Pathology Grand Rounds: May 18, 2023 - Jian Jin PhD

May 18, 2023

Information

Discovery of Novel Degraders and Development of New Approaches to Target Undruggable Proteins - by Jian Jin, PhD

ID9939

To CiteDCA Citation Guide

00:00So my name is Qingyan.
00:02It's my great pleasure to
00:04introduce our ground on speaker,
00:06Doctor Chen Jin from Mountain Sinai.
00:09Doctor Chen Jin currently is a Mountain
00:12Sinai endowed Professor in Therapeutic
00:14Discovery and he also direct the
00:17Center for Therapeutic Discovery.
00:19He's also a Co Leader of Cancer
00:23Clinical Investigation program
00:24for Tisch Cancer Institute,
00:26which is an NCI designated cancer.
00:31The microphone is on.
00:38It's on. It is on.
00:42Maybe I need to be closer. OK. OK.
00:46So I think I just skipped that part.
00:49Now I just want to give you
00:52some background of Doc Tianjin.
00:54Doc Tianjin received his best
00:56degree from the University of
00:58Science and Technology of China.
01:00Which is one of the best universities
01:02in China because I also went there too.
01:06And then he did his PhD training
01:09at the Penn State and did a
01:11one year postal in Ohio State.
01:13Then he was recruited to GSK.
01:16He stayed there for 10 years before he
01:20was recruited by UNC in 2008 as a social
01:24professor And then 24 to 14 he moved to.
01:29Mountain Sinai as a as a full professor
01:33and he stayed there since and he has
01:36country or not and he's initially
01:38with his discovery in epigenetic drug
01:41discovery and more recently he's
01:43more interested in doing degraders.
01:47He has published more than 200 papers
01:50and have more than 70 pattern.
01:53Pattern.
01:55And he has advanced 5 compounds
01:58to clinical trials and one of them
02:01it's has been FD has been approved
02:04by in the US and Japan.
02:07This is a drug called the Dapro to stat
02:10Hypoxo inducible factor polyhydrosis
02:13inhibitor for treating anemia.
02:17He's well funded 50 ones and
02:19many other grants and he couldn't
02:22count how many counts he has.
02:24And he is also currently a member
02:29of Dem PB study section for NIH,
02:33and he's also a cofounder of a degree
02:37the company called the Cogen and
02:40he has been inactive to National
02:42Academy of Inventors in 2022.
02:46He's going to tell us about his
02:48discovery of Nova degraders and
02:50development of new approaches to
02:53target undrugable proteins today.
02:54And without further ado,
02:56please join me to welcome Dr.
02:57Tianjin.
02:58All
03:08right, So I hope everybody
03:09can hear me. OK, bye.
03:14OK, Getting closer. Better. OK. All right.
03:18So I'm gonna try to use my cursor here, so.
03:22People on the zoom can actually see the
03:25arrow here and what I'm pointing to.
03:29Thank you very much team for very
03:33generous introduction and for inviting
03:36me and delighted to be here and and so
03:40very much enjoyed talking to everybody
03:42so far and pretty much look very much
03:45look forward to the rest of the day.
03:47So I'm going to try to be a a rock
03:50star today. Wearing the dark glasses.
03:54But seriously, my eyes are a
03:57little bit sensitive to the light
03:59and so dark glasses help.
04:01Hopefully this is okay with everybody.
04:04So it's my great pleasure,
04:06a pleasure to talk about my laps,
04:11recent progress on discovering
04:13of novel degraders,
04:15and development of new approaches to Target.
04:19And drug bar proteins.
04:28So I may need a on the outside,
04:33not quite sure. Click on the sitting
04:36area here, Okay. All right. Thank you.
04:43So here's my conflict interest
04:46disclosure as Jim mentioned.
04:48I am a cofounder of Cogin,
04:51a San Diego based biotech which is
04:55dedicated to developing normal degraders.
05:00So as some of you may know,
05:03my my labs have taken a target class
05:06approach to generating selective
05:08inhibitors of histomestal transferases,
05:11HM T's, for well over a decade.
05:16In collaboration with the Structure
05:18Genuine Consortium SGC Toronto,
05:20we have discovered a number of
05:24novel and selective HMT inhibitors,
05:28some of which have been widely used
05:31by research community such as the G9A
05:35GLP inhibitors UNC638 and UNC642,
05:38the easy issue inhibitor UNC1999
05:42and the type 1:00 PM T.
05:45Inhibitor MS-23,
05:47so I will not talk about them today.
05:52Since 2014, my lab has also been
05:56active in discovering novel small
05:59molecule degraders including
06:02Protex for oncogenic proteins.
06:05For some of you who are less familiar
06:07with the field with this field,
06:10Protag which stands for
06:12protelysis targeting Chimera.
06:14Is a hydro a hydro bifunctional small
06:17molecule with one end that binds
06:21to ESV light upicane ligas and the
06:24other moiety binds to the protein of
06:27interest by simultaneously binding
06:31the the ESV ligas and protein of
06:36interest protax induce our brains.
06:39The the E3 ligas into close proximity
06:43of the protein of interest,
06:45leading to selective Poly recognition
06:48of the protein of interest and its
06:51subsequent degradation at a proton.
06:56Since the protect concept was first
07:00reported by Craig Kuz and Radisha in 2001,
07:05numerous significant advancements.
07:08Have been made in this field,
07:11some of which are highlighted here
07:13and summarized in this review article.
