmp goes up in time

dependent manner and the p44/p42 erk is tyrosine phosphorylated. low doses of nitric oxide stimulate psych limb gmp levels leading to erk tyrosine phosphorylation, leading to proliferation. we'll look at what high doses of nitric oxide do.

and high doses of nitric oxide induce cellular stress. and might mitochondria can be disrupted, activating caspases causing cellular apoptosis. >> (inaudible). >> i think this would happen in every cell. that's why you have to be

careful about the way in which you deliver your nitric oxide. so if you were just to give it intravenously it would go all over the body and create havoc. so sper/no causes phosphorylation of p53, resulting in less g1 to s transitions, the s-phase is

where we have proliferation, duplication of the genetic material, and when we compare this g1 to s transition the cells can undergo apoptosis. the goal is always you want to kill the cancer cells and leave the normal cells alone. so nitric oxide donors you find

in the literature the snap, and snp, and this causes apoptosis. my particular specialty is lung cancer, so we'll see a lot of data today about lung cancer cells. and so in lung cancer, when you add snap, the proliferation is decreased in a dose-dependent

manner, and the reason is with time these cells will undergo apoptosis, but we see when we add more snap we get more essentially nitric oxide being delivered to the cells and in turn it's metabolized then to nitrite. and another proliferation assay

is mtt assay, so we get a dye being absorbed in the cells, the dye turns blue and gets color and so you get high absorbance then at 540 nanometers. when you add dea/no or papa/no it goes down, undergoing apoptosis, this assay takes a couple days to perform.

and what you can do is you can load your cells with dyes such as the daf and then in turn you could measure fluorescence and you see then when you're adding the papa/no fluorescence goes up, nitric oxide is being delivered and stays up for an hour in the cell.

and another assay is called the clonogenic assay, it takes weeks, add increasing amounts of papa/no the number of colonies is impaired. and another thing you can do is macrophages will deliver nitrous oxide, and so you can add this in the colony assay and you see

the more macrophages you add, the number of alcohol anies that you get is -- colonies is reduced significantly. so you can get nitrous oxide from delivery agents or you can get it from macrophages. and it will impair the growth of the lung cancer cells.

so one thing that happens is snpl activate a different map kinase, p 38 map kinase, mediator of caspase 3 associated apoptosis, you can have a map kinase inhibitor sb 202290 protecting the cells from nitric oxide mediate the cellular death.

another thing that happens is survivin is reduced, sighful is critical for cell cycle progress regulars. in summary when we have high doses of n-o we can compare bcl-2, reactivate p 38 map kinase cytochrome c activation,

will then cause caspase 9 activation, which leads to caspase 3 activation, and cell death by apoptosis. so we see low doses ofen isgood for proliferation, high doses cause apoptosis but in patients we don't have anything that can deliver the amount of n-o

required to inhibit cancer growth in people. so that's still an area of investigation. so in terms of oxidants then, we can activate the map kinase with low doses of reactive oxygen species leading to or we can activate nf kappa b

which will cause release of cytokines leading to inflammation and cancer cell survival. so the erk pathway various growth factors and reactive oxygen species reactive nitrogen species can activate raf, a kinase which can activate mek, a

tyrosine kinase, and in turn that can activate erk, that gets phosphorylate and can go into the nucleus, causing proliferation, expression of nuclear oncogenes, high doses of reactive oxygen species then it will activate mek 3 and mek 4, lead to p 38 activation, and jnk

activation and stress responses. this is a cartoon to illustrate what goes on in the cells, the reality is the reactive nitrogen pathway and reactive oxygen pathway go on at the same time. if we have high doses of n-o it will react with oxygen, leading to nitrogen dioxide, and then in

turn this can lead to nitrogen, dinitrogen trioxide and target nucleofiles. these can occur outside the cell. we have our nitrous oxide synthase reacting to produce nitrous oxide and cytrillene. there's the cellular membrane.

and then this is going to produce more n-o and we get the nitrous oxide synthase going. and the n-o across the membrane and go into the extracellular component. and then again the n-o will form nitrogen dioxide, and that can form n2o 3 causing nitrosilation

of protein, extracellular or membrane protein. so these reactions all occur very rapidly within a minute. and then numerous cellular proteins are nitrosylated such as ras that gets nitrosylated, cysteine, resulting in

activation of map kinase cycle. then in terms of caspase 3 it can get nitrosylated to cause apoptosis it has to be denitrosylated. both the nitrosilation and denitrosilation are important. here we're looking at the effects of hydrogen peroxide on

cytosolic pre-calcium and so when we add hydrogen peroxide we see the nitric oxide activity increases, and also what increases is extracellular calcium. and a control for calcium is ionomycin, we see ionomycin elevates calcium and increases

nitric oxide synthase. and this is a complicated slide that we're going to spend a little bit of time on. and this gets along without a cell and is transduced cytoplasmic oxidative event. n-o synthase activation leads to nitrous oxide which can cause

complexes and then cytochrome c oxidase also gets nitrosylated. we'll talk about these shortly. caspase can also get nitrosylated. and then reactive nitrogen species can cause tyrosine nitration of nf kappab and keep nrf2.

basically we have a signal then and the signal is often stimulated by an increase in cytosolic calcium, and then the calcium can lead to depolarization of the cells, sending a signal. and so with nitric oxide synthase activation, the n-o can

affect hemebinding protein, we can get nitrosylation, and can react with superoxide to form proxy nitrite and this results in tyrosine phosphorylation proteins. when we start an event such as increasing nitric oxide a whole

cascade of reactions can occur in the cell. and one of the ones that is pretty important is with the protein tyrosine phosphatase. we'll be discussing soon about the egf receptor and how it can tyrosine phospho-late, very important in regulation of

growth in cancer cells, and one thing that happens when you get activation of the egf receptor is reactive oxygen species are released such as hydrogen peroxide, and then this in turn impairs the activity of the protein tyrosine phosphatase, so that more proteins get tyrosine

phosphorylated. with nitrous oxide this will react with the cysteine on the protein tyrosine phosphatase and again enzymatic activity will be greatly reduced. and here you see a sequence of amino acids and here is the crucial cysteine that gets

affected. so when we get activation of the egf receptor then that can lead to activation of cyclooxygenase 2, and when that is increased in cancer cells, prostaglandin e2 goes up and can bind to the ep2 receptor, and this can then increase vegf which leads to

angiogenesis. and ep2 receptor then can cause transitivation of the egf receptor. so these little g protein coupled receptors, they can't tyrosine phosphorylate protein but the kinase domain can tyrosine phosphorylate proteins

and kick in the growth cycle. so every cell has leukotriene, and prostoblandins, it can be metabolized to leukotrienes and to prostoblandins. so some cancer cells prefer choose to use this pathway. the aracidonic acid can get metabolized to form

prostobldins g2 and the synthase produces intermedia forming vgi 2 but the cancer cells that i've studied it leads to pge2. and there's two types of cyclooxygenase present, 1 and 2. 1 is expressed in many normal cells, especially the g.i.

