Skip to Main Content

"Clinical Research Opportunities in Brain Tumors" and "Molecular Basis of Gut Microbiome-associated Carcinogenesis"

 .

"Clinical Research Opportunities in Brain Tumors" and "Molecular Basis of Gut Microbiome-associated Carcinogenesis"

May 05, 2021

Yale Cancer Center Grand Rounds | May 4, 2021

Antonio Omuro, MD and Seth Herzon, PhD

ID
6567

Transcript

  • 00:00We have two speakers today covering
  • 00:02quite a diverse array of topics,
  • 00:04which is terrific. Both internal
  • 00:05speakers and I encourage people.
  • 00:07If you have questions to
  • 00:09type them into the chat.
  • 00:11And then we'll get those questions
  • 00:13answered when the talks are are finished.
  • 00:15So our first speaker is Antonio Omuro.
  • 00:17You may you may know he he is a professor
  • 00:19of neurology and the chief of Neuro
  • 00:22Oncology here and Clinical Leader Program
  • 00:24leader of the shin of your family,
  • 00:26can bring tumor center,
  • 00:28which is a new program here.
  • 00:30He received his initial
  • 00:31medical training in Brazil,
  • 00:32then worked at Memorial
  • 00:33Sloan Kettering for a while,
  • 00:35and began his faculty career
  • 00:37at University of Miami.
  • 00:38He joined us in 2012.
  • 00:40He's an international leader in their
  • 00:42clinical care and research on brain tumors.
  • 00:45Leading leading pivotal research
  • 00:46programs and treatment of these cancers,
  • 00:48the Genevier family Brain Tumor Center is a
  • 00:50new yellow initiative for the Comprehensive,
  • 00:52multidisciplinary brain tumor.
  • 00:53Karen, perhaps you might hear a
  • 00:55little bit about that from Antonio,
  • 00:57so Antonio, the floor is yours.
  • 00:59Thank you for speaking today.
  • 01:03Thank you very much,
  • 01:04then super like to thank the
  • 01:06organizers for inviting me to
  • 01:07talk to you today and for today.
  • 01:10Specifically, I was asked to
  • 01:11share with you what's happening.
  • 01:13Our division in terms of clinical
  • 01:15trials and how we're tapping into
  • 01:16Yale talent to build our portfolio,
  • 01:18but I would also like to share
  • 01:20with you the state of our fields
  • 01:22and the spirit of almost like an
  • 01:25invitation to even more investigators
  • 01:26and labs to join us in this task.
  • 01:33So today we're going to concentrate
  • 01:35on gliomas and the reason for that is
  • 01:38that they account for the vast majority
  • 01:41of the brain tumors and as you can
  • 01:44see here this is a fight chart showing
  • 01:47all malignant intracranial tumors,
  • 01:48and the vast majority of the patients
  • 01:51have either glioblastoma or other
  • 01:53forms of gliomas which for the
  • 01:55most part our IDH mutants,
  • 01:57which account for grades two and three
  • 01:59others like Thomas and grades too.
  • 02:02In three algorithms, this is 3 or
  • 02:04by semester form as many Germans.
  • 02:07In order rare tumors.
  • 02:09But the bottom line here is that
  • 02:11this even the most common tumor,
  • 02:14which is unfortunately the great
  • 02:16for glioma or glioblastoma,
  • 02:18still is a relatively rare disease
  • 02:20with only three point 1 patients
  • 02:22for each 100,000 people.
  • 02:24So it is again relatively rare disease
  • 02:27Fortunately, but it is, as you know,
  • 02:30a very devastating disease and.
  • 02:32The reason why this is such a charming
  • 02:35diseases that you know the anatomic
  • 02:37location really doesn't help us.
  • 02:40So these are places that presents
  • 02:42with these large tumors with
  • 02:44lots of surrounding edema,
  • 02:46an infiltrative microscopic disease.
  • 02:47These terms are highly vascularized,
  • 02:49so we're at the same time dealing
  • 02:52with an uncle logic disease,
  • 02:54but truly we're dealing with
  • 02:56a neurologic disease as well,
  • 02:58and you can imagine how challenging
  • 03:01it is to manage.
  • 03:03All of these symptoms were still
  • 03:04trying to make a difference in
  • 03:06terms of uncle logic treatments.
  • 03:11Ends reflecting this challenge is
  • 03:13the fact that the only drug that
  • 03:16has shown to improve survival so
  • 03:18far is this alkylating alkylating
  • 03:20agent that is more than 20 years old.
  • 03:23So this is the most dolomite and in
  • 03:25controllers here is saying that the
  • 03:28Missouri might improves both work
  • 03:30for survival and overall survival,
  • 03:32but even the experimental arm
  • 03:34in the pivotal trial,
  • 03:35which was published in 2005.
  • 03:38Survival remained only 15 months for again,
  • 03:41newly diagnosed disease and further
  • 03:43analysis of this data has shown that
  • 03:46this survivor benefit is mostly driven
  • 03:49by tumors that have this afternoon.
  • 03:52Check silence of the Mt gene
  • 03:54promoter by methylation so these
  • 03:56patients with math Laden GMT tend
  • 03:59to respond better to Tim's or mine,
  • 04:02but they account for only
  • 04:04about 30% of the patients.
  • 04:06So for the remainder of the patients.
  • 04:09The only real treatment that
  • 04:11is available is radiation.
  • 04:15And we did try a lot of agents,
  • 04:18and here you're looking at a slide
  • 04:21from 2005 where we were talking
  • 04:23about all of these clinical trials in
  • 04:25glioblastoma and in other diseases.
  • 04:27Testing these novel target therapist.
  • 04:29So we're very excited that for
  • 04:31the first time in would be able
  • 04:33to treat these patients with
  • 04:35therapies that would carry minimal
  • 04:37toxicities and tremendous efficacy,
  • 04:39but as you know, the story was much more.
  • 04:44You know, harder than than what we
  • 04:47originally thought, and one by one.
  • 04:49All of these stars went on to
  • 04:53fail in recurrent disease.
  • 04:55The sad thing is that or maybe the
  • 04:57lucky thing for other diseases is that
  • 05:00the majority of these drugs ended up
  • 05:02being approved for other indications,
  • 05:04but all of that rise in glucose
  • 05:08Thomas have failed.
  • 05:10And more challenging is the fact that
  • 05:12we're not really sure what is it about
  • 05:15the omens that all of these drugs
  • 05:18actually fail one after the other,
  • 05:20is that because we are targeting
  • 05:23the wrong targets,
  • 05:24maybe they're not sufficiently
  • 05:25relevant for Uncle Genesis,
  • 05:27or there are too many feedback
  • 05:29loops and redundant pathways were
  • 05:31now more and more learning about
  • 05:34temporal spatial variations?
  • 05:35Or is it be 'cause these
  • 05:37are the wrong drugs and?
  • 05:39We have problems you know,
  • 05:42of achieving adequate concentrations,
  • 05:43especially for drugs.
  • 05:45They are not very potent.
  • 05:47We do need to have better blood
  • 05:51brain barrier penetration because a
  • 05:53lot of these microscopic disease is
  • 05:56behind an intact blood brain barrier.
  • 05:59Also,
  • 05:59we still don't know how to
  • 06:01select basis for these drugs.
  • 06:03We're still not sure if it should select
  • 06:06based on specific mutations or should
  • 06:08we go through transcription subgroups or not?
  • 06:10Do any selection whatsoever and treat
  • 06:12a large number of patients then then
  • 06:15identify the responders and then go after
  • 06:17the phenotypes that predict response.
  • 06:19So in other words,
  • 06:20regardless of what we do,
  • 06:22we certainly need to improve translation
  • 06:24components within our trials,
  • 06:25improve the science before, during,
  • 06:27and after the trial and this is.
  • 06:30Actually,
  • 06:30paradigm that we have
  • 06:31been following artificial.
  • 06:36So the low hanging fruit is to try
  • 06:39to use the genomic information
  • 06:41that is now widely available.
