Therapeutic Targeting of Epigenetic Mechanisms in Cancer

January 29, 2024

Information

Yale Cancer Center Grand Rounds | January 26, 2024

Presented by:

ID11232

To CiteDCA Citation Guide

00:00Good morning everybody.
00:01It's so nice to see everybody here.
00:03So let's get started.
00:04So this is a special occasion
00:06and actually nobody better than
00:08Doctor Armstrong to present the
00:10Lecturer in honour of Lance Tallman.
00:12So this series was established
00:15in 2012 by Doctor Marvin Sears.
00:17Dr. Sears was a long time chair
00:19and founder of the Ophthalmology
00:21and Visual Sciences at Yale,
00:23and he established this lecture
00:24series series in honour of his mother,
00:27Lance Tallman,
00:27who passed away from leukemia.
00:29And it really was the first lecture series
00:31dedicated solely to hematologic malignancies.
00:33So hematologists are always
00:35delighted and it's really intended
00:37to bring to Yale pioneers,
00:39you know,
00:40who study malignant hematologic diseases
00:43and then bring treatments to the patient.
00:46There's actually nobody better than to give
00:48today's lecture than Doctor Scott Armstrong.
00:51Doctor Armstrong is the
00:52President of Dana Farber,
00:54Boston Children's Cancer and Blood
00:56Disorder Center and the Chairman of
00:59the Department of Pediatric Oncology
01:00at Dana Farben Cancer Institute.
01:02And since 2016,
01:05he serves as the Associate Chief of
01:07the Division of Hematology Oncology
01:09at Boston Children's Hospital and
01:11was previously the Director of the
01:13Center for Abidinex Research at
01:15Memorial Sloan Kettering Cancer
01:17Center and Professor of Pediatrics
01:19at Weill Cornell Medical College.
01:21I'm not going to go back to
01:23medical degrees and all this,
01:24so good to have you here.
01:25So Roger Armstrong really has you know
01:28pioneered research in in epigenetics
01:33and studying pediatric cancers,
01:34right.
01:35And we always learn that studying
01:37cancer for example in Pediatrics can
01:39then really enlightened mechanism
01:41of disease also for adult patients.
01:43And I think it's super,
01:44super exciting to hear your talk
01:46today really bringing basic mechanism
01:48all the way from the lab to benefit
01:50so many of our patients and we're
01:52incredibly excited to have you here.
02:01Thank you for the nice introduction
02:02and and for the lectureship and
02:04the plaque that's very nice.
02:06And thank you for coming today.
02:08It's not the most beautiful
02:09day to be out walking around.
02:10So I appreciate you you making it here.
02:13And as Stephanie said,
02:15I'm going to talk to you today
02:17about work we've been doing over
02:20the past couple of decades now
02:23focused on originally a relatively
02:26rare subset of leukemia and then
02:28move to more common leukemias
02:31and maybe even beyond leukemias.
02:33And as many of you probably know
02:35that the concept of targeting
02:37chromatin or epigenetic based
02:39mechanisms been around for quite
02:41some time and there have been some
02:43therapeutic advances in that regard.
02:44But it's been it's stops and
02:46starts I would say along the road
02:48and hopefully I can convince you
02:50that we're maybe finally starting
02:52to make some significant go get
02:54some significant insights there.
02:56So these are my disclosures.
02:57I do consult for a number of biotech
02:59companies trying to convince them that
03:01these mechanisms are relevant that's
03:03and sometimes I'm able to do that.
03:05And then this patent on amended inhibition,
03:08NPM one AML.
03:08I'm going to talk about NPM one AML.
03:11And more important disclosure is these
03:12are the people that do the work.
03:14I don't do the work and I'm very lucky
03:17to have tremendous fellows in the lab.
03:20And actually all of these fellows other
03:22than Emily who soon will leave the
03:24lab to go start her own have started
03:27their own independent lab based careers now.
03:30So this is the outline of the
03:31talk I'm going to introduce you.
03:33Many of you probably know a lot of
03:37this to the MLL or CAT or MLL or KMT
03:41two ACI can't even remember what the
03:44other name is for complex and MLL
03:47rearranged leukemias and then move to
03:50the therapeutic development of small
03:52molecules that target those complexes.
03:54Talk a little bit about the clinical
03:56translation and resistance mechanisms
03:57that we're starting to see to those
04:00molecules and then talk about the
04:02potential role in other cancers.
04:03And as I already mentioned the,
04:05the concept of the OR the
04:08relevance of epigenetics.
04:10And just for those of you that
04:11are that are purists in the
04:13epigenetic and chromatin space,
04:14I will interchangeably use
04:16epigenetic and chromatin biology.
04:17Actually there there is a group of
04:20people that think those two things are
04:22not the same thing and the concept
04:24that these mechanisms are relevant and
04:26cancer has been around for quite some time.
04:28This is not a new idea and epigenetics
04:31really encompasses many different
04:33types of modifications of chromatin,
04:36DNA methylation,
04:37histone modifications,
04:38complexes have proteins that
04:41read those histone modifications.
04:43The nucleos,
04:44there's nucleosome remodeling complexes
04:45that you've probably heard about.
04:47The bath complex also are
04:49frequently mutated in cancer.
04:50So we've known that this is relevant,
04:52but what to do about it has been
04:54a little bit harder to understand
04:56and the the kind of simple
04:58concept is, is these mechanisms
05:00control developmental gene expression
05:02and if we were smart enough we'd
05:05figure out how to therapeutically
05:06target them and reverse those cancer
05:09causing gene expression mechanisms and
05:11hopefully we're starting to get there.
05:13There are some FDA approved drugs that
05:15you probably know about H TAC inhibitors
05:18and DNA methyl transferase inhibitors.
05:19I would say whether or not those molecules
05:22are working via epigenetic mechanisms,
05:25still a little bit of a question,
05:26but indeed those were the first ones
05:29that could be working via these some of
05:32these mechanisms that were FDA approved.
05:34So this is the leukemia that I became
05:37most interested in as a fellow
05:39back in the late 1990s.
05:40Now those are leukemias with
05:43rearrangements of the KMT 2A or
05:46or MLL gene and in Pediatrics as
05:49mentioned I'm a pediatric oncologist.
05:52This rearrangement when found in
05:53infants with AOL predicts a very poor
05:56prognosis and in pediatric leukemia
05:58therapy we're actually not used to that.
06:00We cure most of our patients with
06:02AOL and we find a subset that has a
06:05less than 40% long term survival.
06:07We that's unusual and this subset
06:09is that if an infant comes in that
06:12has ALL and has a rearrangement of
06:14this gene that it's probably in
06:16the 4050% long term survival now.
06:23And so back in the late 90s I joined
06:26Stan course Meyer's lab to start to
06:28learn about that and then obviously
06:30ultimately to start my own lab.
06:32So this is the the wild
06:34type MLL or KMT 2A protein.
06:37It's very large, it's in the nucleus,
06:39it's about 500K Daltons,
06:41makes it has made it difficult to study.
06:44It has a number of different
06:46domains and is bound to chromatin.
06:48We've known that for over 20 years.
06:50And when the translocation occurs that
06:52in terminus of MLL is fused to the
06:54C terminus of what's 100 different
06:56fusion proteins also making a little
06:59bit complicated to understand.
07:00And this is just the history that
07:02I won't go through in too much
07:04detail other than to say that the
07:06wild type MLL protein is was shown
07:08in the mid 90s by Stan Course,
07:10Mars Group and others to control
07:13development of blood system of
07:15hematopolysis through presumably
07:16control of the homeotic or hox
07:18genes that are important in many
07:20types of development and but
07:22in blood development as well.
07:24And that was really actually
07:26pointed to that concept by studies
07:28done even before that in fly,
07:30in fruit flies and Drosophila,
07:32showing that the trithorax gene,
07:33which is the Drosophila homologue is
07:36important for development as well.