tract. it's inhibited by nsaids such as aspirin. cyclooxygenase 2 is induced in inflammation and neoplasia by many factors such as - we'll focus on up determinal growth factor receptor, but other things can do it as well such as

tgf beta, hypoxia and uv light. and 2 is inhibited by aspirin and celecoxib, and a lot of clinical trials are done with celecoxib in cancer patients. it has 600 amino acids, and a heme group, so it interacts can a distal heme at 193, and an adjacent heme at 374, and then

aspirin can actually acetylate this enzyme, totally inactivate and gets degraded. cyclooxygenase has a few less amino acids but it's got a heme that binds to a distal histodene at 206 and adjacent heme at 387 and acetylates at ser 529. the problem is with aspirin, it

can cause stomach ulcers, so it's a chemical that you have to utilize very carefully. whereas the celecoxib has less side effects associated with it. an animal model that we've used for lung carcinogenesis is a/j mouse and you can give a carcinogen and we often gave it

urethane, which is varnish, if you ever varnish your floors be careful. and you see in the mouse then after two months you start getting these nodules, adenomas in the lung. and so the assay is simply you count the number of nodules in

the lung, it's a function of time, after administration of drugs. so here is an amino cyto chemistry experiment staining, the brown color shows it in the alveoli and bronchi. so i was surprised when i first came to nih someone said you

should try looking at nonsteroidal anti-inflammatory drugs in the lungs, and indomethacin, it's like a strong aspirin, it slowed the tumor growth. this is in the a/j mouse, the number of tumor nodules was reduced.

and we actually wrote up an abstract for this, for the aacr meeting in 1996, and i was astounded. i had to do a press release on drugs in non-small cell lung cancer, shown to be effective in colon cancer but in lung cancer they are not quite as good.

and this press interview i recall there were are lots of people there, and one of the reporters said, well, does this mean if you take aspirin every day you can smoke cigarettes? i said, no, that's not a good idea, because it's not going to stop the cancer from coming, but

if you don't smoke then you won't get the cancer. so you don't have to worry about it. and then the reporters asked me, they said, doctor, do you take aspirin every day? and i just sort of laughed. i said, no, i'm allergic to it.

i can't take it. i'll get stomach ulcers. [ chuckles ] so it's amazing sometimes when you do these press interviews what you have to say and where they are trying to lead you. but as i mentioned then, in colon cancer these drugs are

more effective, and they found in an animal model oral celecixib, cox2 inhibitors reduced 79%, lung cancer patients treated serum vegf declined, key for angiogenesis. so here we're looking at a series of drugs, and their effects on prostoblandin levels.

aspirin is effective but you have to use very high doses. and then ten years ago someone decided, well, let's couple nitrous oxide to aspirin, you see it's potent, strongly reduces the pge2 levels, we were working with some sulfur drugs and they were also quite

effective, valueproate and sulindac. dup-697, inhibits peg2. and peg2 when it's produced it interacts with this g protein coupled receptor, ep2 receptor, this has about 360 amino acids, and what it does is when it's activated by the pge2 it

elevates cyclic amp, goes up within minutes, we looked at the mrna for the ep 3 receptor, present in many of our lung cancer cell lines. and then in terms of binding, the pge2 bound with pretty good affinity about .04 micromolar. and other components worked

nearly as good. such ales pge2, pgi2, an antagonist ah68 on6809 thatwould block the receptor and antagonize it. pges2 increases cyclic amp, ah6809 is reversible antagonist that blocks the receptor. so here we're looking at the

cyclic amt level, we see when we use 1 1 micromolar there's elevation of the cyclic amp three-fold and this is blocked by the ah6809. and then the cancer cells can actually make vegf and secrete it, and then when the cancer undergoes metastasis, one of the

tumors become big, they undergo and tumor vessels from the host grow in to provide more oxygen and nutrients and the vegf will stimulate the endothelial cells that have vegf receptors. and so we can stimulate with tma which elevates protein kinasec which delivers n-o, that will

activate the cox2, cox2 will increase pge2, which will then bind to the receptor, and increase the vegf. and so the vegf mrna is increased when we add pge2, or egf, and it's blocked by h89 which inhibits the protein kinase.

and so here again we're looking at these mice that get colon cancer. and if we do a knockout mice, getting rid of all the cox2, the vegf protein is reduced by 94%. and non-small cell lung cancer patients, cox2 mrna expression is associated with vegf mrna,

increased microvessel density, decreased patient survival and early relapse. so the cox2 then we see it can get turned on by nitric oxide, but it can also get turned on by the egf. and so here we have a dose-response curve, when we add

egf we see the cox2 mrna readily increasing, and this is a monoclonal antibody that blocks the receptor, and the cox2 increase caused by egf is so the epidermal growth factor receptor is a large protein, a acids, we have an extracellular domain of about 500 amino acids.

this binds the egf, and then the tyrosine kinase activity is turned on, and it leads to tyrosine phosphorylation of protein. cancer cells can use other growth factors as well, such as the igf 1 receptor, to stimulate their growth and then here we

see the vegf receptor which is present on the endothelial it's got a tyrosine kinase that can be activated, causing various proteins as well. so the tyrosine kinase receptors then are key to stimulating proliferation in the cancer cell and the efg receptor we see 1186

amino acids, crosses the membrane just once. g protein coupled receptors cross the membrane seven times. and there's a 542 amino acid intracellular domain that contains tyrosine kinase activity. and what happens is the slicing

at 721 binds atp, takes the phospate and transfers it to various tyrosine amino acids. so various proteins get tyrosine phosphorylated, such as pi3 kinase, phospho-liepase c gamma and will tyrosine phosphorylate itself when done with other things.