  • 06:43Only streamers to see if we
  • 06:46can improve our outcomes.
  • 06:47So as you know, global someone was the very
  • 06:51first tumor sequenced by the TSJ effort,
  • 06:54and since then gene sequencing
  • 06:55has become the norm when managing
  • 06:58these patients and here looking
  • 06:59at all types of biome's and these
  • 07:02different colors here represents
  • 07:04the different subtypes of gliomas.
  • 07:06And you have no difficulty to see that.
  • 07:10The genomic signatures are very
  • 07:13distinct across the different
  • 07:15histologies you can see here.
  • 07:17The quintessential signature of the
  • 07:19algal blooms, which is more penalty?
  • 07:22Q coalition ideate mutation Sir
  • 07:25promoter mutation and see I see
  • 07:27and if you put P1 and here is the
  • 07:31quintessential signature of Astros
  • 07:33with guided meditation AT Rex,
  • 07:35Magician or lost interpretive fermentation.
  • 07:37And here's this essential signatures
  • 07:40of global stoma.
  • 07:41Now we start to see Jeff Farm
  • 07:44to fication or mutation Pete.
  • 07:46Then loss or mutation and lots
  • 07:49of formalities in CD case.
  • 07:51So putting those patients now,
  • 07:54arranging them into what kind of pathways
  • 07:57ended up being abnormally disturbers,
  • 08:00we can see the vast majority of
  • 08:04patients follow this cake recipe.
  • 08:06So basically 1000 kinase pathway with
  • 08:10PKU KTM Tor pathway activation and F1
  • 08:13you see also a lot of these patients
  • 08:17with arousing in the T3 pathway leading to.
  • 08:22Abnormalities in senescence and a pop
  • 08:24ptosis and a lot of these patients
  • 08:27having a normality's in cell cycle control.
  • 08:31But then when we put all of these
  • 08:34patients a match to actually which
  • 08:36mutations have a track record of
  • 08:39being drug and what you can see is
  • 08:42unfortunately each of these mutations
  • 08:44is actually very where we're not
  • 08:47being very good at identifying
  • 08:49drugs for those specific phenotypes
  • 08:51we heavily rely on basket trials.
  • 08:53But unfortunately basket trials
  • 08:55typically exclude patients with
  • 08:57brain tumors were left with no
  • 09:00trials or very trials that address.
  • 09:02These questions we do have
  • 09:03some low hanging fruits.
  • 09:05Of course ideas mutation will
  • 09:06talk a little bit more about that,
  • 09:09but again,
  • 09:09the message here is that it is
  • 09:12very difficult to run start
  • 09:14therapy trials of these days.
  • 09:16Because you really need to have
  • 09:18strategies to tackle the rarity
  • 09:20of each of these phenotypes.
  • 09:24And adding to our challenges are how
  • 09:27these tumors evolve overtime and how
  • 09:30they are heterogeneous to begin with.
  • 09:33So this is a patient, for example,
  • 09:36that at diagnosis she was enrolled in
  • 09:39one of my trials of a notch inhibitor
  • 09:42and she had a very typical signature
  • 09:46of astrocytomas with identification
  • 09:48interaxon to 53 mutations and several
  • 09:51potential target targetable abnormalities
  • 09:52with other abnormalities, but.
  • 09:54When this patient again,
  • 09:57she received the nearly diagnosed.
  • 10:01Trial and then when she recovered,
  • 10:03she was operated on again,
  • 10:05even though she had a very small tumor.
  • 10:08And what we found is that all of those
  • 10:11potential target mutations are actually gone.
  • 10:13We're seeing some passengers here.
  • 10:15But the reality that's what's driving
  • 10:17this tumor now is actually probably about
  • 10:20Melanie's at the OBGYN attic level,
  • 10:22and you can imagine that if at this
  • 10:25point in time of her disease we work
  • 10:28to enroll her in a clinical trial.
  • 10:31Most patients do not have another brain
  • 10:33surgery to have another sequence,
  • 10:35so you go to archive tissue and we
  • 10:38would have selected her for trials
  • 10:40that probably were irrelevant for
  • 10:42her at this point in time.
  • 10:44Again,
  • 10:45those are males that we thought
  • 10:47were present were actually gone.
  • 10:50This is another example of potentially
  • 10:54targetable mutations that actually were
  • 10:57very different at the time of recurrence.
  • 11:00And.
  • 11:00Another difficult challenge are
  • 11:03these patients here.
  • 11:05So these are patients that we serve created.
  • 11:08These is a result of the use of the
  • 11:12Mozilla might that can cause mutations in
  • 11:15mismatch repair genes at typically MSH.
  • 11:186 and what happens is that these patients
  • 11:22with mismatch repair defects start
  • 11:24accumulating all of these mutations
  • 11:26and you can imagine that developing
  • 11:29target therapies for these folks.
  • 11:32Is much harder.
  • 11:33And one of the surprising findings
  • 11:36of our studies have been that
  • 11:39these are actually much more common
  • 11:41than we previously thought.
  • 11:47So in moving forward what we're trying to
  • 11:49do is to again improve the science linked
  • 11:52through the early development trials,
  • 11:55so we more and more relying Phase
  • 11:57Zero tries to show us if our drugs
  • 12:00are actually getting into the brain,
  • 12:02especially in areas with intact
  • 12:04blood brain barrier.
  • 12:05We also want to see if the.
  • 12:09The drugs are hitting their targets and
  • 12:12we like to look at the pharmacodynamic
  • 12:14effects in these resected specimens.
  • 12:17Be more and more have we have to
  • 12:19work with their companies to have
  • 12:21basket trials that actually include
  • 12:24patients with our rare phenotypes.
  • 12:26There's a shift towards more
  • 12:28of a newly diagnosed disease.
  • 12:30Be 'cause these are easier patients,
  • 12:32and the genomics information is
  • 12:34actually fresh, and where we're
  • 12:36dealing with recurrent disease,
  • 12:37we typically like to re
  • 12:39sample specials for target.
  • 12:41Therapies, if anything,
  • 12:42at least to exclude the hypermedia phenotype.
  • 12:44And we also like to of course update the
  • 12:46gene sequencing and the Uncle Genic trimers.
  • 12:49Another trend in our field, this try
  • 12:51to target these strong communications,
  • 12:52but that's not an easy task.
  • 12:54And again,
  • 12:55we're going to talk a little
  • 12:57bit more about that.
  • 12:58But the vast majority of trials right now
  • 13:02is actually trying to find alternative
  • 13:06strategies that address more stable targets.
  • 13:09So the low hanging fruit of stable
  • 13:12targets is actually immuno therapies.
  • 13:14So we do know that blue,
  • 13:17blasphemous do grow in a very
  • 13:19human suppressive microenvironment.
  • 13:20And we have identified several
  • 13:22emergent points that seem to be very
  • 13:25important in this disease.
  • 13:27But on top of identifying
  • 13:29the right even checkpoint,
  • 13:30we have the challenges of the
  • 13:33anatomic location itself.
  • 13:34So you can imagine that it's much harder
  • 13:37to trigger him and logical response.
  • 13:40In the brain,
  • 13:41which is,
  • 13:41you know,
  • 13:42traditionally considered the so-called
  • 13:43sanctuary sites for the immune system.
  • 13:46And we have to get these email
  • 13:48responses to act fast because these
  • 13:51are tumors that grow very rapidly
  • 13:54and they cause symptoms and we
  • 13:56don't have the luxury of waiting
  • 13:59several months or years to react.
  • 14:01The benefits of the email of therapies.
  • 14:04And of course,
  • 14:05if you're triggering inflammatory
  • 14:07responses in the brain,
  • 14:09we have to deal with the risks of
  • 14:11new log symptoms and neurotoxicity.
  • 14:14An another important thing is.
  • 14:16That this information could
  • 14:18potentially mimic some aggression,
  • 14:19so managing these patients can
  • 14:21be challenging because we have
  • 14:24to learn to how to recognize,
  • 14:26see the progression versus real
  • 14:27tumor progression on the MRI.