07:38And then Mike Cleary and Terry,
07:40rabbits showed in very nice mouse
07:42studies in the late 90s that
07:44the MLL fusion proteins indeed
07:46do directly induce leukemia.
07:48And then David Alice's group showed
07:50that in the wild type MLL is a histone
07:53modifying enzyme modifies histone H3
07:54on lysine 4 through this enzymatic
07:56domain here at the C terminus.
07:59So this was really the first well
08:02characterized chromatin regulator
08:03that is known to drive tumor genesis.
08:07And so that was really why in the
08:10early early 2000s a lot of labs
08:12jumped on this to thinking maybe
08:14this would give us some insight into
08:16chromatin based mechanisms and cancer.
08:19So to summarize a lot of work that
08:21we did talk thinking about cells
08:23of origin of of this type and
08:25other types of leukemia.
08:26Not going to get into that too
08:27much today because I want to
08:29get to the therapeutic part.
08:30But we were able to show that the
08:31MLO fusion when we put it into either
08:33stem cells or progenitor cells,
08:35either mouse or human,
08:36that the MLO fusion can drive
08:38the development of leukemia from
08:40multiple different cell types
08:41in hematopoietic development.
08:43And these concepts have now been
08:45shown in other types of tumors.
08:46But at the time that was a
08:48relatively new concept.
08:49But what more importantly what it
08:50let us do is really characterize
08:52the gene expression program that's
08:54driven by this MLL fusion protein
08:57when we put it into in this case
08:59a mouse progenitor cell.
09:00And we could look very quickly to
09:02see what types of gene expression and
09:05chromatin based changes happened when
09:07the MLL fusion binds to chromatin.
09:09And our work and many,
09:11many people's work across the
09:13the world with this model
09:14originally developed by Mike Cleary's
09:16group have I would say that MLO
09:19fusion driven leukemia now is about
09:20as well characterized mechanistically
09:22this as any type of leukemia,
09:25probably any type of cancer
09:27largely because of this model.
09:28So we know where the fusion binds throughout
09:31chromatin, which genes it controls.
09:33We now have mechanisms of turning the fusion
09:35off and we can see what genes get shut off.
09:37We understand now quite a bit
09:39about what the MLL fusion does.
09:42Exactly how it does it is
09:44still a little bit unclear,
09:46but that is what we and others
09:48are really working on now.
09:50And of course that is what we
09:51need to know and under in order
09:53to develop hopefully therapeutics
09:55that can target these mechanisms.
09:57So this is a very simple actually
10:00summary of how the MLL fusion works.
10:03So here in Gray is the ML AF9 fusion,
10:06the in terminus of MLL bound to
10:08some of the proteins normally found
10:10in the MLL complex Menon and Ledge
10:13F here and those help localize
10:15the fusion protein to chromatin.
10:17And then the C terminal part of the
10:19fusion brings in a number of complexes
10:22is histone methyl transferase .1 L,
10:23which is a histone H3 lysine
10:2679 methyl transferase And this
10:28so-called super elongation complex,
10:30which is really a fundamental complex
10:33for controlling transcription broadly,
10:35not just in this setting.
10:36Certainly the MLL fusion drags these
10:40chromatin regulatory and transcriptional
10:42control proteins and complexes to its
10:45target genes to drive gene expression.
10:48So with that level of understanding,
10:51it became easier to go to pharma and
10:53biotech and to get them interested
10:55in developing small molecules that
10:57might target these mechanisms.
10:58Even though at the time no one knew
11:01if these were mechanisms be relevant
11:03beyond this relatively rare disease
11:05which is probably A couple thousand
11:07patients per year in the United States.
11:09But I'll show you,
11:10I think we think that indeed and
11:12in fact we have now shown in in
11:14patients that it it actually is.
11:16So each, as I mentioned,
11:17each of these labels in red is a small
11:20molecule that's been developed to target
11:22various components of this large complex.
11:25And to summarize broadly on molecules
11:28that target the complexes on the right,
11:30the kind of general
11:32transcriptional complexes,
11:33the problem there has primarily been
11:35toxicity that actually not too surprising
11:38you turn off transcription broadly,
11:40we can do that with some of
11:41our chemotherapeutic drugs.
11:42It's relatively toxic on the
11:44left side and to date the problem
11:47has actually been efficacy,
11:49meaning we can do pre clinical studies
11:51and I'll briefly load to this with
11:53.1 actually get really impressive
11:55changes in gene expression and such.
11:57But then we go into patients with
11:59the .1 inhibitor get a little bit
12:02of clinical signal and but unable
12:04to maintain that clinical response.
12:07But but toxicity has not primarily
12:09been a problem for the molecules
12:11on the left side of this figures
12:13and that's kind of where we've
12:15focused. And I'm going to talk a
12:17lot about this protein Menon here,
12:19which is really a scaffolding protein
12:22that's bound to the MLL fusion
12:24and helps keep it on chromatin.
12:25I'll show you more about that in a minute.
12:28Men and as an important part of the
12:30Amylo complex was first demonstrated
12:32in 2004 by Michael Cleary's group.
12:34And now there are many small molecules.
12:36I'll talk about the one we've been
12:38working on that disrupt this interaction.
12:40I'll show you how in a minute.
12:42And that really that concept,
12:44the first chemical biology done around
12:46this was done by Yolanda Grimbeck is due
12:49in Michigan where they developed the
12:51first small molecule to bind to minute.
12:53So just quickly I'm going to
12:55this is kind of jumping ahead,
12:56but it's a concept that I think is
12:59important in terms when we think
13:01about targeting chromatin complexes
13:02in leukemia or other diseases.
13:04So if this is a even more simplified
13:07view of the ML AF9 bound to chromatin
13:10through its adapter proteins and
13:11bring it in this case the .1 complex.
13:14I told you that we worked actually
13:16over a decade ago now with a
13:19company called Epizyme to make small
13:21molecule enzymatic inhibitors of .1.
13:23We've shown with genetic studies
13:25that that is an important component
13:27of this complex and that histone
13:29modification that it put places
13:31is important for controlling MLO
13:33fusion driven gene expression.
13:35And that went into clinical trials and
13:37we were able to a couple of patients
13:39actually went into to remission
13:41and many had some response but it
13:43they were all pretty transient.
13:44And I'm going to show you that
13:47that's not the case with the min an
13:49inhibitor and this summarizes why
13:51we think that is that if you the .1
13:54L inhibitor inhibits the enzymatic
13:55activity of this protein,
13:57but the whole complex remains stuck
13:59on chromatin.
14:00So it's actually doesn't take a
14:02lot for the cell to figure out how
14:04to deal with the fact that that
14:06enzymatic activity is no longer there.
14:08Whereas what I'm going to show
14:09you with the MLL Menin inhibitor,
14:10it actually disrupts the whole complex
14:13and much of the complex lifts and
14:15the fusion protein lifts off of
14:17chromatin and we think that probably
14:19is why the the Menin inhibitor
14:22approach is more dramatic.
14:24So in 2019 we published the small
14:26molecule that we were working on which
14:28was at the time called BTP 5O469.
14:31The version of it that's in the clinic
14:33is called Syndex 5613 or Revu Minib.
14:35Get to that in a minute.
14:37And this is a very potent small
14:39molecule that binds to this pocket on
14:41men and this is the men and protein.
14:43This is the crystal structure of this
14:45protein here in purple and blocks
14:48the interaction between men and
14:51and the MLL fusion and that leads
14:53to as I mentioned the loss of this
14:55complex on chromatin,
14:56but interestingly enough only
14:58does it at very selected sites.
15:00This complex actually remains on on
15:03chromatin and the wild type complex which
15:06would also be potentially disrupted
15:08by this molecule remains on chromatin
15:11at many sites throughout the genome.
15:13But there's a very selected group
15:16of genes loci where it's lost.
15:18The importance of that is,
15:19is that we know genetic inactivation of MLL
15:23broadly is toxic that to hematopoietic cells.
