Dr. Timothy Thompson Video (Text Version)
Title of Talk: Glioma Pathogenesis-Related Protein 1, a Tumor Suppressor That Stimulates Oncoprotein Destruction
Speaker
It is always interesting and inspiring to hear a productive and creative researcher describing what she’s discovered about the mechanics of prostate cancer progression. And we thank you for all your hard work. If you have any questions I know you do; you’re—can you hold them until the end when we’re going to have a general question and answer period? Our next speaker, Dr. Timothy Thompson is Professor in the Department of Genitourinary Medical Oncology Research at the University of Texas MD Andersen Cancer Center and Director of Prostate Cancer Basic Science Research.
He holds the Joan and Stanford Alexander Distinguished Chair for Prostate Cancer Research. Dr. Thompson graduated from the University of Colorado at Boulder with a PhD in Pharmaceutical Sciences, Pharmacology in 1985 and did a Research Fellowship at the Imperial Cancer Research Fund in London in—between 1985 and 1988. He has over a decade of research and publications on the subject of caveolin which plays a very important role in metastatic and advanced prostate cancer. But the topic that he’s going to talk about today is GLIPR1, glioma pathogenesis-related protein-1, a novel p53 target gene with pro-apoptotic activities that could eventually lead to a novel systemic protein therapy; Dr. Thompson.
Timothy Thompson, Ph.D.; Professor, Department of Genitourinary Medical Oncology Research, University of Texas MD Andersen Cancer Center; and Director of Prostate Cancer Basic Science Research
Okay; I think it’s actually closer to two decades. I’d like to thank the organizers, in particular Dr. Kyprianou for inviting me to speak in this session. It’s just a fantastic meeting; it always is and it’s an honor to be here. So today I’d like to talk about a gene, a protein called GLIPR1, sort of a tongue-twister, glioma pathogenesis-related-1 gene and protein. We originally detected this CDNA actually using differential display many years ago and it sat in the refrigerator without any work being done on it for several years. And we finally began to work on it and determined that it actually was a p53 target gene. And not only was it a p53 target gene but it was part of a—a gene sub-family that both in human and mouse contained other genes, GLIPR like-1 gene and GLIPR like-2 gene on human chromosome 12 that are clustered within a very short or very small region I should say—170 kilo-base pairs on chromosome 12 and its [intenic] region on male chromosome 10.
All of these genes have p53 binding sides and many of them appear to be direct targets of p53. Interestingly the—most of them are secreted proteins or cleaved from the membrane, so they get into the extracellular milieu. So this—the GLIPR-1 sub-family is part of a larger gene family called the CAP-super family, cysteine rich, secretory proteins in the mammalian cells, antigen five in insect cells which include some insect venoms interestingly, pathogenesis-related 1 proteins which include plant hormones and so you can see here the phylo-genetic relationships of this sub-family that’s written up a couple of years ago in this review. So you know based on this homology and based on this conservation, we believe that the functions of—of this gene family were very important. We began to really pursue it aggressively.
So the first experiment we did was a total knockout of GLIPR-1 and this led to reduced tumor-free survival in wild type and also heterozygote p53 mice and interesting—tumor spectrum including lung carcinoma was seen. In subsequent studies, we showed that GLIPR-1 is down-regulated in human primary prostate cancer through gene methylation, and we also demonstrated that GLIPR-1 induces p53-independent cancer cells selective growth suppressor and apoptosis. It is a p53 target gene, but it had independent activities.
Because of these growth the suppressant properties, we began to look at its relationship to c-Myc. And the first thing we did was to assemble some tissues of normal adjacent and primary prostate cancer from 34 pairs—from the same sample of prostate cancer. We showed that GLIPR-1 mRNA levels were down-regulated in 27 of those 34 pairs and—and c-Myc mRNA levels were up-regulated in 29 of those 34 pairs. And this was a statistically significant inverse correlation.
We did comparable studies using immuno-staining, GLIPR-1 going down in—in primary tumors; c-Myc going up in primary tumors and using the DeMarzo criteria here in the nuclear area, we showed again that there was an inverse correlation of GLIPR-1 staining and c-Myc staining.
So this led to a lot—a large series of expression arrays and pathway analysis and if all the data were consistent with c-Myc being down-regulated and a sub-set of c-Myc genes being down-regulated by GLIPR-1, and c-Myc is I think sort of a neglected oncogene in prostate cancer—was one of the first identified as being important and certainly it’s one of the most important oncogenes in human malignancies deregulated in at least 30% of cancers and—and actually regulates 15% of the human genome.
So you can see here a sub-set on a heat map of some of the genes that we identified in these expression arrays, c-Myc, some c-Myc target genes being down-regulated by GLIPR-1. This is over-expression of the gene in three human prostate cancer cells. We validated this with RTPCR and on western blots and you can see here just the profound down-regulation of c-Myc by over-expression of GLIPR-1, less than one-tenth of the levels of c-Myc by over-expression of GLIPR here. When we over-express GLIPR we see reduction in cells in S-phase; we see an increase in cells in G2 and we see in many of these prostate cancer cell lines a sub-G1 component that indicates apoptosis.