  • 14:34But we did try and here you're
  • 14:36looking at the very first results
  • 14:39of the very first phase one trial
  • 14:41utilizing image checkpoint inhibitors
  • 14:43in global stoma and this was done in
  • 14:47with VMS and in this trial we treated
  • 14:5040 patients both with nivolumab or
  • 14:53two combinations of Nivola Bintulu
  • 14:55map and what we found is that yes,
  • 14:58the target definitely was present
  • 15:00in the majority of patients,
  • 15:02so 60% of the patients had
  • 15:05PDL one expression.
  • 15:06But we didn't see any brain toxicities
  • 15:08which is good and perhaps bad because
  • 15:10this could potentially reflect the
  • 15:12fact that we are not achieving
  • 15:14much and overall survival was very
  • 15:15similar to historical controls,
  • 15:17although some places it seemed to
  • 15:19Mount more of an email response.
  • 15:21But this went on to be tested in
  • 15:24randomized trials and we are now
  • 15:26reporting the final results of
  • 15:28these shows and one by one they
  • 15:30all failed to improve survival,
  • 15:32both newly diagnose and recurrent disease.
  • 15:37So we're not giving up on immunotherapy,
  • 15:39so I think our task now is to try to
  • 15:42send what is that about the brain?
  • 15:45That in spite of PDL one expression
  • 15:48we're not seeing any help from Anti
  • 15:51PD one or anti PDL one therapies and
  • 15:54I think for for this question I think
  • 15:57it is great to have a helping hand of
  • 16:00people that study the CNS immunology
  • 16:03and in this project what we did is to
  • 16:07partner with Doctor David Hoffer and
  • 16:09I'll also Liliana Luca to look at how
  • 16:12can we actually come up with better.
  • 16:16Image checkpoint inhibition that is
  • 16:18relevant for for this Mike environment
  • 16:21and what the heifers lab came up with
  • 16:24is that this image of point called digit
  • 16:27seems to be much more relevant in the brain.
  • 16:31It was very interesting that in
  • 16:34their studies they found a lot of
  • 16:36teacher expression in DBMS and not
  • 16:38so much digit expression in the
  • 16:41quintessential inflammatory disease
  • 16:43which is multiple sclerosis and.
  • 16:45They went on to perform
  • 16:48several studies utilizing,
  • 16:49so sequencing that sort of confirmed
  • 16:52that T cell dysfunction was being driven
  • 16:55by digit in this particular disease.
  • 16:58So to test this hypothesis in
  • 17:00patients with design,
  • 17:02this clinical trial where we are looking at.
  • 17:06A different cohorts of patients prior
  • 17:09to surgery where they will be treated
  • 17:12with either infected or anti PD one
  • 17:14or the combination or placebo and
  • 17:16then these patients will be brought to
  • 17:19surgery and then we will do a tumor
  • 17:22single cell RNA sequencing with an
  • 17:24axe as well as some studies to produce
  • 17:28some spatial validation of the findings.
  • 17:30And there will also follow these
  • 17:33patients longitudinally to see if we
  • 17:35can monitor what's happening in the tumor.
  • 17:38By analyzing the T cells in the periphery.
  • 17:41So it's a very exciting trial.
  • 17:43So I wish we had started the
  • 17:45development of Inter PD one this way
  • 17:47by understanding the science before
  • 17:49going to more or larger studies that
  • 17:52would end up being negative,
  • 17:53but we're very excited about this mechanism.
  • 17:56Action also is important to emphasize
  • 17:58that this combination of anti teachers
  • 18:01and in fact PD one is very hot in
  • 18:03the fields right now as you know it
  • 18:06is already in phase 3IN.
  • 18:07Non small cell lung cancer.
  • 18:09And we're very excited to bring
  • 18:11this trial here to you.
  • 18:16Also, to understand a little
  • 18:18bit more of the immune system,
  • 18:20we need a good models that are
  • 18:24immunocompetent and one of the.
  • 18:26A richness of our environment
  • 18:28here is Doctor City chains work
  • 18:30producing these jam models of global
  • 18:33stomas where he can pretty much
  • 18:35produce avatars for all of these
  • 18:37phenotypes that I just showed you,
  • 18:40and one of the ideas here is to see
  • 18:42how these different phenotypes respond
  • 18:45to these different immunotherapy's.
  • 18:47So this is very exciting data
  • 18:49which again illustrates how we can
  • 18:52concomitantly to the development in
  • 18:54the clinic to also try to understand.
  • 18:57Are these treatments in parallel in the lab?
  • 19:02Now another barrier for.
  • 19:07For the development of effective
  • 19:09even responses is the work
  • 19:12being done by the Iwasaki slab.
  • 19:14So Akiko has been working with
  • 19:17Eric Song and General Thomas,
  • 19:20and she has recently had this nature
  • 19:24paper where they showed that.
  • 19:27There is a defective lymphatic
  • 19:29drainage from the brain that you
  • 19:32can correct utilizing the GFC.
  • 19:34So in her models that the combination
  • 19:36of Veg FC and Anti PD one actually
  • 19:40improves survival and was also
  • 19:42interesting that they also produce
  • 19:44some experiments by injecting anti
  • 19:47PD one directly into the CSF and also
  • 19:51the results seem to be better than
  • 19:54systemic administration of anti PD one.
  • 19:57So this is all giving rise to another
  • 20:00generation of characterizing and
  • 20:02some new compounds that we hope to
  • 20:05bring to clinic in the mid term.
  • 20:10Now also again another important
  • 20:12barrier in Spanish solid tumors,
  • 20:14but particularly in glomus,
  • 20:16is the role of tumor associated macrophages
  • 20:19and how they produce these emails.
  • 20:22Suppressive tumor convergence
  • 20:23and one of the ways that we could
  • 20:27potentially intervene in this was
  • 20:29discovered by an item here at go,
  • 20:32where she's looking at the role of
  • 20:35this little Robo one, signaling
  • 20:37which seems to attract and polarize.
  • 20:40Save Microfridge is in in the
  • 20:42brain microenvironment and
  • 20:43Livingstone my confirming.
  • 20:45And when she did experiments to knock down,
  • 20:48slid to, or to block this pathway,
  • 20:51she achieved better immune responses
  • 20:53and inflammation of anti PD one.
  • 20:55She had a really significant improvement
  • 20:57in survival or in this tumor bearing mice.
  • 21:00So the idea here is now to generate
  • 21:03enter Robo Nanobodies one of the
  • 21:06barriers project would be then
  • 21:08how can we get this number?
  • 21:10At least to penetrate into the brain.
  • 21:13And since she's very resourceful,
  • 21:15she has the answer.
  • 21:17It looks like.
  • 21:18If you block antibodies,
  • 21:20if you use antibodies blocking
  • 21:22this receptor called UNC 5B,
  • 21:24you conserve produce an on demand
  • 21:26blood brain barrier opening,
  • 21:28so this is less a few hours and it's
  • 21:31great for drugs up to 40 kilodaltons.
  • 21:34So the idea here is that if this is
  • 21:37successful, we could combine this.
  • 21:39These agents with many of the
  • 21:41chemotherapies in order target therapies
  • 21:44that we are trying to use to treat
  • 21:47these patients in a more efficient way.
  • 21:50And overcome the problem of living
  • 21:51there countries so very exciting
  • 21:53work that we hope to see more of.
  • 21:57No moving on into.
  • 21:59It's still sticking to the Mockingbird,
  • 22:02but moving on to partnerships with pharma.
  • 22:05One of the our partnerships
  • 22:08is with this drug.
  • 22:11This company called In Pharmaceuticals
  • 22:13and these folks have discovered
  • 22:16a novel receptor with within the
  • 22:18Alpha V beta three integrin that
  • 22:20is started by this FP PMT drug
  • 22:23that seems to have an amazing
  • 22:25activity in their mouse models.
  • 22:27Really melting the mice.
  • 22:29And this was the first.
  • 22:31Now we're now designing the 1st
  • 22:33in human trial here GAIL,
  • 22:36that will start in a couple of months.
