Silvia Ottaviani, PhD
Research Associate, Imperial College London, UK
Connect with Silvia
Silvia is an expert in non-coding RNAs (ncRNAs) biology in cancer. She is currently a post-doctoral researcher at Imperial College London, looking at the role of ncRNAs in cancers, mainly pancreatic and breast cancers. Silvia's research has been predominantly focused on microRNAs (miRNAs), however more recently she expanded her work on long non-coding RNAs too. Silvia has recently published her main post-doctoral project in Nature Communications journal where she investigated novel miRNAs involved in the TGF-β response in pancreatic cancer. She is also co-author of several publications on miRNAs in cancer and she is author of two comprehensive reviews on the role of ncRNAs in health and disease.
Silvia obtained a BSc in Medical Biotechnology from University San-Raffaele in Milan in 2007. During her undergraduate studies she was awarded a competitive scholarship to undertake a summer research internship in the RNA laboratory of Dr Elisa Izaurralde at EMBL in Heidelberg. Silvia then obtained an MSc in Molecular Medicine in 2008 and a PhD in cancer biology in 2012 both from Imperial College London. Silvia's PhD was funded by the Prostate Cancer Charity and was carried out under the supervision of Dr Laki Buluwela and Prof. Simak Ali. During her PhD she investigated the function of glycine N-methyltransferase, a newly identified androgen-regulated gene in prostate cancer. With a strong background in RNA biology and cancer biology, Silvia joined the laboratory of Prof. Justin Stebbing and Dr Leandro Castellano where she is currently conducting her post-doctoral research on ncRNAs in cancer.
Silvia is also extremely passionate about teaching and she has always endeavoured to be actively involved in education alongside her research. She has supervised several Master students, visiting undergraduate students and she has recently become a deputy module leader for teaching Molecular and Cellular Biology to BSc students at Imperial College London.
Learn about Silvia's research
Title: Investigating the role of microRNAs in pancreatic cancer progression
- Understand the role of miRNAs in pancreatic cancer
- Learn how miRNAs affect tumourigenesis and metastasis using in vitro and in vivo models
- Learn how we can efficiently identify miRNA functional targets using an experimental approach
Pancreatic ductal adenocarcinoma (PDAC) is a deadly disease with a 5 year-survival rate of approximately 6&. Despite recent efforts in developing novel therapies, chemotherapy remains the standard of care for advance pancreatic cancer patients. MicroRNAs (miRNAs) are short non-coding RNAs that act as post-transcriptional negative regulators of gene expression. Our laboratory is interested in investigating how miRNAs dysregulation contribute to PDAC tumourigenesis. One of the major signalling pathways that drives tumourigenesis in PDAC is the TGF-β pathway. This pathway promotes epithelial-to-mesenchymal transition (EMT), metastasis and stemness. In this study we aimed to find out whether miRNAs were involved in the TGF-β response, as this was yet largely unknown. We used miRNA expression profiling and identified miRNAs regulated by TGF-β. We then focused on the top two miRNAs upregulated by TGF-β and characterized their role in PDAC. We used several strategies to inhibit the function of these miRNAs, including the genome editing approach CRISPR. With these systems we showed that silencing of these miRNAs impaired EMT, motility, stemness in vitro and tumourigenesis and metastasis in vivo. Furthermore, we identified globally the targets of these miRNAs by integrating AGO2-RIP sequencing with RNA-sequencing upon overexpression of the miRNA of interest. We found that the candidate miRNA targets significantly overlap and mainly inhibit p53 and cell to cell junctions’ pathways, which are all important in PDAC progression. We also showed that the candidate miRNAs were up-regulated in PDAC patient samples, and their specific tumoral expression strongly correlated with reduced overall survival and disease-free survival. These findings demonstrate a fundamental role of miRNAs within the TGF-β response and represent potential novel targets for therapeutic intervention in PDAC.
Watch the webinar
Presenter: Silvia Ottaviani - Investigating the Role of MicroRNAs in Pancreatic Cancer Progression
00:00 – Slide 0
Moderator: Hello, everyone and welcome to today's live broadcast Investigating the Role of MicroRNAs in Pancreatic Cancer Progression presented by Silvia Ottaviani, research associate, Imperial College, London. I'm Alexis Corrales of LabRoots and I'll be your moderator for today's event. Today's educational web seminar is brought to you by LabRoots and sponsored by Thermo Fisher Scientific. For more information on our sponsor, please visit Thermofisher.com. (00:32)
Now, let's get started. Before I begin I'd like to remind everyone that this event is interactive. We encourage you to participate by submitting as many questions as you want at any time you want during the presentation. To do so simply type them into the "Ask a Question" box and click on the send button. We'll answer as many questions as we have time for at the end of the presentation. If you have trouble seeing or hearing the presentation please click on the Support tab found at the top right of the presentation window or report your problem by clicking on the answer question box located on the far left of your screen. (01:09) This presentation is educational and, thus, offers continuing education credits. Please click on the Continuing Education Credits tab located at the top right of the presentation window and follow the process to obtain your credits.