07:17In particularly,
07:20this field have seen explosive
07:23growth over lot 8 years and
07:28more than 20 protests have been
07:30advanced to clinical development
07:35so since 2014.
07:37My lab has discovered normal degraders
07:40of epigenetic targets such as WDR 5 is,
07:44H2, PMT 5, NSD 3, NSD 2:00 PM, PRC one.
07:50We have also developed novel degraders for
07:54other oncogenic proteins such as kinases AKT,
07:58CDK 46, EGFR, Mac, and ALK.
08:05In addition, we have developed a
08:08number of new technologies for
08:10advancing the target degradation,
08:13target protein degradation field.
08:15So today I will talk about our WDR
08:195 and they should the greater work.
08:22First to give you a flavor for our
08:26protect discovery effort followed by
08:28the new technologies we have developed
08:31to target and drug ball proteins.
08:37So our WDR 5 degree
08:41work was done in collaboration with
08:44Greg One's lab at University of
08:47North Carolina at Chapel Hill and
08:49a new Agua's lab at Mount Sinai.
08:52So this work was spearheaded by three
08:55extremely talented young scientists.
08:57She Feng Yun, She's a former.
09:01Instructor in my lab.
09:03She's currently have independent Pi
09:05position in Fudan University in China
09:09and don't really a former post out in
09:12Greg Von's lab and Jatish Contour.
09:17He is a former joint post out in
09:19a news lab and my lab is currently
09:22a senior post out in a news lab.
09:25So why do we want to develop a
09:28WDR 5 degrader?
09:30And WDR 5 is an important UNCLE protein,
09:34but it is not an enzyme so it is
09:39a important scaffolding protein
09:42and acts as a functional subunit
09:45of MML mesotransferase complex.
09:48WDR 5 is critical for a 3K four
09:53methylation SL as well as MML complex
09:56mediated regulation of gene transcription.
10:00WDR Five also interacts with CIMIC.
10:03The interactions between WDR 5 and
10:06its binding partners are essential
10:09for sustained oncogenesis in MML
10:12range leukemia and in solid tumors
10:15such as pancreatic, pancreatic,
10:18ductal adenocarcinoma pedac,
10:21a number of small molecule inhibitors
10:24that blocks the protein protein actions
10:26between WDR 5 and its binding partners.
10:29Have been developed.
10:31However,
10:32this WDR 5 PPI inhibitors exhibit
10:37modest cancer cell killing effect
10:39and the lack in vivo efficacy.
10:42Likely due to this PPI inhibitors
10:46block only some but not all of
10:49WDR five's on cogenic functions,
10:52so we therefore pursuit
10:55pharmacological degradation of WDR 5.
10:58As a novel therapeutic strategy for
11:01treating WDR 5 dependent tumors,
11:06So we used O ICR 9429 as the WDR
11:115 binder because O ICR is the
11:14best known WDR 5 PPI inhibitor,
11:18which is highly potent and selective
11:21for WDR 5 O ICR was previously
11:24developed by Raymond Olivar's lab at.
11:27The Ontario Institute for Cancer Research
11:29has the name of OICR in collaboration
11:33with Structured Genuine Consortium.
11:36Based on the crystal structure
11:38of the WDR 5 OICR binary complex,
11:41we identified A solving exposed
11:45region shown here and design,
11:48synthesize and evaluate a initial set
11:52of compounds which contain various
11:54linkers and E 3 legacy ligans.
11:57From this study,
11:58we identified MS33 as the initial
12:02lead which contained this relatively
12:05relatively long linker and this
12:08classic Wehl 1 Wehl ligand.
12:14So we solved a high resolution crystal
12:17structure of WDR Five MS33 and Wehl Elongan,
12:22see Elongan B VCB ternary complex.
12:25Which is the first crystal
12:27structure of any WDR 5 protect
12:30history Ligus ternary complexes.
12:33So as illustrated here,
12:35the linker of MS33 was relatively
12:42extended in the ternary structure in
12:47the ternary complex and MS33 induced
12:51limited protein protein interactions.
12:54Between WDR 5 and Wedgel.
12:57So based on this critical
13:00structure insights, we design,
13:01synthesize and evaluate another set of
13:04compounds which contain much shorter
13:08linker linkers and simultaneously
13:11enhanced bonding to both WDR 5 and Wedgel.
13:16So from this study we identified
13:20MS-67-A highly effective WDR 5 degrader.
13:24Which contain this very very short
13:27linker and this modified WDR 5 binder
13:31and this mesolated VHL ligand.
13:34We also developed MS-67N as a negative
13:40control of MS-67 which contain the
13:45identical WDR 5 binder and the linker
13:48for the dyster isomer of the VHL ligand.
13:51Which abolishes the binding to VHL.
13:55So we also saw the high revolution
13:58crystal structure of WDR 5MS67 and VCB
14:02ternary complex which confirmed MS-67
14:07induced much more extensive protein
14:10protein interactions between WDR 5 and
14:14VHL and enhanced binding productivity.
14:19The enhanced binding correctivity
14:21between WDR 5 and which L induced
14:24by 67 was also confirmed using
14:28isothermal titration calorimetry
14:32and 67 but no not O ICR or 67 N
14:36totally and a selectively degraded
14:38WDR 5 in number of MML range range,
14:43leukemia cell lines and
14:44impedex cell lines in a time.
14:49VHL natilation and prezone dependent manner
14:5567 but not OSCR or 67 N effectively
15:00suppressed transcription of WDR 5 regularly.