so the egf receptor forms a dimer when it binds egf, and then in turn we can tyrosine phosphorylate the pi3 kinase, and this will activate a survival cycle whereby akt gets phosphorylated, vcl 2 is then expressed, leading to cellular survival.

so we mentioned egf, it increases pi3 kinase, but one of the things egf does is causes a transient increase in the hydrogen peroxide which inactivates the protein tyrosine phospho-phase, addition of egf causes hydrogen peroxide.

the disulfide then can be reversed by the chemical thioredoxin. egf receptor causes reactive oxygen species to be released when activated. and we see here that the egf receptor is normally tyrosine phosphorylated by various

entities in cancer cells, they often make transforming growth factor alpha. but then a very curious thing is what's called transitivation. here you have a g protein coupled receptor when it's activated it can lead to tyrosine phosphorylation of the

egf receptor, and this is very rapid. it occurs within minutes. so we're adding pge2 then and we see that we're getting tyrosine phosphorylation of the egf receptor, we know it's going to the ep2 receptor because we add antagonists, ah 6809 and the

blocks it, no effect on total, just tyrosine toes phosphorylation of the egf receptor. what we're studying is tyrosine phosphorylation of py1068 of the egf receptor. so for the egf receptor then they came up with a tyrosine kinase inhibitor gefitinib, it

causes phosphorylation, but it's blocked by gefitinib. and the reason being is what's happening is the transforming growth factor alpha that's present in the cells is being released and then that's binding to the egf receptor and causing the tyrosine phosphorylation.

so egf will also tyrosine phosphorylate by itself and gefitinib will block that tyrosine phosphorylation. so gefitinib then is being used to treat patients but it doesn't work in all of the patients. it turns out in lung cancer about 15% of the patients have

mutated egf receptors. and so these mutated receptors are sensitive to gefitinib, it's only 15% of the patients, 160,000 die each year from lung so 15% of that is over 20,000 patients that respond, so these are big numbers then. so this sort of shows a

schematic of what's going on in the cells so we mentioned before it's not just reactive nitrogen species, not just reactive oxygen species. all of these things are going on in the cell at the same time. and so the egf receptor we know can stimulate cox2 expression,

so can nitric oxide. when cox2 expression is increased, you get more pge2, and that can bind to the ep2 so what can happen then is this receptor can activate a protein called sarc and matrix with alloprotease causing release which stimulates egf receptor,

this is very rapid. this whole thing occurs in a minute. but when the ep2 receptor is acid cyclic anp can go up, causes phosphorylation of creb, this takes several hours. this cycle is very rapid. we have all these things going

on in the cell but the time is different. here we're talking about minutes, here we're talking about hours. so going back to these nsaids we studied these sulfur valproate and sulindac tear dicofenac wasn't as good, this

is a xenograft study in nude mice. in cancer you use nude mice, they don't have fur, that's why they look nude. but the key thing about them is they don't have an immune system. so then you can put in these

human tumor cells and human cancer cells and tumors will then form and they will grow exponentially. and so if we inject these drugs then into the animals, sulfur valproate and sulfur dicofenac tumor growth is dramatically slowed.

so then what you can do is you can measure the tumor volume in the animal, the tumor also are just under the surface of the skin. and so you see the volume in millimeters cubed is almost 2000. and if you give enough valproate

the tumor volume is reduced 80%. this is a control that the drug is dissolved in, no effect. and the valproate is much more effective than the dicofenac. what you also try to show when you do these animal studies is that you're not causing toxicity so the animal weight should be

the same. this is what's done pre-clinically, we have the capacity at at nci if the drugs look good in animal models we can develop them and apply for investigational new drugs with the fda, test them in phase phase 1 clinical trials in building 10

to make sure they are not toxics, and phase 2 clinical trial looking for effectiveness, if something is then effective you can go to phase 3 clinical trials but what happens is nci sells the license to a big pharmaceutical firm and then the pharmaceutical firm does the

phase 3 clinical trials at about a dozen different institutions. all over the u.s. and sometimes in different parts of the world. so at nci we've developed about 30% of the drugs that are used to treat cancer patients. nsaids we were testing. and so the valproate is a

cytostatic agent, but the nitric oxide aspirin is cytotoxics and we're getting apoptosis. in the end it causes toxicity to normal cells as well, and so it failed clinical trials in patients. another thing that can happen with these nsaids is what's

called epithelial mesenchymal transition, cancer cells initially you have the primary tumor such as lung, it can undergo metastasis to other organs, lung cancer goes to the liver, lymph nodes, bone and brain. and these epithelial mesenchymal

transitions then are facilitated as the cancer progresses. and so one thing we found is how are these nsaids inhibiting proliferation, well, we found they were impairing these epithelial mesenchymal transitions, so when added sulfur valproate, e-cadherin

went up and also we found fomentin went down. cadherins play a major role in epithelial cell adhesion. in cancer, epithelial cells are sensitive to drugs such as chemotherapies. here we see the e-cadherin, and it's in the membrane and

interacting with itself and holding the cells together so that they form an epithelial shape. and the e-cadherin interacts with beta-catenin. mesenchymal shape cells are more rounded. here we see dose-response curve,

this drug, s-valproate, zeb3 is going down, indicative of impairing mesenchymal transition. here we're see it's a function of pge 2, if e-cadherin goes down when activated by zeb-1 then we go from the epithelial phenotype to the mesenchymal

think oh phenotype. great structural composition, mesenchymal the cells are rounded so they can migrate. and then what was found is if the cells are of the epithelial type they respond better to tyrosine kinase inhibitors,

gifitinib in the u.s., erlotinib in europe. , transfects to cell line, increasing responsiveness to gifitinib. we added inhibitors, increased potency of gifitinib, and this is an mtt assay, clonogenic assays with lots of colonies.