Moderator: I'd like to now introduce our presenter, Dr. Silvia Ottaviani. Silvia is an expert in non-coding RNAs biology and cancer. She is currently a postdoctoral researcher at Imperial College London looking at the role of non-coding RNAs in cancers, mainly pancreatic and breast cancers. (01:41) Silvia's research has been predominantly focused on microRNAs. However, more recently she expanded her work on long non-coding RNAs, too. Silvia has recently published her main postdoctoral project in Nature Communications journal where she investigated novel microRNAs involved in the TGF-beta response and pancreatic cancer. She's also co-author of several publications on microRNAs in cancer and she's author of two comprehensive reviews on the role of non-coding RNAs and health and disease. (02:11) For complete biography on Dr. Ottaviani, please visit the Biography tab at the top of your screen. Dr. Silvia Ottaviani, you may now begin your presentation.
02:26 - Slide 1
Hello, everyone. Thank you for the introduction and thank you everyone for signing in today. So, I am Silvia Ottaviani, a postdoc researcher studying the role of non-coding RNAs in cancer, in particular pancreatic cancer. So, I'm based at the Division of cancer at Imperial College London in the U.K., and I work in the laboratory of Dr. Leandro Castellano and Professor Justin Stebbing. (02:55) Today I'm very excited to talk to you about how our lab is studying the role of microRNAs in pancreatic cancer and, in particular, I will present the results of my first postdoc project that has been recently published in Nature Communications. So, let's get started.
03:12 - Slide 2
So, pancreatic cancer is a lethal disease with a 5-year survival rate of approximately 6 percent. It is currently the fourth highest cause of cancer deaths and it is predicted to be the second leading cause of cancer deaths in the next decade if outcomes are not improved. Specifically, ductal adenocarcinoma is by far the most common malignancy of the pancreas. It is indeed the focus of the majority of the research in the field, certainly, from our laboratory. (03:47) So, sometimes we refer to this as PDAC. So, the reasons why pancreatic cancer is associated with very poor prognosis are primarily the lack of early diagnosis and ineffective treatment for advanced tumors. Additionally, pancreatic cancer is characterized by remarkable resistance to most conventional treatments, so chemotherapy and radiotherapy as well as molecular target therapy. (04:15) Therefore, every research effort to improve our understanding of the molecular mechanisms that drive this aggressive disease is crucially important, and will allow development of more effective treatment.
04:32 - Slide 3
So, I'm sure you're all familiar with microRNAs. Here, I just would like to highlight some important points. So, microRNAs are small non-coding RNAs that negatively regulate gene expression through mRNA degradation or translational inhibition. (04:49) This is just an overview of the biogenesis pathway. So, briefly, microRNAs are transcribed as primary microRNAs by RNA polymerase II. The pri-microRNAs is then cleaved by the micro-processor which is a complex including DROSHA and DGCR8. This cleavage produces the 60 -70 nucleotide precursor microRNAs. (05:16) The pre-microRNA is then exported to the cytoplasm by XPO5 and further process by DICER1 to produce the mature microRNA. So, one strand of the mature microRNA is loaded into the RISC silencing complex which directs the microRNA to the target mRNA by sequence complementarity and maybe gene suppression, as we said, by either mRNA degradation or translational repression.
05:47 - Slide 4
So, microRNAs are fine-tuned modulators of gene expression and they regulate a variety of biological processes. So, perhaps, it is not a surprise the fact that this regulation of microRNA expressions is involved in cancer. So, indeed microRNAs can function as tumor suppressor and oncogenes. So, here on the left panel we can see a schematic representation of the biogenesis of microRNA that we just discussed in normal tissues. (06:22) So, the overall result is normal rates of cellular growth proliferation differentiation and cell death. In the middle panel we have an example of a reduction of the microRNA, the functioning as tumor suppressor. So, that leads to tumor formation. So, the reduction of microRNA levels can occur because of defects at any stage of microRNA biogenesis — that's why we have this question mark — and ultimately lead to the inappropriate expression of the microRNA target oncoprotein. (06:59) So, the overall outcomes may be increased proliferation, invasion and angiogenesis and reduce cell death leading to tumor formation. In the last panel we have an example of amplification or over-expression of a microRNA that has an oncogenic role. So, this would also result in tumor formation. In this case, increased amount of the microRNA would significantly decrease the levels of the target tumor suppressor gene and lead to cancer progression. (7:31) So, increased levels of mature microRNA might occur because of amplification of the microRNA gene, or a constitutively active promoter, or increase efficiency microRNA biogenesis, or indeed increased ability of the microRNA. And those are all indicated by these question marks.