15:03Genes in RN6 studies and effect
15:07of 67 significantly overlapped
15:09with that of WDR 5 knockdown
15:1467 but not OSCR or 67 N.
15:17Effectively reduced chromatin bonds,
15:20CMIC and MML complex components
15:24and 67 but not 67 N decreased H3K4
15:30dimethylation in both Western Black
15:32analysis and in Chipsique studies.
15:37Phenotypically 67 but not OACR
15:41or 67 N effectively suppressed
15:44in vitro cell growth and.
15:47Induced cell cycle arrest and
15:49apoptosis in number of MML range
15:52leukemia cell lines as illustrated
15:54here and also in PDEX cell lines and
15:58not showing here and importantly 67
16:02but not OACR significantly inhibit
16:05tumor tumor growth in vivo and
16:09improved survival in multiple in
16:12vivo mouse models including this.
16:15MML re enriched AML PDX model
16:19even though the much higher drug
16:22levels were achieved for OICR
16:25than 67 in tumor samples.
16:28We also established PKPD relationship
16:31for MS-67 for this in vivo models
16:36and Greg Moss lab and my
16:39lab also discovered MS-40.
16:41A novel CRBN recruiting WDR 5
16:45degrader which effectively degraded
16:48WDR five in a concentration time
16:52CRBN and UPS dependent manner.
16:57Interestingly we find using a mass back
17:01based global proteomic studies we find MS-40.
17:07Can effectively degrade not
17:09only WDR 5 but also IKZF one,
17:13which is the CRB and new substrate.
17:15We subsequently confirmed using Western
17:18Black analysis that MS-40 and effectively
17:22indeed effectively degraded CRB and
17:26new substrate IKZF one and three,
17:28but not GSPT one in addition to WDR 5.
17:34So next we developed 2 control compounds.
17:38MS-40 and TWO which effectively
17:41degraded IKZF ONE and Three but not
17:45WDR 5 and MS169 which degraded WDR Five
17:50with a similar potency as MS-40 but
17:54did not degrade IKZF ONE and Three.
17:57Interestingly,
17:58we find MS-40 which degrade both WDR
18:025 and IKZF 1/3 was more effective.
18:07In suppressing the proliferation
18:09of MML range leukemia cells,
18:11then MS-40 and TWO alone,
18:15which degraded IKZF 1 THREE only, ALL,
18:20MS169 alone which degraded WDR five only,
18:25and as expected,
18:27the cold treatment of MS-40 and TWO and
18:31MS169 displaced similar effectness as MS-40.
18:36So in addition,
18:39IMS 40 but not 40 and two all 169
18:43significantly inhibit tumor growth in vivo.
18:46Even though all three compounds were similar,
18:50drug levels were achieved
18:52for all three compounds.
18:54So taken together,
18:56this results suggest pharmacological
18:59degradation of a WD R5 as a
19:02novel serial pedic strategy.
19:04Is superior to pharmacologic logical
19:07inhibition of protein protein
19:09actions between WDR 5 and its
19:12binding partners for treating WDR 5.
19:14Dependent tumors do degradation of
19:18WDR 5 and IKZF 1/3 and could be more
19:22effective than degradation of WDR
19:245 or IKZF 1/3 alone in suppressing
19:27the proliferation of MML rearranged
19:31leukemia in vitro and in vivo.
19:33Another key take away from this study
19:36is the ternary complex structure
19:39based design which is extremely where
19:42in the protect field is a powerful
19:45approach and can lead to highly
19:47effective protect degradation and
19:50lastly the degradation of CRBN new
19:53substrates by CRBN recruiting protects
19:56needs to be monitored very carefully.
20:00And such NEO, St.
20:01neo St.
20:02degradation could potentially be
20:05exploited to yield more effective
20:07anti cancer therapeutics.
20:11So now I'm going to talk about
20:13discovery of easy two inhibition.
20:15So canonically easy two is the
20:18main catalyst subunit of Hollycomb
20:21repressive complex 2P R C2 which
20:24catalyzes A3K27 trimethylation
20:27and mediating gene repression.
20:30An easy issue is overexpressed
20:32in many cancers including TNB or
20:34triple negative breast cancer.
20:36TNBC and it's a high expression level,
20:40correlates with the poor prognosis and
20:44knocked out easy issue effectively
20:47inhibit the growth of breast cancer cells,
20:51including TNBC cells. However,
20:53all easy issue inhibits are ineffective.
20:57In suppressing the provision of TNBC
21:00cells even though they effectively
21:04reduced H3K27 trimesolation mark,
21:07we therefore pursued development of
21:09easy shoot degraders to Pheno copy anti
21:13tumor effect of easy to knock down.
21:19So in collaboration with Roman
21:20Parsons Lab at Mount Sinai,
21:22we discovered the first easy
21:25shoot selective degrader.
21:27MS1943 So this project
21:29was spearheaded by Anjima,
21:32a former poster in my lab
21:34alias a former poster in Ramon
21:36Parsons lab and Kwongsu park
21:38instructor in my lab currently.
21:44So in in contrast to the WDR 5 protax
21:48degraders I just talked about it,
21:51MS1943 is not a protag.
21:54Is the hydrophobic tag based
21:57degrader which links a selective
22:01easy issue. Remember with this bulky
22:05hydrophobic hydrophobic elementing group.
22:09OK and as illustrated here MS1943 is
22:13highly selective for easy issue and
22:16some of you probably know Craig Ku.