15:26But there's some reason that this
15:28mechanism seems to be only important
15:31at certain developmental loci.
15:33And we think that may be why we're not
15:35seeing tremendous toxicity that you
15:37could imagine by lifting this chromatin
15:39complexes off the genome broadly.
15:41And this is just one example of
15:43a preclinical PDX study that we
15:46did with an MLL rearranged AML.
15:48All of you probably know what
15:49PDX studies are,
15:50inject the human leukemia into
15:52immunodeficient mice and treat the mice here
15:55with two cycles if you will of of therapy.
15:57And the beauty of this experiment is
16:00from a technical standpoint is Syndax
16:03actually supplies us with mouse chow
16:06that's impregnated with the drugs.
16:08So you don't actually have to dose the mice.
16:11People in the lab love that fact.
16:13So basically you inject the leukemia,
16:15change the food and come back in
16:16a month and see what happened.
16:18And it's obviously not quite that simple,
16:19but so we were able to do these experiments
16:22with much more speed and much less pain
16:25than they than they often can take.
16:27And so you can see here that the the
16:29mice that were treated with the the MIN,
16:31an inhibitor essentially the AML
16:33in this PDX model was eradicated,
16:36which as most of you know who
16:38do these types of experiments.
16:40That's not the how these experiments
16:42usually go.
16:43You might see a prolongation of survival
16:46but not this degree of response.
16:49Same thing in a model that a lot
16:51of people use,
16:52this retroviral model of ML AF9
16:54that we've used a lot as well,
16:55which does lead to a very aggressive
16:59mouse AML.
17:00And you can see here,
17:02we've treated mice that have gotten
17:04these cells injected with the,
17:06the min,
17:06an inhibitor for this period of
17:08time and in fact,
17:08it eradicated the disease in this model.
17:11So that is again very different from
17:13what most of the graphs will look like
17:16when you do an experiment like this.
17:18So really indicating that
17:20there's significant potential.
17:22So I'm going to switch diseases or
17:24switch subtypes of leukemia and tell
17:26you a little bit about a different
17:28subtype that one is interesting and
17:31two actually kind of changed the
17:33level of interest from biotech and
17:35pharma because it's way more common
17:38and that's NPM one mutant AML.
17:40And I keep saying that because as
17:43you can probably tell in paediatrics
17:44and other rare cancers,
17:46we're constantly and many of you
17:48probably recognize this when we're
17:50talking about drug discovery and development,
17:52having to convince people to work
17:54on the diseases we're interested in
17:56because they're indeed quite rare.
17:58When you go from 1000 to 2000 patients per
18:00year in the United States to 15 into 17,000
18:03patients per year in the United States,
18:05you actually get a lot more interest.
18:07And so that's one of the
18:09reasons beyond just the the,
18:11the mechanistic interest that
18:12I think this is important.
18:14So Michael Kuhn, when we were in New York,
18:17a fellow at the time asked
18:20the question of well, OK,
18:21we know that the MLO rearranged leukemias.
18:23This is a gene expression data for
18:25the HOX genes and MIS ones from Tim
18:28LAYS group in a bunch of AM LS,
18:30and you probably know how to read these.
18:31Each row here is AG in each
18:33column with leukemia sample.
18:34And here's the MLL rearranged group OX,
18:37A cluster we know that's long known.
18:39In fact, some of the microarray data
18:41that I did as a fellow demonstrated
18:43that the Hox A cluster is expressed in
18:46MLL rearranged leukemias as is MIS one.
18:48Those are two targets of the ML effusion
18:51that have been worked on for many decades.
18:54It turns out in the NPM one mutant
18:56leukemia as you can see here on the right,
18:58they have the Hox A cluster as
19:01well and B cluster,
19:02another Hox cluster and MIS one.
19:04So this has also been known for some
19:07time that this subset of AML and PM one
19:09mutant also tends to express the Hox genes.
19:11So Michael using CRISPR which in
19:152016 was still relatively new did
19:18a what's called a domain scan.
19:19I won't get into the details of that,
19:21but the point being that the inactivation
19:24of MLL and ultimately I mean of men
19:27and MLL led to differentiation and
19:30ultimately death of NPM one mutant AML cells.
19:33And the concept being at the time
19:36that the disruption of MLL and
19:38Menon in the NPM one state somehow
19:41leads to a decrease in the Hox gene
19:44expression and therefore ultimately
19:46a enemies one expression and therefore
19:49ultimately changes in gene expression.
19:52Hannah Okleman,
19:52who was a a fellow in the lab just
19:54recently moved back to Germany
19:56to start her own lab,
19:57then demonstrated that similar to
20:00the MLL fusions that NPM 1 mutations
20:04in mouse models could transform
20:06multiple hematopoietic cell types,
20:09stem cells a little bit more significantly
20:11more efficiently than progenitors,
20:12but also could transform progenitors
20:15as well and.
20:17Needless to say,
20:18there was still a question as to kind of
20:21mechanistically how all this was working.
20:23I'm going to summarize some chip seek
20:26data here that really shows that at
20:29least at this level of understanding,
20:32it's quite similar to what we see
20:34with the MLL rearranged leukemia.
20:37So these are NPM one mutant cell
20:39lines where we're doing chromatin
20:41immunoprecipitation for men and for
20:43example either in a control setting in
20:47black PMSO treated cells or the min
20:49inhibitor treated cells and you can see
20:51that min and comes off of chromatin.
20:53We know that when we treat
20:54with the min inhibitor
20:55min and lifts off of
20:57chromatin MLL shown here.
20:58Now this is wild type MLL because
21:01it's a mutant in PM one setting
21:03doesn't come off the Hawks loci.
21:04We'd seen that in the MLL fusion setting
21:07as well but does come off of MES 1.
21:09So this is that concept that I show you
21:12that certain loci respond differently to
21:14the min and inhibitor and this is showing
21:17the RNA SEC or the gene expression.
21:20And you can see that when there's
21:22a correlation between loss of MLL
21:24occupancy and loss of gene expression
21:26and this is that looking at that more
21:28broadly by what's called a GSDA analysis.
21:30So the point being that much like what we saw
21:33in the MLL fusions in the NPM one setting,
21:36you treat with the men inhibitor
21:37men and comes off of chromatin.
21:39The MLL protein itself comes off of
21:41chromatin at about 100 to 150 loci
21:43and those genes lose their expression
21:46when the cells differentiate.
21:48And then if we go into PDX studies to
21:51NPM one mutant flip 3 ITD Co mutant,
21:54these are relatively aggressive AM LS.
21:57You can see again pretty impressive
22:00response in the PDX setting.
22:02And for the people who start do these
22:04types of experiments in the audience,
22:07we actually as you probably know the
22:09standard way to do this is put in
22:11leukemia into an immunodeficient mouse,
22:13the first sign of any leukemia in
22:14the peripheral blood of those mice,
22:15you start treating them.
22:17We actually waited in this experiment
22:19on the right until the mice were ill
22:21and started treating them and actually
22:25we could recover essentially 4 out of
22:27five of the mice and eradicate the disease.
22:30So we stacked the deck against us and
22:32we're still able to to make that or
22:35extend the the survival of those mice.
22:37So to summarize this at this point
22:39we the MIN inhibitor and NPM one
22:42mutant AML induces differentiation,
22:44reverses leukemia,
22:46genic leukemic gene expression,
22:48certain genes like MIS one removes
22:51MLL from those loci and we get
22:54dramatic responses.
22:54So that was exciting and that was
22:57enough to get Janssen and Ichi
22:59and the various other large
23:01pharmaceutical companies interested in,
23:03in small molecule development
23:05of Menon inhibitors.
23:06But it's still there was still
23:08something here we don't quite or didn't
23:12quite understand is why is it that
23:14the NPM one mutant AML is depending
23:17so much on the MLL Menon complex.
23:20So Hannah dug into that as well.