  • 22:39But to understand this better we
  • 22:41did bring Yellow Labs into the mix
  • 22:44to better define how is this rug
  • 22:46really working and who are the
  • 22:48best candidates by understand a
  • 22:50little bit more about the effects
  • 22:53on cell invasion signaling networks
  • 22:55and gene expression.
  • 22:56So one of the assets that.
  • 22:59We're realizing in partnership with
  • 23:02under left ankle is looking at these.
  • 23:05The use of his integrated platform,
  • 23:07which is the so called race essay
  • 23:10which is a disrupted analysis of cell
  • 23:13phenotype extremes where he uses
  • 23:15the cell migration as a surrogate
  • 23:18marker of tumor aggressiveness and
  • 23:21and then you can test the multiple
  • 23:24drugs utilizing this essay as
  • 23:26a form of drug screening,
  • 23:28and he's applying this rug.
  • 23:31Your days were very interesting results
  • 23:33and we hope to then identify partners.
  • 23:37Which are the best genomic candidates
  • 23:39and then see if we can optimize
  • 23:42the trial as we go by in reaching
  • 23:45with either best candidates or
  • 23:48potentially novel combinations.
  • 23:50So again,
  • 23:51that's just to illustrate that
  • 23:53it's very important to really
  • 23:56involve our laboratories.
  • 23:57Even in trials that are
  • 24:01being conducted by pharma.
  • 24:03Now sticking again,
  • 24:04not now moving on to other
  • 24:07more stable targets,
  • 24:08and one of them is ideas with patient
  • 24:11and this story came out of Doctor
  • 24:14Kendra's lab where he found it ideas.
  • 24:17Mutations change DNA repair through the
  • 24:19production of two hydroxy obliterate,
  • 24:21which is the byproduct of this mutation,
  • 24:24and this results in sort of brokenness
  • 24:26that then can be targeted by PARP inhibitors.
  • 24:29So he has several clinical
  • 24:32trials of these park inhibitors.
  • 24:34And we are now hoping to see
  • 24:37if this will actually improve
  • 24:39outcomes for these patients.
  • 24:44Also, again sticking to the DNA repair thing,
  • 24:48we recently submitted a United team led
  • 24:51by Mayo Clinic and John Jennifer Correa
  • 24:54in partnership with even colonies.
  • 24:56Doctor Bindra and I.
  • 24:58So we have two projects.
  • 25:00One is trying to optimize MDM two
  • 25:03inhibition for these patients and 80
  • 25:06Rd in ambition for these patients.
  • 25:08And this will again bring two other
  • 25:11Phase 0 slash 1 clinical trials.
  • 25:14To our portfolio, hopefully soon.
  • 25:21Now we don't have time to
  • 25:23review all of our portfolio,
  • 25:25but we do have partnerships with
  • 25:27industry for opening other tries
  • 25:29to fill in gaps in our portfolio.
  • 25:31Doctor Blundin has activated the
  • 25:33Agile trial which is a multi drug
  • 25:36multi arm clinical trial that is
  • 25:39happening worldwide so we have access
  • 25:41to these drugs for our patients
  • 25:43and have a bunch of other choice.
  • 25:45But the theme here is really to focus on
  • 25:48early therapeutic development and then
  • 25:50participating inside cooperative groups.
  • 25:52Please, for those extremely rare phenotypes,
  • 25:55for example,
  • 25:56byref mutant craniopharyngioma switch,
  • 25:58again very difficult to find patience
  • 26:01and for those we do need to partner
  • 26:05with other places around the country.
  • 26:09And I could go on and on talking
  • 26:11about all of the years signs
  • 26:14that is going in brain tumors.
  • 26:16I select a few stories that
  • 26:18are closest to clinic,
  • 26:20but all these people in this picture
  • 26:22and many others that I'm not even
  • 26:25mentioning today are producing amazing
  • 26:27size that we can actually use into
  • 26:30our portfolio and bring it in anymore.
  • 26:32Let's say intelligent trials
  • 26:34ranging from data science,
  • 26:35junior imaging and all sorts
  • 26:37of therapeutic approaches.
  • 26:39So in conclusion,
  • 26:40so we're lucky enough to have this
  • 26:42unique breath of scientific expertise.
  • 26:45Our focus is really on investigating
  • 26:47shaded trials that are home grown
  • 26:49and our other focuses on early stage
  • 26:52development with former partners,
  • 26:54but also bringing along our
  • 26:56own labs email collagen,
  • 26:57a repair have emerged as leading teams,
  • 27:00but here there we have many patents about,
  • 27:03although many are not ready
  • 27:05for complication and need a
  • 27:08lot of help for development.
  • 27:10We certainly need more work
  • 27:12on existing available drugs,
  • 27:13for example coming from Seatac
  • 27:15and Pharma and a lot of work in
  • 27:17functional genomics so that we
  • 27:19can figure out finally how to
  • 27:22target these undruggable targets.
  • 27:23So that I would like to finish
  • 27:26by thanking all of the people.
  • 27:28So when we talk clinical trials,
  • 27:30really the merit is all of others of the
  • 27:34labs of the all of the infrastructure.
  • 27:37I would also like to acknowledge
  • 27:40our division attendings and aips,
  • 27:42or actually managing treating
  • 27:43these patients in the trials.
  • 27:45I would like to thank again the Cito staff.
  • 27:49They're going through rough times,
  • 27:50but right Decker is navigating and it's
  • 27:53going to get us out of this situation.
  • 27:57A big thanks to the PRC reviewers
  • 27:59'cause one of my hats is actually as
  • 28:02the Pearcey chair and we we acknowledge
  • 28:04along with Barbara Burtness that there
  • 28:07was a lot of work that goes into
  • 28:09this and that I would like to thank
  • 28:12them publicly at this opportunity.
  • 28:14Lots of things.
  • 28:15So why CCI that help us with
  • 28:17investigating share clinical trials.
  • 28:20All of the people that have been
  • 28:22enabling this research and finally
  • 28:24a big thank to the YCC and Smile
  • 28:26leadership with more Pickens.
  • 28:28Kevin versus Kevin Beans loosely,
  • 28:30and neither will all understand importance
  • 28:33of our clinical trials portfolio.
  • 28:35Last but not least, again,
  • 28:37I would like to thank the show her
  • 28:39family for their generous gift.
  • 28:41In fact, then I'm not going to
  • 28:43talk about this today because we're
  • 28:45still working on the details,
  • 28:47but the word is out of the receive a
  • 28:49generous gift from that foundation,
  • 28:51and we're hoping to put together a
  • 28:54nice program that will again enable
  • 28:55and expand on our research efforts.
  • 28:58Thank you very much and I'll take
  • 29:00some points if you have time.
  • 29:03OK, thank you very much Antonio.
  • 29:05Very interesting work.
  • 29:06Are there any questions that people
  • 29:09want to enter into the chat?
  • 29:11While we're waiting,
  • 29:12I have a quick question.
  • 29:13You mentioned this idea
  • 29:14of opening up the bread.
  • 29:16The blood brain barrier
  • 29:18by targeting a molecule.
  • 29:20Is it worth going back to some
  • 29:21of the earlier drugs that
  • 29:23weren't terribly effective to
  • 29:24see that whether or not that
  • 29:26might help them work better?
  • 29:28Yes, I think there is a whole list of drugs
  • 29:33that perhaps will need to be revisited.
  • 29:37Although most of these drugs would
  • 29:40actually be again in rare phenotypes,
  • 29:42because I think those are,
  • 29:44we still need to select them
  • 29:47by those specific mutations.
  • 29:48The problem of copy number remains
  • 29:51regardless of flipping connectors.
  • 29:52I don't think blood brain barrier penetration
  • 29:55was the reason why we couldn't target
  • 29:58EGFR amplification or Pete and loss.
  • 30:01I think that is a different question,
  • 30:03but if we are to even answer
  • 30:06those we still need.
  • 30:08This kind of approach,
  • 30:09'cause it makes our life so much easier.
