Cell Cycle Conference October 2021

Video Transcript

• 00:00 - 00:12:  So welcome to the Cell Cycle Club.
• 00:12 - 00:20:  You know, this meeting is coming on for 10 years old now, and so it's really great that it's continued.
• 00:20 - 00:36:  We've always had a focus on trying to, you know, encourage early career researchers to give talks, particularly for the short talks, and try and focus as much as we can on early career researchers.
• 00:36 - 00:40:  So if you'd like to ask a question, you can put that in chat.
• 00:40 - 00:58:  Just put the letters ECR to indicate that you're a student or postdoc, and we will try and prioritize those questions because we value those questions most of all, much more than from old fogies like myself.
• 00:58 - 01:07:  We also have an opportunity for sort of networking and feedback, and we keep this meeting, we try to keep it quite informal.
• 01:07 - 01:10:  So please don't hesitate to ask questions.
• 01:10 - 01:14:  And of course, we have a great keynote at the end of the meeting.
• 01:14 - 01:22:  I want to thank Abcam. Abcam has sponsored this meeting from its inception all those years ago.
• 01:22 - 01:28:  And although sadly, we're not allowed to have beer anymore, we really value the organizational support.
• 01:28 - 01:34:  I doubt Rob and I would have the technical know-how to sort out the Zoom meeting from quite this format.
• 01:34 - 01:41:  So thanks very much to Abcam. They've been such great sponsors over the years.
• 01:41 - 01:46:  A lot of the young folks presenting their work today are presenting unpublished data.
• 01:46 - 01:56:  So please do not advertise the contents of their talks to your colleagues unless you ask the speakers first.
• 01:56 - 02:01:  Some of the work should be treated as a personal communication. It is confidential.
• 02:01 - 02:06:  We try to keep the meeting as interactive as possible. So the short talks are 20 minutes.
• 02:06 - 02:11:  That gives, obviously, we don't have interruptions anymore in the virtual format,
• 02:11 - 02:16:  but there's plenty of time for discussion and questions after each of the short sessions.
• 02:16 - 02:24:  So and as I say, I will try and prioritize and Rob will try and prioritize early career researchers.
• 02:24 - 02:28:  The breakout session, I'll discuss a bit more at the end.
• 02:28 - 02:34:  Polly's already mentioned it. And perhaps I'll do that at the end.
• 02:34 - 02:39:  Our next meeting will be again on the virtual format because it's working so well.
• 02:39 - 02:49:  And that will be in March. If you have a there will be a poll at the end because this meeting is evolving each time.
• 02:49 - 02:55:  It changes every time. So if you have an idea for a great keynote that you'd really like to see,
• 02:55 - 02:59:  please let us know and please do fill in the poll.
• 02:59 - 03:05:  Finally, we would say this if we're in an auditorium, if the fire alarm goes off,
• 03:05 - 03:09:  since you're all in your own homes, that's kind of your problem.
• 03:09 - 03:16:  OK, so with the introductions done, I'm going to introduce our first speaker for today.
• 03:16 - 03:25:  Let me just pull up my running schedule. And that is Denise, who's from the Ewald Lab in Tübingen, Germany.
• 03:25 - 03:30:  And she'll talk about start is not the metabolic commitment point of the cell cycle,
• 03:30 - 03:34:  which is odd because I always thought it was. But there you go.
• 03:34 - 03:37:  So welcome. Welcome, Denise. And we very much look forward to your talk.
• 03:37 - 03:43:  Thank you so much. OK, I'm sharing my screen.
• 03:43 - 03:52:  You can all see it now. And thank you so much for giving me the chance to talk about my PhD project today.
• 03:52 - 04:02:  So in this project, we're using budding yeast as a model organism to study the eukaryotic cell cycle, basically.
• 04:02 - 04:10:  Budding yeast divides asymmetrically. But other than that, its cell cycle is quite similar to more complex organisms.
• 04:10 - 04:17:  And during G1 phase, budding yeast passes a point called start, which is our main interest in this work.
• 04:17 - 04:22:  So the start point is analogous to the restriction point in mammals.
• 04:22 - 04:31:  And it was first described as the point where the cell no longer responds to mating pheromones and goes to division instead.
• 04:31 - 04:40:  And that start was accepted as the cell cycle commitment point where the cell no longer responds to metabolic signals and stress signals.
• 04:40 - 04:44:  And from this theory, our lab is interested in the metabolism part.
• 04:44 - 04:54:  And in this project, more specifically, nutrient signals. On nutrient signals, the current model states that nutrients regulate cell cycle,
• 04:54 - 05:02:  mainly during pre-start and post-start cells complete division independently of nutrient signaling.
• 05:02 - 05:14:  But what happens after start? In this work, we are challenging the current model and we hypothesize that start is not the one and only metabolic decision point.
• 05:14 - 05:24:  And to test our hypothesis, first, we need to understand if and how nutrient signaling affects cell cycle progression in post-start cells.
• 05:24 - 05:31:  And for that purpose, we did starvation experiments with unsynchronized cells using a cell growth platform.
• 05:31 - 05:41:  We first grew them in glucose media, then imposed total starvation without any amino acids or any carbon source that they can metabolize.
• 05:41 - 05:51:  And to monitor the cell cycle, we fluorescently detect cell cycle proteins and then we image them using fluorescent microscopy in long-term experiments.
• 05:51 - 06:01:  And then after that, we did single-cell analysis. And in order to monitor start, we used a protein called V5.
• 06:01 - 06:06:  So V5 is a transcriptional inhibitor. And when it's in the nucleus,
• 06:06 - 06:14:  it binds to the transcription factors, SBF and MBF, and inhibits the transcription of cyclins, Cln1 and Cln2 transcription.
• 06:14 - 06:27:  And as the cell progresses through G1, V5 is diluted and its inhibition is triggered by the Cln2-CDK1 complex, which leads to its export from the nucleus.
• 06:27 - 06:39:  And at the point at which 50% of the V5 is exported from the nucleus, this positive feedback loop becomes self-sustaining and start is defined as that exact point, basically.
• 06:39 - 06:47:  So here, V5 is not only a good marker to monitor start; it actually defines start.
• 06:47 - 06:59:  And after passing that point, the cell progresses in the cell cycle, goes through S phase and then this is repeated until the cell dies.
• 06:59 - 07:08:  And to show you how we use V5 in our experiments and how the cell is affected by nutrient switches, I will show this video.
• 07:08 - 07:17:  So here, the upper picture is the phase contrast image in our experiment, and this one will be the V5 mCherry channel.
• 07:17 - 07:24:  And in the graph, the x-axis is the time course of one of my starvation experiments.
• 07:24 - 07:32:  So the cell is only starved in between these red bars, and before and after that, the cell is getting enough glucose.
• 07:32 - 07:38:  And the y-axis will be nuclear V5 intensity. So I will start playing the video.
• 07:38 - 07:47:  Now we see the cell progresses through the cell cycle. And now the cell passes start, so we know that this is a post-start cell.
• 07:47 - 07:57:  And as you can see, under starvation, the cell either delayed or paused the cell cycle and then went into G1 after finishing it.
• 07:57 - 08:02:  And then after glucose supply, it continued the cell cycle normally.
• 08:02 - 08:11:  And also, when we compared the cell cycle before starvation and after starvation, we can clearly see that during the starvation,
• 08:11 - 08:21:  the cell cycle is definitely longer and we can see here that the cell is affected by the nutrients.
• 08:21 - 08:30:  And then we looked at hundreds of cells and we saw the cells tend to behave in three different kinds of ways.
• 08:30 - 08:39:  Firstly, we observed that only 15% of the cells continued the cell cycle and 14% of the cells permanently arrested during starvation,
• 08:39 - 08:44:  but then continued their cell cycle after we gave the glucose back.
• 08:44 - 08:53:  And 71% of the cells delayed or paused the cell cycle and the cell that I've just shown you is a part of that group.
• 08:53 - 09:06:  And while analyzing the cells, we noticed that these behaviors are also cell cycle phase dependent and we wanted to analyze it a little further.
• 09:06 - 09:13:  For that, for each single cell, we measured the time passed from start to starvation,
• 09:13 - 09:18:  which will be the x-axis of the graph, which I will explain in a minute.
• 09:18 - 09:27:  And then for the same cells, we measured, sorry, we measured the overall cell cycle duration, which will be the y-axis of this graph.
• 09:27 - 09:31:  So in this graph, all these dots represent one single cell.
• 09:31 - 09:37:  And this green line represents the average G1-S phase duration for unperturbed cells under glucose.
• 09:37 - 09:41:  And this red one is the threshold for permanent arrest.
• 09:41 - 09:53:  So what we get from this graph is if the cell is caught to starvation earlier in the cell cycle, the more they tend to basically react to nutrient changes.
• 09:53 - 09:58:  And we see here that the cells delayed their cell cycle more compared to this area.
• 09:58 - 10:07:  And we have more cells that permanently arrested the cell cycle under starvation conditions.
• 10:07 - 10:18:  However, other than these three groups of behavior, we saw a different type of behavior which came as a surprise to us.
• 10:18 - 10:24:  And I will call that the fourth group. And to explain it, I will also use a video.
• 10:24 - 10:32:  So the experimental setting and the old setting is the same, and the y-axis is still nuclear V5 intensity.
• 10:32 - 10:40:  So in this video, I want you to focus on the V5 mCherry channel mostly and the area around the starvation.
• 10:40 - 10:48:  And I will play the video now. So we see the cell is in G1, it passes start and then it will complete its cell cycle.
• 10:48 - 10:53:  After completing, we see that V5 is in the nucleus and out.
• 10:53 - 11:00:  And when the cell is exposed to starvation, we see that the V5 is translocated back basically.
• 11:00 - 11:07:  And to make it more clear, I will show this part of the video a little slower.
• 11:07 - 11:13:  So here in the mCherry V5 channel, we see that it's clearly in and completely out.
• 11:13 - 11:25:  And just after a couple of time points of exposing to starvation, we see that V5 is translocated back into the nucleus.
• 11:25 - 11:36:  And after seeing this, we asked two questions. Firstly, when does it happen in the cell cycle and how frequent will it be?
• 11:36 - 11:46:  And to answer this question, we analyzed 557 post-start cells and ended up in this graph, which I will explain now.
• 11:46 - 11:52:  So here, the x-axis is again the time between start and starvation basically.
• 11:52 - 12:00:  And this time we divided it into five-minute time frames and created bar charts for each five minutes.
• 12:00 - 12:08:  And the y-axis is just the total number of cells we analyzed for each bar, which is quite coincidental.
• 12:08 - 12:14:  And in these bars, the red part is the cells that translocate V5 back.
• 12:14 - 12:18:  And the black ones are the cells that go through the cell cycle.
• 12:18 - 12:25:  So what we get from this is firstly, we saw all cells in the first five minutes after passing start,
• 12:25 - 12:32:  they basically translocated V5 back into the nucleus instead of continuing with the cell cycle.
• 12:32 - 12:39:  And the second thing that we noticed is the first five minutes are really crucial in the cell's decision.
• 12:39 - 12:44:  When we look at this area here inside this red line,
• 12:44 - 12:52:  we see that the ratio of the cells that translocate V5 back is quite high compared to the cells that go through the cell cycle.
• 12:52 - 13:00:  And this ratio starts decreasing and decreases and reaches zero.
• 13:00 - 13:06:  And then we wanted to investigate this a little bit more.
• 13:06 - 13:12:  And then we ask what the mechanism is and what drives this V5 translocation basically.
• 13:12 - 13:21:  As a first step, we decided to stay in this first feedback loop to see if the first feedback loop is really activated.
• 13:21 - 13:25:  And we decided to check it using Cln2.
• 13:25 - 13:34:  So in this feedback loop, we already know if it's activated, Cln2 peak should come up after V5 transport.
• 13:34 - 13:42:  So what we basically did is to check if we will have any Cln2 peak or not.
• 13:42 - 13:45:  So I will also show this with a video here.
• 13:45 - 13:53:  So the one on the right is the Cln2 neon green channel and the one in the middle is the V5 mCherry basically.
• 13:53 - 14:05:  And this video will show again one of my experiments, but only the last part of the glucose period and really early starvation.
• 14:05 - 14:09:  So I will play the video here.
• 14:09 - 14:14:  As we see now, we have the V5 peak, so it's in the nucleus.
• 14:14 - 14:20:  And after a couple of time points, we will see the Cln2 peak here, which is now.
• 14:20 - 14:25:  And then starvation starts at 160.
• 14:25 - 14:34:  And after a couple of time points later, we will see the V5 going back into the nucleus again, which is now.
• 14:34 - 14:40:  So with this, yeah, the translocations happen with high Cln2.
• 14:40 - 14:45:  But now the question is, how about CDK activity?
• 14:45 - 14:56:  So are V5 translocations driven by CDK1 or is there a completely different mechanism that regulates V5 reentry?
• 14:56 - 15:02:  And to check this, we looked at how V5 is phosphorylated basically.
• 15:02 - 15:07:  So V5 has 19 phosphorylation sites in total.