07:54 - Slide 5
What about microRNAs in pancreatic cancer? Obviously, as in many other cancers, many microRNAs have been identified and shown to have a role as oncogenes or a tumor suppressor. So, this is a table of selective microRNAs identified that might be clinically relevant in the management of pancreatic cancer. So, as you can see, microRNAs have been identified acting as oncogenes in the top part of this table or tumor suppressor in the bottom part. (8:29) And they are involved in many different biological processes. I won't go through the list in detail. I will just mention a couple. For example, a very well-known oncomiR in pancreatic cancer is miR-21. Over-expression of miR-21 contributes to gemcitabine CAM resistance and enhance malignancy of pancreatic cancer cells. As for a tumor suppressor, a very well-known tumor suppressor is the let-7 family of microRNAs. (09:01) Let-7 was found down regulated in a number of pancreatic cancer cell lines and it induces reversion of EMT in gemcitabine resistant cells.
09:15 - Slide 6
So, our laboratory is interested in studying the role of microRNAs in pancreatic cancer tumorigenesis and progression. Here, I'd like to point out two important studies that were published in the laboratory. (9:30) So, in the first study we performed a microRNA expression profiling in pre-malignant pancreatic tumors compared to PDAC, in order to identify microRNAs deregulated during PDAC development. So, we found many microRNA that were down regulated in PDAC compared to low malignant tumors. And we showed that amongst those microRNAs down-regulated, miR-16, miR-126, and Let-7d regulate known PDAC oncogenes. (10:05) So, BCL2, KRAS, and CRK. Before the down-regulation of miR-16, miR-126, Let-7d, promotes PDAC transformation by post-transcriptional upregulation of crucial PDAC oncogenes. (10:26) In the second study, we combine data from microRNA and then mRNA expression profiling in PDAC cell lines and also in PDAC samples from patients to identify the microRNA that contribute most to tumorigenesis. So, we identify three microRNA; miR-21, miR-23a, and miR-27a that acted as cooperative repressors of a network of tumor suppressor genes, that included PDCD4, NEDD4L, and BTG2. (11:02) In addition, we have shown that inhibition of miR-21, miR-23a and miR-27a had synergistic effect in reducing proliferation of PDAC cells in culture and also tumor growth in mice.
11:20 - Slide 7
So, moving on, we decided to concentrate on specific pathways that are relevant in pancreatic cancer. So, this is an overview of the aberrant signaling pathways in pancreatic cancer. It is important to mention that the majority of tumors, around 95 percent of all cases, are driven by mutational hyper-activation of KRAS. So, this downstream signaling is excessively activated. We decided to focus on the TGF-beta pathway, as TGF-beta has a vital role in PDAC. (11:59) And at the time there wasn’t a comprehensive study of microRNA regulated by TGF-beta. So, briefly, TGF-beta binds to the type 2 TGF-beta receptor which activates the type 1 receptor, and this leads to the activation of SMAD2/3 transcription factor, and they couple then with SMAD4 and they translocate into the nucleus to regulate SMAD regulated genes. (12:31)
12:33 - Slide 8
So, TGF-beta signaling has a vital role in pancreatic cancer as well as in other cancers, and it has a dual role. So, it acts as a tumor suppressor in the pre-malignant stages of tumorigenesis, and then switches to an oncogene role at later stages of the disease. (12:58) So, the most dominant functions as tumor suppressor are growth inhibition and apoptosis and the dominant functions as oncogenes are EMT invasion and metastasis and stemness. So, we decided to focus on the oncogene functions of TGF-beta.
