Yale Cancer Center Distinguished Lecture Series: "T Cell Lifestyle in Chronic Viral Infection and Cancer: Implications for Immunotherapy"

October 09, 2024

Information

ID: 12187
To Cite: DCA Citation Guide

Download Transcript

00:00Buddy,
00:01thank you for being here.
00:03I'm, Barbara Burtness. I'm the
00:05associate director for translational science,
00:07and it's really my honor
00:08to be kicking off our,
00:11distinguished lecture series for, this
00:14year at the at the
00:15cancer center
00:16by having,
00:18the privilege of hosting doctor
00:19Rafi Ahmed.
00:20So,
00:22Rafi is a professor in
00:24the department of microbiology and
00:25immunology at Emory.
00:27He's also the, director of
00:29the Emory vaccine center, coleader
00:32of the cancer immunology research
00:34program at Winship Cancer Institute,
00:36and an investigator in the
00:38Emory Center for AIDS Research.
00:41He is really
00:42a pillar of the immunology
00:45revolution,
00:46that has impacted
00:48the the care of, cancer
00:49patients.
00:50He's,
00:52as I said, a professor
00:53in the department of microbiology
00:54and immunology, a member of
00:56the National Academy,
00:57and a world renowned immunologist.
01:00His work in the past
01:00decade has been highly influential
01:03in shaping our current understanding
01:04of memory t cell differentiation
01:06and antiviral
01:08t and b cell immunity.
01:10His goal is to use
01:11this information for vaccine development
01:13and insights into virally mediated
01:15cancers.
01:16A major area of focus
01:17for him has been identifying
01:19cellular molecules that regulate the
01:21generation and maintenance
01:22of CD eight and CD
01:24four T cell in humoral
01:25immunity.
01:26One such molecule is mTOR
01:27that his lab recently identified
01:29as a major regulate
01:30regulator of memory CD eight
01:33T cell differentiation.
01:35Before we, get to hear
01:37him, and I I promise
01:37you he's a brilliant speaker,
01:39I do want to, have
01:40the opportunity to present a
01:42plaque to you. So when
01:43I come up?
01:48So this honors your your
01:50role as a distinguished lecturer
01:51here. Okay? Thank you for
01:53coming.
02:22It will it fall off?
02:29Yeah.
02:31Okay. Yeah.
02:33Okay.
02:44Ah, okay. Yeah. What I'm
02:47missing
02:49was
02:51glass.
02:53Okay. Can you all hear
02:54me?
02:55Thank you, Barbara, for this,
02:57very kind introduction and for
02:59inviting me.
03:00Did we succeed on our
03:01third try or the fourth
03:02try? We we try to
03:04arrange the
03:05my seminar and something or
03:07the other happens. And I
03:08know, Barbara, that you have
03:09a flight to take, so
03:10please don't miss your flight
03:11because of me. Feel free
03:12to leave.
03:16Alright. So I'm gonna talk
03:17to you about,
03:19T cell exhaustion, and I
03:21thought I'd start by really
03:22asking the question,
03:24what what is T cell
03:24exhaustion? What does it really
03:26even mean?
03:28I think in the early
03:29days,
03:30there were very few labs
03:32that were studying T cell
03:33exhaustion. And although we didn't
03:34know everything,
03:35I think we kind of
03:37knew what we were doing.
03:38But this has become now
03:40a very important issue for
03:42in many areas, especially in
03:44cancer.
03:45And I think so I
03:46thought I'd spend if,
03:48the first five, ten minutes
03:49of the talk really kind
03:50of taking you through,
03:52some historical
03:54observations that were made,
03:56that really first pointed out
03:59the
04:00the phenomena that during chronic
04:02viral infections,
04:03the t cells are not
04:04very functional. Just give you
04:06just a brief,
04:07summary of that and then
04:09bring you up to date
04:10in terms of what we
04:11currently mean,
04:13by t cell exhaustion.
04:14And then I will, spend
04:16some time
04:17describing,
04:19these very critical resource,
04:21CDAT cells or the stem
04:23like cells,
04:24which maintain CD eight responses
04:26during
04:27conditions of chronic infection.
04:29After that, I'll switch to,
04:31describing our recent work with
04:33HPV positive head and neck
04:35cancer.
04:36And if there is time,
04:37I'll present,
04:38a summary of our recent
04:40work with
04:41the combination therapy of PD
04:43one plus IL two. Yeah.
04:44But if I don't, Barbara,
04:45I can just stop after
04:47the HPV one. Yep.
04:48Yeah.
04:50So the question is what
04:51is T cell exhaustion?
04:53So the history of this
04:55really was started first, I
04:57would say, in the nineteen
04:59seventies,
05:00when people
05:02who were studying mouse models
05:03of chronic viral infection,
05:05in particular from LCMV,
05:07but also from mouse hepatitis
05:09virus and a strange virus
05:11called lactate dehydrogenase
05:13virus,
05:14which is also
05:15used to be. Now nobody
05:16uses it, but was a
05:17model for chronic infection.
05:19And there were these observations
05:20that,
05:22based on functional assays,
05:24that the t cells were
05:25not as responsive as one
05:27would expect.
05:29And interestingly, at that time,
05:30the major
05:32t cell assay that was
05:33used was chromium release, chromium
05:35fifty one release. And what
05:36was noticed was that when
05:37you took
05:39t cells or you took
05:40spleen cells
05:42from, during the acute phase
05:43of infection,
05:45there was good killing of
05:46the target cells, which these
05:48would be virally infected target
05:50cells. But when you took
05:51spleen cells from a chronically
05:53infected mouse, there was minimal
05:55chromium release,
05:56that is minimum killing. That
05:58really was the first beginning
06:00of those. And, of course,
06:01later on, cytokine assays
06:04also confirmed that.
06:06And then these observations made
06:08in mice, chronic infections were
06:10quickly extended
06:12to several human chronic viral
06:14infections.
06:15To HBV,
06:16people again found that,
06:19people people who are HBV
06:21carriers, again, it was difficult
06:23to detect good c d
06:24a t cell responses in
06:25the blood. But sim similar
06:27with HCV,
06:29some of the herpes viruses,
06:31and in particular HIV. Some
06:33of the early beautiful work
06:34done in human,
06:36infection showing, T cell dysfunction
06:39actually was in HIV.
06:41And then,
06:42we were at this stage.
06:44The question still remained.
06:49Is the decreased CDF T
06:50cell response
06:52that we are seeing by
06:53some functional assays, is this
06:55due to deletion
06:57of the virus specific cells,
07:00or are the t cells
07:01still present
07:02and truly dysfunctional?
07:04And this is an was
07:05an important question because often,
07:08it was thought that when
07:09we don't detect
07:10a t cell response, that
07:12either the t cell response
07:14was never generated
07:15or the t cells had
07:16been deleted.
07:18And we should thank Mark
07:19Davis for the
07:21development of the MHC tetramer
07:22technology
07:24in this paper by John
07:25Altman,
07:26Michael, McKheiser, Williams,
07:28and Davis in Science in
07:29nineteen ninety six. This was
07:31a major breakthrough because this
07:33allowed us to, for the
07:34first time,
07:36to actually physically visualize,
07:38the t cell that we
07:39were studying.
07:40Not just a functional asset,
07:41but now actually you could
07:42see it.
07:43And this allowed actually
07:45our lab and Rob Zinkernagel's
07:47lab, and these were the
07:48papers published by Alan Zayak,
07:50who was a postdoc in
07:52the lab, and Gallimore was
07:53in Rob Zinkernagel's lab to
07:55show, that when you compare
07:58an acute infection,
07:59LCMV infection with the chronic
08:01LCMV infection,
08:03in both instances,
08:05you find tetramer positive cells.
08:07They are there.
08:08But when you do functional
08:10assays
08:11like ICS assay into cytokine
08:13or Elispod, then the functionality
08:15of those cells is much
08:17less. K? And this was
08:19the the clear first demonstration
08:21that the cells are present,
08:23but that they
08:24lack or are not as
08:26functional as the memory cells
08:28or effector cells you get
08:29during acute infections.
08:31And, of course, this is
08:32very very quickly extended to
08:34human chronic viral infections
08:36and and to cancer.
08:39So then the
08:41next question that was addressed
08:43by us and by many
08:44other people in the in
08:45the in the field was,
08:47what is the gene expression
08:49profile
08:49of these
08:51exhausted cells
08:52versus the acute cells? It
08:54was shown that, actually, they're
08:55strikingly different.