• 15:07 - 15:14:  And 12 of those are CDK phosphorylation sites that are mainly responsible for V5 transport.
• 15:14 - 15:17:  And seven are the non-CDK phosphorylation sites.
• 15:17 - 15:24:  And from the seven, four are the G1 non-CDK phosphorylation sites.
• 15:24 - 15:34:  So our first guess was these four G1 non-CDK phosphorylation sites as a candidate for the V5 translocations.
• 15:34 - 15:43:  Because these four G1 non-CDK sites, what role they're playing is still not known completely.
• 15:43 - 15:51:  And to check this, we created a 4A phosphorylation site mutant.
• 15:51 - 16:00:  And before explaining the results for 4A, I want to go back to this bar chart that I've just shown you as a reminder,
• 16:00 - 16:03:  because I will be comparing the results with this.
• 16:03 - 16:14:  So what I did basically is I repeated the same starvation experiments and reached this bar graph again.
• 16:14 - 16:21:  So when we compared this to the wild-type V5, we saw a similar pattern.
• 16:21 - 16:29:  So we can still see a really high ratio of V5 translocating cells in the first five minutes.
• 16:29 - 16:34:  And still, the first 25 minutes are crucial.
• 16:34 - 16:42:  So these phosphorylation sites are clearly not responsible for the V5 translocations.
• 16:42 - 16:49:  And the G1 non-CDK phosphorylation sites are definitely not the answer.
• 16:49 - 16:56:  And now that we know the non-CDK sites are not the answer that we're looking for,
• 16:56 - 17:04:  we wanted to go to this second feedback loop and check the downstream regulators of start basically.
• 17:04 - 17:09:  And the first protein that we checked is a protein called Sig1.
• 17:09 - 17:14:  So what Sig1 is that it's a CDK inhibitor.
• 17:14 - 17:22:  And for the cell to progress through S phase, it needs to be degraded in the cell.
• 17:22 - 17:26:  With Sig1, we hope to answer two questions mainly.
• 17:26 - 17:32:  So is CDK activity inhibited by Sig1 during V5 translocations?
• 17:32 - 17:42:  And is Sig1 stabilization required for V5 translocations?
• 17:42 - 17:49:  So here, this graph shows an example cell that re-enters V5.
• 17:49 - 17:53:  The x-axis is the time course of the experiment.
• 17:53 - 17:58:  And the starvation period is the time in between the spread bars again.
• 17:58 - 18:04:  And the y-axis will be the nuclear fluorescent intensity.
• 18:04 - 18:10:  And the orange graph shows Sig1 intensity and the blue one is V5.
• 18:10 - 18:16:  So from this, we see that Sig1 is not stabilized at all.
• 18:16 - 18:23:  And it's completely out from the picture when the cell is translocating V5 back.
• 18:23 - 18:30:  And let me check if Sig1 can be predictive of the cell's decision in any way.
• 18:30 - 18:39:  And to do so, we ask if there's any difference in the cells that continue their cell cycle?
• 18:39 - 18:44:  And for that purpose, for each single cell in our data set,
• 18:44 - 18:48:  we looked at the minimum Sig1 intensity before the starvation,
• 18:48 - 18:51:  which is pointed as number one here.
• 18:51 - 18:58:  And we looked at the maximum Sig1 intensity before starvation, which is number two.
• 18:58 - 19:03:  And then we looked at the Sig1 intensity during the V5 re-entry.
• 19:03 - 19:07:  And we basically ended up with this.
• 19:07 - 19:14:  So if we take point number two as 100% and point one as 0%,
• 19:14 - 19:18:  what we saw is in the cells that continue,
• 19:18 - 19:22:  7.8% of Sig1 is not degraded under starvation.
• 19:22 - 19:27:  And in the cells that translocate V5, this was 10.9.
• 19:27 - 19:30:  And these numbers are really close to each other.
• 19:30 - 19:33:  And we only had 60 cells in our data set.
• 19:33 - 19:38:  So they will probably get even closer when we analyze more cells.
• 19:38 - 19:45:  And here, we concluded that Sig1 is definitely not stabilized during the V5 re-entries.
• 19:45 - 19:48:  And it's not predictive of the cell's decision.
• 19:54 - 19:59:  And after Sig1, we looked at Cln5, which is...
• 20:00 - 20:07:  B-type cyclin. And naturally, we ask, what happens to Cln5?
• 20:09 - 20:14:  And I will show you a couple of videos to see different types of Cln5 behavior.
• 20:15 - 20:22:  And this is the first one. So we have phase contrast and V5 mCherry channels on top.
• 20:22 - 20:26:  And the one on the bottom is the Cln5 neon green channel.
• 20:28 - 20:36:  And in the graph, the x-axis is the time course of the experiment.
• 20:36 - 20:38:  And before the red bar, the cell is getting its glucose.
• 20:39 - 20:43:  And after that, the cell is completely starved. So let's play the video.
• 20:47 – 20:55:  And now the cell goes through its cycle, normally under glucose. And now V5 is going in.
• 20:55 - 21:02:  And it passes start. And right after start, the cell is exposed to starvation.
• 21:02 - 21:10:  So it's a cell that's really early in the cell cycle. So what we get here is Cln5 synthesis has not been induced
• 21:10 - 21:18:  yet before starvation. And it basically stays like that during starvation. So it's not induced at all.
• 21:20 - 21:27:  And after starvation ends, even though you cannot see it in this graph, the Cln5 starts being
• 21:27 - 21:34:  synthesized as the cell leaves G1. And then the cells continue in their normal course.
• 21:36 - 21:43:  And this is the second example cell that I want to show you. So this setup is completely the same,
• 21:43 - 21:49:  and they're cells in the same experiment. So here, the cell is a little further in the cell cycle,
• 21:49 - 21:58:  as we can see from basically here. And the Cln5 synthesis is induced, but hasn't reached its max
• 22:00 - 22:03:  when we compare it to a pre-start phase.
• 22:03 - 22:09:  And the important thing to keep in mind is the maturation time of the fluorescent proteins while you're doing microscopy.
• 22:10 - 22:20:  So most likely, we most likely observe the real intensity of Cln5 when the starvation and the re-entry is exposed here.
• 22:20 - 22:26:  So this is the point that we look at. And probably the cell hasn't synthesized any Cln5
• 22:26 - 22:33:  after the re-entry, basically. So let's play the video to see what's happening.
• 22:35 - 22:40:  Now the cell is going through the cell cycle. I think it's good. And the starvation starts.
• 22:41 - 22:49:  So what happens is, so during starvation, the Cln5 intensity starts decreasing, and it
• 22:49 - 22:55:  reaches probably a baseline. As you can see from here, the intensity is really low,
• 22:55 - 23:00:  even though you cannot see a really high decrease. And after we give the glucose back,
• 23:00 - 23:08:  and like, as you can see at the end of the video, when V5 is exported, we have this really sharp
• 23:09 - 23:17:  Cln5 video, Cln5 signal, sorry. So what we saw here, that we saw two Cln5 peaks in the
• 23:17 - 23:25:  same cell cycle. And with these two cells, we thought they seem to be going back to a pre-start
• 23:25 - 23:33:  phase. But to be sure, and to confirm it, we will be using Cln2 in the future.
• 23:36 - 23:43:  And I also want to show you a second group of cells that confused us a little,
• 23:43 - 23:51:  and I will also show it with a video. So this is what makes this cell special or different.
• 23:51 - 23:56:  So this is a cell that's more further in the cell cycle, as you can see here.
• 23:58 - 24:04:  And another thing that's different is the slope of the V5 re-entry is lower here,
• 24:04 - 24:11:  compared to the other cells that we saw. And let's play the video.
• 24:16 - 24:22:  And now we press start, and then starvation starts. And as you see here, Cln5 reached,
• 24:23 - 24:32:  probably its maximum when we compare it to a pre-starvation phase. And here you can see,
• 24:33 - 24:42:  it stays really high. So our first assumption is that, so,
• 24:43 - 24:49:  this is a different type of re-entry, basically. And maybe it can be a post-start S-phase arrest.
• 24:52 - 24:55:  And we're still thinking about what this Cln5 state means, basically.
• 24:54 - 25:01:  So is it actually a post-start S-phase arrest, or something completely different?
• 25:01 - 25:10:  Or is Cln5 responsible to determine the irreversibility? And if you also have any ideas or recommendations on that,
• 25:10 - 25:17:  I will be really happy to take all of them. And as I come to the conclusions,
• 25:18 - 25:24:  so we saw that post-start cells respond to nutrient signals by delaying or arresting
• 25:24 - 25:31:  cell cycle progression. We also observed that the cells translocate V5 into the nucleus,
• 25:31 - 25:38:  under starvation, in 25 minutes after start, which implied that the start can be irreversible.
• 25:40 - 25:48:  And so far, we know that the reason for V5 re-entries are not G1 non-CDK phosphosites,
• 25:48 - 25:51:  it's not Sig1 stabilization, and it's not Cln2 repression.
• 25:54 - 25:58:  And as next steps, we will be trying to answer two questions, basically.
• 25:59 - 26:05:  The first one is the mechanism that's causing the V5 re-entries. And to do so,
• 26:05 - 26:14:  we want to check additional phosphorylation sites of non-CDK or CDK sites. And we will test
• 26:14 - 26:18:  a protein called KIF1, which is a stress-activated CDK inhibitor
• 26:19 - 26:25:  that is thought to act similarly to mammalian p21 and p27 proteins.
• 26:26 - 26:32:  And then we plan to understand until when it starts reversible. Is it
• 26:33 - 26:37:  very quick to pass the threshold, or is it at the beginning of replication?
• 26:38 - 26:45:  So, if you also have any recommendations on which proteins to use to see the reversibility of start,
• 26:46 - 26:49:  I'm also willing to and happy to listen to all of them.
• 26:50 - 26:55:  And in the project, I first want to thank my supervisor, junior professor Jennifer Ewald, who
• 26:57 - 27:04:  helped me a lot with her ideas, and my whole lab. Also, my second supervisor, Professor Donald
• 27:04 - 27:10:  Rappaport, and Saj Peli from the University of Lausanne, the Donchich Lab from UT Southwestern.
• 27:11 - 27:17:  Also, Gregor, Fabian, and Andreas from ETH Zurich. And also, thank you for your interest and
• 27:19 - 27:22:  if you have any questions or recommendations, I will be happy.
• 27:27 - 27:33:  Well, welcome back to the second session. So, the breakout session with all the parallel talks
• 27:33 - 27:37:  was something new that we tried, and I think they work really well.
• 27:37 - 27:41:  It was very easy to get into them, and the one that I attended was very interactive, and a lot of people attended.
• 27:41 - 27:45:  So, that's something that we're definitely going to continue to develop. But now we're going back
• 27:45 - 27:51:  to the main program, and we'll start it off with another selected short talk. This one is from
• 27:51 - 27:57:  Camilla from the Sorensen Lab at the University of Denmark. So, before the break, we had a talk
• 27:57 - 28:02:  about the G1-S transition, then we kind of got stuck in S-phase, and I'm afraid we're not going
• 28:02 - 28:07:  to go much further than S-phase, especially if we're going to start talking about WI-1. We're
• 28:07 - 28:13:  definitely not going to go into M-phase this time. So, this is a talk about the WI-1 kinase,
• 28:13 - 28:19:  of course, that limits entry into M-phase, but it's going to be on the work that has been carried
• 28:19 - 28:25:  out in Sorensen's lab that the role of how it supports DNA replication. So, Camilla, if you
• 28:25 - 28:31:  can share your screen, then we can start the presentation. Thank you.
• 28:33 - 28:38:  All right. Thank you so much for that introduction, and thank you everyone for attending my talk at
• 28:38 - 28:44:  this Digital London Cell Cycle Club. I'm honored to present my work for you today, and, of course,
• 28:44 - 28:48:  a big thank you to the organizers for allowing me to talk at this digital conference.
• 28:49 - 28:55:  So, as Rob just mentioned, my talk is called WI-1 Kinase Supports DNA Replication Fork Stability by
• 28:55 - 29:02:  Limiting CDK2 Activity, and I'm currently a PhD student at the Technical University of Denmark,
• 29:02 - 29:08:  but this talk today will be from my work that I did in my master's project in the Sorensen lab
• 29:08 - 29:15:  under the supervision of Klaus Sorensen at the Biotech Research and Innovation Center
• 29:15 - 29:23:  at the University of Copenhagen. All right. So, first, hang on. Yeah. First, I want to tell you
• 29:23 - 29:29:  a little bit about WI-1 and why this kinase is both interesting and relevant, and why I spent
• 29:29 - 29:36:  almost two years studying its functions. So, WI-1 is a conserved kinase, and it's conserved
• 29:36 - 29:42:  in eukaryotes all the way back to yeast, where it was first discovered, and actually the loss of WI-1
• 29:42 - 29:50:  is embryonic lethal in mammals, so it's quite an important kinase. A handful of WI-1 inhibitors
• 29:50 - 29:57:  have been discovered in the last decade, and one of them, called AZD1775, or in a little bit more
• 29:58 - 30:05:  friendly terms, Adversatib, has garnered interest as a chemotherapeutic drug,
• 30:06 - 30:11:  and it's currently in Phase II clinical trials. And this inhibitor has shown
• 30:12 - 30:17:  effects both as a single agent and also in combination therapies to treat cancers.