23:22You guys probably know this,
23:23but this is just a little bit
23:25about the mutant NPM one protein.
23:27It's shown here.
23:29Schematically,
23:30it's found mostly in the nucleolus
23:33in the wild type setting,
23:35but when the mutation occurs,
23:37it's a mutation in this nuclear
23:39or localization signal that then
23:41leads to a nuclear export signal.
23:43So the mutant in PM one is largely
23:45found in the cytoplasm and that
23:47was recognized by the people who
23:49initially described this mutation.
23:51However,
23:52there is some
23:54that remains the mutant in PM one
23:57in the nucleus and we've taken
23:59advantage of a system that many of
24:01you probably know about where you
24:04can now by CRISPR mediated homologous
24:06recombination actually tag if you will,
24:10whatever gene or protein of interest with a
24:13degradable version of FK PP12 shown here.
24:16And actually this cell line was made
24:18by Peggy Goodell's group in at Baylor
24:21and you have a mutant NPM one protein
24:23that has this degradable version
24:25of F KBP 12 and you can treat them
24:27with a small molecule Protac that
24:29will degrade the whole thing.
24:30So you can degrade the mutant oncoprotein
24:33and look fairly rapidly after degradation
24:35as to what's happening and here's
24:37how rapidly you get degradation.
24:39By 60 minutes you've got about half
24:41of the mutant protein gone and by 120
24:44minutes essentially all of us gone.
24:45So it's relatively rapid and these
24:47types of experiments are quite
24:49illuminating because you really have
24:51very tight control over over the system.
24:56And what we see is that we get
24:59differentiation when we degrade the
25:01mutant onca protein and ultimately
25:03apoptosis of the cells as well.
25:05And this is just the Western blot looking.
25:07We can actually separate the mutant
25:09protein from the wild type because
25:11it's tagged,
25:11it's here and you can see that here.
25:14This is the cytoplasmic prep,
25:15the nuclear prep and the chromatin
25:17prep and here's the mutant in PM one
25:19and we can control that by degrading
25:21it to show that that signal actually
25:23is what we think it is.
25:25And in fact there is a fair amount
25:27of the mutant in PM one in the
25:28nucleus and on chromatin.
25:30Then if we do chip seek to say
25:32where is it in the nucleus and
25:34where are where is it on chromatin.
25:36You can see here with with two different
25:39antibodies in black that the NPM one mutant,
25:42NPM one protein is bound to many
25:44of the similar genes that we've
25:46learned about with the MLL fusion,
25:48the Hox cluster MIS one.
25:50And when we degrade it that signal goes away.
25:52And I keep saying that because
25:54particularly with chip seek experiments,
25:55the opportunity for background signal
25:57is real and this is makes you feel
26:00much better that the signal that
26:02you're looking at is indeed the
26:04signal that you are interested in.
26:05And now you actually can go to primary
26:07patient samples with those antibodies
26:09and see the NPM one protein mutant
26:11NPM one protein bound there as well.
26:13And here's the list of the top 50
26:15or so genes to which the mutant
26:17NPM one protein is bound.
26:18And you can see as some of these genes
26:20that I've already talked about OX,
26:22A&amp;B cluster and a number of other genes
26:24that we tend to pay attention to and
26:26stare at when we're talking about MLL or
26:29looking at MLL rearranged leukemias as well.
26:33So,
26:34but is it controlling gene expression.
26:36So now we have mutant
26:38NPM one protein bound to
26:40interesting sites in on chromatin
26:42and we can degrade it and show
26:44that that signal is specific.
26:46And now what happens to transcription?
26:48So this was 24 hours later,
26:50quite a bit later.
26:51Most of those genes that I just showed
26:53you where the NPM one protein is bound,
26:55their expression is down and
26:57this is an approach called pro
26:59seek which I won't get into the
27:00details as to how one does this.
27:02So many of you,
27:03some of you who work on transcription
27:04probably know this technique.
27:06But essentially it measures the
27:08amount of bound RNA polymerase 2 out
27:10throughout the length of the gene as a
27:12surrogate for transcriptional activity.
27:15And what we can see that as quickly
27:17as 30 minutes after treatment of the
27:19cells with the NPM one protein degrader,
27:22you're already seeing a decrease in
27:24transcription at those sites where
27:26the NPM one was previously bound.
27:28And and if you look at it across
27:30the the the all express genes
27:32you don't see those changes.
27:33So in fact it's the NPM one protein's
27:37bound there and somehow controlling
27:40transcription of these genes.
27:42And one of the ways it's doing
27:44it is by keeping RNA Pol two,
27:46CK nine that that super elongation
27:48complex that I told you is critical for
27:50transcription bound to those genes.
27:52So when we treat with the degrader
27:54NPM one comes off and then much
27:56of the transcriptional apparatus
27:58comes off of those genes as well.
28:00So it's maintaining a state that allows
28:03for those critical complexes including
28:06pole two to to bind to those low side.
28:09I'm going to go through the details of
28:12this but just to kind of summarize it
28:15that when we degrade mutant in PM one,
28:17we lose RNA polymerase two occupancy
28:19where the NPM one was previously
28:21bound within an hour.
28:23So off goes NPM one,
28:24off comes pole two and a number
28:26of other histone modifications
28:28like H3K27 acceleration,
28:29some of you know associated with various
28:32types of gene expression decreases
28:34and then the histone modification,
28:36other histone modifications start
28:38to decrease somewhat later.
28:39For those of you interested in transcription,
28:41we can talk more about this in detail later,
28:43but it looks like it's like when
28:45we degrade the mutant in PM one,
28:47the decrease in gene expression
28:48is actually biphasic.
28:49There's so initially there's a decrease
28:52of about 50% probably because pole 2
28:55is not quite as there's not as much
28:58pull two and other complex occupancy.
29:00And then after about 3 days we
29:02see a dramatic another dramatic
29:04decrease in gene expression.
29:06We think that's because now the
29:09histone modifications are starting
29:12to come in and and work together
29:14with whatever the previous mechanism
29:15was to fully shut off transcription.
29:20So how does this connect to to Menon?
29:26So
29:31there we go. So we now if you
29:33treat with the MIN inhibitor,
29:36what happens to this chromatin bound in
29:39mutant in PM one and I'll just quickly
29:42summarize it by saying you can see here
29:45here's the mutant in PM one we treat
29:47with the MIN inhibitor doesn't come
29:49off the hogs locus but it actually
29:50does come off of the mis one locus
29:52exactly where we're seeing MLL come off,
29:54same thing down here and if you compare
29:57that the gene expression again those are
29:59the genes that are losing expression.
30:02So to summarize what I'm saying here is
30:04that when we degrade the mutant NPM one
30:07protein with this degrader molecule,
30:09we lose RNA pole two CDK 9 and
30:12ultimately .1 at those loci.
30:14When we treat with the min inhibitor we
30:16do the same thing but at a subset of
30:19the loci where the NPM one is bound.
30:21So very similar to what's happening
30:24with the MLL fusion.
30:25However, we're now looking to see if
30:28indeed the mechanisms are identical.
30:30And it turns out that while some
30:31of the complexes are overlapping,
30:33these are the mechanisms are
30:35not perfectly identical.
30:37That is,
30:37some complexes are important
30:39in the MLO fusion,
30:41not an NPM one and vice versa.
30:42So we're trying to work through those
30:45details because as you can imagine,
30:47the next step that we want to do
30:48is come in and target some of these
30:50other complexes with small molecules.
30:51So to summarize this part before,
30:53now I move to the clinical translation.
30:56There are a subset of leukemias that have
30:59high level Hox gene expression MIS one,
31:01another transcription factor
31:02called PBX three.
31:04This actually accounts for about
31:0640% of human AML and it's ones
31:10with these genetic abnormalities,
31:13MLO rearrangement,
31:14NPM 1 mutation and I'm actually more
31:19relatively rare but more common than
31:21in adults rearrangement called Newt 98
31:24rearrangements in both pediatric and adult.