the valproate a little bit, the gifitinib, when you combine them you get dramatic reduction in the number of colonies. so in summary then the key point is low doses of nitrous oxide as well as reactive oxygen species can increase proliferation so the nitrous oxide leads to

increased phosphorylation of the erk, at high doses of n-o p53 is phosphorylated, p38 map kinase is activated, leading to cancer cell apoptosis, cox2 activated, activate of ep2 receptors, and cge 2 increasing vegf causing transactivation of the egf receptor leadng to increased

pro livation. the s valproate inhibit cox2 and decrease proliferation of non-small cell lung cancer cells and tumors and increase e-cadherin expression, s valproate, cells have a wild type egf receptor. that's about it for this

lecture. any questions? yes? >> (inaudible) >> right. so as dave wink mentioned, last week, basically if you have nitrous oxide level .1 micromolar or less can stimulate

proliferation, 7 to 1 will inhibit proliferation and cause apoptosis. so especially it impairs with the electron transport chain, high doses, and so the cells don't have energy and cancer cells need lots of energy because they are programmed to

do one thing, and that's grow. so they need lots of energy. and if they don't have it, they will die. those just undergo apoptosis. that's their weakness, they are programmed to do one thing. but the key to cancer cells is they are mutating all the time

and so in patients they often find that one therapy is effective, and then a year later the tumor regrows. and it's because the tumor is mutated into something else. so with cancer you're always trying to hit a moving target. and that's the thing.

you know, the clinicians they find a great drug and they think the patient is cured of cancer. but they are always worried that it can mutate and come back in a different form and usually the further mutations the cancer is even worse and it will respond less to the drug.

>> repeat the question. >> it's more stable but as you saw, they did the nitric oxide aspirin, and it certainly was more effective at killing cells, but the problem is there was too much toxicity and it killed normal cells as well.

so that's always the trick with the cancer drugs. and that's why we do the clinical studies. you're looking first to make certain it's not toxic, and then secondly you're looking at efficacy. will it stop the cancer growth?

and so with the clinician there's always this balance of therapeutics versus toxicity, you you want to maximize therapeutic and minimize toxicity. at the clinical center now there's a big redoing of the organization because basically

there was sort of too much toxicity going on and so now they are redoing the clinical trials to get back on track. and maximize the therapeutics and minimize the toxicity. so you're always under a lot of pressure with all these new drugs the pharmaceutical

companies develop and they want us to test them and they want their drug to ultimately be approved by the fda, so that it can be used in patients and they will make big money. but we have to sort of evaluate the therapy versus the toxicity and make certain that, you know,

the drugs aren't going to harm our patients too much. okay. we'll move on then. dave roberts is here. [off-mic discussion] ph.d. from university of michigan, good football team

this year, he came to nih first he was niddk and then joined the nci, chief of biological pathology section. and he's pathology. dave. >> okay. thanks, terry.

can everyone hear me? so fortunately terry went through a lot of pathways also involved in angiogenesis so maybe in addition to starting a few minutes early i can get you out early because we don't have to spend that much time on some of the signaling pathways.

so what i want to convey to you today is the idea that angiogenesis exists in a much broader physiological context and terry and i are both interested in the perspective of cancer but in pharmacologically trying to target angiogenesis and for cancer you want to

inhibit angiogenesis to limit nutrients and oxygen that are provided to tumors. but so far one of the disappointments of angiogenesis inhibitors is they have side effects. what i'd like to show you is how redox signaling can explain some

of these side effects. and so i broadened the topic to angiogenesis tissue perfusion and stress responses because understanding how agents intended to block one are actually altering other physiologic pathway to understand the physicallality

and pharmacology. i'll introduce some of the angiogenic factors, and angiogenesis inhibitors, how each of these interactions with redox signaling. then go more in depth into the first identified physiological angiogenesis inhibitor, affects

redox signaling and how we can therapeutically target angiogenesis inhibitor to control responses and then as i mentioned take you outside of the angiogenesis box to understand how both endogenous and therapeutic angiogenesis inhibitors may have other

effects on physiology that are relevant to their use. finally extending to stress as an endpoint which can also be controlled through the thrombospondin pathway. terminology, terry mentioned angiogenesis which in adults is important for existing blood

vessels to grow into a tumor or wound beds to regain profusion during wound repair but developmentally you have to first generate a vascular plexus, and this process early in development is known as vasculogenesis. this is de novo, in the mouse,

there's no circulation, just a mass of cells and the primitive mesenchyme differentiates to form the first endothelial cells which can form blood vessels. this is vasculogenesis. they have to polarize, the heart starts beating to move blood through the plexus, and so it

has to polarize into arteries and veins and arterials, that's the arterio genesis step. once a plexus is formed, then so later in development and in the adult for wound repair and for cancer angiogenesis is important. and because vascular system is

pressurized, in higher animals, and the endothelial cells in vessel walls are somewhat leaky, you have diffusion, ultrafiltrative plasma that leaks out of the vessels and in higher animals from fish on up we have a second vascular system which collects what leaks out of

these vessels and returns it to the central circulation, and this is lymphatic system and so the formation of that tree is a second independent process known as lymphangiogenesis. these processes are controlled by a series of growth factors. terry mentioned vegf.

vegf is -- i'm sorry -- is the first member of the vegf family, vegf-a, it and other members of the vegf family can bind to vegf receptor 1 and 2, kinase receptors. and binds to another set of receptors important for lymphangiogenesis, and the focus

for cancer drugs has been on vegf, as you'll see, but there are also other factors important for angiogenesis and turns out the vegf signaling through receptors is necessary for forming blood vessels but not sufficient to stabilize blood vessel, you need secondory

signaling including angio poietin which binds to tyree septemberors. t y e c p o

r s . the angio poietins mediate stabilization of vessels, and then the lymphatics respond to vegf c and d but they uses a different vegf receptor, vegf receptor 3.