08:57The gene expression program of
08:58functional memory cells and exhausted
09:00cells is very different. This
09:02was work that John Biri
09:04did
09:05in in two thousand seven.
09:06Many other people followed up
09:08on it, extended it to
09:09humans. It was also shown
09:11that, PD one is a
09:13major regulator
09:14of the CDT cell exhaustion.
09:18And so at this time,
09:19basically, we knew that the
09:21gene expression profile
09:23of the
09:24exhausted cells was very different
09:26from what you see,
09:27in terms of memory cells.
09:29It was known that there
09:30are PD-one and also many
09:32other,
09:33inhibitory receptors that regulate,
09:36T cell, function.
09:38But
09:39what was not known until
09:41about five years or so
09:42until twenty sixteen or so
09:45was really a clear definition
09:47of the different
09:49subsets which are there. So
09:51exhaustion wasn't a single cell
09:53that was functionally
09:55exhausted.
09:56But, basically,
09:57there was very interesting heterogeneity
10:01of
10:02the pool of exhausted t
10:03cells,
10:04in terms of the functionality.
10:06So
10:08this was work that was
10:09done
10:11in my lab by Sejgan
10:12I'm and also Uttschneider
10:14published a paper around the
10:15same at the same
10:17time, working with,
10:19Dietmar and and
10:21Werner Held.
10:22And there was also a
10:23paper by Lin and Yee.
10:25These three papers came out
10:26in twenty sixteen,
10:28and they defined,
10:30what I'll refer to as
10:31the stem like CDAT cells.
10:34These cells are also often
10:36referred to as, as t
10:38pect cells as
10:40standing for precursors of exhausted
10:42cells.
10:43And there are several
10:45interesting features of this cell.
10:46And, actually, my talk today
10:48really is about this cell.
10:50K? And I'll tell you
10:52more about this also in
10:53the context of HPV.
10:56So there were very unusual
10:57features of this cell. The
10:59first unusual feature was that
11:00even though
11:01this is
11:03in the conditions of a
11:04chronic infection, these cells are
11:06mostly quiescent.
11:08They're not rapidly dividing. There's
11:10some slow self renewal.
11:14And, also, their location was
11:16interesting.
11:17Even when you have a
11:18situation where virus is in
11:20multiple tissues with lymphoid and
11:22non lymphoid, these cells tend
11:25to be mostly
11:26in lymphoid tissues.
11:27And within the lymphoid tissues,
11:30they are in the t
11:31cell zones. And that's an
11:32interesting point because these cells
11:34express CXCR five,
11:36but they still don't
11:38go to where the b
11:39cells are. There's remaining more
11:41where the t cells are
11:42in from. So they are
11:44in the t cell zone,
11:47and they're also resident. That
11:49is they're not circulating.
11:51So this is the population
11:52that's within,
11:54the spleens and lymph nodes
11:55of a chronically infected mouse
11:57and has has been shown
11:59in humans. They're also in
12:01the tumor,
12:02environment.
12:03But, again, they are mostly,
12:07not proliferating and they are
12:08they are quiescent
12:10and also resident.
12:12So what do these cells
12:13express? They, of course, express
12:14PD one because PD one
12:16as
12:17has been shown by many
12:19people
12:19is a marker that comes
12:21up upon TCR stimulation. If
12:23you take a naive T
12:24cell,
12:25you stimulate it
12:27in vivo or
12:28or in vitro within twenty
12:30four hours, p d one
12:31comes up. It's a marker
12:33that comes up immediately upon
12:34upon TCL activation. So these
12:36cells are seeing antigen. They're
12:38p they're p d one
12:39positive.
12:41But this stem like population
12:42also expresses TCF one
12:44and expresses BCL six,
12:47has high levels of post
12:49stimulatory molecules
12:51other than PD one and,
12:54and a little bit of,
12:55TIGIT that doesn't have too
12:57many other inhibitory receptors,
12:59these cells.
13:02They have no effector molecules.
13:05They express interesting chemokines and
13:07chemokine receptors.
13:09They have CXCL five, so
13:11they respond to CXCL thirteen.
13:13But they also have sufficient
13:14amounts of, CCR seven
13:17that they can respond to,
13:19CCL nineteen and twenty one.
13:21They express interesting chemokines that
13:23bring DCs around them.
13:26They're one of the few
13:27cells that express XCL one.
13:29XCL one attracts,
13:31CDC one. That's the XCLP
13:32positive DCs, so they are
13:34right, next to, so, basically,
13:36they've they have a nice
13:37niche that they've surrounded themselves
13:39with.
13:40And once and what these
13:42cells are doing is that
13:43at any
13:45in a slow rate,
13:47continuously, they're generating
13:49these transitory effector cells. So
13:51the step in this differentiation
13:52is downregulation of TCF one,
13:55up regulation of TIM3, TBET,
13:57they start proliferating,
13:59express effector molecules
14:01and go out in the
14:02circulation.
14:05And then eventually these cells
14:06will get
14:08what we call and others
14:09have called the termally differentiated,
14:12cells with minimal proliferative capacity.
14:14So if you look at
14:16to go back to my
14:17question, what is exhaustion?
14:19So, basically, exhaustion
14:20is all three of these
14:21cells at any given time.
14:24The ratios of these three
14:25can differ.
14:27Stem like cells usually are
14:28a very small percentage
14:30and only in the lymphoid
14:31tissue. They'll be about maybe
14:33ten to twenty percent in
14:34the lymphoid tissue.
14:36If you go to non
14:37lymphoid tissues, there's hardly any,
14:39less than one to five
14:40percent.
14:42Affector cells usually are also
14:44on the lower side,
14:46not a large number of
14:47them.
14:48Can be about twenty percent,
14:49ten percent, especially in a
14:51established long term chronic infection.
14:53They're only about ten percent.
14:56Most of it are these
14:57more differentiated exhausted cells.
14:59So when we and others
15:01in the past were looking
15:02at all three of these,
15:03because if you use a
15:04tetramer,
15:05you sought all three of
15:06these populations.
15:07We were mostly measuring
15:09this cell in terms of
15:11the image that we had.
15:13But there are clearly cells
15:14there which are too important
15:16and functional,
15:18and you have this effector
15:20cell that's being generated,
15:21and then you have this,
15:23terminal differentiation.
15:25So the answer to the
15:26question is that
15:27T cell exhaustion is actually
15:29a
15:30ongoing immune response.
15:32It is tightly, highly regulated
15:35by inhibitory receptors
15:37and also by the immunosuppressive
15:40myeloid cells that surround that.
15:43So it's
15:44suppression by the myeloid cells,
15:47inhibitory receptors on the t
15:48cells, plus probably some others
15:50keys inhibitory cytokines that are
15:53doing it. But it's active.
15:54It's active, and it's ongoing
15:56because you're always generating
15:58a low number of effector
16:00cells from that stem like
16:01population.
16:04And the key cell
16:06to maintain this,
16:08whole
16:09exhaustion program if you want
16:10to use that term is
16:12the p d one positive,
16:13Tc f one positive, Tox
16:15positive, and I prefer to
16:16use the term resource cell
16:18because this is the cell
16:20that's feeding
16:21the T cell response under
16:22condition of chronic infection. If
16:24you don't have the cell,
16:26the system just collapses. That
16:27is because the effector cells
16:30are short lived. They go
16:32further, become exhausted. Those cells
16:34also have a it's the
16:36the lifespan is very, is
16:38not very long.
16:41So what does p d
16:41one do? Basically, p d
16:43one changes the ratio
16:44of these cells.
16:46When you do p d
16:47one blockade, you get much
16:49faster differentiation
16:50of the stem like cells
16:52to give you more of
16:53these effective cells.
16:55So the main
16:56thing the p d one
16:57does
16:58in terms of changing the,
17:01is the it changes the
17:02ratio. That is, you get
17:03many more effector cells,
17:05which will contribute to the
17:07the killing of these target
17:09cells.
17:10And then, eventually, it will
17:12it will go into this
17:13thing.
17:18So when are these cells
17:19generated? Actually, fairly early
17:21in in the program. You
17:23get generation of these, stem
17:25like cells.
17:27And then, as I told
17:28you, this is what keeps
17:29the the engine going.