13:17 - Slide 9
So, how does TGF-beta induce EMT metastases and stemness in pancreatic cancer? This is a very simplistic representation of the molecular pathway, but I wanted to point out some of the important factors. So, as we said before, TGF-beta activates SMAD2 and 3, which in turn form a complex with SMAD4 to simulate the expression of pro-EMT genes such as Snail, Twist and Zeb. (13:47) This leads to repression of epithelial genes such as E-cadherin to activate the EMT program. And at the same time the miR200 family is a very important family of microRNAs, and they've been shown to be very highly expressed in epithelial cells. So, they maintain the epithelial phenotype, and they do this by repressing Zeb 1 and 2. (14:17) It has been also shown that miR200 family microRNAs act as a stemness inhibiting microRNAs because they suppress pluripotent respect transcription factors such as BMI1, Sox2 and KLF4.
14:35 - Slide 10
So, simply, the aim of our study was to identify the microRNAs regulated by TGF-beta in pancreatic cancer that had not been identified before. So, we wanted to uncover whether microRNAs could add an additional layer of regulation in the complex signaling of TGF-beta responses.
14:58 - Slide 11
So, to discover novel microRNAs implicated in PDAC progression through TGF-beta, we created an in vitro cell line model of EMT spectrum. Specifically, we use BxPC-3 cells that are very epithelial, then we use PANC-1 cells that are part epithelial and part mesenchymal, then we treated PANC-1 cells with TGF-beta for 72 hours. And, as you can see from the picture, the cells treated with TGF-beta, they adopt a more spindle shape and mesenchymal-like morphology. (15:36) And, finally, we have S2-007 cells that are highly mesenchymal. So, here on the bottom left, you can see a western blot for E-cadherin in these lines, and as expecting you can appreciate that the expression of E-cadherin are inversely correlated with the mesenchymal status of the cells.
15:58 - Slide 12
So, next we performed microRNA expression profiling using encounter system from nano-strain, and this is a heat map that shows a selection of up and down regulating microRNAs from the analysis. So, we confirm that the miR200 family members are down-regulated in mesenchymal-like cells compared to the epithelial BxPC-3 cells. (16:26) And excitingly, from the analysis we found that only 2 microRNAs, miR100 and 125b increased proportionately with the mesenchymal status of the cell. And so, we validated this result by qPCR. So, here we can see this stepwise increase of the level of both miR100 and 125b with the mesenchymal status of the cell. And, also, we can see that these microRNAs were significantly up-regulated by TGF-beta.
17:00 - Slide 13
So, next we wanted to investigate the mechanism of speed to speed regulation of these microRNAs. So, firstly, we looked at the genomic location and found that miR100 and miR125b, together with let-7a, came from the same primary transcript, the long non-coding RNA miR100hg. So, miR100hg is a tricistronic host gene for these microRNAs. So, our first question was, is the TGF-beta regulation of these microRNAs transcriptional? (17:36) So, in order to answer this question we performed RNA-seq in PANC-1 cells treated with TGF-beta. As expected, TGF-beta significantly up-regulated pro-EMT factors and down-regulated cell to cell junction proteins like E-cadherin. Interestingly, miR100hg was amongst the RNAs significantly up-regulated by TGF-beta. (18:04) This, therefore, indicates that the induction of miR100 and 125b by TGF-beta is indeed transcriptional. So, the answer is yes. So, next we wanted to investigate whether this regulation was occurring through SMAD 2/3 transcription factors. (18:25) We, therefore, performed chip-seq for SMAD2/3. As you can see from these tracks, you can appreciate that SMAD2/3 interact with several regions of the miR100hg genes with strongest binding around the transcription start site. However, we later appreciated that even in the absence of this main site, TGF-beta was still able to induce these microRNAs, suggesting that SMAD2/3 may use other sites more intensively to regulate miR100hg expression in the absence of the main ones. (19:05) So, even though the regulation of this transcript was more complicated than we initially saw, the regulation indeed happens through SMAD2/3 transcription factors. So, the answer of this question is yes as well. So, to summarize, we think the TGF-beta activates SMAD2/3 that in turn binds to miR100hg locals and induces other expressions. The primary transcript is elevated that eventually the mature microRNAs are elevated. This all makes sense except for one aspect.
19:45 - Slide 14
So, I told you before that miR100hg is tricistronic host gene for miR100, 125b, but also let-7a. And here is the problem. We are saying that miR100 and 125b are oncogenic microRNAs up-regulated by TGF-beta and let-7a, as discussed before, is a very well-known tumor suppressor microRNA. So, any rise of let-7a levels following miR100hg induction would serve to counteract this effect. (20:23) So, what is going on? We decided to look at the kinetics of those three microRNAs upon TGF-beta treatment. As you can appreciate from this figure, why is miR100 and miR125b increased throughout the time course? Let-7a initially rise and then goes back to the initial untreated levels. (20:49) So, indeed also in our microRNA profiling with 72-hour TGF-beta treatment, let-7a was not significantly up-regulated. So, this later suggested that let-7a is repressed at the post-transcriptional level by a factor that is possibly regulated by TGF-beta.