although vegfs are considered to be dominant angiogenic factors, at least during development because if you knock out vegf or even one copy of vegf or its receptor, this is embryonic lethal. so this is clearly a necessary pathway for embryonic

angiogenesis in development, but in cancer there often is upregulation of other angiogenic factors which can ultimately bypass the role of vegf, and that will turn out to be a problem for inhibiting. physiologically all of these angiogenic factors are balanced

by a large number of angiogenesis inhibitors, we'll be talking specifically about thrombospondin 1, i want to leave you with a list so you're not thinking this is all so simple. that it would be easy to control.

in fact there are multiple angiogenesis inhibitors physiologically. there's a lot of excitement about ten years ago with the fda approval of the first angiogenesis inhibitor for treating cancer. and this was avastin.

generic name is bevacizumab. this is an antibody that binds to vegf and sequesters it. this is approved in 2004 for colon and subsequently other cancers, antibodies bind to receptor and block vegf binding, and all of these initially had some good activity, effective in

subset of patients, but they tend to prolong life by up to six months, sometimes a year, but then they stop functioning and terry mentioned that cancer is very adaptable, and so it turns out that there are ways of getting around this. there are also vegf kinase

inhibitors, small molecules that have been approved for treating cancer, recently approval of decoy vegf receptor, a sponge to absorb vegf and prevent it to getting to its receptor on endothelial cells. another approach realizing there are other tyrosine kinase

inhibitors important for driving angiogenesis to use what was officially considered failed drugs because they are not so specific, all inhibit vegf receptor and other tyrosine kinase inhibitors, a number of these have gained use in treating several types.

another strategy is downstream, mtor, there are drugs that target this that proved to be useful. and back in the 1950s a drug thalidomide was developed and given to women for treating morning sickness and turned out the side effect were deformities

in the babies that were born, and although the drug was withdrawn from that use, it got reborn in 1990s with the discovery the reason this caused developmental abnormalities is because it's inhibiting angiogenesis. and so thalidomide itself which

has fairly non-specific activity but also some derivatives of thalidomide which is much more specific for angiogenic activity is now approved for treating several cancers. now, in addition to the fact the tumors can get around the activities of the angiogenesis

inhibitors, another problem has been that they share certain types of side effects, and what i'd like to show you is why redox signaling can help us understand side effects. so looking at bevacizumab through various studies as well as kinase inhibitors, very

common side effect is hypertension. they increase blood pressure. they also induce thrombosis. and different degrees with different inhibitors but a common theme for all of these. and in the past couple of years it's become clear they also have

effects in inducing cardiac dysfunction. so in general, these are meant to just inhibit vegf receptor 2 on endothelial cells inhibiting angiogenesis but why are they having all these other effects? so let's go off and look first at redox signaling and then come

back to that question. so redox signaling has direct effect on angiogenesis, both from the perspective of angiogenic factors and angiogenesis inhibitors. if you look at the factors, this is true for vegf and some others that i'll show you, these pro

angiogenic signals true tyrosine kinase inhibitors such as terry discussed in the first hour can lead to activation of the number of intracellular signaling pathways, changing intracellular calcium, and inducing the synthesis of a number of redox-active transmitters.

vegf receptor for example activates enos, and this is important for angiogenesis, but the n-o dependent signaling is important for regulating blood pressure and hemo stasis, at the level of vegf. they also change level also of

superoxide synthesis, which is important for controlling nitrous oxide signaling and also as terry mentioned for regulating the protein tyrosine phosphatases. we'll go into this in the upcoming slides. finally calcium signaling

activation of hydrogen sulfide synthesis controlling angio general is -- angiogenic responses for several pathways. terry introduced this clearly, at low levels of nitrous oxide signals tend to be pro angiogenic because activation -- (indiscernible) pro angiogenic

signals, and regulate through the prostoblandins pathway, another way n-o works is through inactivated pro hydroxylase, which elevates vegf, so there are multiple ways that n-o controls synthesis of vegf. and again high concentrations which are typically made by

endogenously by inducible nitrous oxide synthase which is nos2 rather than enos, which is nos3. terry introduced peroxide has biphasic effects, stimulatory at low concentration, inhibitors at high concentrations, and this holds true in endothelial cells,

i won't belabor it, i discussed inactivation of protein tyrosine this can occur as terry mentioned by nitric oxide but another important way of inactivating the tyrosine phosphatases is reactive oxygen species which oxidize to

sulpenic acid, net effect is it limits the duration of pro stimulatory effects through tyrosine kinase receptors such as vegf. and i don't think you've had any introduction to hydrogen sulfide. some last week.

so just a little bit about it, hydrogen sulfide evolved over the past decade from early estimates were that circulating levels were on the order of 10 to 100 micromolar, that turned out to be an artifact of the message used, and now circulating levels in plasma are

mechanized to mechanic recognized to be on the order of 1 micromolar. it is made by three enzymes, beta synthase, gamma lyase and sulfur transferase, several pathways can form hydrogen sulfide, it gets metabolized, reverse pathways can be

converted back to hydrogen and like nitric oxide it doesn't exist on its own very long, it tends to form complexes with iron sulfur clusters and to polymerize and modify reactive glut athone, signaling involves a number of targets so it's not as clear as nitric oxide

signaling with the most sensitive target for nanomolar levels. a number of signaling targets have been identified, it's still being worked out which is the physiologically developed. angio general vegf receptor 2 by hydrogen sulfide.

so there are multiple biochemical targets, metals that ligate thiol and modification and formation and the downstream effects involve a lot of organ systems, it's important for neuro modulation, and some animals but not humans, can induce hibernation, important in

ischemic response, oxygen sensing and angio angiogenesis,you may hear more later in this lecture series. we don't have time to talk about it today. focusing on angiogenesis this is one of the early studies that showed that administering

hydrogen sulfide systemically to angiogenic ananimal, here compared to one of the classic pro angiogenic factors, this is from one of our early papers that showed that increasing levels of hydrogen sulfide increased proliferation of endothelial cells and

increased hsp27 which can regulate survival and stress responses. and as i mentioned, vegf receptor is a hydrogen sulfide target, published in 2013, they identified cysteines 1040, 1045, at which form a disulfide and hydrogen sulfide can react with

these and this modulates the confirmation of vegf receptor, and controls signaling. so a mutant form of vegf receptor where 1045 alanene couldn't form the disulfide bond shown to have higher kinase so hydrogen sulfide reacting with the disulfide and