• 30:18 - 30:24:  Right. So, preclinical studies showed increased sensitivity to
• 30:25 - 30:32:  Adversatib in TP53 mutated cancer cell lines. However, P53 status does not always predict
• 30:32 - 30:39:  the response to WI-1 inhibition. So, in conclusion, I think WI-1 is an interesting
• 30:40 - 30:44:  kinase to study. It's definitely garnering interest as a chemotherapeutic drug,
• 30:45 - 30:50:  and it has already shown results as an anti-cancer agent. However, we still lack
• 30:50 - 30:56:  some knowledge on exactly how this kinase exerts its functions and what inhibiting it
• 30:56 - 31:02:  means for cells, and this includes identifying synthetic lethal relationships or biomarkers.
• 31:04 - 31:09:  Right. So, WI-1 is best known for its G2 checkpoint function, I'd say.
• 31:10 - 31:19:  So, this checkpoint, of course, is like the last sort of barrier before you go into mitosis,
• 31:19 - 31:25:  so it's a really important checkpoint, and it's activated in response to, for example,
• 31:25 - 31:31:  DNA damage or replication stress, and it involves the activation of the ATR and CHK1 kinases,
• 31:31 - 31:39:  which we also heard a little bit about before. And, of course, in response to DNA damage,
• 31:40 - 31:50:  the activity of WI-1 ensures that CDK1, its activity is limited, and downstream mitotic
• 31:50 - 31:57:  factors also are inhibited, which induces a strong cell cycle arrest. And when WI-1 is not present,
• 31:57 - 32:04:  there's a checkpoint override, which may cause cells to prematurely enter into mitosis,
• 32:04 - 32:11:  which can, in turn, cause genomic instability or mitotic catastrophe if there is DNA damage
• 32:11 - 32:18:  present. WI-1 also has a function in the intra-S phase checkpoint, where it, through
• 32:18 - 32:26:  the same mediators, ensures that there's a limitation to origin firing, which, in turn,
• 32:26 - 32:33:  slows forks and delays the cell cycle. And when WI-1 is not present, there's massive initiation
• 32:33 - 32:40:  of replication and widespread DNA damage. So, I will go more into the replication functions
• 32:40 - 32:46:  of WI-1, because WI-1 actually also functions in the unperturbed normal cell cycle.
• 32:48 - 32:54:  Right. So, a little bit about the functions of WI-1 in replication. So, again, WI-1 inhibits
• 32:54 - 33:02:  CDK activity. This is mainly CDK2 in the S phase. And what it does is that it limits origin firing
• 33:02 - 33:10:  of the dormant origins that are licensed. And this ensures that cells have enough
• 33:10 - 33:17:  building blocks, essentially, to carry out replication in a timely manner or in a
• 33:17 - 33:26:  faithful manner. And in the case of WI-1 being absent, this means that CDK activity goes massively
• 33:26 - 33:34:  up. And basically, a lot of origins are fired untimely, which, in turn, leads to nucleotide
• 33:34 - 33:41:  shortage and also shortage of other replication factors that may be limiting to the replication
• 33:41 - 33:49:  machinery. And actually, it's also been shown that this massive replication stress that occurs when
• 33:50 - 33:56:  forks no longer have the building blocks to continue replicating also exhausts RPA pools.
• 33:56 - 34:06:  So, RPA covers single-stranded DNA strands in replication when there's replication stress.
• 34:06 - 34:14:  So, in conclusion, the loss of WI-1 leads to widespread replication stress and
• 34:14 - 34:22:  RPA exhaustion of other replication factors and potentially replication catastrophe.
• 34:23 - 34:35:  All right. So, we already know of all these functions of WI-1 in regards to replication
• 34:35 - 34:42:  initiation and replication progression, but we hypothesized whether WI-1 also functions behind
• 34:42 - 34:53:  the fork. So, it was shown back in 2005 that BRCA2, which is mainly known as a factor involved
• 34:53 - 34:59:  in homologous recombination repair, but also is known to have an important role in fork protection,
• 35:00 - 35:06:  is regulated in a cell cycle-dependent manner through CDK activity. And what happens is CDK
• 35:06 - 35:13:  activity phosphorylates a serine in the C-terminus of BRCA2, and this in turn
• 35:14 - 35:20:  inhibits the interaction of BRCA2 with RAD51 filaments. And RAD51 filaments are important
• 35:20 - 35:31:  at stalled forks to protect from nucleolytic degradation, basically. So, we hypothesized that
• 35:31 - 35:36:  maybe WI-1 also has a role in protecting stalled replication forks, potentially by governing
• 35:36 - 35:43:  BRCA2 serine 3291 phosphorylation status through its limitation of CDK activity.
• 35:44 - 35:50:  So, my aim was to investigate whether WI-1 is protecting stalled replication forks from
• 35:50 - 35:58:  excessive degradation. And for this, I utilized a single molecule DNA fiber technique,
• 35:59 - 36:05:  DNA fiber assays, to investigate what happens at stalled forks. And the setup is,
• 36:05 - 36:10:  I think Charlotte also mentioned a little bit about how this is performed. You basically
• 36:10 - 36:17:  label your cells with a thymidine analog, CLDU, and then you switch to labeling with another
• 36:17 - 36:26:  analog, IDU. And for the fork protection assay, you then add, we at least, we use hydroxyurea
• 36:26 - 36:34:  to stall replication forks that are replicating and incorporating these analogs. And then you can
• 36:34 - 36:42:  assess how well these stalled forks are being protected by looking at the ratio between the
• 36:42 - 36:47:  length of the green tract and the length of the red tract. And in the case of less protection of
• 36:47 - 36:53:  forks, there may be nucleolytic degradation, which means that the IDU tract basically gets
• 36:53 - 37:01:  shorter, and thus there is a decrease in the IDU to CLDU ratio. So, this is the setup for
• 37:01 - 37:08:  the data that I'm going to show you. And, right, this is a representative image of how these fibers
• 37:08 - 37:14:  would look. So, yeah, you have, this is in the presence of hydroxyurea after the labeling,
• 37:14 - 37:20:  and you can see that we have nice incorporation of CLDU and then IDU. And then we test what
• 37:20 - 37:27:  happens when you add WE1 inhibitor to these IDU tracts. Are they being degraded?
• 37:28 - 37:35:  So, I tested first my hypothesis that WE1 activity protects H2 stalled forks.
• 37:35 - 37:39:  And what I could see in both U2OS and HEX cells, two different cell lines,
• 37:39 - 37:47:  is that upon WE1 inhibition, I see a decrease in the IDU to CLDU ratio, which would indicate that
• 37:47 - 37:54:  there's excessive degradation of nascent DNA at these stalled forks. See this in both cases.
• 37:55 - 38:04:  And I also look at the effect of WE1 inhibition on CDK activity, and I can
• 38:05 - 38:11:  see that there is a massive increase in this signal, which is recognizing
• 38:12 - 38:18:  CDK substrates, phosphorylated CDK substrates. So, it's a readout for CDK activity, and I see
• 38:18 - 38:24:  a massive increase here. I also have a readout for replication stress, which is this phospho-RPA
• 38:24 - 38:30:  that Charlotte also mentioned. And I can also see that upon WE1 inhibition for four hours,
• 38:30 - 38:38:  this is massively increased. So, indeed, it does seem that WE1 activity is important for the
• 38:38 - 38:46:  protection of stalled forks. So, I wanted to go into a little bit more detail about these CDKs.
• 38:46 - 38:54:  So, I first wanted to test whether this was really through the increased CDK activity. So,
• 38:54 - 39:04:  I used a pan-CDK inhibitor, Roscovitine. So, I did the same assay, but I added a CDK inhibitor
• 39:04 - 39:08:  as well to see if this would rescue the fork degradation that I saw. And this was indeed the
• 39:08 - 39:17:  case. And I could also rescue the antibody, the Western blot signal. So, I wanted to look
• 39:18 - 39:27:  more into whether this is CDK1 or CDK2-mediated, because WE1 phosphorylates both of them,
• 39:27 - 39:35:  and CDK1 is very important for the mitosis part, whereas CDK2 could be more important for
• 39:35 - 39:41:  replication, since that is the main S phase CDK. So, I utilized some inhibitors that show
• 39:41 - 39:48:  some increased sensitivity or selection to one CDK over the other. And what I could see,
• 39:48 - 39:56:  interestingly, was that inhibition of WE1 and CDK1 did not rescue the degradation seen with
• 39:56 - 39:59:  just WE1 inhibition. However, when I used a
• 40:00 - 40:07:  more CDK2-specific inhibitor, I could see a rescue of the phenotype. And I also validated
• 40:07 - 40:12:  this by depleting CDK2, and I could see the exact same thing. So, this would suggest to
• 40:12 - 40:22:  me that the fork protection that WE1 carries out is through its limitation or its inhibition
• 40:22 - 40:29:  of CDK2 activity. Then, to get a little bit more into the mechanism, we thought, okay,
• 40:29 - 40:38:  this could potentially be a result of chromosome pulverization, which has been shown to happen
• 40:39 - 40:46:  when WE1 is inhibited for prolonged amounts of time. So, MUS81 and the SLX4 nuclease complex
• 40:47 - 40:57:  is usually only activated at mitosis to make sure that chromosomes are segregated. And WE1
• 40:57 - 41:05:  activity keeps the MUS81 and the SLX4 apart from each other through CDK1 regulation.
• 41:06 - 41:12:  So, it could be that these chopped up fibers are a result of chromosome pulverization. And
• 41:13 - 41:19:  you can see here from the paper that showed nicely that, okay, upon WE1, loss of WE1,
• 41:19 - 41:24:  there is massive pulverization of chromosomes. And this is indeed through MUS81 activity.
• 41:25 - 41:31:  So, I did the setup again. And this time, I depleted cells for MUS81. And what I found
• 41:31 - 41:39:  interestingly was that this is actually not a result of MUS81 activity. So, it must be something
• 41:39 - 41:44:  happening more upstream of this, something happening more directly at the fork protection
• 41:44 - 41:52:  state. So, then I will talk a little bit about these pathways for protecting stalled forks.
• 41:53 - 41:58:  So, upon replication stress, there are lots of factors that work at the stalled fork
• 41:58 - 42:08:  and perform a process known as reversal, which is a protective way to our protective mechanism,
• 42:08 - 42:16:  basically. And when you have protection factors such as BRCA2 and a plethora of other factors,
• 42:16 - 42:23:  you ensure that these DNA strands are protected from excessive degradation.
• 42:24 - 42:30:  So, it's already known that in the absence of BRCA2, there is excessive degradation by
• 42:30 - 42:37:  a nuclease called MRE11. And later on, this serves also as a substrate for MUS81
• 42:37 - 42:43:  in complex with ME1. And since our initial hypothesis was that
• 42:44 - 42:49:  WE1 could potentially be inhibiting BRCA2 through this serine phosphorylation,
• 42:50 - 42:57:  we thought, okay, maybe inhibition of WE1 will also lead to fork degradation by MRE11.
• 42:57 - 43:04:  So, I tested this by adding an MRE11 inhibitor, and I could see some
• 43:08 - 43:13:  increase in the ratio, but still, this is not rescuing entirely the phenotype.
• 43:14 - 43:21:  So, there is another pathway through which forks are processed, and this is mainly carried out by
• 43:21 - 43:32:  DNA2. So, these two sort of branches are somewhat exclusive, and cells either have one being active
• 43:32 - 43:40:  or the other, depending on which protective factors are present. So, I tested whether WE1
• 43:40 - 43:49:  activity is perhaps limiting DNA2 at stalled forks. And indeed, it seems that way. So, I depleted
• 43:49 - 43:55:  DNA2 and performed the same assay, and I could see a complete rescue of the phenotype, which
• 43:55 - 44:00:  suggests that WE1 is basically keeping DNA2 from degrading stalled forks.
• 44:03 - 44:10:  Right. Then I took a step back to see whether also other CDK regulators could potentially
• 44:10 - 44:18:  be working in the same way through the CDK2 inhibition. So, as I mentioned in the beginning,
• 44:18 - 44:27:  there is also CHEK1 kinase, which also limits CDK activity, and I wanted to see if inhibition of
• 44:27 - 44:34:  this kinase could also lead to degradation of excessive degradation of stalled forks.
• 44:34 - 44:40:  And interestingly, I found that both using two different inhibitors, I didn't see a degradation
• 44:41 - 44:49:  in or a decrease in the ratio. Actually, I saw an increase, which is quite hard to explain,
• 44:49 - 44:57:  actually. So, this is some intriguing results that could perhaps also serve for some discussion later.