31:26So again accounting for
31:28about 40% of patients.
31:31So that all while all that was happening,
31:35Syndax and another company called cure
31:37oncology that many of you know about,
31:40we're developing small molecules
31:42here Syndax 5613 and here Cure's
31:46what's now Zyftominib and these
31:47I'll show you some of this,
31:49the data from the Cindex trial
31:52and some mechanism of resistance.
31:54And as I mentioned essentially
31:56when the NPM one story came out,
31:59we were called by Janssen and actually
32:01another couple of other pharmaceutical
32:03companies and at least these three
32:05now have MIN inhibitors that are
32:06right on the tails if you will of
32:08the cure of development and Cindex.
32:11So this is so now we're into
32:15patients with the Syndex 5613.
32:17This is just date some of
32:19the data from the phase one.
32:21This is one of the first patients
32:23that was treated at Dana Farber and
32:26we were able to get the peripheral
32:27blood and here are the blast.
32:29And you can see this is with
32:31Revumenib that day three,
32:32not much has happened.
32:33Day seven start to see a decrease
32:35in peripheral blast,
32:36day 14 further decrease and by day 30
32:39at least the peripheral blood blasts
32:41are in this case essentially gone.
32:44Florian Perner is a postdoc
32:45who was doing this.
32:46He sorted these cells and looked
32:48at gene expression and in fact the
32:50gene expression changes that we see
32:52here look very similar to what we
32:53had seen in the preclinical studies
32:56not being a clinical trialist.
32:58I'm going to summarize the whole
33:00phase one right here with a lot
33:02of work from a lot of people and
33:04this was published last year.
33:07This is the Revue Minib Phase
33:09one with Syndex 5613.
33:10The other name for it,
33:1268 patients with relapsed
33:15refractory leukemia.
33:16As you probably know,
33:17many of these patients have had
33:19tremendous numbers of cycles of
33:21various types of therapies and the
33:24CR rate depending on how you count
33:26CRS and for those of you who do
33:27clinical trials and A and all we can
33:29talk about that is somewhere in the
33:3040% range with an overall response
33:32rate of about 50% and a median
33:35duration of response about nine months.
33:37So for in relapse refractory setting,
33:39these are actually pretty impressive
33:42numbers and the cure oncology
33:44small molecule seems to be doing
33:47something having similar activity.
33:48And in fact some of that data from
33:51Janssen was also just printed presented
33:53at ASH and it looks like the activity
33:55of that molecule is is similar.
33:57So indeed there looks like
33:59there's significant clinical
34:01activity of this approach.
34:02This lot slide just reminds me to
34:05point out and and then now I'm talking
34:08to people who are doing AML clinical
34:10trials in the in the audience.
34:12An interesting phenomenon that
34:13you know better than I do,
34:16but that is influencing how these
34:17drugs are are able to be developed
34:19and that is when you treat patients
34:21with them in an inhibitor,
34:23many of them will develop this syndrome
34:25called differentiation syndrome.
34:27But in this setting it looks a little
34:30different clinically I'm told,
34:31than the differentiation syndrome
34:32that you usually see when you
34:34treat patients with acute per
34:36myelocytic leukemia with ATRA.
34:37In fact patients have died from
34:40this differentiation syndrome and
34:42so that has prompted the FDA to
34:44call this a dose limiting toxicity.
34:46Think about what that means.
34:47It means that your your your dose limiting
34:51toxicity is actually occurring as a
34:53result of efficacy of your molecules.
34:56So we can talk about what So I think
34:58the FDA is fighting against the the
35:00the some things that they shouldn't be,
35:02but that's a that's a whole other soapbox
35:05that we can talk about if we want to.
35:07That would be like for those of you
35:08who treat patients with ALL saying you
35:10start to see a little tumor lysis syndrome,
35:11we'd better stop treating them
35:13because that's bad. No, that's good.
35:16So with all this in mind and the
35:19clinical activity looking interesting,
35:21we figured that it was much like
35:23any single targeted agent,
35:24there was likely to be some mechanism
35:27of resistance to that targeted agent.
35:29And right about the time we started
35:31thinking about this the Broad Institute
35:33developed this screening based on or
35:36screening approach based on single
35:38nucleotide base editing which in fact
35:40what you can do is tile in this case
35:44Menon the whole length of the gene
35:47with guides that will mutate not
35:49every nucleotide because of the way
35:51that it's designed but where you can
35:53mutate the majority of amino acids
35:55across the length of that protein
35:59to to basically do an in a
36:02cellular mutagenesis screen
36:03to see if you can phenotypes.
36:06So what Florian decided to do is to
36:08get that base editor library made
36:10for the minute gene treat cells with
36:13the min an inhibitor and see if there
36:16were mutations that made the cells
36:18resistant to the min an inhibitor.
36:21And in fact there were there shown here
36:23in two different MLL rearranged lines
36:25and interestingly enough we we kind of
36:29looking back this was probably silly.
36:31We looked at this,
36:32the new technique and there's a
36:34little bit of noise and we didn't
36:36know exactly what to make of
36:38it looked kind of interesting.
36:39Florian put it in the drawer and kind
36:41of didn't do too much more with it
36:43until we got a call from Ross Levine
36:46and Etan Stein at Memorial Sloan Kettering.
36:48Actually,
36:49I got a like emergent text from Ross
36:51which I thought something really bad
36:52had happened and he said we have to talk now.
36:55So I called him and he said we
36:57found mutations in Menin in samples
36:59from patients that have progressed
37:01on the Menin inhibitor.
37:03And great, you know,
37:05what are they?
37:06And in fact the first one was this
37:08mutation 3 andine 349.
37:10So we dug Florian's data out and
37:13we're like holy cow,
37:14the patients are getting the same
37:16mutation that the base header screen
37:18had suggested they might get even though
37:21we weren't confident in our in our data
37:23to go ahead and start studying that.
37:25But Needless to say with that information,
37:27we started studying this in quite a bit
37:30of detail and we went to send X and
37:32got samples from a number of patients.
37:34And it looks like within two to
37:37three months about 40% of the
37:40patients had developed.
37:41They weren't in fluorid
37:42relapse or progression,
37:43but had developed a clone with
37:46this a min and mutation in them.
37:49That's what this is showing here and
37:50here is like and those mutations
37:52were not present at screening.
37:53So this is just a pie chart in
37:55red here showing you the size of
37:57the clone that has developed the
37:58min and mutation.
37:59So this is acquired selective
38:02mutational resistance to the men,
38:04an inhibitor which as most of
38:06you probably know is considered
38:08a validation of the therapeutic
38:10targeting the kinase world when this
38:12happens and essentially we think
38:14it's saying the same thing here and
38:16we found a both in patients with MLO
38:18rearranged an NPM one mutant AML.
38:20So this is the, if you will,
38:22the gold standard for that validation
38:24of a therapeutic target in patients
38:26that that that you put so much pressure
38:29on the target that the cancer mutates.
38:31It's such that it's no longer effective.
38:33I'll show you why in a minute.
38:35And of course as most of you know,
38:37we can have to have combinations
38:39anyway and we were able to show that
38:41we can do the same thing in PDX models.
38:43So we take our MLL rearranged
38:45or NPM one mutant models,
38:46treat them with in an inhibitor and
38:49in some cases but not all they will
38:52develop the mutations that we see in
38:55in the patients and mechanistically
38:57we we know how this is working.
38:59I won't go into all the details but
39:01this is just one chip seek experiment.
39:03In the wild type setting you can
39:05see Menin comes off of chromatin,
39:06this is chip seek increase in
39:08concentrations of the Menin inhibitor.
39:10But if you have mutated Menin in that
39:12cell line, it no longer comes off.
39:14And we know now biochemically it's
39:17because the binding affinity of the
39:19Menin inhibitor has been shifted
39:21significantly as a result of those mutations.