preventing the intrasubunit formation is a way of activating vegf receptor 2 and accounts for part of the pro angiogenic activity of hydrogen sulfide. in terms of tumor growth, an interesting study from 2014 that showed that cancer cells can be proangiogenic through production

of hydrogen sulfide. they showed that exposure to the cancer cells at a pro angiogenic activity caus migration of endothelial cells, an this inhibitor inhibits an important pathway, showing this proangiogenic activity at the cells could be blocked by

inhibiting the synthesis of hydrogen sulfide through the enzyme. t cells are infiltrating at the tumors, when they become activated through the t-cell receptor we found that this induced mrna and protein levels of these

synthetic enzymes, then released in the t tumors, there are multiple source also of hydrogen sulfide that can drive tumor angiogenesis responses. so looking more specifically at vegf receptor and vegf, i've shown you pathways that nitric

oxide and other reactive oxygen species with control synthesis of vegf, and vegf in turn induces the synthesis of nitric oxide, there's a positive feedback. i want to put this in the context of a big picture of what vegf and vegf receptor 2 are

doing in proangiogenic signaling. it binds to receptor, recruits adaptor proteins, that brings src in, src is important for migration and permeability and it's a tyrosine kinase, and that activity phosphorylates enos on tyrosine 8 which regulates

activity and phosphorylates hsp 90, this phosphorylation of tyrosine 300 is stimulatory. vegf receptor through pi3 kinases induces other kinases able to phosphorylate akt, 308, atkt phosphorylates enos at 1177, an important activating site.

vegf receptor 2 also through plc gamma and amp kinase can also phosphorylate this site so we're seeing there's redundant pathways activating enos downstream of vegf receptor 2 and plc gamma regulates levels of cytoplastic calcium, binds to calmodulin, and when has calcium

binds to enos, further involved in its in its activation. vegf is turning on efos enos synthesizes, as terry discussed earlier. n-o can also come from the exogenous source, macrophages associated with cancer or other

cells, and so n-o can be produced in endothelial cells and diffuse in from other cells producing externally, crosses the membrane and directly activates guano cyclate. we've looked at n-o signaling but we don't have time to go into this in great detail.

the takehome message is that many pro angiogenic factors also are able to activate the nitric oxide and some probably hydrogen sulfide pathways. but what i want to focus on next is how angiogenesis inhibitors also interact with the same redox signaling pathways.

and in fact are antagonists of the activating pathways that we just discussed. so we'll be focus on thrombospondin 1, others have similar activity. what is thrombospondin 1? this is a large secreted protein, you can see it by

rotary em, shadowing em, it's a fairly extended protein, weighs half a million. and it's released by cells in response to normal physiological responses, it's also turned on by growth factors. and so levels of thrombospondin normally circulate in plasma at

about 100 to 200 picomolar, these are important for physiological function, thrombospondin 1. during injury, it gets released by activation of platelets, elevate levels ten-fold in serum. but in conditions of acute and

chronic inflammation, levels can also go much higher. conversely in cancer, levels of thrombospondin tend to decrease, and so tumor cells tend to shut down thrombospondin production. so what does this mean in terms of angiogenic signaling and specifically in the context of

redox? well, early studies had shown thrombospondin was an inhibitors but required fairly high concentrations, higher than physiological concentrations to inhibit this signaling. but we published in 2005 that in the presence of n-o donors, or

substrates such as arginine which are used to make physiological nitric oxide, this switch shifted to the left from being a 10 nanomolar to 10 picomolar inhibition, in addition thrombospondin became a much more potent angiogenesis inhibitor and conversely work

from the wink lab in collaboration shows nitric oxide is a regulator of thrombospondin regulation, for vegf there's also feedback regulation of thrombospondin expression. of the redundant activation of n-o i showed for vegf? well, one thing that

thrombospondin does through binding to its receptor cd47 is inhibit ability of nitric oxide to sivate, blocking by exogenous donors or drugs designed to activate sgc. we subsequently showed that this signaling also inhibited ability of cgk.

cd47 through our work and from a group in arizona was shown to regulate calcium levels, calcium is important for enos activation, suggested that cd47 was also controlling activation of enos, and subsequent studies in our lab showed cd47 is also able to control the direct

activation of vegf receptor 2, by vegf so this controls all the redundant upstream signaling. mechanistically, we showed these are interacting with each other, in the presence of vegf this stays intact and the receptor gets phosphorylated, if thrombospondin binds first its

associate cd47 from the complex and then vegf fails to phosphorylate to receptor. so this is another way that thrombospondin controls the activation of pathways that are going to be in turn activating nitric oxide demand. as i mentioned, i'm not going

into detail throw thrombospondin is able to redundantly inhibit cascade, but other an this is an experimental drug made by abbott laboratories, went through phase 1 and phase 2 and then was withdrawn, endostatin is a protein fragment identified, also has gone through trials in

statin. bevacizumab and tyrosine kinase inhibitors that inhibit receptor are acting upstream and also inhibiting nitric oxide signaling as does that, intended to inhibit angiogenesis but the signal is important for platelet activation to explain why

hypertension and hemo stasis effects are seen as side effects of the therapeutic angiogenesis inhibit respect yo. i want to i n h b

broader cardiovascular effects. the thrombospondin levels can be modulated from physiologic level, by a variety of pathologic conditions, either higher or lower, and i told you that at physiologic level also of thrombospondin we see tonic regulation of nitric oxide

so that predicts in these conditions where levels of thrombospondin are elevated, this might ablate nitric oxide and conversely see why we could get rid of thrombospondin removing limits on nitric oxide signaling causing more angiogenesis of the tumor which

is what the tumor wants to grow. and just very briefly, it turns out that this signaling through cd47 also controls hydrogen sulfide signaling. we published a few years ago this signaling through cd47 in addition to controlling the mek erk pathway limits induction

involved in hydrogen sulfate synthesis. what does thrombospondin do under injury conditions? it's induced in response to injury. so one type here, ischemic injury known as mcfarland skin flaps, the back much a mouse,