• 44:58 - 45:05:  And I also checked whether P21, which is also a CDK regulator, would have some
• 45:05 - 45:13:  similar function, and this was also not the case. So, this seems like somewhat specific to the
• 45:13 - 45:22:  function of WE1. Just of note, I could also see that upon CHEK1 inhibition, I wouldn't see the
• 45:22 - 45:29:  same, like, massive increase in the CDK activity. There is still massive replication stress, but I
• 45:29 - 45:34:  don't see this really high explosion of CDK activity. So, this could also maybe explain why
• 45:35 - 45:39:  these two kinases, although they share similar functions, they're still quite separate.
• 45:42 - 45:50:  Right. So, to summarize my findings, we know that when replication forks stall,
• 45:51 - 45:56:  and we have active WE1, then there is controlled activity of CDK1 and CDK2,
• 45:57 - 46:03:  which ensures a proper DNA damage response, replication slowing, and checkpoint activations,
• 46:03 - 46:09:  and this all serves to give some time for the DNA replication to, or for DNA damage
• 46:09 - 46:14:  or replication stress to be dealt with properly and ensures DNA integrity.
• 46:15 - 46:22:  However, when WE1 is not active, we know that there are some functions of WE1 with regards to
• 46:22 - 46:30:  the checkpoint and limiting the replication, as I just said, but now I believe I've shown
• 46:30 - 46:36:  that loss of WE1 also deregulates CDK2, which in some way
• 46:41 - 46:54:  increases DNA2 and basically, yeah, so DNA2 has access to a stalled replication fork and
• 46:54 - 47:03:  excessively degrades it, which can give rise to DNA double-strand breaks and genomic instability,
• 47:03 - 47:10:  basically. Yeah, sorry. So, basically, yeah, the decreased fork protection may explain
• 47:11 - 47:16:  a rapid accumulation of DNA double strands that we see after WE1 inhibition.
• 47:18 - 47:23:  So, I don't know how I am on time. Maybe one more minute. Basically, I just wanted to
• 47:23 - 47:31:  mention a few other pathways that could be very interesting to look into. So,
• 47:33 - 47:40:  WE1 inhibition has already been shown to synergize with PARP inhibition, and this
• 47:40 - 47:47:  is very toxic to cells, both normal cells and cancer cells, because this creates a lot of
• 47:47 - 47:53:  replication stress and cells, like no cells, are basically able to deal with it. However,
• 47:54 - 48:01:  a paper by Fang et al. showed that if you do sequential therapy with PARP inhibition
• 48:01 - 48:09:  and WE1 inhibition, you target the cancer cells over the normal cells due to the cancer cells
• 48:09 - 48:15:  having high endogenous levels of replication stress. And this could be interesting to also look
• 48:15 - 48:24:  whether there's some synergy in the fork protection. Also, it has been found that WE1
• 48:24 - 48:33:  inhibition synergizes with, or is especially potent in cancers that have mutations in DNA
• 48:33 - 48:39:  repair pathways, such as the Fanconi anemia proteins or the BRCA proteins. So, these cells
• 48:39 - 48:45:  have also less fork protection, and it would make sense that inhibition of WE1,
• 48:46 - 48:54:  in addition to the effects it has on the checkpoint overrides and the premature mitotic
• 48:55 - 49:03:  entry, also has some toxicity coupled to its even further decreased stabilization and protection
• 49:03 - 49:10:  of forks. And finally, WE1 inhibition has also been shown to synergize with CHEK1 inhibition
• 49:10 - 49:15:  and ATR inhibition, which are also some interesting directions that you could go into
• 49:16 - 49:23:  to explain why these show such potent toxicity in combination. Right. And with that, I will
• 49:23 - 49:28:  conclude my talk. And I want to just thank the Samson group, especially Valdemar Petrosius,
• 49:28 - 49:34:  who has been the PhD supervisor overseeing much of my project in this group, and of course, Klaus Samson.
• 49:35 - 49:43:  So, thank you all for listening. Thank you. Just me clapping. Everybody, you can only hear me
• 49:43 - 49:49:  clapping. So, we already have a few questions. So, we'll start with the Q&A one, which is from
• 49:50 - 49:56:  Jumana Bhattar. Can you unmute and ask the question?
• 50:03 - 50:06:  Yeah, there we go. You just have to unmute. You're still on mute. Yeah, there we go.
• 50:07 - 50:10:  Hi, everyone. Actually, I'm not Jumana. I'm Diego. I'm here with Jumana.
• 50:11 - 50:20:  So, I have a curiosity about what you show with CHEK1 inhibition is that CHEK1 inhibition
• 50:21 - 50:24:  not only doesn't cause fork degradation, but it actually
• 50:26 - 50:32:  seems to stabilize it somehow. So, I am kind of curious about it because I thought
• 50:33 - 50:39:  CHEK1 inhibition could be associated with fork breakage. So, could you comment on that? Wouldn't
• 50:39 - 50:44:  you see nascent strand degradation in that case? Thank you.
• 50:45 - 50:51:  Thank you for that excellent question. Indeed, we were a bit confused when we saw these results
• 50:51 - 51:01:  first. So, the CHEK1 inhibition associated with DNA damage is not something that we're
• 51:01 - 51:07:  contesting because there is a lot of replication stress going on with the CHEK1 inhibition due to
• 51:08 - 51:14:  massive origin firing and nucleotide depletion as well. So, you do have a lot of replication
• 51:14 - 51:20:  stress, which can be very stressful for cells because of this replication stress, which can
• 51:20 - 51:26:  turn into DNA damage. However, at this particular setup where we're looking at just the stability
• 51:26 - 51:32:  of these forks, CHEK1 does not seem to really matter much in terms of that. So, it is very
• 51:32 - 51:42:  curious and I think the answer has to do with the massive activation of the CDKs because CHEK1
• 51:43 - 51:51:  also functions through this Treslin to exhibit its functions in the replication to basically
• 51:51 - 51:56:  fire origins. And it could be that this is more dependent or this is a pathway that cells are
• 51:56 - 52:03:  more dependent on, just use more of, and that WE1 is really the one hardcore controlling the
• 52:03 - 52:11:  CDK2 activity in the replication. Okay, we're just going to go for a question in the chat
• 52:11 - 52:19:  from Marcel van Voogt. Marcel, if you're there, unmute yourself and you can ask the question.
• 52:20 - 52:28:  There we go. Yes. Hi. Hi, Camilla. Great talk. I was wondering, CDK2 hyperactivation
• 52:29 - 52:37:  does many things. Have you checked whether inactivation of DNA2 makes cells resistant
• 52:37 - 52:42:  to WE1 inhibitor to see if it actually contributes to the sensitivity that we observe?
• 52:45 - 52:53:  So, the viability, you mean? Yeah. So, if you inactivate DNA2, are these cells still sensitive to WE1?
• 52:54 - 53:02:  I believe so. I believe so that there's still the other effects of entering into mitosis
• 53:03 - 53:08:  prematurely and pushing cells quicker through the S phase. I believe those are still
• 53:09 - 53:12:  problems, but it seems that the cells would be better off
• 53:12 - 53:22:  at least stabilizing the forks that are stalled. But I have no solid data on this, on the viability,
• 53:22 - 53:28:  but this would be my hypothesis that at least the cells, when I look at them after depleting DNA2
• 53:28 - 53:35:  and treating with the WE1 inhibitor, they're still, they're not, they're not extremely happy.
• 53:35 - 53:36:  Thanks. Okay, then we have three questions in the chat, and then we're going to move on to
• 53:36 - 53:40:  Okay, then we have three questions in the Q&A. We're going to start with one
• 53:40 - 53:44:  that is an early career researcher, because we're going to prioritize those,
• 53:44 - 53:55:  which is Athanasios Atsalias. There we go. I pronounced that correctly. We're close enough.
• 53:56 - 54:01:  You can ask your question. Yeah. Go ahead.
• 54:05 - 54:09:  I think you're breaking up. We can't hear your question.
• 54:11 - 54:17:  So maybe you can work on your connection and then, and then we'll go first, we'll go to
• 54:18 - 54:24:  Maritz Rurda, and then after that, Luis Toledo. So Maritz.
• 54:25 - 54:27:  Oh yeah, we can hear you now. Okay, go ahead.
• 54:27 - 54:35:  Okay. Thank you. Excellent pronunciation of the name. So this is Athanasios from
• 54:35 - 54:41:  University of Oxford. So I would like to ask if there is any data available on the concept of
• 54:41 - 54:47:  synthetic lethality after WE1 inhibition with the traditional cytotoxic chemotherapy?
• 54:49 - 54:56:  Yes, the short answer is yes, but not from our lab. There are multiple papers out that show
• 54:56 - 55:04:  both preclinical and clinical trials utilizing this adversity in combination with platinum drugs
• 55:04 - 55:15:  such as cisplatin. And it definitely shows that inducing DNA damage and replication stress and
• 55:15 - 55:20:  premature entry into mitosis and checkpoint overrides is definitely very toxic to cells.
• 55:20 - 55:27:  So it's very promising. Yeah. So thank you very much. The reason I'm asking is because
• 55:28 - 55:32:  if you are going to use different checkpoint inhibitors, for example, you must specify the
• 55:32 - 55:40:  cell cycle point at which the WE1 inhibition acts on promoting the DDR response. Because
• 55:40 - 55:47:  otherwise, you're going to have some controversy. Whereas the cytotoxic chemotherapy acts regardless
• 55:47 - 55:52:  of the cell cycle point where the cancer cells are. So this is why the reason I'm asking,
• 55:52 - 55:56:  but thank you very much for your answer. Okay, thank you very much for your question.
• 55:57 - 55:58:  Thank you.
• 55:58 - 56:01:  Then we'll go to Maritz Rurda. If you could.
• 56:01 - 56:02:  Yes, thanks.
• 56:02 - 56:03:  Already there. Yeah, go ahead.
• 56:03 - 56:08:  Thanks for the talk. I'm wondering, do you know of other processes that could
• 56:08 - 56:12:  explain differences in ratios between IDU and CLDU?
• 56:14 - 56:23:  Yeah. So since WE1 has a lot of functions in the replication, mainly on the initiation
• 56:23 - 56:32:  and the origin firing, we try to design this assay so that we only look at the stalled fork.
• 56:32 - 56:36:  So behind the replication fork, basically. So we designed it this way that we see the
• 56:36 - 56:44:  CLDU and the IDU ratio, and then we stall the forks using hydroxyurea. So there shouldn't
• 56:44 - 56:51:  be much more going on at this point. So the effects of the initiation, we've tried to cancel
• 56:51 - 56:56:  out. So this is why we look at it this way. Yeah. Thank you.
• 56:56 - 57:02:  So we'll finish it with a question from Luis. He's already there. So you can ask your question.
• 57:03 - 57:03:  Hi, Camila.
• 57:04 - 57:05:  Hi, Luis.
• 57:05 - 57:10:  Nice data, especially the DNA2 is very impressive. I was just wondering,
• 57:10 - 57:13:  because I'm a bit concerned about the conditions that you use for experiments,
• 57:14 - 57:23:  because during the development of the whole fork degradation field, we've always been looking at
• 57:23 - 57:28:  the time points or something quite relevant here. But I think in your conditions,
• 57:28 - 57:33:  you're actually pushing the limits to the other extreme, because I mean, for our HU plus WE1
• 57:33 - 57:40:  inhibitor, I mean, I can assure you that this is just 100% replication catastrophe in SPI cells.
• 57:40 - 57:45:  So I mean, I think you're just looking at broken forks. So it might be a good model to study,
• 57:46 - 57:51:  maybe the section of broken forks, but this is just broken forks. I mean, all your data points
• 57:51 - 58:07:  in that direction, especially the RPA phosphorylation that you see. So I mean, you can just take in the lab, I can assume that would be the case. So are you doing any controls on that or what are your thoughts?
• 58:07 - 58:22 : We’ve also looked at earlier time points to see if we could see an effect and after two hours of WE1 inhibition we could already start to see an effect of the four protections so I don’t know if you would think that is
• 58:23 – 38:32: it's also very, it was almost 100% replication catastrophe, basically we formed WE1, so
• 58:33 - 58:40: you can of course take this study to Copenhagen one day. I think that would be nice if you have
• 58:40 - 58:46:  some good suggestions. Definitely, that would be nice. We can talk anytime. Yes. Cool. Thanks.
• 58:47 - 58:50:  I think he's inviting himself over for lunch next time. Yes.
• 58:55 - 59:00:  Okay. Thank you very much. Lots of questions. Thank you for the presentation. That was very
• 59:00 - 59:08:  interesting. So we'll move on, of course, to the end of our meeting with our keynote. So it is a
• 59:08 - 59:13:  great pleasure to introduce our keynote today, Sir David Lane, who needs no introduction. So I
• 59:13 - 59:20:  could stop right here, but I won't. I'm looking actually at a list of all his awards and his
• 59:20 - 59:26:  contributions to research in the UK and worldwide. The thing is, if I would list them here, we would
• 59:26 - 59:32:  have no time left for his keynote. So I'm not. So I'm going to just say one thing is that David is
• 59:32 - 59:38:  a real inspiration because very early on in his career, he made a very fundamental biological
• 59:38 - 59:44:  discovery, which of course, is the discovery of p53. And then the rest of his career, he used it in
• 59:45 - 59:54:  developing treatments based on that information and insight, developing clinical relevance of his work
• 59:54 - 59:59:  that will benefit cancer patients and their families, which is, of course, really inspirational because
• 59:59 - 60:04:  most of us do very fundamental work, but you can really see from his career that could lead to great
• 60:04 - 60:08:  benefits in the clinic. So we set up the cell cycle club to provide a platform for early
• 60:08 - 60:13:  career researchers to build a network and to share their unpublished work and get feedback
• 60:13 - 60:18:  from a knowledgeable audience. But in addition to that, we always invite a keynote speaker
• 60:18 - 60:23:  to inspire the early career researchers and maybe the whole cell cycle field as a whole
• 60:24 - 60:29:  in that respect. I think we couldn't have done better with Sir David Lane. So I'm very
• 60:29 - 60:31:  much looking forward to the keynote. So without further ado, David, the5 stage is yours.