17:31The next, slide is just
17:32giving you a quick look
17:34at the gene expression program,
17:35again, published by many people
17:38multiple times. Just to give
17:39you a flavor of how
17:40these three cells differ in
17:41gene expression.
17:43Again, all three so this
17:44is using a MHC class
17:47one tetramer to sort the
17:48antigen specific cells and then
17:49doing single cell RNA seq.
17:51You see these three subsets.
17:52You've got the stem, the
17:54transitory, and the more exhausted
17:55or the term we differentiate,
17:56all expressed p d one.
17:59All expressed stocks.
18:01Only the stem like cells
18:02expressed t c f one.
18:04These two don't.
18:05TEM three is expressed by
18:06these cells and not that,
18:08not the stem like. Again,
18:10thirty nine, another inhibitory
18:11marker expressed by the more
18:13differentiated cells and not by
18:15the stem like cell. I've
18:16spiked in here just what
18:18the gene expression program is
18:19of, naive cells. So naive
18:21cells, of course, have no
18:22p d one. They have
18:23no talks, but they have
18:24TCF one. In fact, every
18:26good t cell has TCF
18:28one. Memory cells have TCF
18:29one.
18:31The stem like cells have
18:32TCF one, and also you
18:34have Tcf one in all
18:36all naive cells.
18:38PIM three and thirty nine
18:39don't have it. I mean,
18:40they've not not except the
18:42naive cells. If you look
18:43at effector molecules,
18:45effector molecules are not expressed
18:47by these cells. You have
18:48them in the differentiated
18:50transitory and exhausted. Again, granzyme
18:52b comes up. CD twenty
18:54eight is higher in these
18:55cells,
18:57and
18:58CD one twenty seven is
18:59expressed by the stem like
19:01cell. Again, CD one twenty
19:02seven, as is well recognized,
19:04is essential for long term
19:06survival of the t cell.
19:08IL seven signals are essential.
19:09So if you don't have
19:10CD one twenty seven,
19:12then a cell will not
19:13be persisting or surviving long
19:15term. And And it's the
19:16stem like cells that have
19:18and seven as do as
19:19naive cells, of course. And
19:20XCL one purely only made
19:22by these cells, not by
19:24these or by the naive
19:25cells.
19:30I want to spend a
19:32few minutes asking the question,
19:33are these stem like cells
19:35actually receiving TCR signals actively?
19:39And I've told you the
19:40express PD one, so it
19:41suggests that they might that
19:43that they should be. But
19:44then we ask this question
19:45more
19:46directly by using, node seventy
19:48seven
19:49reporter,
19:51cells.
19:53This is a transgenic mouse
19:55where the nodes if the
19:57cells get TCR signaling, expressed
19:59new seventy seven, which is
20:00downstream of TGA signaling,
20:02GFP will come up. So
20:03this shows you the new
20:05seventy seven staining in this
20:06case is where you're looking
20:07at GFP.
20:09So if you look at,
20:11a mouse that has cleared
20:12the infection and is a
20:13LCM b immune mouse,
20:15here's the tetramer staining.
20:16And, of course, these cells
20:17have neither new seventy seven,
20:20minimal to none, and they
20:21don't have PD one. But
20:22if you look at the
20:23antigen same specificity, antigen specific
20:26cell during chronic infection,
20:28you see that they're all
20:29new seventy seven positive,
20:31and they're all p d
20:32one positive.
20:34If you then ask the
20:35three subsets,
20:36the three clusters,
20:38and we can subdivide it
20:40into the the three clusters
20:41by using these MARCOS TIN
20:43three and CD one zero
20:44one, these are the stem
20:45like cells. These are the
20:46transitory effectors, and these are
20:48the more thermally differentiated cell.
20:51And then you ask that
20:52do Samsung's spinning. You see
20:53that all three of them,
20:55like, are node seventy seven
20:56positive.
20:57So even though these cells,
21:01this so all of these
21:02cells are getting TCR signals,
21:05continuously.
21:06But in spite of the
21:07fact that you're getting the
21:08TCR signals,
21:11the
21:12stem like cell
21:14is not dividing.
21:15There's only minimal proliferation.
21:18The transitory cell which have
21:20recently emerged from it are
21:22all mostly in cycle.
21:24So the recently emerging
21:26effector cells from the stem
21:27like cells at any given
21:29time, over fifty percent of
21:30them will be divided.
21:32Because they are just emerging
21:33from it.
21:35The terminally differentiated cells
21:37in terms of proliferation are
21:39truly exhausted.
21:40They have minimal to no
21:42capacity
21:43to further proliferate.
21:44They don't do that after
21:45PD one blockade.
21:46They don't do that after
21:47any cytokine that you give.
21:49So these are
21:50not divided.
21:52But this cell,
21:54which is
21:55very poorly potent
21:57and is able to divide
21:59after you give PD one
22:00blockade or other cytokine,
22:02under these conditions actually is
22:04quiescent.
22:07And then if you ask
22:09which cells express effector molecule,
22:13this stem cell expresses
22:15no granzyme b. Even though
22:17it's getting TCR signals, it
22:18expresses no granzyme b. The
22:20transducer cells, of course, have
22:22lot of granzyme b and,
22:24more exhausted cells also have
22:25granzyme b.
22:27So I find this quite
22:28fascinating
22:29that this cell is getting
22:31TCR signals,
22:34but and is it we
22:36know that this cell is
22:37functional to report and it
22:39can after you do the
22:40PDN blockade, it will
22:42differentiate to give more transit
22:43v cell. But in its
22:45own state,
22:47it does not express granzyme
22:49b.
22:51Basically,
22:52think of it as a
22:53TFH cell
22:54because the TFH so it
22:56has b c l six.
22:57It has, in some ways,
22:59hijacked
23:00some aspects of the TfH
23:02program.
23:03So when you have CD4
23:05cells that differentiate, you get
23:06Th1,
23:07you get TfH.
23:08The TfH cell, which has
23:11Bcl six, and so on,
23:12it shuts down the effector
23:14program.
23:16So TfH cells will never
23:18have effector molecules.
23:19It's the t h one
23:20cells that have it. So
23:22the stem like cell has
23:23taken some interesting aspect
23:26of that biology. It certainly
23:28is not a TfH cell.
23:29It doesn't even go near
23:30b cells. It's not involved
23:32in helping,
23:34b cells, but it has
23:35captured a very interesting aspect
23:38of TFA biology
23:40in its, differentiation program.
23:43And but when it then
23:44differentiates, you remove the PD
23:46one break. When it differentiates,
23:48now you get
23:49the the t h one
23:50type cells, which is your
23:51transitory effector cells, which express
23:53all the molecules
23:55that the cytotoxic t cell
23:56expresses.
23:57I think this is really
23:58quite,
24:01interesting biology for us to
24:02work out. So the stem
24:03cell niche,
24:06so this is
24:07not expressing these things
24:09is actually an active program.
24:11It's not a passive program.
24:12It's not due to lack
24:14of TCR stimulation,
24:16not due to lack of
24:17antigen. That is the program.
24:18That is the transcription program
24:20that changes after you remove
24:22the p d one break
24:23or can do it by
24:24some others,
24:25cytokines that you provide.
24:28Okay. So let me now
24:29get to,
24:30human,
24:32stem cells. Many people have
24:34shown this.
24:35Elegant studies done looking at
24:36neoantigen
24:37specific cells in melanoma,
24:39in many other,
24:41cancer systems have shown that
24:42you have these stem like
24:44cells also
24:45in,
24:46in human cancer.
24:48We we have done shown
24:49this in lung cancer and
24:51head and neck cancer.
24:52I'll just show you couple
24:54of slides from lung cancer
24:56just to tell you that,
24:57yes, you find these cells,
25:00and this was the paper
25:01we published,
25:02last year, but I had
25:03also shown this in another
25:05paper earlier.
25:06But the more interesting thing
25:07from this study,
25:09in addition to showing the
25:10stem like cells in lung
25:11cancers,
25:13was the location of these
25:14cells
25:15within the lung cancer.
25:17So if you look in
25:18the lung cancer
25:20studies,
25:21in lung cancer,
25:23samples,
25:24these cells are not where
25:25the tumor is. They are
25:27away from the tumor.
25:29So on the tumor, you
25:30will find the more
25:32sector like cells, the transitory
25:34cells, and you find cells
25:36which are the more exhausted
25:38or terminally differentiated.