incision made in u-shape, one by two or three centimeters on the back, skin flap lifted up, mice if you worked with them have very loose skin. so unlike humans, where the perfusion is coming from penetrators that come from the under lying muscles in a mouse

it's in this bag of loose skin, except around its head, and so the perfusion of blood into the skin is mostly lateral, coming from the side. so i'm making incisions, destroying the blood vessels supplying blood flow to the flap, and if you look several

days later in the wildtype mice distal regions of flap have become necrotic, and if we look at perfusion as i'll show you in a minute they have lost blood perfusion but interesting this in the thrombospondin-1 null mouse flaps maintain viability, we cut the blood supply but they

seem to adapt to that. another model was to ligate femoral artery, as you can see by laser doppler on this leg, this foot perfused, this lost blood perfusion, this is immediately after surgery in the wildtype or mice that lack cd47, around the spine and receptor,

and by seven days later it's not much better in the wildtype mouse, but cd47 null mouse has regained perfusion. so these are two injury models that show how well mice can handle the ischemic response. skin grafting, remove and plop

it on the wound bed and somehow that has to survive long enough to get perfused. and skin grafting can be successful if it's done well, but full thickness skin grafts in these mice, loose skin, always fail. the tissue dies.

but in the null mice, lacking thrombospondin or its receptor, we saw they survive. so is this just angiogenesis? going back to terry's theme of time, this is a process that happens on the scale of hours to days, forming new blood vessels doesn't very, very quickly.

hemostasis happens fast. vasodilation and changes happen in seconds to minutes, changes in vascular permeability. all of these depend on nitric oxide. so here we look at an acute blood flow response in our mice. so we've taken either a wildtype

or a thrombospondin null mouse, and we've performed a dynamic mri imaging, bold mri which is blood oxygen level and so this is just one cross-section of the mouse, lying on table, this is a section where you have right leg, left leg, and the tail. and so the gray is the t1,

image, colors are indicators of the bold signal which is telling you how much oxygen is flowing through, what the oxygen levels are in the tissue. so you start at baseline, with these mice, and then we administered in tail vein a small injection of a rapid

nitric oxide donating drug. and in a wildtype mouse you see that within 5 minutes after receiving this the bold signal goes up, increase in blood flow, but in the thrombospondin null that increase was doubled. this is healthy tissue now which has physiologic levels of

thrombospondin versus lacking thrombospondin is inhibiting this n-o signaling by 50%. so in the absence of injury the level of thrombospondin is normally in your blood vessels is limiting how much nitric oxide can vasodilate. extending this to the flap model

in the previous slide, we did laser doppler to look at blood flow, and instead of waiting several days when the tissue is going to be dead we looked immediately after the surgery and so imaging was done here five minutes and 60 minutes after the excision, sort of see

here where the surgery was done. and what we see quantified here on the right is that in a wildtype mouse as you expect it's a very happened fall in perfusion in this injured flap. but in the thrombospondin null, also very rapidly within minutes we see that this curve diverges,

this can't be an angiogenesiswhich happens over days. but what happens in a cancer? well, tumors also are trying to maximize flow, and an interesting observation made a number of years ago, the steal effect, this is true in animal models and in patients that if

you look at responses to challenge with nitric oxide, i showed you normally healthy tissue that transgenic exposure to nitric oxide increased blood flow, well, this is true in the control in the mouse that doesn't have a tumor. but if we image the limb that

has a tumor implanted, we see that rather than blood flow going up as it does in the normal muscle, in the tumor, nitric oxide challenge reduces blood flow in the tumor, transient effect lasts for half hour and returns to baseline. and the mechanism for this has

been established by others, but i wanted to show you a schematic here. if we simplify tumor to model where we have one blood vessel going to the tumor, one going to a normal tissue adjacent, assume blood flow is balanced, it turns out vasodilators such as nitric

oxide are able to dilate out of normal tissue but in the tumor empirically found to be different, you have more flood flow going out, decreasing in the tumor. conversely if you treat with a vasoconstrictor such as epinephrine this is going to

preferentially constrict the blood vessels that are outside of the tumor which then passively directs more flow into and that's the explanation for the steal effect. so does this apply to thrombospondin also? thrombospondin is as i showed

you in healthy tissue able to limit the vasodilatory activity of nitric oxide, when we did the experiment with the bold mri method looking at the tumor tissue we saw that whether the tumor is in a wildtype versus null mouse, this difference goes away.

so the ability of endogenous thrombospondin in typical blood flow is present in normal tissue but limited in tumor tissue. so this is physiology and pathophysiology, but you also want to extend this to therapeutics. so we've looked at a variety of

ways that we could disrupt this pathway. we've used thrombospondin blocking antibodies to suppress the expression of cd47, blocking antibodies that prevent cd47 binding and small molecules. the significance of this is that we would want to inhibit this

signaling cascade thrombospondin through this receptor and increase nitric oxide stimulation. you want to maintain flow after ischemic injury and with diseases of aging such as diabetes to limit that and it turns out that in diabetes

hypoglycemia induces thrombospondin eression so one of the causes for chronic ischemia in diabetics is accumulating thrombospondin which shuts down nitric oxide in the lower extremities especially. conversely if you're a cancer

researcher you want to inhibit tumor angiogenesis, so we would like drugs that can increase this inhibitor signal to decrease nitric oxide signaling and also for bleeding and acute injuries such as shock where you want to accelerate platelet aggregation.

other companies have thought about this also and just in the past year there's been launch of five clinical trials, and the approach that's been taken so far has used either humanized anti-cd47 antibody, one made at stanford, and licensed to 47, inc., another, both in human

clinical trials, trillion in canada took another approach, cds 47 binds to a counter receptor, so they use this as a decoy receptor, made a soluble recombinant form and that's also in the phase 1 trials. and there are a number of other antibodies and small molecules

that are currently in development. so how could these be used? well, i'm going to show you an example that we've done in our lab. i think this will apply to some antibodies in other therapeutics as well.