• 60:35 - 60:40:  Thank you. Thank you very much indeed, Rob. And it's a very nice introduction.
• 60:40 - 60:45:  Actually, I’m going to talk a little bit about my career and give a little bit of interaction
• 60:45 - 60:50:  to when I was a young, early career researcher. So, yes, it's a great pleasure to be able
• 60:50 - 60:54:  to talk to you. And thank you very much for the kind introduction. As I said, I'm going
• 60:54 - 60:59:  to talk a little bit about the history of P53, partly to give that inspiration because
• 60:59 - 61:05:  your early observations can lead on to many other things. So we celebrated in the
• 61:05 - 61:13:  lab last March, 40 years of P53 research, because it’s 40 years since the first
• 61:13 - 61:19:  paper came out in Nature. And what was happening 40 years ago, we've been told not to be political.
• 61:19 - 61:22:  So I just put this slide up, and then I go on to the next slide. But keeping making things
• 61:23 - 61:29:  great again. So I'm really going to split the talk into sort of two sections, really.
• 61:29 - 61:35:  Phase one and two is sort of how we discovered P53 and the discovery that the mutations
• 61:35 - 61:40:  in the gene encoding this protein are found in nearly every cancer. And then the phases that
• 61:40 - 61:45:  are still very much ongoing, still trying to work out exactly what P53 does to be a
• 61:45 - 61:53:  tumor suppressor and how we can use this to find new treatments for cancer. And it's a big field.
• 61:53 - 61:59:  I think there's about 90,000 papers published on P53. And it’s actually still full of controversy.
• 61:59 - 62:04:  So though there's been a lot of studies published, some basic issues are still
• 62:04 - 62:07:  under intense investigation. And I'll try and highlight those as well.
• 62:09 - 62:13:  So if you want to read a fun account of the early discovery, this is a book written by
• 62:14 - 62:20:  Cathy Weston, and it's really about the labs at ICRF, the old LRI labs, which are one of the
• 62:20 - 62:27:  precursor labs for the Crick Institute. And it gives a kind of personal account of
• 62:27 - 62:33:  the work I was doing there with Lionel Crawford in the late 70s. It's very flattering
• 62:33 - 62:42:  to me. So I like that. But it's just one version of events. So I left my PhD with Av Mitchison
• 62:42 - 62:49:  working on autoimmunity and went to join the ICRF at Lincoln's Inn Fields and started working
• 62:49 - 62:56:  with Lionel Crawford on SV40 T antigen. This is the oncogene of the small DNA virus SV40
• 62:56 - 63:01:  that was widely used as a model for transformation then because the virus could transform
• 63:02 - 63:08:  cells and tissue culture. It could cause tumors in some animal models. And very importantly,
• 63:08 - 63:15:  there was a temperature-sensitive mutant of SV40 that could switch between being active as
• 63:15 - 63:21:  an oncogene or inactive as an oncogene on temperature change. So that allowed you to
• 63:21 - 63:27:  show that not only was T antigen necessary to turn a normal cell into a transformed cell,
• 63:27 - 63:33:  but also that the protein needed to be continually expressed to do that. So you could move the cells
• 63:33 - 63:40:  between 32 degrees and 39 degrees and at 32 degrees they would grow as cancer cells, and at
• 63:40 - 63:48:  39 they would not. So this was a very powerful system. We used an assay, the first paper
• 63:49 - 63:56:  published in 1977, soon after I arrived to kind of characterize the only tool we had at that stage,
• 63:56 - 64:02:  which was the serum from animals that bore SV40 induced tumors. And we were able using that assay
• 64:02 - 64:10:  to identify a few individual animals that had made a stronger immune response. And then using
• 64:10 - 64:17:  protein A beads and heat-killed Staphylococcus aureus, which was a big breakthrough around that
• 64:17 - 64:24:  time, we were able to get very clean immune precipitates of T antigen. And we were able to
• 64:24 - 64:29:  immunize with the purified protein that we had purified essentially by immunoprecipitation
• 64:29 - 64:35:  and cut out a band from the gel. And that led to the production of a very large amount of a very
• 64:35 - 64:40:  strong specific antibody to T antigen. And it was the finding that that antibody
• 64:40 - 64:46:  co-immunoprecipitated a host protein that was really the basis of the discovery of P53 that
• 64:46 - 64:53:  we reported in Nature. This is the paper. And as you can see, it came out in March, 1979. In those
• 64:53 - 64:59:  days, things were very different. I mean, we sent in the paper, I think towards the end of 1978,
• 64:59 - 65:04:  they asked for a couple more experiments. I think it had a total of four figures in it. And,
• 65:04 - 65:08:  you know, it was published pretty easily. So life was very different to what it is now.
• 65:09 - 65:16:  And the key gel really is this one shown here. So this is an immunoprecipitation of a SV40
• 65:16 - 65:22:  transformed cell line with this new antibody that I've made that was very specific to T antigen.
• 65:22 - 65:26:  And I've just shown exactly a diagram of what we did. So this is a classic
• 65:27 - 65:34:  co-immunoprecipitation. And you can see these two bands, CoIP, T and 53K, as we called it then now,
• 65:34 - 65:40:  P53. I mean, this is now such a routine kind of method that nobody thinks about it. But at the
• 65:40 - 65:46:  time, it was very much criticized. And people said, oh, this is a background band, or it's a
• 65:47 - 65:51:  fragment of T antigen. And we had to do quite a lot of work to convince people that actually,
• 65:51 - 65:58:  this was a host protein. And technically, you know, it's always these small innovations that
• 65:58 - 66:03:  make a difference. So the real difference for us was having this much better polyclonal antibody
• 66:03 - 66:09:  at the time, and using these heat-killed bacteria, and then being very, very careful about how we
• 66:09 - 66:15:  washed and processed the sample that gave us this very clear data, which made it very, very easy to
• 66:15 - 66:24:  interpret that this was a co-immunoprecipitation. I had an interesting time. I was working in
• 66:24 - 66:28:  London with Lyle, and then I got an offer to go and be a postdoc in America. So actually,
• 66:28 - 66:34:  the paper was written when I was in America. And this is me and Keith Willison in Cold Spring
• 66:34 - 66:41:  Harbor in 1979, when the paper was written. I'm working on a very tiny bench, because in Cold
• 66:41 - 66:44:  Spring Harbor, the benches were shared by everybody. It was a very crowded environment.
• 66:45 - 66:51:  And I came back from Cold Spring Harbor and started to work at Imperial College. And I was,
• 66:52 - 66:59:  as a new, young researcher, I think your colleagues are terribly important. And I was extremely lucky
• 66:59 - 67:05:  to work with a group of four of us, Gene Beggs, Peter Rigby, and David Glover, who were all of us
• 67:05 - 67:10:  still extremely good friends. And actually, eventually, all of us got elected to the Royal
• 67:10 - 67:16:  Society. So we've all done well. And we really worked together. So I think the message for me was
• 67:16 - 67:21:  to have close colleagues who can support each other. We were all very much in the same phase
• 67:21 - 67:26:  of trying to get our labs going and having our first PhD students, et cetera. So that kind of
• 67:27 - 67:32:  environment of a group of us was very good. And we also started a kind of a bit like the
• 67:32 - 67:37:  Cell Cycle Club. We had a molecular biology club in London that was a very good place for
• 67:37 - 67:43:  people to meet each other. And what emerged from all of that is still something very striking.
• 67:43 - 67:51:  And that is this finding that all of the DNA tumor viruses, notably SV40 and the Polyoma viruses,
• 67:51 - 67:57:  but also adenoviruses, and most importantly, the human papillomaviruses, all very specifically
• 67:57 - 68:02:  target these two proteins, which are at the heart of the regulation of the cell cycle and
• 68:02 - 68:06:  the regulation of the stress response. And I still find it quite an extraordinary piece of
• 68:06 - 68:13:  biology that this amazingly specific interaction across huge evolutionary distance picks out these
• 68:13 - 68:22:  two host proteins so precisely. P53 T antigen was discovered in 1979. And then the RB interaction
• 68:22 - 68:28:  came some years later in 1988, actually from a colleague of mine who'd been working also with
• 68:28 - 68:36:  Lionel, Ed Harlow. And the same kinds of approaches were used looking for co-immunoprecipitation of
• 68:36 - 68:42:  host proteins with viral oncogenes. It's always nice in the end, because you always have doubts
• 68:42 - 68:47:  as a scientist, you know, to see finally the structure. And this is a very nice structure of
• 68:47 - 68:54:  P53, which is here in the sort of purplish color surrounding the hexamer of T antigen. T antigen,
• 68:54 - 68:59:  of course, is the protein that not only is a transforming oncogene, but it also is a helicase.
• 68:59 - 69:06:  that drives the replication of the SV40 virus. And this interaction is a very potent one.
• 69:06 - 69:12:  And it works by blocking P53's function as a transcription factor because the T antigen
• 69:12 - 69:20:  binds on P53 in exactly where P53 normally interacts with DNA. So it blocks P53 from
• 69:20 - 69:26:  binding to DNA. And some of the key residues here are, as we'll see later, hot spots for
• 69:26 - 69:33:  mutations that occur in P53 in human cancer. The next sort of phase really was just as I
• 69:33 - 69:40:  was transitioning again. So this was at the, after my time at Imperial College, I left Imperial
• 69:40 - 69:49:  College in 1985 and went to Clare Hall and worked there for the next five years. And this was a
• 69:49 - 69:54:  branch of the Imperial Cancer Research Fund that was near Putter's Bar, became very important for
• 69:54 - 69:59:  Thomas Lindahl and others' work in DNA repair. But I was one of the first recruits there.
• 69:59 - 70:05:  And it was while I was there that the sort of explosion in P53 really happened in the sort of
• 70:05 - 70:12:  very sort of '89, '90. The realization that it was a tumor suppressor, not as some people thought
• 70:12 - 70:19:  of an oncogene, that point mutations occurred in P53 in nearly every type of human cancer as
• 70:19 - 70:25:  studied and were associated with loss of heterozygosity. And also that the mutant protein
• 70:25 - 70:32:  was often overexpressed in cancer cells. And this staining just shows it classic immunohistochemistry
• 70:32 - 70:39:  of a formalin-fixed paraffin-embedded section of a human skin cancer stained with a monoclonal
• 70:39 - 70:44:  antibody that we make called DO1 that people still use a lot that recognized the accumulation
• 70:44 - 70:51:  of nuclear P53 in the cancer cells. And I think the, this ability to see accumulation of protein
• 70:51 - 70:56:  really spurred the field a lot because we were able to examine huge numbers of cancer cells
• 70:56 - 71:02:  very quickly and many, many archival examples. So we even had materials that had been
• 71:02 - 71:08:  fixed in Victorian times and they still, they still worked. And we had a very, very large
• 71:08 - 71:14:  survey. So we tested more than 200 different types of cancer looking with these reagents
• 71:14 - 71:20:  for the accumulation of P53. And so we're now in the genomic age. And of course there's been
• 71:20 - 71:27:  an enormous amount of work done on cancer genome sequencing. And there's a fantastic P53 mutant
• 71:27 - 71:32:  database held at IARC and there are also cosmic databases. And essentially what's emerged from
• 71:32 - 71:39:  that is that there are, you know, a number of cancer genes that are frequently mutated and
• 71:39 - 71:46:  obvious examples are the RAS oncogene, but P53 is the most frequently mutated and in the most
• 71:46 - 71:51:  different types of cancer. So it's a very striking observation that I think keeps everybody in the
• 71:51 - 71:58:  field very focused on trying to work out exactly what P53 does. If you look at those patterns of
• 71:58 - 72:05:  mutation, they are quite distinct and unusual for a tumor suppressor because typically when you look
• 72:05 - 72:10:  at mutations in tumor suppressor genes, there are lots of functional mutations and you tend to see
• 72:10 - 72:16:  a lot of nonsense and frameshift mutations, but strikingly in P53, you see a lot of missense
• 72:16 - 72:23:  mutations, about 74% of missense, and they do occur repeatedly at certain codons. So there are
• 72:23 - 72:30:  about 10 particular codons, which account for about 50% of all the mutations that are seen.