25:39The stem like cells are
25:41away from where the actual
25:42tumor is,
25:44and we found that some
25:45of them are actually in
25:46tertiary lymphoid structures
25:48within the tumor.
25:50And they're closer to the
25:51c d four t cells
25:53there than to the b
25:54cells.
25:56So, essentially,
25:58a similar kind of compartmentalization
26:01that we saw in the
26:02chronic viral infection model, that
26:04is they're not where the
26:06actual
26:07virally infected,
26:09cells are, but a little
26:10bit away from it. So,
26:11like, if you look in
26:12the spleen of these chronically
26:14infected mice,
26:15the major infection is in
26:16the red bulb of the
26:18f four
26:19eighty macrophages and other cells.
26:21These cells are in the
26:22t cell areas,
26:23but there isn't much infected
26:25cells.
26:26So they're hiding this beautiful
26:28niche that they have. It's
26:30keep it keeps them away
26:31from the major. It's just
26:33part of their whole program
26:34that keeps them there. It's
26:35when they differentiate
26:37that they will send the
26:38effectors out there. And then
26:39there was a beautiful paper,
26:41much nicer paper than ours,
26:43paper by,
26:45by Lee Hakowen,
26:47who did this published this
26:49paper,
26:50this this year.
26:51And in this one, he
26:52shows that again in lung
26:54cancer,
26:56if you look at the
26:57stem like cells, they are
26:58in what he calls stem
27:00immunity hubs, which are again
27:02removed from where the tumor
27:03is, and they're close to
27:05where
27:06the t cells are. In
27:07fact,
27:09he makes the analogy that
27:10these cells are found in
27:12areas
27:12that look like lymph node
27:14areas.
27:15K.
27:16And this is all within
27:17the lung tumor. He's not
27:19looking at a lymph node.
27:20He's looking right at the
27:21tumor. And we had done
27:22the same thing. We found
27:23that where you see the
27:25cancer,
27:26the tumor antigen, very few
27:28of these cells are there.
27:29They're always away to the
27:31side.
27:34Okay. Let me now go
27:35to,
27:37head and neck cancer. And
27:38one of the reasons we
27:39did this,
27:41Barbara, not because we knew
27:43anything about head and neck
27:44cancer, but because
27:45we wanted to look at
27:47antigen specific cells. Because in
27:49our lung studies,
27:50when you're looking at total
27:52cells,
27:53and we did not have
27:54the bandwidth to be doing
27:55the sequencing and identifying the
27:58precise mutation in each individual
28:00and doing it. So we
28:01said, let's go to
28:03a viral mediated cancer
28:05where we could would be
28:06easier for us to identify
28:08the epitopes which are being
28:10recognized, and then we can
28:11use tetramers and look at,
28:13HPV specific cells in the
28:15tumor.
28:16So we started out by
28:18collaborating with Nabil Saba and
28:21Mihir Patel
28:22and what they gave us
28:23was the following material.
28:26They gave us,
28:28in this study, we have
28:29now extended it and we
28:30have more data,
28:32But I'll share with you,
28:33some of our published work.
28:35So, basically, we had,
28:37seventeen patients
28:39who were HPV positive head
28:40and neck cancer patients. These
28:42were all treatment naive. By
28:44what I mean is that
28:46they had not had previous
28:47treatment. They had just been
28:48diagnosed
28:49as having head and neck
28:50cancer.
28:51And we I we got
28:52from them at the time
28:53of surgery, we got the
28:55sample of the blood,
28:57and we got the primary
28:58tumor, which you all know
28:59is in the tonsils. And
29:00then we also got,
29:02a metastatic lymph node. Not
29:04a draining lymph node, but
29:06a metastatic lymph node. Yeah.
29:09Now at this stage, we
29:10had no idea what
29:12responses these,
29:15individuals were making
29:16in the tumor, and the
29:17tumor material, as you know,
29:19is very limited.
29:20So we first had to
29:21identify
29:23what potential epitopes were being
29:25seen by these seventeen patients.
29:28So what we did
29:29to address that issue,
29:31we did the following,
29:33experiment. So we took the
29:34blood from these,
29:36HNSCC,
29:38patients
29:39and then expanded these cells
29:41for two weeks, basically
29:43expanding the cells in vitro.
29:46And we used, about two
29:47hundred fifty predicted HPV
29:49peptides.
29:50We, of course, included e
29:52six, e seven, which is
29:53the canonical oncogene. But we
29:55also included e two and
29:56e five because there were
29:58some reports
29:59saying that one can get
30:01CDH responses to e two
30:03and e five. These were
30:04not in tumor samples, but
30:05in people who were HPV
30:07positive.
30:08And, so we included e
30:09two and e five.
30:11And then expanded this population,
30:14looked at,
30:15reactivity if we could expand
30:17any cells from these individuals
30:19from the blood. We did
30:20first interferon gamma ELISpot, then
30:22did ICS.
30:23And then if the response
30:24was was good enough, we
30:25then went ahead
30:27and identified the precise tetramer.
30:30So here is an example
30:31of one of the good
30:32responders.
30:33It's a very nice after
30:34its expansion.
30:35We're looking now at the
30:37pool of peptides, all five
30:38of all four of them,
30:39all two fifty. We see
30:40nice response.
30:41By ICS, you see very
30:43nice,
30:45interferon gamma and tNF alpha
30:46production.
30:47And then by tetramer again,
30:49a very nice staining. This
30:50is the same tetramer, double
30:52tetramer staining. You see that.
30:55And, interestingly, all seventeen patients
30:57that we looked at
30:58were positive.
31:00So there is all of
31:01these individuals.
31:02There was no one that
31:03was negative. Some were higher,
31:05some were lower in the
31:06response, but all of them,
31:08gave us a response.
31:10That is, there was some
31:11low number of cells present
31:12in the blood that were
31:14that we could expand.
31:17And then we ended up
31:18identifying several,
31:20tetramers that we made.
31:22Note that e two and
31:23e five tetramers,
31:25there are several came responses
31:27came from e two and
31:28e five, some from e
31:29six. This is not by
31:29any means comprehensive.
31:31We had these,
31:32tetramers. So we used these
31:33tetramers.
31:34Now we could go back.
31:35We knew which individual
31:37was responding to which cell
31:39based on the expansion from
31:41the blood. So we can
31:42now go back
31:43and do the stain of
31:44the TILs themselves.
31:48And the results were really
31:50I never expected this. I
31:51did not expect to see
31:53that in these individuals,
31:56this is now no expansion.
31:58This is just staining the
31:59TILs,
32:00with the tetramer.
32:02So here is PD one
32:03on this axis, tetramer on
32:05this axis. This is a
32:08tetramer
32:09recognizing peptides one fifty one
32:11to one fifty eight from
32:12the e two. It's a
32:13o one restricted.
32:14And the primary tumor,
32:16and here's the metastatic tumor
32:19From the metastatic,
32:21the the the lymph node
32:22that also has the tumor,
32:24three percent
32:25point one percent.
32:27Here's another one,
32:29five percent, ten percent.
32:32It's really quite amazing
32:33that these
32:35treatment naive individuals who are
32:37coming in for surgery
32:38have such a vigorous response
32:41ongoing
32:42in their, in their tumor.
32:44This is a
32:48what if we use the
32:49same tetramer and look in
32:50the blood?
32:51Very low.
32:53The response is barely detectable.
32:56Yeah. Because, again,
32:59I didn't I didn't talk
33:01about this earlier, but the
33:02but in the,
33:04when you we did some
33:06parabiosis experiments
33:08in our chronic LCMV infection
33:09mice to see which cells
33:11are in the blood and
33:12which will actually
33:13migrate. It's only the transitory
33:15effectors
33:17that will come out. So
33:18the more differentiated cells
33:21are resident at the sites
33:22of infection,
33:24and the stem like cell
33:25is resident
33:27at the
33:32circulates
33:33are the recently
33:35generated effector like cells.
33:37So so it's not surprising
33:40that in the PBMC,
33:42there are very few it's
33:43not it's not zero because
33:45you know that's where we
33:46expanded these cells from. So
33:48we know that the cells
33:49were there, but when you
33:50do a stain,
33:52the frequency is very low.
33:53I highlight this part because
33:55we often look at,
33:57new antigen specific cells, we
33:59want to look at tumor
34:00specific cells in the blood
34:02of somebody who's not had
34:04any treatment.