  • 30:14Are there any questions from the audience?
  • 30:23I was also struck by the lots of different
  • 30:26mutations upon recurrence. He showed.
  • 30:31What is that thought to
  • 30:33be due to is just so high
  • 30:36perforation rate of these tumors.
  • 30:38Yeah, well, I think so.
  • 30:40First of all, these tumors are very
  • 30:42heterogeneous to begin with, right?
  • 30:44So these are guns that are
  • 30:46were there to begin with,
  • 30:48but it looks like the treatment process
  • 30:51ends up eliminating a lot of this so
  • 30:54called cancer associated mutations.
  • 30:56Another unknown mutations emerge and also
  • 30:58a lot of these are actually epigenetic.
  • 31:01Changes.
  • 31:01So there's a whole line of research
  • 31:04trying to then understand this and
  • 31:07more canals are interested in in that
  • 31:10line of research and other labs to
  • 31:12see how we can target these tumors
  • 31:15at recurrence that are sort of,
  • 31:17you know,
  • 31:18very simple from a genomic standpoint,
  • 31:20but not so simple at the epigenetic level.
  • 31:23Well, thank you very
  • 31:25much. Very interesting.
  • 31:26We have to move on to the second speaker.
  • 31:31In our second stewartii climb down
  • 31:34from the Hill from Science Hill,
  • 31:36is Seth hairs on who's the Milton
  • 31:39Harris professor of Chemistry
  • 31:41received his PhD at Harvard and then
  • 31:44post Doc at University of Illinois,
  • 31:46and he's interested in natural
  • 31:49product's particular products that
  • 31:52affect the synthesis or damaged DNA.
  • 31:55And he's received numerous
  • 31:56multiple Young Investigator awards
  • 31:57and working with Jason Crawford
  • 31:59is a terrific collaboration.
  • 32:01Looking at them,
  • 32:02the metabolites made by the human
  • 32:04microbiota and identified some of
  • 32:06them that actually damaged DNA and
  • 32:08therefore contribute to cancer.
  • 32:09So Seth, we're looking forward
  • 32:11to hearing about your work.
  • 32:13Thank you.
  • 32:15OK, thanks Dan,
  • 32:17thanks for the introduction.
  • 32:19And thanks to all to
  • 32:21everyone for the invocation.
  • 32:22Comment for attending the lecture.
  • 32:26I will talk today about work we've
  • 32:27been doing in the human microbiome,
  • 32:29but actually ignore it.
  • 32:33Just calling on ever. Snap because.
  • 32:38Cave against drug resistant TMZ resistant.
  • 32:40GBM that we're very excited about but
  • 32:43that will be a story for another day.
  • 32:49And so right. So today I'll talk
  • 32:52about a project that's been ongoing
  • 32:54in my group for about 6 years.
  • 32:56And we've been looking to understand
  • 33:00the molecular basis of a carcinogen
  • 33:03carcinogenic phenotype that was
  • 33:06observed from certain gut bacteria so.
  • 33:09I'll go through sort of the sequence
  • 33:12of events to kind of outline sequence
  • 33:15of discoveries to outline the problem,
  • 33:17and so in 2006 this was the paper that
  • 33:21set off a lot of interest in this area.
  • 33:25Eric Oswald and coworkers identified
  • 33:28certain strains of commensal and
  • 33:31pathogenic E coli that had a biosynthetic
  • 33:35gene cluster known as the CLB cluster.
  • 33:38So by that I mean that gene genetic
  • 33:42locus contains the coding for enzymes
  • 33:46that make a secondary metabolite and
  • 33:49he took these CLB containing bacteria
  • 33:53and did a transient infection.
  • 33:56HeLa cells with them and then
  • 33:58looked at the effect on the cells,
  • 34:02and he found that they underwent
  • 34:04cell cycle arrest.
  • 34:05Meglos cytosis and using a comet assay.
  • 34:09Another Gamage to XD he saw that
  • 34:12they accumulated double strand
  • 34:15breaks in their DNA.
  • 34:17And so this is a very interesting phenotype.
  • 34:19It's not the first time.
  • 34:22Microbes, have, you know,
  • 34:23produced Gina toxins,
  • 34:24but it was was was a very interesting
  • 34:28example and I'll come to in a second wait.
  • 34:31Why it's attracting so much attention?
  • 34:34Subsequent to that report,
  • 34:37there's been numerous studies
  • 34:40trying to ascertain whether or not
  • 34:43there is a role for these bacteria
  • 34:47in colorectal cancer formation
  • 34:49and from the same group in 2010.
  • 34:54It was shown that in in
  • 34:58intestinal loop models of.
  • 35:01My step or infected with CLB bacteria
  • 35:04they observe DNA damage in vivo.
  • 35:07They observed gamma H2 X they observed
  • 35:11increased mutations in the HP RT&TK
  • 35:14loci and then also hyperproliferation
  • 35:16following exposure to the bacteria.
  • 35:19So they seem to be driving tumorigenesis
  • 35:22and then there were subsequent
  • 35:24studies following up looking at
  • 35:27similar types of in vivo effects.
  • 35:30So using IL.
  • 35:32Knockout mice,
  • 35:33it was shown that infection with
  • 35:36these bacteria leads to a higher
  • 35:39rate of tumor formation and then
  • 35:42there were three groups that did.
  • 35:45Meta analysis of of ***** samples
  • 35:48from from CRC patients and what
  • 35:51we find is that about 60 to 70% of
  • 35:55CRC patients have these bacteria
  • 35:58and that's and that's versus about
  • 36:0120% in the healthy population.
  • 36:02And the other sort of bit is that
  • 36:05the preponderance of these bacteria
  • 36:07tracks with the severity of the cancer,
  • 36:10so people with more advanced CRC were
  • 36:12at the high end of that correlation,
  • 36:15whereas people with early stage
  • 36:17CRC were more at the lower end.
  • 36:21And so it wasn't really until last
  • 36:23year that a causal relationship
  • 36:26was unequivocally established.
  • 36:28There were two studies from mayor
  • 36:31and then Boxtel and Cleavers an
  • 36:34in the Cleavers study.
  • 36:36They generated an organoid and
  • 36:40infected that organoid chronically for
  • 36:43about three or four months with the
  • 36:48CLB positive bacteria and what they
  • 36:52showed is that you get the mutational
  • 36:56signature transformation and proliferation.
  • 36:59We also find that that mutational
  • 37:04signature is found.
  • 37:06Enriched in in CRC patients as well,
  • 37:08and so the mayor study came
  • 37:11to similar conclusions,
  • 37:12and essentially these two papers you know,
  • 37:15this is a rare example in the microbiome
  • 37:18where you actually establish causation.
  • 37:20So these two papers brought this
  • 37:23phenotype to the two sort of
  • 37:26a causal level and what my lab
  • 37:28has been trying to do of course,
  • 37:31is understand the molecular
  • 37:32basis for all of this OK and so.
  • 37:39Oswald, in his initial paper,
  • 37:42had done a series of very nice and you know,
  • 37:48robust control experiments to establish
  • 37:51that this genotoxic phenotype.
  • 37:54Is due to the final biosynthetic
  • 37:56product product of the CLB cluster.
  • 37:59In other words, if one modifies any
  • 38:01of the enzymes in the CLB pathway,
  • 38:04you lose this genotoxic phenotype OK,
  • 38:07and so the implication then is that
  • 38:10it's the fully elaborated molecule.
  • 38:13That is the active toxin,
  • 38:15not something in route to
  • 38:18another another product.
  • 38:20And we call that molecule Coley bactine.
  • 38:23And. So the field basically
  • 38:26set out to do what we do best,
  • 38:30which is isolate compounds and the
  • 38:33classic way of isolating natural
  • 38:35product secondary metabolites is to
  • 38:38culture the Organism of interest.
  • 38:40In the case of bacterial
  • 38:43secondary metabolite,
  • 38:44you might grow it in liquid culture,
  • 38:47growing on scale.
  • 38:50Extract start to fractionate by
  • 38:52HPLC and then we typically do.