• 72:30 - 72:35:  So there's some selection here. I'll come back to later to how one interprets that, but it is
• 72:35 - 72:41:  this observation that's behind the strong immunohistochemical staining for P53 that
• 72:41 - 72:45:  you see in cancer because the cancer cells are continuing to make the protein and in fact make
• 72:45 - 72:53:  it in much larger amounts than normal cells. So the other real breakthrough that happened was
• 72:53 - 73:00:  again in 1990, so all of these things happened very quickly, was the discovery of the basis of
• 73:00 - 73:07:  Lee-Fraumeni syndrome. So this is a slide of many of us meeting at the Lee-Fraumeni Symposium a few
• 73:07 - 73:14:  years ago, and in the front row is Dr. Fraumeni and David Malkin and Arnie Levine and myself,
• 73:14 - 73:19:  and behind us are, you know, teenagers, all of whom are affected by Lee-Fraumeni. It was a very
• 73:20 - 73:26:  humbling meeting to attend actually because these teenagers often have sarcomas and so many of them
• 73:26 - 73:33:  have had to have amputations, for example. So it's a very devastating syndrome. It tells us,
• 73:33 - 73:39:  you know, how important I guess P53 is because these kids have just a single base change out of
• 73:40 - 73:48:  3,000 million bases in DNA, and it's still a big problem. I mean, how to tackle this kind of
• 73:48 - 73:55:  genetic predisposition to cancer is really difficult, and at the moment David Malkin has
• 73:55 - 74:03:  pioneered a screening program so that the kids get MRIs, and if they see lesions arising,
• 74:03 - 74:07:  they get very early surgery, and that has been extremely successful in maintaining
• 74:09 - 74:15:  the health of these children and young people. But it's very complicated. Not every mutation
• 74:15 - 74:24:  is the same, and we see a lot of challenges now in terms of people being sequenced, you know,
• 74:24 - 74:31:  for their genomes for interest and then appearing to have a P53 mutation, and is this mutation
• 74:31 - 74:35:  important or not? So we have this whole issue, which is true in many areas of genetic counseling
• 74:35 - 74:42:  for cancer of the issue of variance of unknown significance. And then in the late '80s,
• 74:42 - 74:48:  a really extraordinary body of data accumulated about a cohort of people in Brazil. There's now
• 74:48 - 74:54:  about 90,000 people reported who are all descendants of a single founder mutation,
• 74:55 - 75:00:  and they have a very curious mutation in P53. This is not in the DNA binding domain, but it's in the
• 75:00 - 75:06:  oligomerization domain. It affects a single hydrogen bond, and they have a kind of very
• 75:06 - 75:15:  weak phenotype. The kids tend to get particular types of tumors, adrenal cancers,
• 75:16 - 75:21:  but there's a lot of evidence for background genes in this environment. So some families are much
• 75:22 - 75:27:  more badly affected than others, and so in some families it looks like a quite penetrant gene.
• 75:27 - 75:34:  In others it looks very weak, and people are just beginning to pick apart the kind of modifier
• 75:34 - 75:42:  genes that affect P53's function as a tumor suppressor in this human context. So trying
• 75:42 - 75:47:  to work out what P53 does, I mean, there are a number of breakthroughs. I mean, clearly the
• 75:47 - 75:52:  protein that was discovered to be a sequence-specific DNA binding protein and also to be
• 75:52 - 75:57:  active as a transcription factor, it was discovered to be extremely highly regulated by
• 75:57 - 76:05:  the ubiquitin E3 ligase MDM2, which binds it and acts as an E3 to attract E2 and transfer ubiquitin
• 76:05 - 76:18:  to P53, resulting in P53's degradation, and it's a feedback loop because P53 activates the
• 76:18 - 76:24:  transcription of the MDM2 gene itself. And then downstream of that active P53
• 76:25 - 76:28:  are a lot of different genes involved in many different processes. So
• 76:29 - 76:34:  at least 100 genes are direct targets of P53 as a transcription factor.
• 76:38 - 76:46:  And the other main feature of this system, really, is the number of different stress
• 76:46 - 76:53:  signals that can trigger P53. And these include very key components of the cell cycle, such as
• 76:53 - 77:01:  effects on the centriole and mitosis. So PLK4 inhibitors are very, very good stimulators of
• 77:01 - 77:07:  the P53 response, DNA damage itself, but also other signals such as nucleotide depletion,
• 77:07 - 77:11:  ribosome disruption, the action of oncogenes, particularly the work of
• 77:12 - 77:18:  Gerard Evan and others on MYC oncogene and RAS oncogene expression, all triggering the P53
• 77:18 - 77:25:  response. And in addition to that, there's a fairly large literature claiming that P53 does
• 77:25 - 77:31:  other things besides being a transcription factor, that there are transcription activation independent
• 77:31 - 77:40:  effects, so effects, direct effects on DNA repair and DNA replication. And this field is very
• 77:40 - 77:50:  controversial. I think a lot of people are quite biased against these other activities of P53,
• 77:50 - 77:57:  but I must admit, myself, I'm beginning to be a bit more open-minded, partly because there are a
• 77:57 - 78:04:  number of mutants in P53 that have been made in mouse models that have lost a lot of transcription
• 78:04 - 78:09:  activation function. They have some residual activity, but it's very weak, and yet they're
• 78:09 - 78:15:  still extremely good tumor suppressors. So, you know, one argument is, well, those genes that
• 78:15 - 78:20:  they still activate are the important ones, but the other argument maybe is that something else
• 78:20 - 78:28:  is happening. And one of the challenges here is getting mutants that can distinguish these
• 78:28 - 78:33:  different properties because the middle of P53, the DNA binding domain, is the only bit that's
• 78:33 - 78:39:  got any kind of structure, but that's a very, it's not a very tightly folded protein, so its
• 78:39 - 78:43:  whole structure is quite easily disrupted by mutation. So, it's hard to get a mutation that
• 78:43 - 78:51:  would distinguish, for example, one particular function of P53 from another. There's been an
• 78:51 - 78:57:  enormous effort to understand the significance of P53 status for cancer patients, and in general,
• 78:57 - 79:03:  it's clear that having a P53 mutation is a bad thing, and you can see this both in progression
• 79:04 - 79:09:  survival and overall survival. Of course, you're looking here at patients that have received
• 79:09 - 79:14:  treatment as well, so it's a combination of the effect of the mutation, and also the impact of
• 79:14 - 79:21:  the mutation on how the patients are treated that is contributing to these differences in survival.
• 79:22 - 79:29:  And you even can see differences about the type of mutation. So, some mutations in the DNA binding
• 79:29 - 79:34:  domain, for example, seem to be more adverse than mutations in other parts of the molecule.
• 79:34 - 79:40:  This is also a controversial area, and I'll come back to it as I talk a bit more about these issues.
• 79:42 - 79:48:  One thing that has emerged very clearly, though, is that P53 is extremely important for the acute
• 79:49 - 79:54:  damage response, and if you treat mice with a single dose of ionizing radiation,
• 79:54 - 79:59:  you see very, very clear induction of apoptosis in areas...
• 80:00 - 80:04:  of tissues that are highly proliferative. For example, the crypt cells in the small intestine,
• 80:05 - 80:10:  cells in the thymus and in the spleen, in the bone marrow. And this is a P53-dependent effect,
• 80:10 - 80:16:  because when the P53 knockout mice were developed, they were found to be extraordinarily resistant
• 80:16 - 80:22:  to this kind of radiation damage. And this is very important and interesting. It's a big question as
• 80:22 - 80:28:  to whether this acute radiation DNA damage response is actually what's required for
• 80:28 - 80:35:  tumor suppression, because work by a number of groups, and particularly Gerard Evan and others,
• 80:35 - 80:41:  have shown that there is a different sort of P53 surveillance that can be, to some extent,
• 80:41 - 80:46:  uncoupled from this acute response that's also important for tumor suppression. So, it's still
• 80:48 - 80:52:  not as clear as one would like it to be. It's a very attractive model to imagine,
• 80:52 - 80:56:  but it's this acute response that's necessary to eliminate cancer cells. But
• 80:57 - 81:04:  it's clear from very nice experiments using inducible P53 that you can induce P53 a long
• 81:04 - 81:09:  time after the damage and still see protection, suggesting that it's more a late effect of
• 81:09 - 81:15:  radiation that can be detected and eliminated by P53, or the cells containing those lesions,
• 81:15 - 81:24:  eliminated by P53 to reduce cancer development. The key kind of regulation of P53 at the
• 81:24 - 81:30:  biochemical level has been well worked out, though there's still, again, a lot missing from
• 81:30 - 81:35:  the equation. But essentially what's been discovered is that P53 is being continually degraded
• 81:36 - 81:44:  through the combination of these ubiquitin E3 ligase MDM2 and a highly related protein MDM4.
• 81:44 - 81:50:  And both of these proteins form complexes with P53 and complexes with each other.
• 81:50 - 81:58:  And a very striking experiment is that mice that lack MDM2 are embryonically lethal,
• 81:58 - 82:04:  as are mice that lack MDM4. So, they're both independently embryonic lethal, but both of those
• 82:04 - 82:10:  embryonic lethalities can be rescued by deletion of P53. So, you can actually make mice that lack
• 82:10 - 82:16:  P53, MDM2, and MDM4. They're highly tumor prone, but they develop and they're fertile. So, this is
• 82:17 - 82:23:  really very strongly indicating that this group of proteins are very, very highly associated and
• 82:23 - 82:31:  regulate each other. So, when a signal comes in, whether it's oncogene activation or DNA damage,
• 82:32 - 82:39:  you start to see the inhibition of this breakdown of P53 and the accumulation of the protein. And the
• 82:39 - 82:44:  accumulated protein becomes modified by phosphorylation and many, many other
• 82:44 - 82:48:  post-translational modifications. The post-translational modification of P53 could
• 82:48 - 82:53:  be a five-hour lecture itself. It's basically everything that can happen to a protein happens
• 82:53 - 83:01:  to it. So, it's ubiquitinated, methylated, phosphorylated, etc. So, there's a
• 83:01 - 83:07:  tremendous degree of post-translational modifications that presumably are involved
• 83:07 - 83:14:  in controlling exactly which genes it activates and when and how. This diagram sort of talks
• 83:14 - 83:21:  about the modification of P53, but the other really important thing is modification of MDM2 as well.
• 83:21 - 83:27:  So, there are mice with point mutations in MDM2 that have highly defective radiation responses
• 83:27 - 83:36:  because the mutation has prevented the inhibition of MDM2 by phosphorylation. So, again, I think
• 83:37 - 83:44:  still a very, very important area for further study is exactly how these pathways interact
• 83:44 - 83:48:  and exactly what post-translational modifications regulate this degradation pathway.
• 83:51 - 83:56:  Of course, as I indicated, one of the key features of the system is that in human cancers,
• 83:56 - 84:05:  you often still express the mutant protein and often at very high levels. And you always see
• 84:05 - 84:11:  loss of heterozygosity as well. So, that's a very frequent feature of cancer.
• 84:12 - 84:17:  So, the question is, you know, what's the effect of these mutations? So, obviously,
• 84:17 - 84:26:  loss of wild-type activity is very important. We know that P53 is cell-autonomous haploinsufficient,
• 84:26 - 84:32:  which means that the cell kind of knows whether it's got one copy of P53 or two in a quantitative
• 84:32 - 84:39:  level. So, the levels of P53 are very important. So, the other things that you can see with mutant
• 84:39 - 84:45:  P53 are a dominant negative inhibition of wild-type function of P53. P53 functions as a
• 84:45 - 84:51:  tetramer. So, this is a way to inhibit P53 activity. It's talked about a lot. It can be
• 84:51 - 84:58:  shown very nicely in tissue culture models. But actually, how important it is, I think,
• 84:58 - 85:05:  is open to some question. In genetic models, unless you have a strong stress signal, you see that the
• 85:05 - 85:10:  wild-type is still active in the presence of the mutant. And if you look in the Li-Fraumeni families,
• 85:10 - 85:17:  there isn't really a very clear adverse effect of having a mutant allele versus a null allele,
• 85:18 - 85:21:  which you would anticipate if dominant negative was very important.