When we looked at MEF cells from the total knock-out we found that they had increased MIC—MIC expression and these cells also had increased growth potential just in 2D cultures and also susceptibility to transformation by c-Myc, transformation in colony stimulation actually. And this was a very synergistic susceptibility.
We generated then a biogenic mouse; we—we made our own probasin c-Myc mouse in a different strain that did not—does not give prostate cancer but aggressive PIN and crossed that mouse to the total GLIPR knock-out mouse and showed that there was a synergy here also in the generation of PIN lesions and also some of these mice produced prostate cancer; about 10% of these mice produced prostate cancer. But these were very statistically significant differences in—in high—high-grade PIN in these mice.
Just to show you some of the pathology associated with this, some of the high-grade PIN lesions and the carcinomas, one of the three carcinomas that we saw in this system; interestingly on the bottom here you can see c-Myc staining. This particular mouse has a trans-gene for c-Myc yet wild-type GLIPR and these levels of c-Myc activity but when you knock out the GLIPR, c-Myc activity goes up dramatically.
So what’s the mechanism for this down-regulation? So one of the first things we did is to screen a large series of prostate cancers over-expression of GLIPR sometimes by adenoviral vector, sometimes by standard transfection. It seems to work the same way. And we—this is RNA; we saw down-regulation of RNA and in PC3 cells—most of these cell lines have really non-detectable levels of GLIPR but in PC3 there is moderate expression of GLIPR-1, so we knocked down GLIPR with an siRNA and—and actually c-Myc levels increased here and also in this bladder cancer cell line R24 and so we began to consider that there may be some transcriptional down-regulation of the gene.
One of the first places we looked is at active beta catenin; this is using an antibody to de-phosphorylated catenin which recognizes the epitopes that do not have phosphorylations on them that are phosphorylated by GSK3 beta. And these again in three prostate cancer cell lines were down-regulated; you can see here in the nucleus very basically non-detectable here following GLIPR transfection. Then we looked more directly at transcriptional regulation first using CHIP and showed that GLIPR-1 produced tri-methyl H3K27 mark that was increased on the promoter as well as down-regulation of acetylated H3 and interestingly down-regulation of TCF4 which is the DNA binding partner of beta catenin. We directly showed that there was transcriptional activation using co-transfection promoter luciferase assays and then we began to look at upstream of that what really—whether GLIPR was involved in regulation of beta catenin.
And the first place we looked was GSK3 beta; we looked at levels of GSK3 beta in DU145 and LNCap cells and we saw a significant but small increase in—in protein levels for GSK3 beta and DU but not in LNCap cells. We looked for D-repression of GSK3 beta. This is an AKT site that represses GSK3 beta so we would expect to see this go down. We did not see this in DU145 cells or LNCap cells but what we did in fact see was casein kinase 1-alpha in about a two-fold increase on western blots and expression of this protein which actually primes for GSK3 beta leading to ubiquitination and proteasome degradation of beta casein. So this was very interesting; we began to look for cellular sub-localization of CK1-alpha as well as GSK3 beta which is increasingly being shown to be important in regulation of their activities.
So the—this is DU145 system; up-regulation of GLIPR, down-regulation of MYC; we did see a—a moderate to—small to moderate increase in GSK3 beta staining in the cytoplasm following GLIPR. Then we looked at CK1 alpha and surprisingly again casein kinase 1-alpha which primes for GLIPR-1 in un-induced cells the—the protein was co-localized with giantin which is a Golgi marker and very polarized here in the peri-nuclear area. But when we over-expressed GLIPR-1, there was a very significant dispersion of CK1 alpha into the cytoplasm, a redistribution of this protein that we—surprised and are studying intently. And all of this added up really—you can look at beta catenin here and un-induced cells—beta catenin active in the cytoplasm getting into the nucleus and the presence of GLIPR activity here, the cytoplasm is relatively clear and there’s no staining in the nucleus.
So in—for the RNA expression we—we never saw more than 50% reduction. So we suspected because we did see a dramatic reduction in protein, sometimes 10-fold less protein, we suspected that—that GLIPR-1 may also destabilize the protein. So we began to look at this alternative mechanism, sort of a dual mechanism, and you can see here some—293 cells co-transfection basically doing IPs with ubiquitin and blotting for c-Myc. You can see here ubiquitination of c-Myc. We could basically block this reduction in c-Myc in multiple cell lines with multiple proteasome inhibitors indicating that there was proteasome activity so this indicated that GLIPR was having a direct effect on protein stability.
What is the mechanism of this? Well it’s well characterized; Rosalie Sears identified the GSK3 beta, [3 and 858] phosphorylation site on c-Myc a few years ago, a very important finding and so we looked for evidence that GLIPR-1 actually was increasing that phosphorylation on c-Myc which would lead to its ubiquitination and degradation and we found in DU145 cells evidence of this but not in LNCap cells and this was consistent with other data indicating that LNCap cells in response to GLIPR—that this pathway is—is minimal(ized).