21:10 - Slide 15
So, a very well-known inhibitor of let-7 processing is LIN28A/B. LIN28 can inhibit let-7 through several mechanisms. It can bind to pri-let-7 and inhibit the processing by the microprocessor or it can bind to pre-let-7 and inhibit the processing by Dicer, or it can induce degradation of pre-let-7 by promoting a legal irregulation by the TUT enzymes. (21:45) So, we looked at our RNA-seq data and found that LIN28B was indeed significantly up-regulated by TGF-beta. Instead LIN28A was not expressed at all.
22:00 - Slide 16
So, we looked at the expression of LIN28B in the same TGF-beta time course, and found that LIN28B starts accumulating at six hours — you have the RNA at the top and the protein level at the bottom — six hours when let-7a level starts to go down. We also further validated this finding by showing that in knockout cells for LIN28B, TGF-beta was able to induce let-7a levels, validating our findings. (22:32)
22:34 - Slide 17
So, before going into details of how these microRNAs effect EMT and stemness phenotype, I wanted to summarize here the main method of manipulation of microRNAs. So, there are a wide range of methods that can be used to manipulate microRNA levels, and here, at least, is the one that we use in the lab and, particularly, I've used with this project. (23:00) So, in terms of transient manipulation, we have used lipofectamine transfection of precursor microRNAs for over-expression and anti-microRNAs for inhibition. As for stable manipulation we used precursor microRNA expression, Lentivectors sold by System Biosciences. (23:26) And for knock-down we've used the miRZip anti-microRNA expression Lentivectors and, again, from System Biosciences or we use the CRISPR knock-out of microRNA locus. In terms of the miRZip technology and the miRZip Lentivectors expresses a short hairpin RNA that after processing preferentially produces an antisense microRNA. So, this results in a stable accumulation of anti-microRNAs that literally zip the microRNA leading to permanent microRNA inhibition.
24:09 - Slide 18
So, to evaluate whether these microRNAs regulate EMT, we firstly looked at the morphology upon transient over-expression using pre-microRNAs. So, we did this in three different cell lines, and you can appreciate that upon the over-expression, the cells become more spindle-shaped and characteristic of mesenchymal cells. We also employed immunofluorescence painting to see whether these MicroRNAs were acting by reducing E-cadherin levels. (24:40) As you can see from the red staining, E-cadherin is still expressed. However, we observed the miR100 induced disruption of E-cadherin at the level of cell to cell junction and miR125b promoted a complete de-localization from the junctions.
25:01 - Slide 19
And we next looked at the opposite process. So, MET, which is mesenchymal to epithelial transition. And in order to do that we used stable knock-down clones using miRZip technology, and we've used the cells that express these miRs at the highest levels which are S2-007. (25:27) And so, as you can appreciate from the morphology here the knock-down cells, they acquire a more epithelial phenotype and also we used the cells to assess motility, and we performed the wound-healing scratch assay. And you can see that in the knock-down cells for miR100 and 125b the motility is indeed impaired.
25:29 - Slide 20
And next we wanted to investigate the role of these microRNAs in cancer stem cell formation. So, this is because the activation of EMT programs has been associated with the acquisition of stem cell trait, the normal and neoplastic cells. So, and the gold standard technique to look at cancer stem cells in vitro is the sphere formation assay. So, it consists of seeding cells in low attachment conditions without serum at a very low density. (26:32) In these conditions only cancer stem cells survive and they will start dividing, forming spheres. So, the number of spheres is a proxy for stemness. So, top left we have increased number of spheres upon transient over-expression of the miRs. In the last bottom panel we have reduced number of spheres upon transient inhibition of microRNAs. (26:58) And here on the right, again, we have reduced number of spheres in the stable knock-down clones.
27:07 - Slide 21
So, what about in vivo? So, to assess whether miR100 and 125b regulate tumor initiation capacity of pancreatic cancer, we performed a serial dilution assay in nude mice using the S2-007 stable knock-down clones. So, you can appreciate that the knock-down showed a strong reduction of tumorigenesis and here, using the highest dilution, we have formation of tumors in all conditions. (27:39) However, the knock-down cells are for much smaller tumors.