  • 38:55It's known as activity guided fractionation,
  • 38:57where you're essentially testing each of
  • 38:59these fractions for a particular phenotype.
  • 39:02And then you keep purifying and testing
  • 39:05and purifying testing until you get to
  • 39:08a single compound and you characterize it.
  • 39:11The problem is that this
  • 39:13approach does not work for Kohli.
  • 39:15Bakhtin, OK,
  • 39:16so the molecule is very unstable.
  • 39:19It is very difficult to get the bacteria
  • 39:24to express the CLB pathway ex vivo.
  • 39:27And what we find is that because of the.
  • 39:32You know,
  • 39:33primarily anaerobic environment of the gut.
  • 39:35The molecule actually undergoes oxidative
  • 39:38degradation when you attempt to isolate it.
  • 39:41Sort of on the bench under air,
  • 39:44and just to to.
  • 39:47Give you an example of how
  • 39:48challenging this is.
  • 39:49This is not work from our own laboratory.
  • 39:51This is a group at at Berkeley and scripts.
  • 39:55They've been pursuing Cali,
  • 39:57backed in in the isolated this.
  • 39:59Molecule here in 2019 they obtained 50
  • 40:03micrograms from a 2000 liter fermentation.
  • 40:07If anyone can imagine that,
  • 40:10so we're talking about literally
  • 40:13vanishingly small quantities.
  • 40:17And they they advanced this molecule
  • 40:19as a candidate calling back,
  • 40:21and unfortunately this was derived from a
  • 40:23triple mutant Frankenstein like bacteria,
  • 40:25and I wrote a commentary.
  • 40:26If you're interested on this at
  • 40:28the general thinking in the field
  • 40:30is this is probably not relevant
  • 40:32to the genotoxic phenotype.
  • 40:34But the point is,
  • 40:35these are the links that people
  • 40:37are willing to go to to try
  • 40:39and isolate these molecules.
  • 40:44And so, how do we approach this?
  • 40:47So, as Dan mentioned,
  • 40:48we've been collaborating with Jason Crawford.
  • 40:51Jason is one of the leaders
  • 40:54in understanding Kohli,
  • 40:55backed in biosynthesis.
  • 40:56And So what we've been doing is really
  • 40:59taking knowledge from the biosynthetic
  • 41:02pathway and trying to infer what types
  • 41:05of substructures might be within Kohli.
  • 41:08Backed in itself and how
  • 41:10those might interact.
  • 41:12With DNA and so one of the sort
  • 41:15of models that came out of these
  • 41:18biosynthetic studies is that you
  • 41:20have these fully linear products
  • 41:23offloaded from the PKS assembly line.
  • 41:26There's a searing protease that
  • 41:29removes this residue and blew this ACL.
  • 41:32Asparagine residue.
  • 41:34That generates a primary amine,
  • 41:36and once you format that can start
  • 41:39to wrap up and ultimately lead to
  • 41:42this compound on the bottom here,
  • 41:44which has a cyclopropane ring in
  • 41:47conjugation with with it with the Alpha,
  • 41:50beta unsaturated imming and for
  • 41:52those in the audience that have
  • 41:54worked with Gina toxins,
  • 41:56you know that these electrophilic
  • 41:58cyclopropane's are not uncommon.
  • 41:59This is a sort of a pharmacophore that's
  • 42:02found in a variety of genotoxic natural.
  • 42:05Products and so this was,
  • 42:07you know,
  • 42:08sort of very logically following
  • 42:09from that type of precedent.
  • 42:11The problem is that the problem is this.
  • 42:14No one had isolated these imines.
  • 42:16No one had any spectroscopic data on them.
  • 42:19All we had was this.
  • 42:21This kind of this mechanistic hypothesis.
  • 42:24And so we set out to make it,
  • 42:26and I'm not going to have time to go
  • 42:29through all of the synthetic work that
  • 42:31that went into developing these roots.
  • 42:34But the key steps are shown here.
  • 42:36So we start from this linear precursor
  • 42:38and what we find is that if we
  • 42:41concentrate this down from dilute acid,
  • 42:43we can get this.
  • 42:45Carbon and nitrogen to condense
  • 42:47onto the ketone. You found this.
  • 42:49Finally this image.
  • 42:50We then do a bond formation
  • 42:52deprotect the Bach route to get
  • 42:54to this compound on the left,
  • 42:56we isolate this as as it's TFA salt.
  • 43:00But if you neutralize this,
  • 43:02it's snapshot and so this
  • 43:04carbon attacks this ketone,
  • 43:05you lose water, any formatting.
  • 43:08And.
  • 43:11The assay that we use, that's 'cause
  • 43:13it's nice to give us a lot of detail.
  • 43:15It's inexpensive, it's fast, is is.
  • 43:19Taking linearized plasmid DNA
  • 43:21incubating with the molecule,
  • 43:23running a denaturing gel,
  • 43:24and basically if you look at
  • 43:27the right hand lanes here,
  • 43:29you see these streaks on the gel
  • 43:32going down to about 100 animal or what
  • 43:35that tells us is that at 100 animal
  • 43:38or concentration of this compound,
  • 43:41we're getting extensive
  • 43:42degradation of the DNA.
  • 43:44These are smaller fragments that
  • 43:46have higher mobility on the gel.
  • 43:49And so that was very exciting to us.
  • 43:52And you know,
  • 43:54we hypothesized again that it
  • 43:56was this nucleotide addition to
  • 43:58the cyclopropane that was leading
  • 44:00to this degradation of the DNA.
  • 44:03And so to probe that in a little
  • 44:06bit more detail,
  • 44:07we made a couple of control compounds.
  • 44:10So the first one.
  • 44:12Was this dimeric structure up
  • 44:14top here and so the hypothesis is
  • 44:18that if this is alkylating DNA,
  • 44:21perhaps we can induce two fold
  • 44:24alkylation and perhaps we can then
  • 44:27detect and interstrand crosslink?
  • 44:30And when you incubate with that compound,
  • 44:33indeed you can see down here Crosslink Band.
  • 44:36This corresponds to our positive
  • 44:39control for crosslinking cisplatin.
  • 44:41And then the other thing we did was
  • 44:43we made a negative control where we
  • 44:46took that cyclopropane and converted
  • 44:49it to the gem dimethyl substituent.
  • 44:51The hypothesis being if the
  • 44:53cyclopropane is truly involved,
  • 44:54this compound should be inactive
  • 44:56and going up to half millimolar.
  • 44:59We don't detect any damage in our
  • 45:01assay and so without characterizing
  • 45:03the product without even having
  • 45:05isolated the natural products,
  • 45:07we were able to sort of formulate this
  • 45:10proposal for how these things might be.
  • 45:13Might be alkylating DNA.
  • 45:17And. We are, you know,
  • 45:20we sort of at that point.
  • 45:22Got stocks so that was around
  • 45:232018 when we had identified this.
  • 45:25You know, this DNA reactive
  • 45:27substructure in the molecule.
  • 45:28We knew that it was incomplete.
  • 45:30In other words,
  • 45:31there were other functional groups,
  • 45:33other rings and things with in Cali bactine.
  • 45:35But we didn't know what they were.
  • 45:38And as I mentioned in the beginning,
  • 45:41the you know the classical
  • 45:42isolation approach is not very
  • 45:44successful in this in this context,
  • 45:46and so we were stuck.
  • 45:49Until this paper came out,
  • 45:51and so this is also from the Oswald Group.
  • 45:55They did a beautiful experiment
  • 45:57where they took the collie,
  • 45:59backed in producing bacteria,
  • 46:01grew them up in liquid media,
  • 46:04added exogeneous DNA,
  • 46:05isolated that DNA following incubation,
  • 46:07and ran a denaturing gel and the point
  • 46:10is that they observed interstrand
  • 46:13crosslinks in that DNA that was
  • 46:16exposed to these bacteria OK and so.
  • 46:19I have a note here at the bottom to remind
  • 46:22me you know if you're paying attention.