• 85:22 - 85:27:  And then the other area that's been discussed a great deal is whether this mutant protein
• 85:27 - 85:34:  that you're expressing in the tumor cell at high levels has a direct gain of function. That is,
• 85:34 - 85:40:  is the mutant protein acting as an oncogene and in a way beyond its dominant negative effect,
• 85:40 - 85:46:  because we typically see in adult cancers the loss of the wild-type allele. So, the tumor cells are
• 85:46 - 85:53:  just expressing high levels only of the mutant protein. So, the kind of experiments that led
• 85:53 - 86:02:  people to think strongly that there was a gain of function were published in the mid-2000s, 2004,
• 86:03 - 86:09:  from a number of groups. There are two groups here, where they compared the phenotype of mice that had
• 86:09 - 86:18:  P53 null, P53 deleted, or P53 that expressed a mutant allele. So, a mutant allele which would
• 86:18 - 86:24:  be expressed as a protein, a point missense mutation typical of those found in human cancers,
• 86:24 - 86:30:  but in the mouse model. And what they saw, of course, was the tumor frequency was much higher
• 86:30 - 86:37:  in both of the models in which P53 function was lost, that is the mutant over minus and the minus
• 86:37 - 86:43:  minus. But they also saw in the animals that were expressing the mutant protein in their tumors,
• 86:43 - 86:49:  and a different type of a larger spectrum of different cancers, in fact, carcinomas that
• 86:49 - 86:55:  were invasive and metastatic. And so, that then led to a huge number of studies using
• 86:55 - 87:02:  tissue culture models showing these kinds of gain of function of mutant P53. And I would say, you
• 87:02 - 87:08:  know, it's a very, very strong field in terms of the number of papers and the enthusiasm for this
• 87:08 - 87:15:  idea that people have. But as more recently, it's become a little bit more controversial. First of
• 87:15 - 87:22:  all, a number of studies in which mutant P53 has been knocked out by CRISPR-Cas9 from human
• 87:22 - 87:28:  cancer cell lines show no effect of the deletion, even though these very same cell lines had shown
• 87:28 - 87:36:  apparently very strong effects when P53 expression was suppressed using siRNA. And then a more recent
• 87:36 - 87:42:  study published in a less fancy journal, Lab Investigation, has even gone back and questioned
• 87:42 - 87:48:  very much those early studies. They don't really see in their model the same increase in different
• 87:48 - 87:54:  types of cancer and evidence for invasion, as had been reported earlier. And then there's been some
• 87:54 - 88:00:  very, very large studies, which I summarized in this short article in Science, where essentially
• 88:00 - 88:08:  every possible mutant has been tried in tissue culture models to ask, can you see a clear gain
• 88:08 - 88:14:  of function versus dominant negative versus loss of function? And those studies, very clearly,
• 88:14 - 88:20:  you could show dominant negative effects and you could clearly show loss of function, but the
• 88:20 - 88:25:  independent gain of function has not really been clearly demonstrated. So it's, I think, still
• 88:25 - 88:31:  controversial and I'm probably offending about half the P53 world by causing any question about
• 88:31 - 88:38:  this issue. But the other question you could ask, of course, is why is P53 expressed at high levels
• 88:38 - 88:44:  in cancer cells? And this is not as simple as we thought. Initially, we thought, well, it's a
• 88:44 - 88:50:  mutant protein, so it can't turn on the MDM2 gene, so it can't be degraded because there's no E3
• 88:50 - 88:57:  ligase for it. But it turns out that actually the MDM2 gene has a P53 independent promoter as well
• 88:57 - 89:07:  as a P53 dependent promoter. And in fact, the mutant proteins are substrates for MDM2. So it's
• 89:07 - 89:13:  not quite as simple as we thought. And in zebrafish models, for example, you see that the mutant
• 89:13 - 89:18:  protein does not accumulate until you treat cells, treat the animals with a damaged signal. And the
• 89:18 - 89:24:  same thing is broadly true in mice. We recently looked at this in more detail using the
• 89:25 - 89:33:  mouse model where they carry two alleles of the R172H mutant, which is equivalent to the R175 in
• 89:33 - 89:40:  human. And what we saw in this gut role was the very clear segregated expression of P53 only in
• 89:41 - 89:50:  the bottom of the crypt cells in the intestine. So that led us to ask, why do we only see
• 89:50 - 89:56:  P53 in such a limited distribution? Because if you look in tissue culture cells, every cell is
• 89:56 - 90:02:  happily expressing P53. But we started to wonder what's actually going on in animal models. And
• 90:03 - 90:09:  the question is, why don't we see P53 in the other cells in these tissues? Is it that there's
• 90:09 - 90:18:  another E3 ligase that's contributing to P53 regulation that is not acting, that's acting
• 90:18 - 90:23:  still on the mutant protein? Is it control of translation? Is it the control of the promoter
• 90:24 - 90:32:  for the P53 gene? And so what we started to do was to use the drug bortezomib or Velcade,
• 90:32 - 90:38:  which is a very potent proteasome inhibitor, to start to see if we could begin to understand
• 90:39 - 90:45:  these issues. And if you look in the left-hand panel, you can see here the
• 90:46 - 90:51:  lack of P53 staining completely in these hair follicles. They're highly proliferative because
• 90:51 - 90:58:  they stain with this marker Ki67. But if we treat the mouse with bortezomib, now we can see
• 90:58 - 91:06:  the P53 accumulating very nicely in these cells here. So that tells us, yes, P53 is being made,
• 91:06 - 91:13:  it's being degraded by the proteasome pathway, and that's, you know, that's why we can see it here.
• 91:14 - 91:18:  But the question is, why aren't we seeing it elsewhere in all these other cells in the hair
• 91:18 - 91:23:  follicle? Why are we only seeing it in this small proliferative zone? And to answer that question,
• 91:23 - 91:33:  we went to RNA hybridization and found, using RNA techniques, that we could see very clearly
• 91:33 - 91:40:  that there was extremely tight restriction of P53 mRNA to just the proliferating zone.
• 91:41 - 91:48:  So it turns out that P53 is very highly regulated in organisms, in people, in mice, by
• 91:48 - 91:53:  its transcription. Indeed, its transcriptional inhibition in differentiated cells.
• 91:54 - 92:01:  And this actually turns out to be another major source of heterogeneity of P53 staining,
• 92:01 - 92:08:  because if you look in tumors, you often see P53 staining is quite heterogeneous. And you can show
• 92:08 - 92:15:  by doing this RNA in situ technique alongside the antibody staining that essentially there's
• 92:15 - 92:23:  a very tight correlation. So it does really focus, again, one's attention on P53's role in
• 92:23 - 92:28:  proliferation and DNA repair. It's definitely being made in that fraction of cells that are
• 92:28 - 92:34:  highly proliferative. And there's clearly transcriptional regulation of P53 has, I think,
• 92:34 - 92:40:  been under-investigated. Very early on, people made northern blots of all the tissues of mice
• 92:40 - 92:45:  and found P53 mRNA in sort of every tissue. And so people haven't really worried. And then since
• 92:45 - 92:49:  it's expressed in every cell in culture, nobody's sort of worried about what's really
• 92:49 - 92:53:  happening in those tissues. But it's very clear, in fact, that it's very highly restricted.
• 92:55 - 93:01:  So I'm going to finish now by talking about the progress that's been made in treating
• 93:03 - 93:09:  cancer using knowledge of P53. And this has become a very exciting field, really,
• 93:09 - 93:16:  even in the last few months because there's been a few exciting breakthroughs. So there's a variety
• 93:16 - 93:21:  of different approaches that people have taken. So really split in two ways to take those tumors
• 93:21 - 93:28:  where P53 is still wild type, and that's about 40 percent of human cancers, and to try and activate
• 93:28 - 93:34:  the P53 as a transcription factor to try and see if one can bypass whatever mechanism is allowing
• 93:34 - 93:39:  it not to function normally. And a particular class of those are the MDM2 interaction inhibitors.
• 93:40 - 93:45:  People have also tried, of course, P53 gene therapy, and that's been approved in China,
• 93:45 - 93:49:  but not in other countries yet. And then to look at the mutant protein, there's two approaches,
• 93:49 - 93:54:  one which is to try and get the mutant protein to refold correctly, that is to
• 93:56 - 94:01:  essentially use a chemical to achieve second-site reversion of the mutation, or to degrade the
• 94:01 - 94:07:  mutant P53. I won't talk about the mutant P53 degraders, but I'll briefly mention the others.
• 94:08 - 94:15:  So the refolders have had a lot of attention, and a lot of work came originally from Alan
• 94:15 - 94:20:  Fersht's group in Cambridge, and he focused on a particular mutation, which is the mutation from
• 94:20 - 94:28:  tyrosine at 220 to cysteine, so Y220C, and he saw in the crystal structure of the DNA binding
• 94:28 - 94:34:  domain that this created a suitable drug binding pocket, and indeed tried to develop drugs against
• 94:34 - 94:41:  that pocket. And more recently, Arnie Levine and his colleagues have founded a biotech company that's
• 94:42 - 94:48:  attracted very large funding recently, and they have been able to come up with a compound,
• 94:48 - 94:59:  PC14586, that looks very promising. So it's inducing P21 and MDM2 only in the cell line that
• 94:59 - 95:04:  contains this particular mutation, so this is a recovery of P53 transcriptional activity in response
• 95:04 - 95:10:  to this drug, and it's not doing anything in all of these other cell lines with different P53 mutations
• 95:10 - 95:16:  or null. So this is very exciting. I actually found this data not in a paper and not in a patent,
• 95:16 - 95:21:  but in the S1 filing for the initial public offering of the company, and they raised an
• 95:21 - 95:30:  astonishing $200 million for this work. So I think we can watch this space with enthusiasm.
• 95:31 - 95:38:  The question of, you know, intervention and why and how can we intervene also has raised really
• 95:38 - 95:44:  this whole issue of the ARF protein. So the ARF protein is a small protein induced by oncogenes
• 95:44 - 95:52:  like MYC and RAS that acts as a very potent inhibitor of MDM2 and activates the P53 response.
• 95:52 - 95:58:  So this is quite a well understood pathway, and a lot of tumors, particularly notably melanomas,
• 95:58 - 96:04:  deplete ARF, so they've lost their link between sort of the activated oncogene, the BRAF oncogene,
• 96:04 - 96:12:  and P53 activation because they've lost this intermediate. So they are ideal candidates,
• 96:12 - 96:19:  really, for the use of MDM2 inhibitors. So this has been an area of work for myself and my team
• 96:19 - 96:25:  for many, many years, trying to develop molecules that block the interaction between MDM2 and P53.
• 96:26 - 96:32:  So this is, of course, now become a very hot topic in general. That is that E3 ligases,
• 96:32 - 96:38:  ubiquitin E3 ligases, are potentially very good targets for drug development because they could
• 96:38 - 96:43:  be used to regulate the expression level of their downstream targets. And in another approach,
• 96:43 - 96:49:  they can even be recruited to degrade novel proteins through these PROTAC or GLUE-like molecules.
• 96:49 - 96:55:  And one of the key features of them is a very exquisite level of protein-protein interaction
• 96:55 - 97:03:  that mediates the binding of the E3 ligase to its target protein. So in the case of MDM2 and P53,
• 97:03 - 97:09:  we've been able to study this for a long time due to studies using peptide libraries and phage
• 97:09 - 97:15:  display and the crystallographic strategies of Pavletich and his colleagues. And you can see
• 97:15 - 97:22:  in yellow is P53 in these two views, and it has these three amino acid side chains that are
• 97:22 - 97:28:  absolutely critical, and they fit into this pocket on MDM2. And this is a critical recognition event
• 97:28 - 97:35:  that allows MDM2 to degrade P53. And we were able, many years ago, to improve the binding
• 97:35 - 97:43:  constant dramatically by performing peptide chemistry on this small peptide that
• 97:43 - 97:49:  we identified from P53 that could bind to MDM2. So nature's affinity constant is quite low,
• 97:49 - 97:55:  but we could get a very much tighter binding molecule. And when we made that discovery,
• 97:55 - 97:59:  there was a lot of excitement about it because people could see it was kind of druggable, but
• 98:00 - 98:05:  actually it took a long time before anybody could get a small molecule. And the first group to do
• 98:05 - 98:12:  so was Lubov Vasilev and his colleagues at Roche, and they were able to develop the nutlin molecules.
• 98:12 - 98:17:  And once those molecules have been developed, basically every pharmaceutical company has
• 98:17 - 98:24:  been able to make an inhibitor of the P53-MDM2 interaction. But surprisingly,
• 98:24 - 98:29:  these are all now in the clinic, and it's been a bit disappointing. Many of us at the time thought
• 98:29 - 98:35:  this was going to be a tremendously good medicine, and the problem has been relatively weak
• 98:35 - 98:40:  anticancer activity and some side effects. That is, thrombocytopenia has been reported.
• 98:42 - 98:48:  Molecules have relatively poor bioavailability, and they only act on MDM2, but not MDM4.
• 98:49 - 98:54:  And one of the problems, of course, implicit in the P53 system is that these molecules induce
• 98:54 - 99:01:  and stabilize their own target so that when you activate P53, P53 turns on the expression of MDM2,
• 99:01 - 99:09:  so you have more MDM2 that you have to inhibit. But an alternative approach, which we developed
• 99:09 - 99:16:  and with others, is to use the peptide itself as the drug and to use a cross-linking chemical
• 99:16 - 99:20:  within the chemical synthesis of the peptide, a so-called staple, hydrocarbon staple,
• 99:20 - 99:27:  that makes the peptide much more drug-like. It gives it great stability. It makes it possible
• 99:27 - 99:32:  for it to enter cells efficiently. And these molecules are bi-specific. They will inhibit
• 99:32 - 99:40:  both MDM2 and MDM4. And we assembled a team of chemists and biologists and molecular modelers
• 99:40 - 99:46:  to study these molecules. I'm aware I'm probably getting into the danger zone time-wise, so I'll
• 99:46 - 99:55:  go quickly. Essentially, we were able to get staple peptides that could increase P53 levels
• 99:55 - 99:59:  and activate downstream expression here of MDM2 and P21.