34:05That's
34:06it's not that they're not
34:07there, but you'll have to
34:08look very high. K? There'll
34:09be a very, very low
34:10frequency.
34:11So here are some example
34:13showing you the frequency of,
34:15about six six or seven
34:17different tetramers
34:18in different, individuals.
34:20And you can see that
34:22very high frequencies
34:23ranging from point one percent
34:25up sometimes even ten percent.
34:27But in the blood,
34:28by this approach, below detection,
34:30I. Bill below point or
34:31two percent.
34:33So a lot of cells
34:34in the tumor, k,
34:36but very few cells in
34:37the circulation, antigen specific.
34:40Okay. So the next question,
34:41of course, was, what do
34:42these cells look like? Do
34:43we have that stem like
34:44population? So we did single
34:45cell RNA seq, and the
34:47answer is yes. You again
34:48find
34:49within the tetramer sorted cells.
34:51Now this is all on
34:54getting the,
34:56the cells isolating them directly
34:58from the tumor
34:59and,
35:00and then doing single sign
35:02seq. And, again, you have
35:03the canonical cell here, which
35:05has TCF one. It's amazing
35:06that Xcl one
35:08always goes with them.
35:10It was just wasn't just
35:11a mouse phenomena.
35:12It every human,
35:14stem like cell defined by
35:16us or by the others
35:18have shown that XCL one
35:19is there. Again, this attracts
35:21CDC one. It's part of
35:22the niche that they create.
35:24And,
35:26again, no granzyme b in
35:27these cells.
35:29And you have LEF one,
35:31which is another transcription factor
35:32you see in these stem
35:33cells, more I l seven
35:35receptor over here, and, again,
35:37very little perforin and so
35:38on.
35:40So we found we looked
35:41at about fifteen
35:43or so tetrimers
35:45in different,
35:47different per patients, and all
35:48of them have the three
35:49subsets.
35:50So the blue is the
35:52more,
35:54is the more exhausted subset.
35:55This was, in most cases,
35:57the highest number. But all
35:58of them had
36:00this transitory population, which is
36:02shown here in yellow,
36:03and also had the stem
36:06like population. So these three
36:07sub again, so in these
36:09head and neck cancer patients,
36:10you have an ongoing response
36:12there. You have the stem
36:13like cell is generating
36:15some effector cells
36:17and these are eventually getting
36:18glasses. It's a very active
36:20program ongoing.
36:21In this study, we also
36:23looked at the TCR sequences
36:25and the TCR sequences are
36:26shared.
36:27At least the major clonal
36:29types are shared between these
36:30three, again, establishing the lineage
36:33relationship between them.
36:35But and the other interesting
36:36finding was, in some cases,
36:38these
36:40responses are very oligoclonal.
36:42In some cases, about fifty
36:44percent
36:45of the tetra deposits is
36:47very single clonal type. In
36:49others, a little bit more
36:50heterogeneity,
36:51but very oligoclonal.
36:53And I think you only
36:54see this oligochromality
36:56when you have long term
36:57chronic antigen stimulation, which has
36:59selected for certain TCRs.
37:01Yeah.
37:05Then we ask the question,
37:07are these stem like cells,
37:08the ones shown here in
37:10green, are these cells actually
37:11functional?
37:12Can we somehow demonstrate
37:14functionality in these cells? So
37:17for this, we
37:19we sorted the
37:21we we didn't use tetramer
37:22here because this
37:24the frequency is too low.
37:25We had to get as
37:26many cells as we could.
37:27We wanted to do the
37:28CTV labeling experiment.
37:30Frequency is not lower. The
37:31total numbers are not enough.
37:33So we enriched for this.
37:35So, basically, we did a
37:37sorting strategy
37:38that would enrich for where
37:39the stem like cells would
37:40be and where the more
37:42differentiated ones would be. So
37:43for this, we basically
37:44gated on PD one positive
37:46cells
37:47and then,
37:48used
37:50cells which were TIN three
37:51negative and CD thirty nine
37:53negative. So this would enrich
37:54for the stem like population,
37:56and then the TIN three
37:58positive, thirty nine positive would
37:59enrich for the more differentiated
38:01population.
38:02Then we culture these we
38:04we
38:05label the cells with CTV
38:07so we could look at,
38:08division
38:09and then stimulated them, but
38:11now we know exactly which
38:13peptide to use.
38:15So now we could stimulate
38:16with their Cognate peptide,
38:18and you can see the
38:19results here.
38:20So
38:21so we're looking here at
38:23at five days.
38:24After five days in culture,
38:25this is the unstimulated. There
38:27is no peptide stimulation.
38:29Here is the
38:32on this axis is the
38:33tetramer. Now we can stain
38:34with the tetramer. So even
38:36though we didn't sort with
38:37the tetramer, we now stain
38:38with the tetramers. We are
38:39truly looking at the antigen
38:41specific cells.
38:42And what you'd see here
38:44is that there's minimal proliferation
38:46without the peptide,
38:48And here's the tetramer population
38:50here. No division.
38:52And if you look now
38:53here after adding the peptide
38:56with the more differentiated
38:57cells,
38:58very minimal to no proliferation.
39:01Again, the proliferative capacity of
39:02these more exhausted cells is
39:04very minimal.
39:05But look what happens here.
39:07These cells
39:08proliferate beautifully.
39:10So they have this ability
39:12under the right conditions. You
39:13provide the peptide, and they
39:15will proliferate. You get very
39:16nice proliferation here. This is
39:18a summary of all the
39:19experiments
39:19that we did. Minimal proliferation
39:22of the more differentiated cells,
39:24but very nice proliferation,
39:26and CTV dilution
39:27of the stem like cells.
39:29And what happens now? Do
39:30they differentiate into effector like
39:32cell? Yes. They do. So
39:33if you now look
39:35after the if you now
39:36look at the stimulations,
39:38of how the markers change,
39:40TIM three comes up.
39:42Granzyme b comes up beautifully.
39:44K. See, before,
39:45there was no granzyme b
39:46in these cells. Granzyme b
39:48comes up. Three thirty nine
39:50comes up.
39:51IL seven receptor was there,
39:53goes down. K.
39:55So it and you can
39:56see that these cells had
39:57IL seven receptor. These cells
39:59didn't. K. As I've told
40:01you earlier, these cells don't
40:02survive that long. But after
40:04you generate the factors, IL
40:05seven receptor comes down. And,
40:07of course, over here, not
40:08a whole lot is happening
40:10except to point out that
40:11these cells were already TM
40:12three positive. These cells had
40:14granzyme b, whereas these cells
40:15did not.
40:17These cells had,
40:19you know,
40:22had thirty nine. These cells
40:23did not.
40:26All of these cells are
40:28RO. There's some confusion in
40:29the literature that this
40:31stem like cells are RA.
40:32No. It's all RO, and
40:34this is not the stem
40:36like cells described after acute
40:38infection.
40:38Everything is RO because they've
40:40all antigen experience. In fact,
40:41they're seeing antigen.
40:43CD twenty five comes up
40:44beautifully
40:45after,
40:46this
40:48proliferation.
40:49And CD twenty eight,
40:52which I was pointing out
40:53earlier, is only the stem
40:54like cells that have CD
40:56twenty eight, and they require
40:57this costimetry signal for the
41:00expansion that you see after
41:01PD one blockade.
41:02So it's the stem like
41:03cells that have CD twenty
41:04eight. And interestingly, they're retaining
41:06it at least in this,
41:08in this, shot,
41:09after the shot in vitro
41:11stimulation.
41:12Whereas the more term differential
41:14actually never had much c
41:15twenty eight to begin with.
41:20Okay. So this is what
41:21I was,
41:22telling you where they are
41:23located.
41:24So this is now
41:26stain of a tumor,
41:30of the HPV tumor, and
41:32you can see that
41:33this is where the tumor
41:34is based on the cytokeratin
41:36staining.
41:38Lot of CD8 T cells
41:39in this tumor.
41:42And if you look at
41:42PD one, pretty much all
41:44of them or most of
41:45them expressing PD one. This
41:47is the CD eight. There's
41:49the PD one. And now
41:50you look at TCF one.
41:53This they're here.
41:54This is where there was
41:56no tumor.
41:59Again, they segregate.
42:00They're not where the tumor
42:02is.
42:02Their progeny will go to
42:04the tumor, but these cells
42:05will be in these niches.
42:11So let me conclude this
42:13and
42:14and kind of end on
42:15a couple of,
42:17next steps that we're doing.
