AGT’s Former CSO, C. David Pauza, PhD, Gives Keynote Speech at the Cell and Gene Meeting on The Mesa

TRANSCRIPTION

Speaker: C. David Pauza, Chief Science Officer
Duration: 26:50

“Jennifer, thanks so much and I don’t know what the Nobel Prize Laureate looked like but maybe I look a little bit like him. So, I will try to do my best. That’s our management team, I just want to introduce them because our CEO Jeff is also here and if you want to say hello and ask any questions about what do, that’s another person you can talk to. And then, I wanted to give a few comments on the methods for selecting substitute speakers because Mark Twain said that ‘it takes more than three weeks to prepare a good impromptu speech.’ My impromptu speech was three hours, and so, if you bear with me, with rough slides and incomplete fragment thoughts, we’ll get through to the end.

 

The second thing I would urge is, don’t tell the substitute speaker how many people you asked before them, that can be rather depressing, and in my case, it’s very appropriate this unknown author of this quote pick someone with bronchitis so that the audience will be merciful. And so, I’m begging your mercy and your attention, but I will be brief both for my bronchitis and for your attention span.

 

I just want to make some general comments about immunology, I want to put them in contexts of three little case studies. Two things that are actually related to our company because I know them well and I have slides for them. But also just to think a little bit about where immunology has been in the last 30 or 40 years and what aspects of that are really important to carry over into what we’re all doing, trying to develop cures. And we’re trying to, now, be very sophisticated in our manipulation of biological systems at a level that I don’t think was really imagined just a few years ago.

 

I’d like to just remind you of these three important rules of the road for immunology that, you have this rearranging T cell or B cell receptor repertoire which is massive and most of it is useless. And so it’s what we call an anticipatory repertoire. But we are locked into the numerology of this with the complexity of histocompatibility low side and the capacity for rearranging T or B cell receptors creates this enormous diversity. And a diversity where our alpha-beta T cell repertoire has not been repeated in another person in the history of man and will never be repeated.

 

The complexity is driven by this histocompatibility polymorphism, and it’s a very unique problem in immunology. How and why that arose in the first place? There’s been decades of immunogenetics studies, theoretical immunogenetic studies, trying to decipher this relationship and really, I don’t think we know that much more about it than we did years and years ago. It’s a complex dually selected locus and very, very interesting.

 

The other really important thing about the immune system, especially when you want to engineer it, it is a tissue that has two phases. It has a solid phase, those are what we called the lymphoid tissues, and then it has dispersed phase in blood. And of course, the dispersed phase doesn’t really have a function, it’s just basically calm down cells that are circulating for a few days. But what other tissues do you know that continuously samples itself and gives you access to that sample on a ready basis, so we can get a representative look into the entire immune system with the blood sample and occasionally tissues if you really focus on mucosal biology. But that is an extraordinary advantage that people have when they enter this field of work.

 

Really, I just want to say a word about somebody that Jennifer mentioned. And that’s Mel Cohn. The reason is not to laud Mel, Mel was a tremendous guy, and we’ve lauded him many times, but it’s to remind me to tell you that in the middle 80’s and the early 80’s, this area of San Diego, Torrey Pines area, was a phenomenal hotbed for the origin of quantitative immunology. And people were thinking here about numbers, about counting mechanism, about quantities, to take immunology from what had been a very qualitative science into a highly quantitative science. Everything we do, including Novartis’s big success, goes back to these guys who taught us how to count in the immune system. I don’t think you’ll be able to… I can’t, I thought I had it animated. But behind Mel’s picture is a program from his 80^th birthday meeting, which I was happy to attend at the Salk Institute in 1992, and it’s just a list of the pillars that were in quantitative immunology and most of them are from this area.

 

Mel was the first people to talk about a concept called “The Generator of Diversity” and really coined that term and taught us that the immune system is large, it’s got a very interesting function, and it’s highly clonal. These things are not what we expect from most tissues. I don’t know why a hepatocyte seems to have sharp edges or straight lines when I see it in histology, but I know where lymphocytes come and go, and I know how to work on them, and so that’s been my focus for a long time.

 

In Mel’s lab, we started a program on HIV in 1985. This was not welcomed by the Salk Institute. I had a very interesting experience in 1985, where I had been on the Salk faculty for two days, and I was dragged by Mel into a meeting with Frederic de Hoffmann, Clark Gable, Armand Hammer, and Vince DeVita, then the head of the National Cancer Institute, and a collection of people that I can’t remember. Frederic de Hoffmann absolutely refused to allow us to have an HIV program. But Mel was an original fellow, and the original fellows were little kings in the Salk Institute, so they had their fifty thousand square feet of space, and they can do whatever they want, and that saved me and allowed us to get started there. We did a lot of the early work on the molecularbiology of HIV, trying to find out how it got in and out of cells, what kind of DNA forms it had, and all kind of things. It was really interesting. And then I moved the University of Wisconsin because somebody offered me a primate center and a clinical program, and I said I can’t turn it down, that’s where I want it to go in the field.

 

But today I want to give you kind of three case studies about this numeral logic aspect of immunology, and as I mentioned before, these were generated on the fly today, so we’ll see how it goes. I want to talk about, as I said, two projects from the company first and second, and then I want to talk about CAR-T cells and how they play into this whole game as well.

 

So, the functional cure for HIV, this is the first counting lesson that I want to put out because we have such amazing sophistication about the human immune system and the intervention efforts that are going on. To modify lymphocytes, to bring complex vaccines forward, to do all kinds of stimulatory therapies, these have begun to build a storehouse of information about really, what are the limits of success in the immune system. We now can point to the realistic targets and say if we can achieve target X, we should be able to get function Y. And that relationship will continue to evolve. But from the engineering perspective and product development perspective, you absolutely need a quantitative target. Or else you can’t begin the developmental process, and you can iterate your program. We want to reduce the dependence on antiretroviral therapy in people with HIV, and we want to do it by a very specific approach.