27:45 - Slide 22
Next we looked at metastases. So, this is a schematic of the method that we employed. So, firstly, we generated loosely-phrase stably expressing cells so that we could image the cells in vivo. We then injected the cells into the spleen of nude mice and cells through the portal vein we reach the liver which is the main metastatic site for pancreatic cancer. (28:14) And after one week we removed the spleen to avoid growth of a primary tumor there and we track the cells by imaging using the ID system. We then sacrifice the mice and perform some ex vivo imaging. And just at the bottom are our results and so knock-down for miR125b was very effective in reducing liver metastasis. And for miR100 we noticed a trend, but it was not significant. (28:45) And this mirrors the effect we had in our in vitro assays where it seems that the effect of miR125b is always more significant than miR100.
29:00 - Slide 23
Right. Then then the obvious final question was, how important are those microRNAs in the context of TGF-beta response? Meaning, if we inhibit the function of these microRNAs and we treat cells with TGF-beta, are the TGF-beta responses impaired? The answer is yes. So, here on the left you can see a reduction of sphere-formation in knock out microRNA clones treated with TGF-beta compared to wild-type cells treated with TGF-beta. (29:35) And here on the right you can see that TGF-beta induced motility is impaired in knock-out cells. So, in these experiments we use the CRISPR technology to knock-out the microRNAs.
29:51 - Slide 24
And briefly, this was our strategy. So, in both cases we use pair of sgRNA to disrupt the microRNA locals. So, these deletions made impossible for the locus to be transcribed and, therefore, we had no production of mature microRNAs.
30:09 - Slide 25
Just very quickly, I will touch on the clinical relevance of these microRNAs. So, here we perform in situ hybridization from miR100 and 125b in a collection of 100 PDAC samples and quantify the micron expression intensity. (30:27) So, we found that high miR100 and miR125b levels were associated with both reduced overall survival and reduced disease-free survival.
30:38 - Slide 26
So, so far we have shown that these two TGF-beta regulated microRNAs are involved in several and overlapping phenotypes that could be explained by the regulation of multiple targets. So, we've developed a novel approach for target discovery that we called RIP-USE. So, RNA immunoprecipitation followed by unbiased sequence enrichment analysis. (31:06) So, this method is developed in several steps. Firstly, we over-express the microRNA of interest in cell lines. In our case, we over-expressed miR100 and 125b in PANC-1 cell. Then we performed, in parallel, Ago2-RIP-seq to reveal transcripts that were significantly enriched in Ago2. (31:29) And RNA-seq, also to RNA, to reveal transcripts that are functionally repressed by the microRNA Ago2 target interaction. This is followed by unbiased sequence enrichment analysis using Sylamer, as is shown here, or Seward's algorithms to identify the motives of microRNA target interaction. (31:53) Finally, to test whether the motives identified also inhibit the expression of those genes, we performed cumulative distribution analysis using RNA-seq data. And I won't show the results, but basically in our case for miR100 and 125b, the three prime UTR of transcripts that we're loading onto Ago2 upon over-expression of these miRs were also strongly enriched with their seeds motives. As expected, transcript containing the canonical seeds was significantly down-regulated compared to transcripts lacking these motives. (32:37)
32:38 - Slide 27
So, the great advantage of RIP-USE is that we can identify direct targets that come from the overlap of Ago2-RIP-seq data in the down-regulated genes from the RNA-seq data. So, considering this overlap, we performed pathway analysis using the IPA software, and as a result of this analysis we notice that these microRNAs were regulating common pathways, including P53 signaling, apoptosis, and the role of CHK proteins in cell cycle checkpoint control.
33:15 - Slide 28
And this is just a summary of the microRNA target interaction based on RIP-USE, and also they're grouped in the most significant pathways. So, as you can see these microRNAs regulate many targets and are involved in many different pathways.
33:36 - Slide 29
So, finally, this is a schematic representation of our proposed model. So, TGF-beta, by as much as three, induces transcription of miR100hg which drives up-regulation of miR100 and 125B. We have shown that TGF-beta induces also LIN28b to keep let-7a levels unchanged. These microRNAs regulate many common pathways to induce EMT and stemness and before PDAC progression.
34:08 - Slide 30
So, to conclude, I would like to thank, firstly, my university, Imperial College London, that allowed me to share my research with you today. Of course, my mentor, Dr. Leandro Castellano, who is a great supervisor and a brilliant scientist, and taught me all I know about non-coding RNAs in cancer. I'd like to thank, also, my co-supervisor, Professor Justin Stebbing, and all the lab members. (34:34) Those are just the ones that were directly involved in my work and helped me greatly with this publication. I'd like to thank the collaborators that were directly involved in my work, so Dr. Luca Magnani and Prof. Long Jiao from Imperial College London. Professor Christopher Heeschen from Barts Cancer Institute in London and Dr. Elisa Giovannetti from the Cancer Centre in Amsterdam. And, of course, I'd like to thank the funding bodies for this study that were mainly Pancreatic Cancer U.K. and Action Against Cancer. (35:11) And if you want to know more about this work, this is the article describing my project and it was published in Nature Communications last May.