22:19Craig Kus is also a pioneer of.
22:23The hydrophobic tag based degrader approach.
22:27His lab published the first hydrophobic
22:31tag tag based small molecule degraders
22:35of helo tag fusion proteins in
22:38Nature Chemical Biology in 2011.
22:42So back to EH2M S 1943 an effectively
22:48degraded EH2IN multiple TMPC
22:50cell lines as illustrate here.
22:53And in contrast,
22:55two is issue inhibitors which
22:57were ineffective in inhibiting
22:59the growth of TNBC cells.
23:02Our issue DEGRADER MS1943 effectively
23:05suppressed the growth of the growth
23:09in multiple TNBC cell lines and 1943
23:14was hourly by available in mice and in
23:18TNBC cell line xenograph model 1943.
23:22Significantly inhibit tumor growth
23:24in vivo and had no effect on the on
23:27the body weight of the treaty mice
23:30with PKPD relationship established.
23:34So recently our collaborator Greg Wong's lab,
23:37Greg Wong at University of North
23:41Carolina at Chapel Hill and his
23:43post out June Wong discovered his
23:46issue have a novel non canonical
23:48function in activation of uncle genes.
23:51By binding CMIC and P300 through its
23:57hidden transactivation domain and so
24:01this non canonical oncogenic function
24:05differs from the well known canonical
24:09gene repression function of PRC two.
24:13So to target both canonical and the non
24:18canonical oncogenic function of is H2.
24:20SharePoint used you know that in
24:24collaboration with Gregg Von's
24:26lab at UNC discovered MS177A.
24:28Novel CRBN recruiting is H2
24:32protect degrader which is highly
24:33selected for is H2 as shown here
24:39177 totally degraded is H2 in a
24:44time CRBN and UPS dependent manner.
24:49And 177 also totally degraded CMIC in a
24:54Crbn easy shoe and UPS dependent manner
24:59and phenotypically 177 totally inhibit
25:02the proliferation of MML range look,
25:06AML cell lines and primary patient cells.
25:10It was much more effective than easy shoe
25:15inhibitors in inhibiting the growth of.
25:18The proliferation and the tumor tumor
25:21genesis in MML ranged AML cells and
25:26177 also effectively induced apoptosis
25:30in an easy issue dependent manner.
25:34And importantly 177 significantly inhibit
25:38tumor growth in vivo and prolonged
25:41survival in multiple in vivo mass models.
25:45Including this MML ranged AML PDX model
25:50with PKPD relationship established
25:56so in collaboration using 177 Greg
25:58Watts lab and my lab also and covered
26:02a similar non canonical function
26:04of easy issue in multiple myeloma,
26:08an activation of uncle gene
26:11where binding of cimic and P300.
26:15Where Sue the hidden transactive
26:18activation domain of EH2 and we show
26:23MS177 can effectively target both
26:26canonical and a non canonical function
26:29of EH2 and inhibit the growth of multiple
26:34myeloma cells in vitro and vivo.
26:36So lastly using MS177.
26:41Ling Tai's lab at University
26:43of North Carolina Chapel Hill,
26:45Greg Watts lab at UNC and my lab also
26:48and discovered a novel non canonical
26:51function of easy shoe in prostate cancer.
26:55So easy shoe finds both AR and AR splice
27:00wearing AR-7 ARV 7A constituently
27:04active AR variants enriched in advanced
27:08castration resistant prostate cancer.
27:10Where the transactivation domain
27:14promoting Uncle Genesis and Crpc
27:18growth in mutual and in evil and we
27:23show MS177 can effectively target
27:27both canonical and non canonical
27:30oncogenic functions of E day 2IN Crpc.
27:39So the key, the key takeaways from
27:42these studies are first pharmacological
27:45degradation of the issue but not
27:48pharmacological inhibition of the
27:50issue could be a effective short
27:53periodic strategy for treating TNBC,
27:55MML range of leukemia, multiple
27:58myeloma and advanced prostate cancer.
28:02In addition to the another take
28:04away from this from this studies is.
28:07In addition to the protect technology,
28:10the hydrophobic tag based approach
28:12which have been understudied and
28:15under appreciated by the field and can
28:18lead to degraders that are already
28:21by available and advocacious in EVO.
28:24So well this degraders and this
28:27technologies are very promising.
28:32The conventional protect
28:34approach cannot be utilized.
28:36To target and drug work proteins,
28:39which lack a small molecule binders
28:42as a small molecule binder of the
28:45target protein is needed for the
28:48traditional protect approach to work.
28:50So to target and drug work proteins
28:55including transcription factors TFs,
28:57we developed two novel approaches,
29:00First bridge protect,
29:02second TF protect and TF Dub tag.
29:06So now I'm going to talk
29:08about these two approaches.
29:11So we hypothesized and drug
29:13war proteins which lack
29:17lacks a small worker binders could be
29:21targeted by breached protect if this
29:25and drug war protein interacts with
29:28another protein termed bridge protein,
29:31which have has a small molecule binder.
29:35So by exploiting this bridge protein
29:38we're linking a small molecule binder
29:41of this bridge protein to a ES3 ligas
29:45ligand with appropriate link link linker
29:48bridge protein and induces close proximity
29:52between the ubiquitin machinery and to
29:56this and drug protein bridge protein
29:59complex and could lead to a selective.
30:03Of polyclination,
30:04of preferential polyblination and degradation
30:08of this and drug protein over bridge protein.