• 100:00 - 100:06:  And we started to examine how to improve these molecules, and about over 1,000 variants have
• 100:06 - 100:14:  now been tested of these stapled peptides. And they are able to activate P53-dependent
• 100:14 - 100:21:  transcription. We widely used a reporter assay using a P53 promoter and lacZ beta-galactosidase
• 100:21 - 100:26:  as a reporter. And you can see the staples activate the transcription response. Nutlin
• 100:26 - 100:32:  does so as well, though at high doses, Nutlin can be toxic, and we think this might be nonspecific.
• 100:32 - 100:36:  And the early molecules were very, indeed, very, very weakly active. So it took us some
• 100:36 - 100:41:  time to really establish this approach. But in the end, we were able to get molecules
• 100:41 - 100:50:  that could work in xenograft models. And there's really been a tremendous effort now to develop
• 100:50 - 100:56:  these molecules. Clinically, it's been led by a company called Aileron Therapeutics,
• 100:56 - 101:02:  and they have had some very interesting initial clinical results, which is that they didn't see
• 101:02 - 101:07:  any neutropenia, suggesting that the neutropenia that had been seen with the small molecules like
• 101:07 - 101:14:  Nutlin may not be on target, or it may be a feature of exactly how the peptide is delivered
• 101:14 - 101:19:  that's avoiding that particular side effect. But that observation has really led to a very
• 101:19 - 101:26:  exciting area of a new approach to think about how to use these molecules in the treatment of
• 101:26 - 101:32:  cancer patients. And this is something that we've worked on for a number of years, and it's really
• 101:32 - 101:38:  The concept of chemoprotection. So this is very different. This is treating the normal cell,
• 101:38 - 101:44:  if you like, to try and protect it against the effects of conventional cytotoxic chemotherapy.
• 101:44 - 101:52:  So the concept is that if you have a tumor which is mutant for p53, then a person receiving the
• 101:52 - 101:58:  stable peptide or the small molecule drug, if its toxicity could be reduced, and their normal
• 101:58 - 102:05:  cells would go into a cell cycle arrest. And this would make them resistant to the cytotoxic
• 102:05 - 102:10:  chemotherapy or radiation therapy that the patient would receive. So the idea is that this will act
• 102:11 - 102:17:  as a chemoprotectant and protect healthy cells, but not cancer cells. And it's a very
• 102:18 - 102:24:  attractive approach, because essentially, you're exploiting the specificity of these drugs,
• 102:24 - 102:30:  which is extremely good between p53 mutant and p53 wild-type cells. It's not the same issues of
• 102:31 - 102:37:  mutation to resistance, because you're treating normal cells and not cancer cells with the drug.
• 102:37 - 102:43:  And so Aileron has been taking this now into the clinic. And obviously, you can look for a
• 102:43 - 102:49:  whole number of trial endpoints, the loss of anemia, the loss of thrombocytopenia,
• 102:50 - 102:55:  the absence of the need to do blood platelet transfusion, but also things like hair loss
• 102:55 - 103:03:  and mouth sores and vomiting, all of those symptoms of chemotherapy. So the trials have
• 103:03 - 103:10:  started, and the initial data are extremely promising. So that does look to be a clinically
• 103:10 - 103:16:  meaningful protection against the hematological side effects. And the drug has to be given 24
• 103:16 - 103:22:  hours before the cytotoxic chemotherapy or radiation. And so far, it's been mostly done
• 103:22 - 103:32:  using topotecan as the conventional therapeutic molecule. But I think this is, again, very
• 103:32 - 103:41:  exciting. So in the last few months, we've seen tremendous progress from the Y220C, the
• 103:41 - 103:47:  refolder, and now this very interesting clinical data on the concept of chemoprotection. So I'll
• 103:48 - 103:54:  just finish there. I think new things really are this very tight regulation of p53 transcription.
• 103:54 - 104:00:  The field has concentrated enormously on the regulation of p53 protein stability,
• 104:00 - 104:05:  but I think one needs to look very closely again at how p53 transcription is regulated.
• 104:05 - 104:10:  And then just some of the lessons we've learned from the stable peptide inhibitors. Essentially,
• 104:10 - 104:16:  the key issue is how do these molecules get into cells, and how can we work out how to do that
• 104:16 - 104:21:  more efficiently? We've tried to make stable peptides to many other targets, and we've
• 104:21 - 104:26:  managed to get very strong binders, but they have not gone into cells at all well. And in fact,
• 104:26 - 104:33:  some of the published literature has been shown to be a bit artifactual. For those who study my
• 104:34 - 104:38:  papers will know that there's been quite a controversy over the RAS stable peptides,
• 104:38 - 104:46:  but we're pretty convinced that they don't really work inside cells. So I'll stop there and say
• 104:46 - 104:56:  thank you, and I will stop sharing the screen and hopefully reappear. Yes, thank you.
• 104:58 - 105:05:  Okay. Well, thank you very much. That was very interesting. There will be many questions. There
• 105:05 - 105:11:  is a question from an anonymous attendee, so we'll see whether they can submit that question
• 105:11 - 105:16:  not being anonymous, otherwise I'll just read it. But just to start, I'm going to ask a question
• 105:16 - 105:22:  about, there's also an indication that if you go the other way, if you have more p53, that you
• 105:22 - 105:29:  actually increase longevity. If you look at naked mole rats, they have high longevity. Not just that
• 105:29 - 105:35:  they live longer, but also they live well. They're well. Yeah, absolutely. Absolutely. So this is
• 105:35 - 105:39:  very, very interesting. And there's some reports that if you look in the Brazilian cohorts,
• 105:40 - 105:47:  they include among them people with a very long life. And so it's very interesting. I mean,
• 105:47 - 105:53:  the mouse models show that if you have an extra copy of the gene, you can get extended good
• 105:53 - 105:58:  quality lifespan. Clearly, you've got to be very careful. I mean, the number of mouse models where
• 105:59 - 106:06:  truncated p53s have been shown to induce aging, and MDM2 hypermorphs can have some very nasty
• 106:06 - 106:12:  side effects. So it's absolutely about the level. And that's why these molecules are very tempting
• 106:12 - 106:17:  because you can imagine giving people a little bit of a mole rat-like boost, you know, and,
• 106:17 - 106:22:  and, you know, and the Jared Evans experiment with inducible p53 essentially suggests that,
• 106:22 - 106:27:  you know, quite long after the carcinogenic insult, you can give a little boost of p53 and
• 106:27 - 106:33:  you'll get rid of the incipient cancer cells. So, so one of the frustrations has been not having,
• 106:34 - 106:37:  not having good enough molecules, but these new, these very recent molecules,
• 106:37 - 106:42:  these stable peptides can be given to mice now, I think over a prolonged period,
• 106:42 - 106:46:  they don't have the side effects or difficulties of the nut lens. So,
• 106:46 - 106:49:  and everybody's very interested in exactly your idea, you know, that is,
• 106:49 - 106:53:  can you just ease up the level of it and, and reduce the thing.
• 106:54 - 106:58:  So instead of your vitamins, you just take a pill that boosts.
• 106:59 - 107:05:  Or, or, or what is it in our diet? That's a bit of a p53 activator. I mean, it's very interesting.
• 107:05 - 107:08:  I mean, particularly for the kids with Li-Fraumeni syndrome, because there's some
• 107:08 - 107:12:  kids very clear, they've got, they've lost one allele. So they're, they're not a point
• 107:12 - 107:17:  mutation. They're a loss. So essentially, you know, they're living on one allele. So would you,
• 107:17 - 107:20:  you know, would you, if you increase the level of expression, achieve,
• 107:20 - 107:25:  achieve some protection? That's just very controversial experiments on this topic,
• 107:25 - 107:32:  but, but interesting. So we have a question from somebody that's not anonymous. So it's
• 107:32 - 107:37:  Bart Versterdorp. He's here. So you have to unmute yourself, Bart, and ask your question.
• 107:38 - 107:44:  Yes. Can you hear me now? Yeah. Yeah. Great. Yeah. That was a really nice talk. So I was wondering,
• 107:45 - 107:50:  do you know what are the main transcription regulation mechanisms of p53
• 107:50 - 107:57:  in proliferating versus non-proliferating cells? Yeah, I mean, we're looking at that. I
• 107:57 - 108:01:  mean, there are quite a number of transcription factors that have been reported to be
• 108:01 - 108:08:  involved in p53 activation. But we're particularly interested in KLF4 as a potential repressor,
• 108:08 - 108:14:  because there's quite a lot of evidence that KLF4 can switch off the p53 promoter.
• 108:14 - 108:21:  And so we're starting to look actually using this in situ hybridization, the RNA scope technique,
• 108:21 - 108:26:  because it allows us to very quickly at this, at this animal material that we have to look for
• 108:28 - 108:34:  other transcription factors or regulators that are coming on just as the p53 mRNA signals going away.
• 108:34 - 108:40:  But yeah, I think it's really, it's really under-investigated amazingly. So considering how long
• 108:40 - 108:48:  we've been studying p53. Yeah. Yeah. Great. Thank you very much. 90,000 papers. Yeah. Not very many
• 108:48 - 108:55:  on what promotes transcription of the gene. So I'll ask the question of the anonymous attendee.
• 108:55 - 109:00:  So this goes into that. If you give cells radiation, then wild-type cells, if they're p53
• 109:01 - 109:07:  positive or plus, then they are more likely to die. Whereas cancer cells do not. So there was
• 109:07 - 109:14:  this idea they're saying 20 years ago, that if you could drug and have a temporal drug inhibiting
• 109:14 - 109:20:  p53, that you can then increase the dose. And what happened to that idea?
• 109:20 - 109:24:  Yes. So this is a very controversial, I mean, everything is controversial. So there was a
• 109:24 - 109:29:  report of a drug called Fitrin alpha, which is still widely used in the literature that was
• 109:29 - 109:37:  claimed to inhibit p53 and give some sort of radioprotection. And, but I mean, many people can't get
• 109:37 - 109:42:  it to work at all. I'm one of those people, unfortunately. So I, you know, we try and
• 109:42 - 109:48:  reproduce these studies, but we can't. I think a p53 inhibitor would be an interesting
• 109:48 - 109:54:  molecule to have in the same way as if you, you know, you could argue this, this
• 109:54 - 109:58:  protection that would give. I mean, it's a, it came up a little bit in some of the earlier
• 109:58 - 110:04:  talks. I mean, there's this balance. So how much protection against chemotherapy does cell cycle
• 110:04 - 110:10:  arrest give you, you know, and then the same way, you know, if you, if you damage cells in the
• 110:10 - 110:16:  absence of p53, you know, how well would they recover when p53 came back from that? So it would
• 110:16 - 110:22:  be lovely to have a perfect activator and a perfect inhibitor. I think we're nearly there with a
• 110:22 - 110:28:  perfect activator. I don't think we're nearly there with a perfect inhibitor. We did try. I mean,
• 110:28 - 110:33:  I screened a huge number of molecules with Merck looking for inhibitors and we were not successful,
• 110:33 - 110:36:  but, you know, this is one of those things, if you don't get it, you can't say anything.
• 110:37 - 110:42:  Oh, true. So, so, but that would only work in somatic mutations because if the patient is
• 110:42 - 110:49:  p53 either heterozygous, it wouldn't work. Yeah. Yeah. But that's interesting. I think that was
• 110:49 - 110:56:  the last question. And we just reached our time. So thank you so much. Very interesting
• 110:57 - 111:04:  talk, a nice overview of your career. I hope that it has inspired many of the early career
• 111:04 - 111:09:  scientists here. I would like to, there's another question, but we will send that to you later.
• 111:09 - 111:13:  So I would like to thank all the speakers. Of course, you can just hear me clap.
• 111:14 - 111:20:  I would also like to thank everybody that attended our meeting, which has been, you know,
• 111:20 - 111:23:  throughout, which is actually very consistent around 300 people. So everybody stayed around,
• 111:23 - 111:30:  which was great. Also, I want to thank Abcam for organizing it. It went perfectly.
• 111:31 - 111:36:  Our next meeting is in March. So I hope to see you all there. And please, before you leave,
• 111:36 - 111:42:  just finish the poll that you can see popping up now. It's a very short poll. And if you fill it
• 111:42 - 111:48:  out now, then we won't have to send you an email to ask you to complete the survey. And,
• 111:48 - 111:52:  and that's it. And if you have a suggestion for our next keynote speaker in our March meeting,
• 111:52 - 112:01:  please email us. So fill out the poll emails for suggestions for next keynote speakers.
• 112:02 - 112:06:  And that's it. Peter, anything to add to this? You're muted. So yeah.
• 112:06 - 112:12:  I'm muted. No, that's everything. Thank you very much. I think the new breakout
• 112:12 - 112:15:  session was a, at least for me, a big improvement. I think.
• 112:17 - 112:17:  We'll find out.
• 112:17 - 112:18:  Thanks everyone.
• 112:18 - 112:20:  We'll find out. Everybody.
• 112:21 - 112:22:  Maybe a disaster. It was good for us.
• 112:24 - 112:34:  Okay. Thank you very much. Fill out the poll and hope to see you in March.