39:23And in fact,
39:24we know this now at the crystal
39:26structure level and we know
39:28exactly why that's the case.
39:29So here's revuminib bound to Menin.
39:32You can see over here on the right,
39:35these amino acids
39:39M327T349-G331, all these are mutations
39:40that have been found in patients.
39:44Interestingly enough,
39:45the wild type MLL protein does not
39:49use those amino acids to anchor
39:51and that's actually an important
39:53concept because if you develop
39:55the mutation that where men and
39:58the MLL can no longer interact,
40:00those cells won't survive that.
40:02So that that that's not an
40:03option for the cells.
40:04They have to mutate something that
40:06doesn't affect MLL but does affect
40:08the the binding of the inhibitor.
40:10And in fact that's exactly what they've done.
40:12They've mutated this region of men and
40:15that has plays no role in MLL binding.
40:18And the way that that happens is
40:20essentially right here the yellow
40:22is the the min inhibitor bound to
40:24wild type min and the purple is the
40:27min inhibitor bound to mutant min.
40:28And and you can see essentially what's
40:31called a steric clash which so the the,
40:33the min inhibitor is pushed out a
40:35little bit here because of these
40:37changes in the amino acid and that
40:40leads to a 10 to 100 fold decrease
40:42in affinity of this molecule.
40:44So this is I find this amazing
40:47because it's rare that you get to see
40:50molecularly the difference between
40:51response and resistance which is really
40:54essentially a few angstroms here of
40:57this Menon inhibitor binding to to Menon.
41:01So this we're we're continuing to work
41:04on mechanisms of resistance in the past
41:06five or last five or 10 minutes here.
41:08I'll tell you about some others,
41:11but just getting back to a little
41:13bit to that comment I made about
41:15dose escalation and stopping your
41:18dose escalation before you perhaps
41:20get to full potential efficacy.
41:22And now we see that we're developing or
41:25patients are developing mutations that
41:27all they do is shift the curve a little bit.
41:31It does make you wonder if you'd had a
41:33higher dose and a higher concentration
41:35earlier on if you might have prevented
41:38the cells from developing those mutations.
41:40So we with send X luckily providing
41:43us food now with varying amounts
41:45of the min an inhibitor.
41:47We were able to do a dose response
41:49experiment in a PDX model with increasing
41:52concentrations of the min an inhibitor.
41:54And you can see here that at
41:56the lowest concentration point,
41:58O3 3%, you see no response.
42:01At the minimal the medium concentration
42:03here you do see a response and the
42:06leukemias progress and the vast majority
42:08of them will have developed the mutation.
42:10If you then go threefold
42:12more of them in an inhibitor,
42:14you get a much longer response.
42:16In fact,
42:16maybe some of them here are cured of the
42:19disease and when the resistance occurs,
42:21it occurs without the min and mutations.
42:23So in fact,
42:24a higher concentration does at
42:26least in this model prevent the
42:28development of those mutations.
42:30So you can see if you stop your dose
42:32escalation right here because you're
42:34getting differentiation syndrome
42:35and somebody tells you you have to,
42:38you're actually setting up a
42:39situation where you're going to
42:41get acquired resistance mutations.
42:44Having said all that,
42:45we're still getting resistance way out
42:47here with the single agent at higher doses.
42:49So what's that all about and
42:52I'll quickly summarize this.
42:54Essentially what we're seeing here
42:55is that the leukemia cells and
42:57this is another phenomenon that's
42:59known in other settings now are
43:01not mutating the men and they're
43:03actually changing their state
43:04significantly to lead to a state that
43:07we don't understand completely yet,
43:09but where they're now no longer
43:11dependent on that Hawks niece program.
43:14And and in fact interestingly they
43:16look much more differentiated.
43:17The leukemias themselves almost
43:20look like monocytes in terms
43:22of their their flow cytometry,
43:24but definitely will transplant the
43:26disease from 1 bow to the next.
43:28So they're not monocytes,
43:30but so we're trying to understand
43:32this mechanism mode of adaptive
43:33resistance a little bit better.
43:35Now have developed a model,
43:37a cell line model for it and have
43:40developed some PDX models as well.
43:41But the cell line model actually
43:44lets us move to what everyone
43:47likes to do now which is a genome
43:49wide CRISPR screen to say OK,
43:51how did the dependence do the
43:53dependencies change when you
43:54go from one state to the next?
43:56And to summarize,
43:57a lot of analysis in fact they do
44:00appear to and this is on a subtle
44:03transferase CAT6A or MAZ which
44:05also rearranged rarely in some
44:07leukemias now becomes seems to become
44:11relevant in this in this setting.
44:13So here is just an experiment
44:15showing you that.
44:17So we developed resistant leukemia cells
44:20that don't have the MIN in mutation.
44:21They're resents to the MIN inhibitor.
44:23Here it is in blue.
44:24But if you treat with the cat
44:266A in this case guide,
44:28it re sensitizes the the cell
44:31line to the MIN inhibitor.
44:34But interestingly and interestingly enough,
44:36the CAT 6A by itself in the
44:38absence of the MIN inhibitor has
44:40a little bit of an effect.
44:41It's really something about the relationship
44:45between CAT6A and Menon that is important.
44:48So what's CAT6A?
44:49It's a histone has still transferase
44:51as well modifies histone H3
44:53on various lysine shown here.
44:55And if you do now chip seek in either O
44:58sensitive leukemia cell lines or resistant,
45:01the MLL Menon and CAT6A chip seek
45:06data looks very similar and in so here
45:08both in the sensitive or the resistant
45:10and this is just showing that more
45:12broadly so men and Catsix is there,
45:14it's on the scene and it becomes seems
45:17to become much more important when the
45:20cells adapt to the men and inhibitor.
45:23Needless to say,
45:24we're now doing the experiments to
45:26see if this combination in mice will
45:29reverse the resistance in patient samples.
45:33It won't reverse the resistance
45:35to the mutations,
45:36but it might reverse the
45:38resistance to the adapted form.
45:40But the men inhibitor works well enough,
45:42it's hard to generate that adaptive form,
45:44so it's taking us a little while.
45:47I'm going to skip this just for time's sake,
45:50but to and just to summarize saying that
45:54these complexes here, the .1 complex,
45:57the Super elongation complex,
45:58I've been talking about MLL and CAT6A.
46:02If you look broadly,
46:04some of you probably know that Broad
46:06Institute's been doing broad CRISPR
46:08screens on five 600 cancer cell lines,
46:10and they make all this data publicly
46:12available in many different ways.
46:14You can search that data.
46:15If you ask what genes have a
46:18similar dependency to Menin
46:20throughout all of cancer space,
46:23.1 is the next thing on the list.
46:26And then and then these other
46:28proteins here E&amp;L which is part of
46:30the Super elongation complex CAT6.
46:33A point being these,
46:35it's clear that these complexes
46:37are working together somehow
46:39throughout many cancer cell lines,
46:41Leukemia for sure,
46:43but also probably others as well.
46:46And in fact that led us to this
46:50publication for actually for
46:51about a year and a half ago now.
46:54Matt Hemming,
46:56a paediatric or medical oncology
46:58fellow was interested in
47:00gastrointestinal stromal tumors
47:01and just did a genome wide CRISPR
47:04screen because that's how you start
47:06every project these days it seems.
47:08And in fact cat 6A was one of
47:10the top hits in that screen.
47:12And then we looked a little bit more in more
47:16detail .1 and Menin were in there as well.
47:18And the bottom line is,
47:19is it appears that this CAT6,
47:23A .1 Menin complex cooperation is
47:28important in this type of cancer as well.
47:32And we don't understand,
47:33you might ask why should ask why?
47:36We don't completely understand
47:37why it's the case.