30:14OK, so we selected cycling D1 as the
30:18first target for this bridge protect
30:22approach for the following reasons.
30:25So the first second D1 is the
30:29talk cancer drug target ranked by.
30:33Damat however,
30:34it is end druggable as it's lack
30:37a small molecule binder.
30:40Second,
30:40it's well known second D one former
30:44protein complex with CD46 and
30:47highly potent and selective CD46
30:49inhibitors have been developed.
30:52So by testing all of our CD46 cortex,
30:57we identified MS-28 as the first.
31:01Which protect of cycling D1 which
31:05contain this Hubble site clip as
31:09a CDK 46 binder linked to a Wechao
31:13ligan where a short linker.
31:15So this work was spearheaded by Yang
31:19Xun assistant professor in my lab and
31:22we are June and Leah Leah Rin and both.
31:26Both of them are PhD student in the lab.
31:28So Yan did most of the country
31:31work and you and Leah did all the
31:34biological studies for for this work.
31:40So MS-28 for for gradually degrade
31:44cycling D1 over CDK four and CD
31:48CD6 and did not change MRA level
31:52of cycling D1, CDK 4 and CD6.
31:57The cycling D1 degradation induced
32:00by MS-28IS dependent on VHL,
32:04CDK 6 and UPS and MS-28
32:09can induce cycling D1,
32:13CDK 6MS28 and VHL quanary complex formation
32:21and our cycling D1 the greater MS-28.
32:24Is a superior to the the parent CDK
32:2846 inhibitor PABO cyclip and known
32:32CDK 46 degrader BSJ which degree CDK
32:3746 but not cyclin B1 in suppressing
32:41the proliferation in multiple non
32:44small cell lung cancer cell lines.
32:47We also applied this rich cortex strategy.
32:52To target PRC One components
32:55of PMI One and Room 1B, well,
32:58the EED is the core component of PRC 2.
33:03EED also interacts with PRC One components,
33:07PMI one and Room 1B.
33:09So we aimed to develop a EED binding
33:14protect that can preferentially degrade
33:17PMI One and Room 1B over EED indeed.
33:22We discovered MS147 and the
33:27first P RC1 bridge protect,
33:30which is the way child recruiting
33:34and EE D binding protect.
33:37So this work was spearheaded by Kwansu Park,
33:40an instructor in the lab,
33:42and Lee Hui Chin,
33:43a former post out in the lab and MD
33:46Cab Bear APHD student in the lab.
33:48He just actually successfully
33:50defended his PhD.
33:51And Kwansu and MD did the
33:54biological studies and Lee Hua
33:55did chemistry for this work.
34:00So MS147 preferentially degraded
34:03PRC one components BMI one and
34:07room 1B and selectively reduced
34:12H2A Lysing 119 monoclination which
34:15is catalyzed by PRC 1 / P RED.
34:19And the other PRC 2 components is H2
34:24and SUZ 12 and H3K27 trimethylation
34:28which is catalyzed by PRC Two
34:33and the PRC one degradation induced
34:36by MS147 is dependent on Ed,
34:39VHL and UPS and our PRC
34:45One bridge protect MS147.
34:48Is a superior to the parent Ed Ed two
34:53to six and the known PRC 2 degrader
34:57Protech 2 developed by Astrozeneca
34:59in suppressing the proliferation
35:02in multiple cancer cell lines.
35:07So now I'm going to briefly talk about our
35:11TF Protech and the TF dub tech approach.
35:15So as you know many.
35:19Transcription factors are and druggable
35:23due to the lack of suitable small
35:26molecule binding pockets and therefore
35:29this Tf's cannot be targeted by
35:33the traditional pro tech approach.
35:36So to target and druggable oncogenic
35:40Tf's Weiwei's lab at Howard Medical
35:44School and my lab developed TF pro tech.
35:48By conjugating a DNA oligar nucleotide
35:52which is specific to the TF of
35:56interest to a ESV ligas ligand,
35:58in this case VHL ligand,
36:00where a click action,
36:02so the resulting DNA oligar nucleotide and
36:06and VHL ligand conjugates simultaneously
36:09binds the TF of interest and the VHL
36:14ESV ligas this induced proximity.
36:17Leads to selective ubiquitation
36:20of the TF of interest and its
36:24subsequent degradation at Prozone.
36:26So we have we developed a 2 proof
36:29concept TF protects which effectively
36:33degraded NF Kappa B&amp;E to F respectively.
36:39And we are not the only group developed
36:42the TF protect technology and in fact.
36:46Three papers, including ours,
36:48were published around the same time in 2021,
36:52and so this Craig Ku's Trav
36:56Tech paper describing A keynote
36:59A chemo genetic approach,
37:01was published first, and then this
37:06this paper on a legal protag by
37:09Hao JF One's lab and was published
37:13in advance Advanced Science.
37:17In shortly after our papers
37:19was published in Jax,
37:23So similar to the TF protag approach,
37:26Weiwei's lab and my lab also developed
37:30TF dub tech as a general platform
37:34for stabilizing and drugable tumors
37:37suppressive Tf's by hijacking a
37:40deal pickiness a dub so briefly.
37:44We conjugated ADNL organ nucleotide,
37:47which is specific to the target TF,
37:50to a small molecule ligand of a deopinase,
37:53a DUB, in this case OTU B1 ligand which was
37:59previously developed by Danny Morris lab.
38:01We are a click a click reaction,
38:04so this resulting DNA organ
38:08nucleotide OTU B1 ligand conjugate
38:11simultaneously binds that.