  • 46:25The original female phenotype
  • 46:26was double strand breaks.
  • 46:28Now I'm talking about cross links.
  • 46:30Those two lesions are are intimately related,
  • 46:32and I'll talk about that at
  • 46:34the end if there's time.
  • 46:36But basically we're very excited about
  • 46:38this paper because you can imagine that
  • 46:40Kohli bactine is entrained in that crosslink,
  • 46:43right?
  • 46:43You know,
  • 46:44if that's what's causing the crossing at that
  • 46:46point wasn't completely certain that either,
  • 46:49but.
  • 46:49Assuming that it is all we have to do,
  • 46:52all we have to do is isolate that
  • 46:54crosslink can characterize it.
  • 46:57And so at this point in time, you know.
  • 47:0290% of the biosynthetic
  • 47:03pathway had been mapped out.
  • 47:05We had a very good understanding of what
  • 47:09went in of the amino acids that went into
  • 47:13the pathway and where they ended up.
  • 47:16Following sort of offloading
  • 47:18of the biosynthetic products,
  • 47:20and so for example, it was known through
  • 47:24work that Jason did very early on,
  • 47:27and then others that this amino
  • 47:30cyclopropane comes from methionine,
  • 47:32and these thiazole rings
  • 47:34derived from cysteine OK.
  • 47:37And So what?
  • 47:38This allowed us to do is conduct
  • 47:41essentially isotope labeling experiments
  • 47:44where we generated auxotrophic strains,
  • 47:48either deficient in methionine
  • 47:50or cysteine biosynthesis,
  • 47:52and then supplemented those cultures
  • 47:55with C13 labeled amino acid OK.
  • 47:59And so we can take the wild
  • 48:02type strain the oxygen riffs
  • 48:04with their amino acids incubate.
  • 48:06Here we're using linearized puck 19 DNA.
  • 48:09We can run a gel to verify
  • 48:11that we got crosslinking,
  • 48:13and then we can try and isolate that
  • 48:16cross link and characterize it,
  • 48:18and one of the things that's worth
  • 48:21pointing out is that to do these assays
  • 48:24were talking about 250 microliters
  • 48:26of culture versus 2000 liters.
  • 48:28You know using the old?
  • 48:30The sort of the old fashioned method,
  • 48:33and so to give you an idea of
  • 48:35what the data looks like and
  • 48:37why we do this isotope labeling.
  • 48:39I'll show you this slide.
  • 48:41So for example,
  • 48:42we can spot these ions that I've
  • 48:44marked in colored boxes here,
  • 48:46and the top chromatogram is
  • 48:48the wild type strain.
  • 48:50And what you can see in the Sistine
  • 48:53Auxotroph the middle graph is that
  • 48:55those ions are shifted by three units,
  • 48:58and so that's very useful to us
  • 49:00because it tells us two things.
  • 49:03One is that that Ion is probably
  • 49:05contains Poly back in or the vestiges
  • 49:08of Cali Bactine and then two.
  • 49:10It contains one thiazole residue,
  • 49:12there was one cysteine incorporated
  • 49:15into that unit.
  • 49:17And we can play the same game
  • 49:18with the methionine auxotroph.
  • 49:20So here we get a shift by plus four.
  • 49:22So that tells us there's 11 amino
  • 49:24cyclopropane and tells us it's
  • 49:26also related to Cali Bactine.
  • 49:27OK, so this was the initial work that we did.
  • 49:31We had to carry out a lot more
  • 49:34labeling in order to get the full
  • 49:38structure assignment and So what we
  • 49:41did was we generated a series of we
  • 49:44had our cysteine and methionine auxotroph.
  • 49:46We generated steering and glaci Knox
  • 49:49Atros because those are also incorporated
  • 49:52into the into the natural product.
  • 49:54And then we also did Universal labeling C
  • 49:5813 labeling with glucose an 15 labeling.
  • 50:01With ammonium chloride.
  • 50:03And we can run the same experiment
  • 50:06where we incubate with the DNA,
  • 50:09isolate the crosslink,
  • 50:10digest it,
  • 50:11analyze it by 10MM S and we can then
  • 50:14see different shifts in those ions.
  • 50:17And this data turned out to
  • 50:19be very powerful for us,
  • 50:21because without isolating the compound
  • 50:23without getting any spectroscopic data,
  • 50:25we can,
  • 50:26we can glean an incredible amount of insight
  • 50:29into the molecule structure.
  • 50:30So from the glucose labeling,
  • 50:32we get a shift by 37 units.
  • 50:35That tells us, of course,
  • 50:37that it has 37 carbons.
  • 50:39Ammonia shifts by 8 units,
  • 50:41we have eight nitrogens we can see
  • 50:43that in the methionine auxotroph,
  • 50:45and I'm talking about a higher
  • 50:47molecular weight ion here.
  • 50:49At the top we get a shift by 8
  • 50:51carbons and so that told us that
  • 50:53we had two of these cyclopropane
  • 50:55residues or what was left of them.
  • 50:58Two thiazole rings based on A6
  • 51:00carbon shift in this in the
  • 51:02Sistine Extra if you get the idea,
  • 51:04and so we can basically tease out a
  • 51:07lot of structural data to sort of
  • 51:09see what pieces are need to be put
  • 51:12together here to make the molecule.
  • 51:14And so at any rate we found this
  • 51:18higher molecular weight ion at 956.
  • 51:21Using all that data,
  • 51:22we were able to fit it to this structure
  • 51:25here and so it contains one adenine
  • 51:28residue and have explicitly drawn the
  • 51:31Adenine without connectivity to the base,
  • 51:34because at the time that we did this,
  • 51:37we couldn't specify where
  • 51:39it was bonded to adenine.
  • 51:41We now know that that's in three,
  • 51:44but had one adenine on the right hand side.
  • 51:48You have a cyclopropane that's still intact,
  • 51:51OK.
  • 51:51And then you've got the rest of the core
  • 51:54molecule sort of linking it together,
  • 51:56and so it's it's almost C2 symmetric,
  • 51:58it's it's a hetero dimer.
  • 52:00It's not quite C2 symmetric.
  • 52:01If you look carefully at these
  • 52:03thiazole rings they have different
  • 52:04appendages in different connectivity,
  • 52:06but it's very close OK,
  • 52:08and this structure fit RMS
  • 52:09data within one PPM,
  • 52:11so we're very excited about that.
  • 52:15And so if that is simply a mono
  • 52:18adenine addict and we're getting icy,
  • 52:21else, presumably there's a
  • 52:22dinucleotide add up and we went,
  • 52:25and we were able to find the dyad an addict.
  • 52:29OK, and this fits, fits within 1/2 PPM error.
  • 52:33OK, and so working backwards,
  • 52:35if that's the dyad, and in an act,
  • 52:39then this is the structure of Kohli,
  • 52:42bactine on the bottom here.
  • 52:44OK, and so we've got two cyclopropane's.
  • 52:48And in the middle we have this
  • 52:521/2 dicarbonyl residue OK.
  • 52:55There's a detail here
  • 52:56which is worth mentioning,
  • 52:58which is that this is this kind of compound.
  • 53:02On the bottom is what we characterized what
  • 53:05we expect based on the biosynthetic pathway.
  • 53:08Is the self amino ketone at the top,
  • 53:11but we've done work that shown that
  • 53:14this thing is unstable towards aerobic
  • 53:16oxidation to an Alpha keto imine,
  • 53:19and then hydrolysis 212 die ketone and
  • 53:22so working under air on the bench.
  • 53:25This is this is the compound that
  • 53:27you would have expected to get.
  • 53:29And still so no ones isolated
  • 53:32calling back in yet.
  • 53:33And so how do you prove
  • 53:35the structure assignment?
  • 53:36We can go back and try and make it,
  • 53:39and so we spent some time developing a
  • 53:42synthesis of the molecule and it was.
  • 53:45It was not straightforward because of
  • 53:47its instability, but we could make it.
  • 53:50And then we can do an LCMS coinjection and
  • 53:54we see that has the same retention time.