So then Likun Li who is really the—the major investigator in this work began to look at the literature there was—there weren't any papers about CK1 Alpha at phosphorylation sites on c-Myc in mammalian cells but there were in Drosophila. So we pursued that and he identified some potential phosphorylation sites on c-Myc, Syrene 67, Syrene 252 and so we began to pursue the possibility that CK1 through phosphorylation may destabilize Myc. So the first thing we did was knock it down and we saw actually an increase—about a fourfold increase in c-Myc activity in the presence of GLIPR in DU145 cells and a very significant increase in c-Myc activity with CK1SI and LNCap cells. We over-expressed CK1 alpha and basically were able to drive down the protein expression to levels comparable to that of GLIPR.
So then to directly test the biological significance of these phosphorylation sites, Li Koon made some phosphorylation mutants for GSKT58, CK1-Alpha, S67, S252; we coded—we transfected these along with GLIPR and we were able to prevent down-regulation. You can see here by comparison the T58 mutant basically increases somewhat. The S67 a little more but the S252 mutant basically totally prevented down-regulation by GLIPR of c-Myc.
And then to directly show the effect of these mutants on the half-life of c-Myc again transfecting these mutants and treating with cycloheximide we could show that the 252 mutant dramatically increased the half-life from about 25 minutes to 70 minutes and by comparison I believe about 60 minutes was seen for the 358 site that’s the GSK3 beta site.
So just to summarize this there’s a dual mechanism for GLIPR down-regulation of MYC. GLIPR endogenous expression or up-take of GLIPR into cells can up-regulate CK1 and importantly cause its redistribution out of the Golgi and we’re—we’re studying this now that—the mechanisms of this and the—the first pathway, mechanistic pathway here is to prime for GSK, destabilizing beta-catenin and down-regulation c-Myc transcription. The second pathway that was completely unknown is that CK1 phosphorylates c-Myc on S252; this destabilizes the protein and down-regulates the protein, so a dual mechanism.
So what does all this mean? Can we really translate this into any therapeutic—any therapeutic advantage? Very important paper, very—published last year from the Sloan-Kettering Group again identified c-Myc amplification as a major target in prostate cancer. You can see here that together with other genetic alterations, deletions, this particular cluster of prostate cancer is—really marks a very aggressive prostate cancer in a relatively large series of patients. So we began to pursue the possibility that we could do protein therapy using GLIPR. And I hadn’t intended to put these slides in but I saw Marianne’s therapy talk and decided to put this in. There is a—I founded a company, co-founded a company at Baylor a few years ago and there is intellectual property that has been licensed to a company involving this—this protein.
So what—what we did here was modified the protein. We removed the trans-membrane domain so we could actually purify the protein and began to use it and test it. And this—this was—sometimes in your career you find some pretty remarkable findings. And the reason people don’t use protein therapy is because they don’t get into cells—at least selectively don’t get into cells. That’s—that’s why small molecules have a huge advantage. But when we labeled the protein and treated prostate cancer cells non-transformed prostate cancer epithelial cells and prostate cancer cell lines and labeled the protein at first we saw a docent and time dependence of uptake. But importantly we saw a prostate cancer selective uptake and we’re investigating this phenomenon. It is a phenomenon at this point. But it does—it is reproducible and—and so this was very interesting and really stimulated us to further pursue this as a protein therapeutic.
The—the uptake is clathrin mediated. I won't go through the inhibitors here but it’s a clathrin mediated uptake of GLIPR-1. Again this is a secreted or cleaved protein so it is out there. It does have opportunity to get into—into cells. The protein in vitro, when we treat cells just with GLIPR protein at really relatively low doses therapeutically down-regulates Myc and a sub-set of Myc target genes. It’s similar to over-expression of the gene. Also it selectively induces apoptosis or growth suppression data I’m not showing here. You can see here the—the data I showed previously was in DU145 and LNCap. You can see this correlates very nicely with the Sub-G 1. D145 there is some induction of apoptosis but this was really not seen in non-transformed cells.
So we initially developed a bio-luminescent system and showed that basically direct injection of the protein into tumors. These are V-CAP tumors that are transduced with luciferase. We could down-regulate tumor growth using GLIPR-1 Delta TM protein, down-regulate tumor growth in the way it correlated with bio-luminescent, validating the system. But importantly we got basically—got the same result when we treat either with inter-peritoneal or now we—we have very nice IV administration data at really very low therapeutic doses that we can produce—essentially replicate the inter-tumoral results and show down-regulation of—of tumor growth in these V-CAP luciferase models. So this, in fact, is IP; I could show you the same data for IV.
So we’re developing biomarkers and in collaboration with Chris Logothetis and others at MD Andersen we want—and SPORE funding we want to advance this into a proof-of-principle therapy trial in the near future and—and you can see here that this protein, in fact, produces apoptosis, down-regulation of Myc in vivo and other Myc target genes. So I guess I’ll stop here; thank you very much.
Oh, sorry; just to—this is Likun Li who really is responsible for the majority of the mechanistic work. Guang Yang and Chengzhen Ren who really is responsible for a lot of the—the work on the protein, the engineering of the protein and—and also co-founder of the company. So thank you very much.