38:13Target TF and the OTV one dub.
38:18This induced proximity lead to
38:21selective deuption of the target TF
38:26and its stabilization and we have
38:29we developed the three proof concept
38:31TF dub packs which effectively
38:33stabilized tumor suppressors FOX
38:38O3AP53 and IRF 3 respective.
38:43So lastly, I'm just going to very
38:46briefly mention our Keep One work.
38:49As many of you know out
38:52of 600 plus E3 legacies,
38:54only very limited of them have
38:57been harnessed for targeted protein
38:59degradation with the CRBN and we gel
39:02being utilized most extensively.
39:05So we demonstrated the Call 3 E 3
39:08like us Keep 1, can be harnessed.
39:11For protective element by using
39:15potent selective and non covalent
39:18ligand of keep one which was
39:21previously developed by Glaxosmus 1.
39:23So we developed
39:27MS-83 approved concept Keep one
39:30recruiting PRD three, PRD 4,
39:32Protech which degraded PRD four and PRD
39:35three more durably than the well known.
39:39BRD this well known CRB and recruiting
39:42BRD 234 portag DBAT One MS-83 also have a
39:48superior selectivity profile to dbat 1.
39:51Interestingly, 83 selectively degraded
39:54BRD 4 short isoform also over BRD
39:594 long isoform in MBA MB 231 cells.
40:02So we hope this work expands.
40:05The limited toolbox for targeted
40:08protein degradation.
40:12So in addition to the bridged Protag,
40:15TF Protag and TF Dubtech technologies
40:19and the key point work I just mentioned,
40:23we in collaboration with Weiwei's lab
40:25also developed fully caged Protag and
40:29Optoprotag for selectively targeting cancer
40:32cells over normal cells and TeleTech.
40:35For selective devolution of telemeric
40:39binding of telemeric repeat binding factors,
40:43in collaboration with May Hathaway's lab at
40:46University of North Carolina at Chapel Hill,
40:48we also developed a chemo genetic based hydro
40:53bifunctional deaccillators and accillators.
40:56And in the interest of time,
40:58I will not talk about this work today.
41:01So with that.
41:03I thank all our collaborators
41:05for their contributions,
41:07in particularly Greg Wong and his lab at UNC,
41:11a new AGUA slab at Mount Sinai and Alan
41:15Tarka's lab at Arkansas for the WDR 5
41:19Protect work and Ramon Parson's lab,
41:23Samir Parak Lab and Anas Gusiani's
41:26lab at Mount Sinai for the easy
41:29H2 DEGRADER MS1943 work.
41:30And Greg Wang's lab and Ling Tai's
41:33lab at UNC for the easy to protect
41:37MS177 work and you assume for at
41:41college and for for his help on
41:44the 2nd D1 bridge protect work.
41:46And of course when he and his
41:49lab members for the TI Protag,
41:52TF DUB Tag,
41:54TeleTech 40K H Protag and Optoprotag
41:57work and.
41:58Then Shan Chen and his lab for
42:02conducting pretty much all our
42:05mass back based proteomic studies.
42:09So I also thank my current and former
42:12lab members for their contributions.
42:15So mention some of their names
42:17during the talk and thank funding
42:20agencies for the financial support.
42:22Last but not least,
42:23thank you very much for your kind attention.
42:26Happy to answer the questions you may have.
42:34I'm for a good question. Actually
42:41I'm curious to
42:45be mentioned
42:51with the
42:58pro that.
43:08Right. And Don that's a great question.
43:11The the, the new substrate issues
43:16if if you like all opportunity,
43:20it's mainly through the CRBN ligands
43:24of the the CRBN history like this
43:27and to date the no new substrates
43:30have been identified for VHL.
43:32And so, so therefore we
43:35shall recruiting protects.
43:37So far we have not seen the new
43:39substrate issue and having said
43:42that the I should know that more
43:45than 20 protects have been advanced
43:47to clinical trials all but one
43:51are CRBN recruiting compost. OK.
43:53So therefore in those cases the
43:55new substrates of the CRBN really
43:58need to be carefully monitored
44:00and actually to to this day.
44:01And new new substrates of the CRBN
44:04are still being discovered, OK.
44:05So that the field really need to
44:07watch that carefully but like
44:09you said on the other hand we
44:11could have turned this around and
44:13use this as opportunity and two
44:15actually generated potentially more
44:16effective anti cancer therapeutics,
44:25all right. So that part of we
44:27child part of that I I don't but
44:30for the CRBN part of it really
44:32is the CRBN ligand and and the.
44:35It kind of pretty promiscuous and doesn't
44:37matter actually what linker you put it in.
44:39It still binds to crbn,
44:41but the because the linker will change
44:44a little bit that lead to actually
44:47different new substrate got degraded.
44:49So we do Actually we have pretty
44:52good understanding how to change
44:54linkers to eliminate the.
44:56Let's see the IKCF one and three degradation.
45:01Oh how to eliminate GSPT 1 degradation.
45:04And but if you want to incorporate
45:07some of the new substrate into in
45:11addition to the devolution of your
45:13target protein in that case and
45:15more linker exploration is needed.
45:57Akshay.
46:22Thank you.
46:32Right. That's a great question also.
46:34So let me kind of it's a loaded question.
46:36Let me try to answer 1 by 1. OK, right.
46:39So first about MS-40, OK, right.
46:42Then we're going to talk about the
46:44selectivity of the resistance, OK, right.