35:22 - Slide 31
And, also, obviously, I'd like to thank everyone for signing in today and listen to this presentation. I'm happy to take any questions or comments you may have. Also, I welcome you to connect with me on Twitter or LinkedIn if you would like to learn about the research in our laboratory at Imperial College. Thank you very much.
Moderator: Thank you, Dr. Ottaviani, for your informative presentation. We will now start the live Q&A portion of the webinar. (35:58) If you have a question you'd like to ask, please do so now. Just click on the "Answer a Question" box located on the far left of your screen. We'll answer as many of your questions as we have time for. So, let's get started. Our first question is, you've showed the correlation of these two microRNAs with overall survival and disease-free survival in patients. Are high levels of these microRNAs also correlated with great or differentiation of the primary tumor?
Yeah. This is actually a great question. Yes, we have performed this analysis. So we've done the correlation between miR100 and miR125b with grade of differentiation in the hundred patients samples. So, we actually discovered that only high miR125b and not high miR100 was significantly associated with higher tumor differentiation grade. Thanks for the question.
Moderator: Of course. Dr. Ottaviani, what is the next stage for these microRNAs?
Yeah. This is an obvious question, I guess. We are very excited about this work. So, with this work, we found that those two microRNAs could be important targets for therapy. So, now we are really working to try to bring those targets closer to the clinic. (37:30) So, we are setting up a model of nanoparticle deliveries in order to deliver those anti-microRNAs, and we are setting up 3D model cultures as well so we can test this nanoparticle on a 3D model, and eventually we're going to move to in vivo studies.
Moderator: And our next question is, can you expand a bit more on the transcriptional regulation of miR100hg by SMAD2/3?
Yes. So, yeah, apology. I went a little bit quick on that slide. So, we've basically realized that miRNA100hg has a quite complex regulation. So, from the result of the initial SMAD2/3, ChiP-seq we thought it was quite clear those SMAD2/3 were strongly binding at the transcription start site. (38:35) And, in fact, I can just bring the slides back so you can follow me a little bit better. So, that's slide 14. Here we go. So, we should see the ChiP-seq track now. So, yeah. So, we've seen this very strong binding, the transcription start site. So, to ultimately prove that these — this is where the most important region. (39:09) We deleted this region with CRISPR, and thinking that if we don't have this locus, then, obviously, when we treat this with TGF-beta we won't have the induction of the microRNAs anymore. However, when we did this experiment after deleting part of the promoter, we did see low levels of miR100hg and also low level of microRNAs. (39:37) But TGF-beta was still able to induce miR100 and mir125b. So, then, we looked, actually, in detail and we found that actually miR100hg has multiple transcription start sites. According to, we've looked at H3K27 acetylation and markers. And, also, we could see other SMAD2/3 binding throughout the transcript. (40:04) Here you can see in replicate 2, you can see some other binding, not so much in replicate one. But we've checked some ChiP-seq data from other laboratories. For example, from PDAC cells isolated from mice, and we see clearly multiple binding throughout the miR100hg gene. (40:28) So, therefore, we hypothesize that actually SMAD2/3 may use other sites more intensively to regulate miR100hg expression in the absence of the main ones. However, we're pretty sure the regulation of this transcript is occurring through SMAD2/3. In fact, we show that if we knock-down SMAD2/3 by three bias RNAs, we completely abolish the TGF-beta and use regulation of miR100 and 125b levels. (41:05) And, also, we showed that the miR100 and miR125b level, they start increasing very early after TGF-beta treatment, and indicating that this is a very direct mechanism from TGF-beta through SMAD2/3 transcription factors. I hope that it's clear now.
Moderator: We're getting some very, very interesting questions right here. (41:32) So, so here's maybe our next one. How far away are you from conducting clinical trials in patients with this exciting potential therapy?
Yeah. This is a very good question. So, we are really working really hard now to try and, as I said at the beginning, to try to use these targets, eventually, in a clinical setting. (42:00) But I think that realistically, we're talking about around ten years' time, I would say, because we need to set up all the delivery system and we need to test it in vitro, in vivo and then, ultimately, we can start testing in humans, so in clinical trials. Thank you.