47:39But it does appear that when we
47:42inhibit CAT6A and men and in cell
47:44lines or in mice or in PDX or
47:48xenograft models that a program
47:50that Matt had described earlier
47:52driven by transcription factors
47:54like one called hand one,
47:56which is known to be important
47:58for controlling lineage associated
48:00gene expression in this cancer,
48:03certain goes down pretty rapidly.
48:05So there's something about the
48:07developmental program and this
48:08cancer as well that seems to be
48:11dependent on these complexes,
48:12but you don't really see it as
48:15dramatically as in leukemia until you
48:17start to combine the the small molecules.
48:19Many will do a little bit of it,
48:20Mos cats, XA will do a little bit,
48:22but when you combine them,
48:23you really get a a dramatic response.
48:25So the point being that we're
48:27looking at this in leukemia,
48:28but we're starting to move into
48:30some other cancers as well to see if
48:34indeed these developmental regulators,
48:36if you will, might be relevant there.
48:39I just summarized this data and
48:42I'll end with this.
48:44Pfizer just published a paper
48:46about six months ago.
48:48They've now developed a clinical
48:50grade CAT6A inhibitor and in fact
48:52this small molecule is in phase
48:54one clinical trials in estrogen
48:57receptor positive breast cancer.
48:58And it looks like from this paper
49:01and we've now done a number of
49:03experiments as well that somehow
49:05Menon is supporting the ER driven
49:08gene expression program.
49:09And Needless to say, I mean sorry Cat 6A.
49:13Needless to say, we're now looking
49:14at this Cat 6A men in combination.
49:16And again as much as like that we saw
49:18in GIST, it looks like this the ER
49:21driven program is highly dependent on
49:23those two complexes and we're trying
49:25to work through that now as well.
49:28So to summarize what I've told you,
49:31the MEN in MLO complex is a relevant
49:34therapeutic target and snippet subset of
49:37AM LS and that acquired somatic mutations
49:40in men and are a mechanism of resistance,
49:43not the only mechanism of resistance
49:45but that clearly validates men and has
49:48a therapeutic target in this disease.
49:50We're continuing to work on the various
49:52types of resistance and we do think
49:54that rational combinations like men
49:56and CAT6A or men and other things with
49:58a lot going on in terms of trying to
50:01understand which combinations may
50:02prevent development of resistance.
50:03And perhaps the most exciting in the
50:05longer term is if we can understand
50:08where these mechanisms might be
50:09important even beyond leukemia.
50:11And I think they're going to
50:12be opportunities,
50:13but we certainly still have work to do to,
50:15to prove that.
50:16So I've talked much about the people
50:18in in the lab that have done the work.
50:21These are our collaborators at Dana
50:23Farber actually Nathaniel Gray now
50:25at Stanford and Ross I mentioned and
50:28Chang and Richard who work with Ross
50:31at MSK and some of our collaborators
50:34throughout HMS community.
50:35So thank you,
50:37happy to take any questions and
50:39thanks for thanks for staying.
50:41All
50:49right, absolutely spectacular grand
50:51rounds really going from basic science to
50:54the patient and back and forth and it's
50:56absolutely spectacular. Thank you. Yeah
50:59and and great talk. So I think
51:01the main issue and you know as a
51:04clinical investigator in my mind
51:05with all epigenetic therapies is,
51:07is the therapeutic window as you were
51:09saying like how do you actually disrupt
51:11translational or transcriptional
51:13programs that are relevant to the
51:15leukemia but not to the normal tissue.
51:18So for example with this index
51:20particular drug was this as a result
51:23of screening of thousands of molecules
51:25and because as you mentioned it,
51:27it seems like to disrupt only where
51:29it's relevant to the leukemia,
51:30but it's not disrupting the MLL
51:32interactions that are important for
51:34normal hematopoiesis and other functions.
51:36So that how did this transition happen?
51:38It's just a matter of luck or is
51:40it tons of screening of other yeah
51:42you know the old saying
51:43better lucky than good.
51:44I I think that that's what we found here,
51:48meaning that for some reason and
51:50we're looking into this men and is
51:53only critical for MLL wild type.
51:55Obviously this mechanism probably wasn't
51:58developed during evolution for MLL
52:00fusions to localize to certain loci.
52:03So MLL as I mentioned is a monstrous protein.
52:06It has many domains that combine chromatin.
52:09So it's very likely and there's some
52:11data to support this that different
52:13domains or different binding partners
52:16determine localization to different
52:18places throughout chromatin.
52:19And it just so happens in this it
52:21kind of was predicted by some of the
52:23early Cleary work that Menon was a
52:25unique dependency in these leukemias.
52:27Well that's because it's really
52:30intersecting exactly with the
52:34important MLL fusion driven targets.
52:37We don't know the molecular
52:39mechanism for that yet,
52:40but basically it's a long way of
52:42saying we think there's a multi
52:43valent interaction between MLL and
52:45chromatin and Menon is only important
52:47for a subset of that interaction.
52:50Scott, fantastic talk.
52:51And the the question about
52:53the solar cancer part,
52:54so you alluded to that you are it's
52:56great to see that you and other
52:58companies are looking into this aspect.
53:00So, so we know that oxygens are
53:02often deregulated in solar cancer
53:04as well in addition to leukemia.
53:06So what happens to men inhibitor,
53:09the men and MLL inhibitor monotherapies,
53:12do they have any efficacy in
53:14solar cancer or you have to
53:15really using combinations before
53:16you can see something happening?
53:18Yeah, it's good question.
53:20So actually it's a good point.
53:21There are a number of say subtypes of
53:24lung cancer that express HOX genes.
53:26We actually haven't looked
53:27at that probably should,
53:29but in the both in the gastrointestinal
53:32stromal tumors and the ER
53:34positive breast cancer cell lines,
53:36the men inhibitor will slow their growth.
53:40So they have some effect and that
53:42is it looks like through somehow
53:44modulating the ER driven program,
53:46but it's much more dramatic both
53:48the gene expression changes and
53:50the inhibition of proliferation if
53:52you combine the minute inhibitor
53:54and and the cat 6A inhibitor.
53:56So how that works, we don't,
53:58we don't completely understand yet.
54:01It's a way of saying and and predicting
54:04and hopefully get the word out
54:06before all the trials get shut down,
54:08that the single agents might
54:10have some activity,
54:11but I suspect they won't be home
54:13runs and the companies have to
54:15have the wherewithal to actually
54:17move forward to the combinations.
54:19And those of you who've done this
54:22before know that can be difficult.
54:23So we're going to try to get the
54:26word out that you should move
54:28the combinations quickly before
54:30people lose interest.
54:32There's a lot of psychology and
54:34sociology that goes into keeping
54:36the drug companies interested,
54:38so that's a little bit of a soapbox to
54:40say the single agents do something.
54:42The combination definitely looks better.
54:46Manoj, a great talk.
54:48My question is about the specificity
54:50of both the MLL fusion proteins and
54:52the NPM 1C that you alluded to.
54:53And Amar was also asking.
54:56So I think you probably worked on this
54:58on the cancer discovery latest paper,
55:00but most of them also seem to be
55:02overlapping with like say PRC 2
55:04targets or you know are there other
55:06mechanisms you think are relevant to
55:08why these are so tightly overlapping
55:10the fusion proteins and the
55:12yeah, so, so it's a good point.
55:14So they do overlap significantly with
55:16PRC 2 targets and you know as you may
55:20remember the the this has been predicted
55:23for 3 decades from the Drosophila work.
55:25The initial Drosophila work show
55:28Polycom and Trithorax actually are
55:30known to be genetically genetic
55:32antagonists of one another and the the
55:35trithorax slash MLL complex controls
55:38developmental genes that the PRC two
55:40or Polycom complex wants to shut off.
55:43So in absolutely this is A these
55:46proteins complexes MLL probably
55:49Catsix A are inventing.
55:50We've shown this in some other settings
55:53are preventing the Polycom complexes from
55:55coming in and repressing gene expression.