42:18So, basically, we've identified
42:20three transcriptionally distinct states of
42:22HPV specific CDT cells. And
42:24I didn't show you this
42:25data, but this is in
42:26our published paper. And they
42:27share the TCR. The same
42:29TCR chronotype
42:30is found in all three
42:31subsets.
42:32Again, consistent with showing that
42:34one is generating,
42:36to give rise to the
42:37other. And I've shown you
42:38functional evidence,
42:40also because you you have
42:41this,
42:43this thing happening based on
42:44atria chronotype being shared. But
42:46this is the more important
42:46point.
42:47The cellular machinery, essentially,
42:50I e the PD one
42:51positive stem like or resource
42:53cells that are required for
42:55response to PD one pathway
42:56blockade, they exist in these
42:58h three positive hNSCC,
43:00patients
43:01pretreatment,
43:02suggesting that PD one adjuvant
43:04therapy could be,
43:06an effective treatment.
43:08The other interesting finding here
43:10was,
43:11although the study is limited
43:12by limited, I mean, we
43:13don't have hundreds of patients.
43:15We found that e two
43:16and e five
43:18were the dominant
43:20targets, in particular, e two.
43:22E two is a major
43:23target.
43:24And this finding of ours
43:26has been subsequent in the
43:27last since our paper came
43:28out, two more papers have
43:29come out showing the same
43:31thing.
43:32E two is a dominant
43:34antigen in HPV.
43:39And
43:40and so far, there have
43:41been some very
43:43lot of experiments done using
43:46e six and e seven
43:47vaccines
43:48that have gone into many
43:50clinical trials, both for cervical
43:51cancer and maybe some for
43:53but but most of this
43:54has been focused on cervical
43:55cancer, maybe some also head
43:56and neck. But but we
43:58would strongly recommend that if
44:00you want to do a
44:01therapeutic vaccination
44:03of either head and neck
44:04cancer patients,
44:06especially for head and neck
44:06cancer patients where we have
44:07this data that e two
44:09and e five should also
44:10be part of the vaccine,
44:12in particular,
44:13e two, which is a
44:14very, very dominant antigen.
44:19So I want to kind
44:19of end with couple of
44:21oh, so this was our
44:22paper published by today. But
44:23I didn't say anything about
44:26our second paper that was
44:27published with it, which looked
44:29at b cell responses to
44:31HBV
44:31in the tumor,
44:33and they are also very
44:35strong.
44:36So within the tumor,
44:38we find HPV specific memory
44:40b cells.
44:41We have antibody secreting cells.
44:43That is, there are plasma
44:44cells right in the tumor,
44:46which are secreting antibody to,
44:49e six, e seven, and
44:51e two, and e five.
44:52So there's a lot of,
44:54active b cell response ongoing.
44:57So these are
44:58to use the
44:59the the term heart, these
45:01are very, very immunologically
45:03active, tumors.
45:04Both b cell and t
45:05cell response is ongoing,
45:07to multiple,
45:09HPV antigens with, at least
45:11for the end, for the
45:12CDH,
45:13real dominance
45:14of, of e two in
45:15this response.
45:16So what we have started
45:18now, and I was talking
45:19to Barbara about this,
45:21a very small study,
45:22and this is ongoing,
45:25Took us forever to get
45:26it going, but we have
45:28a small study that's being
45:29done by Mihir Patel and
45:32Nabil Saba. It base basically,
45:33it's an adjuvant,
45:36study where we we are
45:37giving
45:40head and neck cancer patients,
45:43two doses of,
45:45anti PD one that we
45:46got from Roche from Genentech,
45:49and Roche.
45:51So they will get before
45:52they have their surgery,
45:54they will get two
45:56infusions of anti PDL one,
45:59and then they will go
46:00in for the surgery.
46:01We've been collecting blood,
46:03from these patients.
46:05The plan is to have
46:06twenty patients who will get,
46:08PDL one treatment prior to
46:10their,
46:11to their surgery and, and
46:12radiation treatment and so on.
46:14K. And so far, seven
46:16patients have been enrolled,
46:18and,
46:19basically, blood is being collected
46:21pretreatment.
46:23Pre infusion of the PD
46:25one, we'll get one blood
46:27sample. And then after that,
46:28and then at the time
46:29of surgery, we'll get the
46:30tissue sample as well as
46:32the blood.
46:34So seven have been recruited
46:36into the thing already.
46:38Two of them have actually
46:39completed their regimen.
46:41And the early data is
46:42only n of two, but
46:44both these patients have shown
46:46a pathological
46:47response,
46:48within this month of, anti
46:50PDL one treatment. We haven't
46:52done the T cell analysis
46:53yet because we have the
46:55samples there. We first need
46:56to identify.
46:57We need to do what
46:58we did with the previous
46:59time
47:00and see, what
47:02they're responding to. But here
47:03I have a prediction.
47:04Now we will see more
47:05cells in the blood is
47:06my prediction.
47:08Because they have gotten the
47:09anti PDL one, treatment,
47:11blood these effector cells should
47:13come out into the blood.
47:14Now we will see cells
47:16in the blood. That's my
47:16prediction.
47:19And,
47:20so that's where we are,
47:21and I'll keep you posted.
47:22And this is my I
47:23will stop after the next
47:25one. I'm just gonna skip
47:26dial two. Okay?
47:27And then then
47:29so we
47:30so far, it's been all
47:31head and neck,
47:32and,
47:33but I've been in discussions
47:35with several people about
47:38maybe trying to do something
47:40similar
47:41in cervical cancer. The question
47:42is, in cervical cancer, is
47:44it going to be all
47:45e six, e seven only?
47:47Will you also have e
47:48two and e five responses
47:50in cervical cancer or not?
47:51So we are just starting
47:52a collaboration. Actually,
47:54I'm also going to India
47:55next week, okay, for for
47:57this. So,
47:59so it it's we're lucky
48:01that we,
48:02the Emory vaccine center at,
48:05has a lab at a
48:06very nice institute there called
48:08ICGV.
48:10So we have a and
48:11we've had this lab for
48:12many years. Our focus actually
48:14has been on dengue dengue
48:16virus infection.
48:17But now we are starting,
48:19we will start a program
48:20on,
48:21HPV
48:22cervical cancer, and this will
48:24be in collaboration with, one
48:26of the largest and the
48:27best hospitals in India, the
48:28AIMS Hospital in New Delhi.
48:30So, hopefully, in the next,
48:33few years, we will have
48:35some data on what responses
48:37look like in HPV
48:39positive cervical cancer patients.
48:41K. I will
48:43skip the IL two part.
48:47I should skip it. Right?
48:50I have time. Okay. I'll
48:51go through it quickly.
48:55Most of it is published,
48:56but it's it's I think
48:57it's relevant because there's some
48:58IL two.
49:00Okay.
49:02Alright. So I will go
49:04through this quickly. So
49:06so this I'm switching now
49:07to, rational design of, immunotherapy.
49:10So as I've mentioned to
49:11you, there are three of
49:13these subsets there.
49:14So what can we do
49:15to increase the number of
49:17stem like cells?
49:18Can we do something to
49:20change the quality of the
49:21effector cells that we get?
49:24And probably the hardest thing,
49:25can we do something to,
49:28get those truly terribly different
49:30to do something?
49:32But
49:33what I wanna talk to
49:34you today a bit I
49:35l two is improving the
49:37quality of vector cells. I
49:39l two doesn't do much
49:39in terms of expanding
49:41stem cells,
49:42and it doesn't do anything
49:44on the more differentiated cells.
49:45So I l two doesn't
49:46do anything with, with the
49:48cell.
49:49But what we found is
49:50that
49:53that when you combine
49:55IL two with PD one,
49:57and this was, published,
49:59a couple of years back,
50:00that we see a dramatic
50:02change
50:04in the transcriptional
50:05program
50:06of the effector cells.
50:10Basically, what we see is
50:12that when you do,
50:13actually, let me go to
50:15the
50:16let me go to this.
50:16Yeah. Let me go to
50:17this slide. Okay. So, basically,
50:19this is what the previous,
50:21data shows you, that when
50:22you do with an untreated,
50:26situation,
50:27you have the stem like
50:28cells, you have the transitory
50:29effector cells, and you have
50:31the term you differentiated. So
50:33after you do PD one
50:34blockade,
50:34you get many more of
50:36these cells, k, which eventually
50:37will go here. But when
50:39we did
50:40PD one blockade plus IL
50:42two, we totally changed
50:45the
50:45the program of these cells.