46:47The the MS-40 uses selective WDR 5
46:52binder as a moiety, as a binder of WD-5.
46:57So the parent inhibitor does not inhibit.
47:01Lead to degradation of the CRBN
47:03new substrate IKZF 1:00 and 3:00.
47:05So this the degradation of IKZF
47:081/3 only happened to WDR 5 degrader
47:12not WDR 5 inhibitor. OK all right.
47:15So that that because the the compound it
47:18does not bind to the the the inhibitor
47:21portion the does not bind to CRBN.
47:23OK, right.
47:24So that that that that part is that
47:27that is so you know you're right.
47:29I mean the.
47:30And and in this case the degraders
47:34could be actually so so-called less
47:37selective than the inhibitor because the
47:41the degrader degrades the new substrate
47:43of CRBN which inhibitor does not, OK right.
47:47But on the other hand the I
47:50just want to point it out,
47:51the degraders sometimes actually could
47:54be much more selective than inhibitor.
47:58Let's see the in the cases.
48:00We have a multiple isoforms
48:02of a multiple subtypes,
48:04for example cilicate four and six.
48:06OK As you know the that D
48:09approved drugs polycyclip,
48:10ribocyclip and a bamocyclip.
48:12They all have a similar potency
48:15for cilicate 4 and six.
48:16OK but but using the same the ligand
48:20which bind to cilicate 4/6 with same
48:23affinity but the cilicate 46 degraders,
48:25the degraders can actually achieve
48:28selective degradation of cilicate 4.
48:30Over 6 and vice versa.
48:33Mainly it's because not so much of A binding,
48:36but the binding is the it's one event.
48:39But the second event is a
48:40ternary complex formation.
48:41OK, so then the so the degrader,
48:44this ternary complex formation and
48:45whether or not the license residues,
48:48appropriate license residues on the
48:49target 14 in this case city four and
48:52six are available for eucalation,
48:54give you basically another dimension
48:56to achieve selectivity.
48:58So people have achieved selectivity that way.
49:01Even you have a ligand bind to
49:04the isoforms of the subtypes of
49:06proteins with same affinity,
49:08you can achieve selective degradation of
49:101 particular isoform over other isoforms.
49:13OK, right.
49:14So then in terms of drug resistance,
49:19the.
49:20The as you know the kines
49:22inhabitation OF646 inhabit the
49:25drug resistance have been observed
49:28in clinical in clinical setting.
49:31And the I mean whether or not the
49:35drug distance going to happen to the
49:37protests is remain to be seen and in a
49:40clinical setting so far have not been seen.
49:42And part part of the reason for
49:45that is the the binding of the
49:48protect to the target protein.
49:50And the that binary requirement
49:51of the high affinity is not it's
49:53not very stringent as long as the
49:56compound binds somewhat even with you,
49:58you have a lose lost the binary affinity
50:01tenfold, twentyfold,
50:02thirtyfold you still could
50:04lead to selective declaration.
50:06OK, right. And having said that,
50:10people have done in the
50:12laboratory setting observed.
50:13The resistance to degraders,
50:16OK and kind of initial
50:18report kind of interesting.
50:19The resistance happened actually not
50:22as it's not a point mutation in the
50:25binding pocket of your target protein,
50:27but actually rather is the
50:29done regulation of the E3
50:31like this complex component.
50:34OK, so people have observed done
50:37regulation of call 2 and call 4 for
50:40HL&amp;CRB and including compounds respected.
51:04That's great question.
51:05I would love to test that we have not.
51:07We tested in the cell lines in the
51:08graph bottle, but not a PDX model and.
51:14Yeah, we tested as a single
51:17agent in the cell lines model,
51:18but not as not in the PDX.
51:21And also the probably would be in
51:23addition to PDX probably test in the.
51:29Not, not in the e-mail compromised mice.
51:31Let's see just you know regular mouse models,
51:34syngenate mouse models probably could also
51:36see some potential additional benefit.
51:39As you know is issue also actually
51:41involved e-mail response and there's
51:43number of reports actually either
51:45inhibition of user two or declaration
51:47of user two could lead to actually
51:50increased e-mail response.
51:51We have not done that.
52:23Yeah. And these are this.
52:25These are great questions and.
52:27We have not seen the the the over
52:30the amplification of the target and
52:34and the acid resistance mechanism
52:36so far and but it could well happen
52:40and and so as the bridge protect
52:43approach the ideal situation I mean
52:45we we here I showed you 2 proof
52:48concept studies and they are they
52:50look promising but they're not perfect
52:52right so the the the city of the
52:54cycling D1 degrade are still degrade.
52:57CDK 46 somewhat but it's a less definitely
53:00less than a second D1 and the the the
53:06MS147 the PRC one degrader it also
53:08degrade a little bit of the Ed.
53:10Ideally we would like to actually
53:12the bridge protect do not degrade
53:14the bridge protein OK if we
53:16don't degrade the bridge protein.
53:18So that may be actually potentially could
53:22actually for cells develop resistant to that.
53:25Maybe actually it's the last kind of a a,
53:28a opportunity there.
53:29I so we've been thinking about what
53:32would be a ideal bridge protein.
53:33So the a bridge protein ideally would be have
53:37like a kind of some kind of lessened desert.
53:39So basically you have a small molecule but
53:41this this protein have lessened desert
53:43actually does not get ubuccinated and
53:45and so therefore do not get degraded.