55:57So the the way we think that this
55:59is working is during hematopoietic
56:00development as you go from stem cells to
56:03progenitors to fully developed myeloid cells,
56:06the Polycom complex at least for a
56:08subset of developmental loci are
56:10shutting those programs off and the
56:12MLL fusion won't let them do that.
56:14So they're antagonizing and then
56:17Newt 98 fusions and probably NPM
56:19one are antagonizing PRC 2 section.
56:22Yeah,
56:23awesome. I'm going to bring
56:24it over to you in a second.
56:26We have an online question which I
56:27think you probably partially answered
56:29and that is what is the mechanism of
56:31gene specific targeting of MLL EF9
56:33and similarly what you think is the
56:35underlying mechanism for the gene target
56:37specificity of men and inhibitors.
56:39So for the online person, yeah.
56:40So it's a good, it's a very good question.
56:43We do it. I don't the bottom line is,
56:45is we don't know the answer
56:46to the second part of that.
56:47I mean that's the that's at the moment.
56:50Probably the most critical question
56:52is why is it that Menon's only
56:54important for localization of the
56:56MLL compacts to certain loci.
56:58So Needless to say,
56:59we're looking at various aspects
57:00of those loci to try to understand
57:03what that's what that's all about.
57:04MLL targeting to chromatin broadly has many,
57:07probably has many mechanisms,
57:09some of its direct there's a domain
57:11on MLL that binds to what's called
57:13a CPG island which is upstream of
57:16many transcriptional start sites.
57:17Menon plays a role.
57:19There are other accessory proteins
57:20that play roles.
57:21So I think the cell has just given
57:23itself many options to figure out
57:25where to put MLL and and each of
57:28those mechanisms slightly different.
57:31We have a trainee question.
57:34So my name is trainee,
57:37I just wanted to follow my name used
57:38to be that at some point you graduated.
57:40I just want to follow up on your
57:42comment about combination therapies.
57:44So I specifically wanted to ask is there
57:46any thought that men and inhibition
57:47could convert these resistant like
57:49subtypes especially in the pediatric
57:50setting to a chemosensitive form if
57:52there's thought of combining with
57:54chemo to then re sensitize them and
57:56potentially cure those patients. Yeah,
57:57it's it's a good short answer is
58:00we don't know the answer to that.
58:02I I think there's a lot to be learned
58:05and the beauty of having now multiple
58:08small molecule selective small molecules,
58:11we can do those types of experiments,
58:14but the short answer is we
58:16haven't haven't gotten there yet.
58:20Yeah, great talk.
58:21Have you seen any phenotypic
58:23differences in the fusion
58:24partners with MML or MLL?
58:26You mentioned there's you know,
58:27100 different ones.
58:27Do they all have the same kind of,
58:29you know, break points?
58:30Does it change expression?
58:31Do you see any variability
58:32in the kind of those fusion
58:33partners? Yeah, it's a good question
58:36that's been asked for many decades
58:38and not been answered very well,
58:41at least in patient samples because
58:42it's hard to get enough patient
58:44samples of these subtypes to to ever
58:46really do that experiment. Well,
58:49at least for the fusion,
58:51the different fusion AM LS we've
58:53assessed and that have been
58:54assessed in patients, it doesn't.
58:56It's not clear that the fusion partner
58:58is determining men and responsiveness.
59:02Is the fusion partner influencing
59:04the phenotype of the leukemia?
59:06I think that question still still
59:10open and there's aren't enough good
59:12models to really answer that question.
59:18If you delete the at least for AF9,
59:21if you delete the C turn much of the AF9,
59:23it will no longer be transforming.
59:25So that fusion partner is important
59:29and and in in the AF9 setting,
59:31we think it's important because
59:32that's the anchor to drop to pull
59:34all those other complexes in.
59:35But it's that's a nice simple answer.
59:37It's not that simple because
59:39some of the fusion proteins
59:40don't bind to those complexes.
59:42So what they're doing is, is less clear,
59:47beautiful talk. Thank you.
59:48Obviously there are lots of
59:50other chromatin complexes,
59:52switch, sniff, polychrome, etcetera.
59:54Any any thoughts on those?
59:56Are you looking at any exploring
59:58any of those other chromatin?
01:00:02Yeah, accessibility.
01:00:04Epigenetic complexes?
01:00:05Sure. So we have over time looked at
01:00:11the Polycom complex mostly in leukemia
01:00:15and it for whatever reason and this is
01:00:18going to be right contradictory to how
01:00:19I answered one of my previous questions.
01:00:21The Polycom complex does seem
01:00:24to be important in the continued
01:00:27proliferation of many types of leukemia.
01:00:30How that's working and why hard to know,
01:00:34but the small molecule PRC 2
01:00:37inhibitors or or EZH 2 inhibitors
01:00:40don't have tremendous activity.
01:00:42So whether or not it's the enzymatic
01:00:44activity versus some other part of the
01:00:47complex at least in leukemia anyway,
01:00:49I don't know.
01:00:50And I think that probably it brings
01:00:53up a good point that and we've
01:00:56done this in the past as well,
01:00:57but we have to be careful about
01:00:59the thought process that enzymatic
01:01:02inhibition of a protein in one of
01:01:04these complexes is the same thing as
01:01:06complete interactivation of the protein.
01:01:08It's not and we now seen that
01:01:09many different times.
01:01:10So the enzymatic part of the
01:01:13proteins is important,
01:01:14but there's probably a structural
01:01:16component to this that when you
01:01:18take the protein completely out and
01:01:19this is the same way for EZH 2 the
01:01:22the changes are much more dramatic.
01:01:23So it gets to the question as to what
01:01:25his some modifications are doing and
01:01:27that gets to even more deep and complicated.
01:01:30But so long answer to yes,
01:01:32we've looked at the complexes,
01:01:33we haven't looked much at the
01:01:36chromatin remodeling complexes.
01:01:37Segal Kadosh who some of you may know is
01:01:40at Dana Farber and we just let her do that.
01:01:42She they're doing a lot in that regard.
01:01:46I'm sure they're playing a role here.
01:01:48What but how and what we don't know.
01:01:51We have one last question and after this
01:01:54is actually a session for the trainees,
01:01:56very private was Doctor Armstrong.
01:01:59Yeah, one last question.
01:02:00This is actually related
01:02:01to the fusion partners,
01:02:03but as you know we often get,
01:02:05you know when we do the genetech
01:02:07sequencing we get MLL deletions,
01:02:08MLL mutations,
01:02:09sometimes Trisom 11 or you know PDD.
01:02:14And is your sense that beyond
01:02:16the fusion MLL Fusion's those
01:02:19alterations also have susceptibility
01:02:21to an inhibition or yeah,
01:02:23as you probably know since you've been
01:02:25important in running some of the trials,
01:02:27the MLLPTD subtype of AML for
01:02:31some of the trials is included.
01:02:33There's been an assumption to some extent
01:02:36that they should be responsive the PD.
01:02:38So we have generated some MLL.
01:02:39So MLLPTDS are actually a partial
01:02:41tandem duplication of a part of the
01:02:44MLL protein and that subset of AML is
01:02:46incredibly difficult to treat and it's
01:02:48also found in MD's those mutations,
01:02:52but that almost tells you immediately
01:02:54it's also found in MDSMLL.
01:02:56Rearrangements are not found in MDS that
01:02:58they're probably something different
01:02:59going on in the minute inhibitor
01:03:01doesn't have at least in our PDX
01:03:03models the same type of activity in
01:03:04those PDX as it does MLL rearranged.
01:03:07So I think it gets back to this issue
01:03:10that we've been discussing that in
01:03:12that setting either because of that
01:03:15duplication or otherwise the that MLL
01:03:18Oncoprotein sticks on chromatin through
01:03:20a different mechanism than Menon.
01:03:22So we're got a project looking
01:03:25at exactly that,
01:03:26trying to understand what that is.
01:03:28We haven't looked at the deletions.
01:03:30I wouldn't predict they would be.