50:47So these stem like cells
50:50are have tremendous pluripotency,
50:53that is they're not fate
50:54locked into the kind of
50:55effectors they will get. So
50:57this creates another opportunity,
50:59not only to just get
51:01more of the blue cells,
51:02but also to change their
51:04transcriptional trajectory and epigenetically
51:07get much better effector cells.
51:08And that's what PD one
51:10plus IL two combination therapy
51:12does. Not only do you
51:14get many more cells, but
51:16they are transcriptionally
51:17and epigenetically
51:18very distinct.
51:20And what do they look
51:21like now? Now they start
51:22looking more like effector cells
51:25that you generate
51:26during an acute infection.
51:29So I think that this
51:30offers really
51:31very interesting opportunities
51:33in terms of further manipulating
51:35that stem like population.
51:40So then this was the
51:41part that was,
51:44at least, to me, more
51:45interesting, but I think more
51:46problematic
51:47for the field,
51:48k, is what we found
51:51in our the qualifier is
51:53that this
51:54is a mouse model. Okay?
51:56What what we found is
51:57that
51:58CD twenty five engagement
52:01was important.
52:03As you all know, the
52:03trimeric receptor
52:05is,
52:06is the alpha chain,
52:08there's the beta chain, and
52:09there's the gamma chain. The
52:11signaling is through the beta
52:12and the gamma chain.
52:15The alpha chain, which is
52:16CD twenty five,
52:17forms the trimeric receptor. The
52:19affinity of the trimeric receptor
52:21is long orders more
52:23than the affinity of beta
52:25and gamma.
52:26And our results
52:29were that if we use,
52:31so we did three kind
52:32of experiments.
52:33One was if we,
52:37if we do
52:38if we if we use
52:39a cytokine that was mutated,
52:42that is did not have
52:43binding to the alpha chain,
52:45we actually saw no synergy
52:47at all. In fact, that
52:49that basically, it was the
52:50sync effect. It just went
52:52and bound every T cell
52:53that was there,
52:54k, and was not targeted
52:56properly.
52:57And then we also showed
52:59that if you do c
53:00twenty five blockade,
53:02then the synergy
53:03was totally gone. That is
53:05using wild type l two
53:07and then doing blockade,
53:09completely gone.
53:10K.
53:12That the the critical way
53:13that cytokine was working was
53:15binding to the trimeric receptor.
53:17K.
53:21We will skip. Yeah. Again,
53:23this is what I told
53:23you. But but there are
53:25ways out of it. That
53:26is you can still use
53:28so as I told you
53:29to, and also that I
53:31skipped data. Twenty five comes
53:32up very quickly on the
53:33antigen specific cells in vivo
53:35after this combination therapy.
53:39Twenty five blockade abrogates this
53:40effect.
53:41But if you target it
53:43so we also did a
53:44study using,
53:46a a,
53:48molecule from Roche, which was
53:49an anti PD one antibody,
53:52which on which there was
53:53IL two. This was the
53:54mutated IL two, and this
53:56works also quite well.
53:58K. So now you're targeting
54:00it to the cells of
54:01interest, and then you see,
54:03you see an effect. K.
54:05And that was published there.
54:08So where do we stand
54:09now with with IL two
54:10therapy? So I think if
54:12you use,
54:14and, unfortunately,
54:15the failure of many cytokine,
54:18IL two cytokines is consistent
54:20with our findings,
54:21is that
54:22if you use a free
54:24cytokine
54:24that is just the cytokine,
54:26but that's mutated
54:28in the alpha chain to
54:29binding to c twenty five,
54:31it's unlikely to work. And
54:33several
54:35phase one
54:36several phase three clinical trials
54:38have failed.
54:39When
54:40a free cytokine is being
54:41used
54:42and the CD twenty five,
54:43either the either by pegylation,
54:46you you block signal binding,
54:48which was the nectar
54:49IL two.
54:50And other IL twos where
54:51this was, mutated,
54:53they have either stopped after
54:55phase one,
54:57or have,
54:58unfortunately, gone further and stopped
55:00later.
55:01But targeted therapy is one
55:03option, and this many people
55:05are using this now.
55:06Some are using CD eight
55:08targeting,
55:09but probably the PD one
55:10targeting is better.
55:11And, currently, there is a
55:13clinical trial that Regeneron is
55:15doing
55:16where they are also using
55:18a p d one targeted,
55:19but they're using wild type.
55:21So Regeneron,
55:23they just started a clinical
55:24trial in lung cancer and
55:25also other cancer.
55:27They're using what we did,
55:29but with the wild type
55:30l two.
55:31And, actually, that might even
55:32be better. Yeah.
55:34The issue with wild type
55:35I l two is the
55:36toxicity issue.
55:37But I think if it's
55:38more targeted,
55:39it can probably be regulate
55:41it can be better. But
55:42in addition,
55:43there's also now regulated IL
55:45twos being used. That is
55:46they will only become active
55:49after they engage with PD
55:50one. That will bring in
55:51another safety aspect to it.
55:54So I think that this
55:56field is going to see
55:57a lot of interesting data.
56:00IL two is, as I
56:01put in the first bullet
56:02point here, it is the
56:03key cytokine for effective differentiation.
56:06There is no other cytokine
56:08that we have
56:09that is better
56:10in terms of generating good
56:12effectors.
56:13In our mouse model, we
56:14have tried many, many different
56:16combination therapies,
56:18and the one that's the
56:19most superior
56:20is the IL two plus
56:22p d one.
56:24So I think that the
56:25field will now, I think,
56:26is in a in a
56:27place where,
56:29some attention is being paid
56:30to targeting the I l
56:32two and not just using
56:33free I l two that
56:34doesn't bind c d twenty
56:35five. That basically is a
56:37washout
56:38because every cell has the
56:40beta and the gamma chain,
56:43chain. So it binds to
56:44every cell.
56:45It's just a major sync
56:46effect. I think that strategy
56:48is pretty much gone.
56:49But targeting
56:50is a is a real,
56:52possibility,
56:53and then additional safety can
56:55come by
56:56regulating it so that it
56:58only becomes active after it
57:00binds PD one. So I
57:01think it's a exciting time
57:02for these studies.
57:04So to end my talk
57:06on the stem like chronic
57:07cells,
57:08I showed you data on
57:09the chronic infection
57:11and in cancer, both human
57:12and many people, others have
57:14shown this, but also in
57:15autoimmunity.
57:16So now there are several
57:17papers that have come out
57:19that in autoimmunity,
57:21also you have this subset.
57:24Part which which part of
57:25the t s seventeen pathway
57:27would be something bad. But,
57:28again, you need the cell.
57:30The key message here is
57:31that without the cell, you
57:33cannot maintain
57:34the t cell response, whether
57:36it's for good or it's
57:37for bad. This cell is
57:39absolutely essential to keep the
57:41engine going.
57:43So a lot of wonderful
57:44people in the lab. I
57:45mentioned their names when I
57:46was there discussing
57:48great collaboration on, head and
57:49neck cancer with Naveed Saba
57:51and Meeha Patel and many
57:53wonderful,
57:54external collaborators.
57:55Thank you.
58:05There are questions.
58:07Yes.
58:08So thank you for the
58:10talk.
58:11You mentioned that it's sub
58:12vector cells that connect the
58:13stem cells
58:14to the tumor differentiated cells,
58:28Ah, that's a good question.
58:30My we don't have direct
58:31evidence for that. K? It's
58:32a very good question. K?
58:34Whether they move before they
58:37differentiate,
58:37my prediction would be that,
58:40they start differentiating before they
58:42move because that program keeps
58:44them there.
58:45The the,
58:47the chemokine receptors change. So
58:50I think it's the differentiation
58:51step that then
58:53brings them out. And the
58:54key difference actually is down
58:55regulation of t c f
58:57one. And then TBET comes
58:59up. All these other things
59:00come up. So but some
59:01the initial signals for differentiation
59:03are happening in those hubs.
59:05K? But then once they
59:07differentiate, part of the program
59:08is you go out, which
59:09actually is not that different
59:11from what happens to a
59:12naive cell after infection or
59:14vaccination.
59:15K? The initial signals are
59:17in the lymph node in
59:18the t cell zones. Once
59:19they become effect like cells,
59:21they will come out.