we showed that in wildtype mice also in pigs we could protect the skin flap injury from becoming necrotic by suppressing expression of cd47, we protect the resistance to ischemic injury with aging and cardiovascular disease in apoe mice fed a diet, extending to

reperfusion and looking at papers with kidney transplant. i want to the show you data from young versus old mice. in young mice we have low levels of thrombospondin, high levels of cyclic gmp, and this is on the y axis quantifying percent survival of ischemic flap, as i

showed in this picture, and as i mentioned about 50% of flaps survive in untreated animal, giving an n-o donor an fda approved nitric oxide donor, significantly improves flap survival but not completely. conversely, inhibiting nitric oxide synthase with l-name which

you can feed in water to the mice worsened survival of flaps but treating with cd47 morpholino gave the best response so suppressing this inhibitor pathway takes the brakes off and allows the flaps to survive well, almost as well as they do in a null animal.

these are young mice. we allowed the mice to age. you see with age that the control levels go down, so older mice as older people are less able to repair skin injuries and their tissue is more sensitive to death from that. the n-o donor now doesn't have a

significant improvement, giving them exogenous levels, if you give l-name as you expect things get worse but remarkably of suppress of cd47 works almost as well as it does in a young animal. so this suggests that a large part of the decrease in

responsiveness to ischemic stress with age can be explained by this accumulation of nitric oxide which then lowers their nitric oxide gmp signaling. because i'm in the cancer institute, we're interested in other stresses besides ischemic injuries mostly interested to

surgeons, so we looked at the ability of these animals to resist injuries caused by ionizing radiation, doses of radiation that you deliver the tumors limitedded by how much damage the normal tissue that surrounds the tumor can sustain. and so it's well known that at

levels sufficient to kill a cancer, the patient loses their hair, alopecia, ulceration of the skin, typical level of radiation sufficient to kill a tumor. in the thrombospondin 1 null there was less damage and cd47 null it looked like we missed,

but if we look at the mitochondria, a better indicator of the primary dna damage, they have the same initial damage but two months later they can recover from that much better. here looking at the skin, hair follicles must be there. and they are in the cd47 null,

but what you can't see from just looking at the surface is the underlying muscle, this light color indicates extensive death of muscle fibers, and if you look carefully at this limb the limb should be extended but it's retracted because the muscle has atrophied.

in the wildtype this doesn't happen, in the null, and we see here that the tissue remains viable in the muscle. to understand how this happens it was fortunate that cells from the mice cell autonomously show radio resistance. a wild type endothelial cell

from the lungs of a wildtype mouse with increasing doses of radiation, progressively dies. measured here by assay, null cell survives. in human endothelial cells treated with morpholino could protect them. is this n-o dependent?

for ischemic blocking the pathway is n-o dependent, but if we simply gave an n-o donor to these cells, we didn't see any protection against death caused by radiation, or improving -- or pharmacologically increasing have minimal effect. this told us something other

than n-o pathway was important to this radioprotection response, and turned out that this was mediated by an induction of autophagy, the decision whether to die, terry mentioned bcl-2 is an important pro survival pathway that controls mitochondrial mediated

we found cd47 controls beclin-1, downstream mediated, if we block these with sirna we remove protective effect of blocking cd47. more recently we've looked more broadly at metabolic responses, and it turns out this protective autophagy has effect on redox

signaling of cells under stress so one of the pathways that's important for controlling redox s glutathione pathway, glutathione reacts with products, which is then detoxified. if we look at levels in a wildtype cell after radiation we

see the normal level also fall down to baseline quickly. but in the cd47 deficient cells, they are maintained and tend to increase, so the ability of the cells to detoxify reactive oxygen species is controlled by levels of glutathione itself tend to fall, to reduce oxidized

norm but this is maintained if you take away cd47, and a number of metabolites that are important for synthesis of glutathione, also are falling in the wildtype but not in the null cell, this results in a very broad protective effect of blocking cd47 on ability of

cells to respond. so could conclude, takehome message is thrombospondin is an important pathophysiologic inhibitor of vascular n-o signaling by cgmp. this is a pathway that controls angiogenesis, explains why thrombospondin back in 1990 was

identified as an theangiogenesis inhibitors, nitric oxide controls muscle tone and blood vessels and hemostasis because the kinase is important for activation of platelets, in addition by nitric oxide. cd47 is also a direct binding partner of vegfr2 giving broad

ability to control nitric oxide in other signaling pathways activated by vegfr2, this controls ischemia, radiation injuries, former mediated by nitric oxide pathway, radiation stress resistance involves other pathways controlled by autophagy, improving survival of

wildtype animals and hopefully wildtype humans in the future that that these type of injuries and that these also have potent radioprotective activities, mediated in part by improving redox homeostasis, which gives the cells improved ability to scavenge free radicals.

i'll stop there. everybody stayed awake. happy to take questions. [applause] >> ubiquitously expressed, also for the antibodies in clinical trials because red blood cells have 25,000 copies of cd47 on them.

and so the first two agents in clinical trials are humanized antibodies, and if you calculate how much antibody you need, this is being given i.v., you need to overcome the buffering of the red cells before you get to the tumor to try to cure cancer, that amounts to about

60 milligrams of antibody just to cover up what's on the red cells. it's a lot. it's going to be expensive. >> yes. i was brought here to talk about angiogenesis and the vascular system so i didn't say much

about that. cd47 is an inhibitor receptor on cdac cells, and so what i didn't show you, i showed you if you irradiate normal tissue that blocking the pathway was radioprotective, and radiation oncologists want radioprotectance, but the only

one that's approved by the fda is amophosphene that doesn't know the difference between healthy and tumor tissue, choosing the radioprotectant hasn't improved survival of cancer patients presumably because as you protect normally

tissue you also protect tumor, so you don't come out ahead. we were surprised when we looked at tumor bearing mice with radiation that if we block cd47 combination with radiation a lot of tumors go away. and that was initially surprising because if we

irradiated tumor cells outside the mouse there was no difference in their response to radiation, whether or not we blocked cd47 in the tumor cells, so this was something happening in the microenvironment and it turns out that required cd8 t cells and we published other

work i didn't have time to talk about that showed direct killing of tumor cells anti-gent depending by cd8 t cells is enhanced if you block cd47, so this is really a two-fer that we protect formal tissue and sensitize the tumor. okay, thank you.