  • 53:58It has the same tandem Ms as
  • 54:01the as the natural material,
  • 54:04and then finally we did a crosslinking
  • 54:07assay where we basically ran that same
  • 54:10experiment that we ran with the bacteria,
  • 54:13except replace the bacteria
  • 54:16with our compound.
  • 54:17And so this thing will crosslink
  • 54:19add up to about, you know,
  • 54:22down to about 500 nanomolar.
  • 54:24And then we can do the tandem Ms
  • 54:26analysis of those cross links.
  • 54:28And so let me explain what's on this slide.
  • 54:32So when we do the bacterial experiment
  • 54:34where we treat the DNA with the bacteria,
  • 54:37we can isolate the crosslink.
  • 54:39You then run tandem Ms,
  • 54:41you get a whole list of ions,
  • 54:43primary and secondary and tertiary ions
  • 54:46that you see from those crosslinks and so.
  • 54:49You know the argument is.
  • 54:51If we're making the same molecule
  • 54:53that the bugs are making.
  • 54:55Are synthetic compound oughta interact
  • 54:57with DNA in the same way and it ought to
  • 55:01blow apart in a mass spec in the same way?
  • 55:05And So what this plot shows on the
  • 55:07X axis or all of the ions that
  • 55:10we found in the tenant and Ms of
  • 55:13the bacteria derived crosslinks,
  • 55:16we see all of those ions with
  • 55:19our synthetic material.
  • 55:20And the Y axis is simply the experimental
  • 55:23minus theoretical error for those
  • 55:25ions using this synthetic material.
  • 55:27And so the point is we get all the same
  • 55:30ions that we get when we use the bacteria.
  • 55:33They're all within with the
  • 55:35exception of 1 within two PPM OK,
  • 55:38and so we don't have an NMR of Cali bactine
  • 55:41to compare to.
  • 55:42But we can say that structure
  • 55:45that we made interacts with DNA.
  • 55:47It crosslinks DNA and then it blows
  • 55:49apart in attend imeson exactly.
  • 55:51In indistinguishable fashion.
  • 55:53And So what about this?
  • 55:56I see LDS be, you know,
  • 55:58apparent contradiction,
  • 55:59so there's been a lot of debate in
  • 56:02the literature between you know,
  • 56:04debating the mechanism of action
  • 56:06because Oswald had originally observed
  • 56:09DNA double strand breaks using a comet
  • 56:11assay and then came along and said,
  • 56:13no wait, it's cross links,
  • 56:15and for any of you that are familiar
  • 56:18with these repair pathways,
  • 56:20you know that these two phenotypes
  • 56:22are intimately linked, right? And so.
  • 56:24When you start to repair an ICL,
  • 56:27you actually form a DSP that
  • 56:29leads to activation of HR and so
  • 56:31you're going to see gamma, H2, X.
  • 56:34You're going to see streaking
  • 56:35in your comment essay,
  • 56:37and so the two phenotypes
  • 56:39are entirely consistent.
  • 56:40And we actually identified another pathway,
  • 56:43which is just a spontaneous pathway.
  • 56:46So it's well known in the old sort
  • 56:49of Gina toxin literature that N
  • 56:513 adenine addicts are unstable
  • 56:54towards depurination.
  • 56:56And if we run our crosslinking assay,
  • 56:59we sort of.
  • 57:00We modify the assay to be able
  • 57:03to sort of get at this data,
  • 57:05but this is the conclusion is
  • 57:08essentially that these these Icl's
  • 57:10undergo a slow, deep urination.
  • 57:12And then there's a second elimination
  • 57:14of the phosphate that occurs to
  • 57:16lead to a single strand break.
  • 57:18And you can imagine then you know,
  • 57:20in tandem with the repair pathways
  • 57:22and also other alkylation lesions.
  • 57:24Eventually you're going to get those
  • 57:27single strand breaks close enough to
  • 57:29each other to get a double strand break.
  • 57:32And so this brings us to where
  • 57:34we're at in the project, and so.
  • 57:39What we're currently doing is working
  • 57:41with this molecule on the bottom here.
  • 57:44And this is not Kohli backed in itself.
  • 57:47It is a analog of Kohli Bakhtin.
  • 57:50The differences are highlighted in green,
  • 57:52and so the dye ketonen Kohli backed
  • 57:54in on the top here is very unstable.
  • 57:58You can't work with this compound,
  • 58:00it would be, you know Suffiecient
  • 58:02asked to try and use this in a
  • 58:05series of essays to examine it.
  • 58:07Sort of cellular activity.
  • 58:09And so we made what we call the
  • 58:12Dez di ketone analog on the bottom
  • 58:14and working with Christian Jobin
  • 58:16at the University of Florida.
  • 58:18Christians been basically
  • 58:20taking this compound.
  • 58:21Through all of the essays that have
  • 58:24been run by by Oswald and Box Tone
  • 58:27Cleavers and others looking at the
  • 58:30genotoxic phenotype of the bacteria and
  • 58:33seeing if the molecule recapitulates
  • 58:35it and up until now, it seems too
  • 58:38so we see we get activation again.
  • 58:41My page 2X.
  • 58:43Fancy D2.
  • 58:45What he's doing right now is basically
  • 58:47looking to see if this induces the same
  • 58:50type of mutational signature that one
  • 58:52gets with the bacteria and that will be,
  • 58:55you know, sort of the end.
  • 58:57The end point.
  • 58:58Hopefully you know it will
  • 58:59be a positive result,
  • 59:01but whatever it is that will be
  • 59:04sort of the endpoint for this.
  • 59:07And so I just acknowledge all
  • 59:08the people that did the work.
  • 59:10I'll just go through this
  • 59:12quickly to save time.
  • 59:13But this is my collaborator Jason.
  • 59:15Many of you know him.
  • 59:17Many people from my group contributed
  • 59:19to this project over the years,
  • 59:20acknowledged the NCI and Yale for funding.
  • 59:23Thanks again for the invitation
  • 59:24and I'm happy to stay on and
  • 59:27take any questions you have.
  • 59:32Thank you Seth, very interesting.
  • 59:34It makes me appreciate
  • 59:35that I'm a microbiologist.
  • 59:36It's not nowhere near as hard
  • 59:38as being a chemist, I think.
  • 59:40Are there other questions for process?
  • 59:49I want question of course,
  • 59:51DNA damaging agents can cause cancer,
  • 59:54but there also used to treat cancer.
  • 59:58Is there any thought? Is there
  • 60:00possibility of using these compounds
  • 01:00:01therapeutically? Absolutely yeah, that's
  • 01:00:03something we're very excited about,
  • 01:00:04so I didn't get into it.
  • 01:00:07You know, one of the challenges that
  • 01:00:09we in my group is always the chemistry.
  • 01:00:12You know, the chemistry work is
  • 01:00:14a general chemistry to make these
  • 01:00:15molecules is very robust and we started
  • 01:00:18to characterize them with Ranjeet,
  • 01:00:20Bindra slab and he's found
  • 01:00:22that in bracket two mutants,
  • 01:00:23these things are hyperactive and so that's
  • 01:00:26the immediate direction we're going in.
  • 01:00:28Long term, we're looking to see if we can,
  • 01:00:31you know, optimize the properties of
  • 01:00:33these molecules a little bit more.
  • 01:00:34You know it's more than just another
  • 01:00:37crosslinker because I didn't get into it,
  • 01:00:39but there's a mechanism by which
  • 01:00:40we can gauge the activity of the
  • 01:00:42molecule and potentially target it,
  • 01:00:44and so there's a lot.
  • 01:00:46Yeah, there's a lot that we can do,
  • 01:00:48and that's sort of the phase
  • 01:00:50that we're entering into with
  • 01:00:51the project traffic we are after
  • 01:00:531:00 o'clock, so if people
  • 01:00:55have other questions for Seth,
  • 01:00:56just email him I'm sure. Be happy too.
  • 01:00:59Talking thank both speakers for
  • 01:01:01really stimulating talk today. Thank
  • 01:01:04you thanks everyone.