Moderator: And it looks like our next question is going to be, why did you inject the cells into the spleen for pancreatic metastatic assay? Could you explain this model over orthotopic injection of cells in the pancreas?
Yes, absolutely. This is a very good question. So, we've decided to do the spleen injection because it's a very accepted experimental model to generate liver metastasis. Of course, you can inject cells also topically in the pancreas and spontaneously they will form — you wait, and then will form metastasis. (43:06) However, this is a quite long process. So, the beauty of the splenic injection is that it's a very quick assay. So, after injection in the spleen the cells reach the liver within minutes. (43:19) So, I suppose the limitation of this method is that with the splenic implantation of cells we can measure the ability of the cells to grow in the liver, but we're not able to look at the detachment of cells from the primary pancreatic tumor followed by invasion in this trauma. So, it is a good method, but it's really looking at the experimental model of liver metastasis. Yes.
Moderator: SMAD4 is frequently mutated in pancreatic tumors. Perhaps one could hypothesize that the downstream mediators of TGF-beta might differ between SMAD4 wild-type in a mutant. Can you tell us the SMAD4 status of the cell lines you have used in the study?
Yeah. This is a great question. It's absolutely right. So, SMAD4 mutation occurs very frequently, in we think around 31 percent of cases. (44:31) So, in our study TGF-beta mediating induction of this microRNA seems to occur only in PANC-1 and also COLO 357 cells. And, also, we had some KPC derived cells. So, cells deriving from pancreatic mouse model. So, all of those cells are wild-type for SMAD4. (44:59) And, also, when we checked, we checked the regulation in BxPC-3 and S2-007 cells. There is no up-regulation by TGF-beta and those cells do not express SMAD4, and we've shown this experimentally as well. Yeah, this is a very good question.
Moderator: And it looks like we are going to have time for one more question. I'm a bit unclear on why RIP-USE would offer a significant advantage over more conventional methods. For example, over just combining RNA sequencing, Sylamer and TargetScan predictions.
Yeah. This is a good question. So, basically, with RIP-USE, so we are talking about combining Ago2-RIP with Sylamer or Seward's and RNA-seq after over-expression of the microRNA of interest. (46:01) We can obtain accurate identification of direct targets and important information, also, about the type of regulation. So, in particular, Ago2-RIP, experimentally identifies the transcripts that are directly bound to the microRNA of interest, and this can happen through canonical or non-canonical seed interaction. (46:28) If you were to use only RNA-seq, let's say, combined with Sylamer and TargetScan, you could be direct targets that act through non-canonical seed binding because TargetScan would only look at canonical seeds. So, in addition, we found that if we use Sylamer in RIP-seq data it gives very, very good results and it is much superior as detection of microRNA interacting sites compared to RNA-seq. (47:00) So, in fact, when we try to use Sylamer after RNA-seq we didn't see any clear significant signal for canonical microRNA interaction sites. So, you cannot identify neither canonical or non-canonical sites. And, finally, I'd say that RIP-USE, you allow to establish whether the identified motive, that are canonical and non-canonical, are actually capable to down-regulate the transcripts. (47:30) For example, alternative method like CLIP and CLASH experiments have indeed demonstrated that some microRNAs could interact with transcripts through seedless regions, but because these methods are not combined with RNA-seq, they cannot indicate whether these non-canonical interactions are functional for target regression. So, yeah.
Moderator: Thank you, again, Dr. Ottaviani. Do you have any final comments for our audience?
No. I'd like to thank everyone and I hope this presentation was useful for you. And, please, get in touch if you have further questions or comments, yeah, at any point, or you want to know a bit more of what we do in the lab. Thank you.
Moderator: Before we go, I'd like to thank the audience for joining us today and for their interesting questions. (48:29) Questions we did not have time for today and those submitted during the on-demand period will be addressed by the speaker via the contact information that you provided at the time of the registration. We would like to thank Dr. Ottaviani for her time today and for her important research. We would also like to thank LabRoots and our sponsor Thermo Fisher Scientific for underwriting today's educational webcast. This webcast can be viewed on-demand through January of 2019. LabRoots will alert you via e-mail when it's available for replay. We encourage you to share that e-mail with your colleagues who may have missed today's live event. Until next time, goodbye.
End Presentation: 49:07
Links to content or other Internet sites should not be construed as an endorsement of the organizations, entities, views or content contained therein. The opinions and/or views expressed on social media platforms represent the thoughts of the individual and online communities, and not those necessarily of Thermo Fisher Scientific.
For Research Use Only. Not for use in diagnostic procedures.