53:47So that would be we're we're actually
53:50working on that try to try to
53:52discover like a a a improved compost.
53:57Let me brief out that some type of
54:00patient to them in the pool, Silent
54:06room
54:08after the time.
54:11What's your generous study?
54:15Find, What's the letters?
54:18Whatever the study.
54:20Yeah, and it's a great question.
54:22I got to ask the the,
54:26the this is the really is actually it
54:30depends on the target and depend on the
54:33target We Chin and I have been working
54:36on his favorite target for a long time,
54:38a long time. And on the other hand the,
54:43I mean for the WD R5 and the AKT
54:47or CD46 Protax, I didn't.
54:50The AKT and four,
54:51six protects I didn't talk about today in
54:54those cases in the first round of compounds,
54:57let's see first a a few dozen compounds,
55:00we already have a good hits, OK, right.
55:02Then as you continue to optimize,
55:04I mean the head rate actually is
55:05very high in those target you,
55:07you and the the when I see the
55:09initial run getting hits is not a
55:11single hit multiple compounds active,
55:13you see the trend of the SER
55:15point to you which.
55:17Linker links is likely favor
55:18those kind of things, right.
55:19But on the other hand we have like a
55:22some of other targets we have a case
55:25that's a lot worse than the the ones
55:27you're we've been working on many,
55:28many years.
55:29We're just having no heads what's that?
55:31OK, right.
55:33But you know we actually trying
55:35to now turn this around to use
55:38those as the bridge protein,
55:40OK because they don't get degraded, right.
55:43So if they can detect with our.
55:46Favored and drug protein, right.
55:48So therefore we we already have a rich
55:51protein which does not get degraded, right.
55:54So so it wears quite a bit.
55:57And then we we typically use a crystal
56:02structure of the binary complex to
56:05identify solving exposed region and we we,
56:08I mean we try to do modeling
56:10and so far the modeling of the
56:12ternary complex formation is quite
56:14difficult and so to this stage.
56:16The example I showed you,
56:17the WDF 5, we have two high
56:21revolution crystal structures,
56:22right?
56:22So the first one really helped
56:24us tremendously to generating
56:26much more effective protects.
56:28Then we confirmed that
56:29using the second structure.
56:47Right. Yeah, it's a great question.
56:48Also just a general speaking,
56:52there are kind of two approaches, right.
56:54So the one approach we we do so this
56:57is in collaboration with Eva's lab,
56:59we call this controllable protects.
57:02So basically we make the protect but
57:06capped caged with a with a cage group.
57:10So the protect itself is not active,
57:12OK, right. So only when we.
57:14End cage that that by that
57:17can be by UV light or by some
57:21hopefully by hypoxic condition.
57:23OK. Right.
57:24So therefore our radiation OK right.
57:27So then the cage group is released
57:30or cleaved so now we have active
57:33protect and that can lead to
57:37you know so, so then we can
57:39basically turn on and off right
57:41another way we've been doing it.
57:42So part of the.
57:44Controllable protag is we published
57:46a couple of papers together with
57:48many labs so-called Foley caged protag.
57:52So basically again the protag
57:54were caged with a Foley.
57:57Foley Foley receptor are receptors
58:00are overexpressed in cancer cells over
58:04normal cells and binding of the Foley
58:07group to the Foley receptor leads to
58:10endocytosis of the entire molecule.
58:12Then the that Foley that Foley cage
58:15group is cleaved in the in the zone
58:19release the active protect lead to
58:21a degradation of a target protein.
58:24OK so in cancer cells over normal cell.
58:27So this kind of a controllable
58:29protect approach.
58:30Another way we've been a lot of
58:32us are trying to do basically is
58:34to find the targets of working
58:36on the targets that are really.
58:38Driving tumors or Tumogenesis are like
58:40I mean the the critical for that but
58:44it's non essential in normal cells
58:47and and so one example is cycling D1.
58:49Cycling D1 is non essential gene OK
58:52normal cells do not care right So then
58:54you you you really you completely
58:56knock out of the 2nd D1 basically
58:58have no feedback and so that that
59:01and but cycling D1 is important in
59:04Tumogenesis so therefore in this case.
59:07We believe cycling D1 degrader would
59:09have very good circuit window.
59:15We have not done that.
59:16People have done that, yes,
59:38yeah.
59:49Yeah those are great points.
59:51The just the the second point 1st and
59:55and actually protest can be actually the
59:57youth can the PGP substrates, OK, right.
01:00:00So some of them actually got
01:00:02Eflex after quite dramatically.
01:00:04And so then it really depends on,
01:00:06so the try to avoid the PGP substrate is
01:00:10important because otherwise you know the
01:00:12the molecules are already very big, right.
01:00:14Very difficult to get in cell.
01:00:15But if you even the small amount
01:00:17get in the cell got, if got, got,
01:00:19got if likes out then you really
01:00:21don't have a don't have active
01:00:22compost in the inside the cell, right.
01:00:24So that's that's kind of important.
01:00:26OK, right.
01:00:30It's great point.
01:00:31We actually currently working on
01:00:33this especially for a brain tumors,
01:00:36the protag's so big right make
01:00:37it already by available already
01:00:40difficult make it already by
01:00:42available and brain penetrant,
01:00:43it's almost impossible.
01:00:45OK, so therefore really using
01:00:48nanoparticle technologies to
01:00:49deliver protax to the brain,
01:00:51I think, I think that's kind
01:00:53of One Direction to go.
01:01:08Thanks a lot.