59:27Before we talk Thank you.
59:28I was wondering about the
59:30tumor like mitogenics that we
59:31see. Do you also look
59:32at other cytotoxic effectors like
59:34natural killer cells? Are there
59:36a then new line stem
59:37cell?
59:38Yeah. We've been very focused
59:39on we've even been ignoring
59:41CD four cells.
59:43So we've been very focused
59:44on because, again, the,
59:48the tumor sample is small,
59:50the number of things we
59:51can do. So we've been
59:52very, very focused. But Andreas
59:53Wieland, who,
59:55was involved in both of
59:56these studies, he's on the
59:58faculty at Ohio State now,
60:00and he's putting a focus
01:00:01on CD four t cells
01:00:03and b cells.
01:00:05Can I ask you,
01:00:07this work was all done
01:00:08in treatment naive Yes? All
01:00:10done into yes.
01:00:11Cancer.
01:00:12Yes. If you know, in
01:00:13head neck cancer, a big
01:00:14problem is that we've removed
01:00:15all the lymph nodes in
01:00:17in many of our patients.
01:00:18Yes.
01:00:21And yet I think from
01:00:22what you were showing us
01:00:23in lung cancer and and
01:00:24from the fact that we
01:00:25see responses in people with
01:00:27recurrent disease,
01:00:29I assume that the effector
01:00:31t cells that we generate
01:00:32with pembrolizumab
01:00:33in a recurrent Yeah. Patient
01:00:35also Yeah. I know that
01:00:36you're raising a very
01:00:38well,
01:00:38you're raising a very important
01:00:39question is that when you
01:00:41do surgery,
01:00:43and you remove all the
01:00:44lymph nodes,
01:00:46how many of these stem
01:00:47like cells are remaining? I
01:00:48don't know the answer for
01:00:49that. You probably are removing
01:00:52a lot of them.
01:00:54But So I think it's
01:00:54very you can't leave the
01:00:56lymph nodes there because for
01:00:57obvious reasons, but you are,
01:00:59I think,
01:01:00because where are these cells
01:01:02as far as we know?
01:01:03They're in the tumor
01:01:05within these niches that,
01:01:07you know,
01:01:08Neil had to find and
01:01:09other people and we have
01:01:11shown,
01:01:12and they're in the draining
01:01:13nodes. That's been shown by
01:01:15many people.
01:01:16So if when you remove
01:01:18both compartments,
01:01:19you certainly are reducing the
01:01:21number of those cells.
01:01:23The the lung cancer that
01:01:24I told you, that was
01:01:25not there was no surgery.
01:01:26That was people getting PD
01:01:27one blockade.
01:01:30I'm just saying that there
01:01:31must be I'm I'm wondering
01:01:32if there's a process that
01:01:33recruits these TCS TCS one
01:01:35positive cells into Yeah. Right.
01:01:38Right. Right. So that brings
01:01:39up interesting questions about how
01:01:41do you in that setting,
01:01:42how do you increase the
01:01:44the frequency? I mean, one,
01:01:46thing would be therapeutic vaccination.
01:01:48And, hopefully, there's still some
01:01:50naive cells with that TCRs,
01:01:52which can be recruited.
01:01:54And I don't think it
01:01:55would be that all of
01:01:56the cells would be gone
01:01:57from the lymph nodes that
01:01:58you removed. K? So I
01:01:59think you may then need
01:02:00to provide some
01:02:02vaccination
01:02:03strategy
01:02:04for increasing their numbers.
01:02:06Yep. Numbers, I think, do
01:02:07go down after you do
01:02:08it.
01:02:09I can't imagine them not
01:02:11going down, right, because you're
01:02:12removing the reservoirs with these
01:02:14cells there.
01:02:31There probably are more states.
01:02:32Okay. I'm sure that somebody
01:02:33will find subsets of effectors
01:02:35and subsets of stem, but
01:02:37these are the three canonical,
01:02:39subsets. Yes.
01:02:47It's also talks positive.
01:02:50Yeah.
01:03:00It's happening continuously because when
01:03:02these
01:03:04cells are differentiating,
01:03:06they're giving rise to the
01:03:07effector cells, which remain, I
01:03:08think, good for several we
01:03:10don't know exactly.
01:03:11Might be
01:03:13days or weeks they'll be
01:03:14there,
01:03:15and then eventually,
01:03:17they will go to that
01:03:18terminal differential state. There's some
01:03:20people who think that they
01:03:21might be,
01:03:23depending on the environment,
01:03:25they might be generating more
01:03:26of the exhausted cells immediately.
01:03:29But our data would is
01:03:30more consistent with the, you
01:03:32know, it going through a
01:03:34early
01:03:35functional stage before they become
01:03:36exhausted, but both are possible.
01:03:39Yep.
01:03:40So in terms of
01:03:45when is each thing happening,
01:03:47once
01:03:48the system has kind of
01:03:50set in, basically, you have
01:03:51all three of these cells.
01:03:53Are you asking the question
01:03:54when are stem cells generated?
01:04:03Oh, the if they don't
01:04:04respond to therapy, then,
01:04:06many things could be wrong.
01:04:08You know, the tumor could
01:04:09be more aggressive. But from
01:04:11a t cell side, I
01:04:12think when a patient doesn't
01:04:13respond,
01:04:14I think the
01:04:16number of t cells to
01:04:17begin with are very low.
01:04:19See, PD one does not
01:04:21create any new t cell.
01:04:23None of the checkpoint,
01:04:25blockade strategies create a new
01:04:27t cell. They are trying
01:04:28to
01:04:30provide,
01:04:31more differentiation
01:04:32from an existing cell.
01:04:36Take one last question maybe
01:04:37from Brenda Emu, and then
01:04:39would would it be okay
01:04:40if people come up with
01:04:41Oh, yeah. I'll just I'll
01:04:41still I'll be I'll yeah.
01:04:55Is there something in the
01:04:56premalignant stage that there's a
01:04:58difference that that some patients
01:05:00lose some Yeah. That's a
01:05:01very good question. Again, in
01:05:02terms of how their initial
01:05:04responses were, how many cells
01:05:06were made. So
01:05:08and then, also, the question,
01:05:09when do you generate these
01:05:10stem? We have a full
01:05:12story on that. That's okay.
01:05:14But they're generated very early.
01:05:16Very early. So in actually,
01:05:18the so I would say
01:05:19that in the HPV
01:05:20infection,
01:05:22that this cell cell, which
01:05:24has some of these characteristics,
01:05:26would have been generated very
01:05:27early during the infection.
01:05:32It's status.
01:05:34If the antigen clears, then
01:05:36it will not have this
01:05:37phenotype.
01:05:38It'll then start looking more
01:05:39like a memory cell. K.
01:05:41But this cell population
01:05:43is generated very early,
01:05:46and, that's a paper that's
01:05:47coming out. I I didn't
01:05:48wanna include that talk in
01:05:49here.
01:05:50Let's see. But it's really
01:05:51quite nice because the immune
01:05:53system is actually
01:05:54preparing for generating that cell
01:05:57before even knowing the outcome.
01:06:00I mean, the paper that
01:06:00we have in press, and
01:06:02that would be
01:06:03a different seminar, is that
01:06:04even in the acute infection
01:06:06setting,
01:06:08you generate the cell population
01:06:10in small numbers.
01:06:12Because at that time, you
01:06:13don't know is the infection
01:06:15going to be chronic or
01:06:16acute. So there's this very
01:06:18early preparation
01:06:19by generating this, so this
01:06:21fake decision
01:06:22to give you the cell,
01:06:24which is, again, I emphasize
01:06:25Granzyme b negative
01:06:27and having the TFH flavor
01:06:29happens very quickly
01:06:30within the first week.
01:06:32And this cell is there
01:06:33even in the acute infant.
01:06:35Then you look later, it's
01:06:36gone. It's become part of
01:06:37the pool of the central
01:06:38memory cells.
01:06:39We don't know how much
01:06:40of that contributes to it,
01:06:41but that cell is gone.
01:06:42If you but but if
01:06:44the chronic infection
01:06:45ensues,
01:06:46then that is the key
01:06:47cell maintaining it.
01:06:49So I think the HPV,
01:06:50what you're asking is, I
01:06:51think this cell was generated
01:06:52when they first got,
01:06:54infected.
01:06:59Thank you so much. Thank
01:07:00you.