 

We know that the critical challenge to curing HIV is that the body retains a reservoir of infected cells that are absolutely durable for years and years of very potent antiretroviral therapy. We know that those reservoir cells have some kind of consequence, right? Long after you get somebody virally suppressed with good therapy, HIV patients still have a 15 percent lifetime risk of cancer. They have extraordinary rates of metabolic disease, we have 28 years old patients with osteoporosis. The drugs are not reversing the comorbidities, and in fact, effective suppression of the virus does not reverse the comorbidities. And the only correlate that we know, between the susceptibility to comorbid disease and the immune system, is the numerical count of HIV-specific CD4 T cells.

 

So, when a virus infects people, it provides a huge burst of antigen. In the acute infection, it’s the highest levels of replication that the person will see throughout their period of time, throughout their life. That replication, of course, triggers a lot of T cell activation. And if those are CD4 T cells, they become prime targets for HIV infection and deletion. If you map the rates of deletion for different types of CD4 T cells, the HIV-specific ones are gone very quickly at a rate much, much faster than the bulk population. We’ve known this for a long time, and we also know the consequence of it. Because once you lose antigen-specific T cell help, you can no longer differentiate and evolve the cytotoxic T cell response, and you will fail to keep up with viral variation.

 

HIV is not a super mutating virus, it’s just like any old RNA virus. But it has this extra activity, it removed help, made the CTL generation inefficient, and you see that in patients as they go through disease. Early in disease, there’s a tremendous variety of the viral sequences, as soon as that CD4 T cell help is gone, you become almost monotonic in your viral sequences and stay that way for the rest of disease. And it doesn’t matter if you take that CD8 T cells out and put them back in therapeutically, the virus has escaped from those responses and they’re no longer functional, and your inability to generate the next one is where the problem lies. We’ve known this, and all of us tripped over this issue on our way to exotic therapies for HIV. And we at the company decided we’re going to make a concerted effort to do this and it’s a concerted effort to defeat this problem numerically.

 

So, we want to produce an autologous cell product that is highly enriched for these HIV-specific cells. So, if you’re in therapy for a long time, you have de novo generation of T cells, you will get some cells back that is capable of recognizing HIV. And they are measurable, but they are miserably small. I’ll show you some examples in a minute. We can expand them, they are functional cells. We can trigger them with vaccines in vivo, or we can expand them ex vivo with peptide stimulation, and I’ll show you that, that’s actually quite an efficient process. And if we can deliver enough of those cells, we should be able to prejudice the immune system against HIV, except for the one major problem. HIV then has the capacity to delete those cells we just spent our money and effort trying to amplify and inject, that’s where lentivirus vectors come in. You amplified these cells, get a highly enriched product, modify with a lentivirus vector that’s going to resist HIV and prevent depletion of that subset, and now you have a product where you have durable T cell help against a viral antigen.

 

It’s really important also just to remember one little fact about the difference between a CD4 response and a CD8 response, we all think of these things as highly parsimonious. The CD8 response is indeed parsimonious and can be knocked off by a justsingle mutation in an epitope. Not so for the CD4 response. The CD4 response is highly degenerate, it tolerates changes within the antigenic epitopes, and single epitopes work across a broad variety of MHC class II molecules. This is a key to our effort because we don’t have to devise or design matched set of stimuli for each person that walks into the door. We can use the same old batch of peptides, the same old vaccines, and it works for just about everybody.

 

So, we have to stack the deck, right? And we know from now, lots of numerology experience and it’s not only viral studies, bacterial studies, it’s also CAR-T studies are contributing massively to this. We know that there’s a magic number for central memory now, and that magic number is greater than about 0.1% of circulating cells. And if you can hit that number of antigen-specific circulating cells, you can reconstitute immunity. And HIV patients are woefully short of this. So, that’s why we devised this strategy.

 

It looks like this, we have the option in the strategy to include a therapeutic vaccination before we actually leukapheresis these individuals. And we will test that in our upcoming trial, but we haven’t committed to this really, but we can immunize then we can leukapheresis. And you’ll see numbers like 0.05% of circulating cells being responsive to gag peptides, and sometimes it can be much lower than that. But going through this process, enriching and making the product, you can get this up to 7.5 or more percent of the cells in the product, and we calculate that. When you infuse this product, that’ll get us to about 0.3% of total cells, a little loss for engraftment inefficiency and we’re in the ballpark of the magic number. So, we were able to generate a process for manufacturing these cells in 15 days that gets us very close to the target number. And certainly, is justified for a clinical study that we’ll start in quarter one of next year.

 

So, in the HIV case study, that was just bulk… you know, that would just smash a head,  just bulk cell delivery. But if you do it a little bit differently, what if you take a different way to overcome the numerological problem? Remember we said something before about the diversity, the massive diversity of the T cell receptor repertoire and CAR-T then comes in and just discards that entire issue. CAR-T then just comes in, modifies cells, gives them single homogeneous signaling molecules on the cell surface, recognizing single antigens, it destroys the recognition requirement for Class I, it destroys the recognition requirement for Class II, and then it has this ability to target tumors. In the recent news and the approval of the Novartis product is really fantastic for this industry.

 

But I want to point out a couple of other things that these guys did. Because immunology is counting cells, it’s counting signal strength, and it’s also counting cell concentration because cells don’t do anything in the periphery. We study them and test them, but it has no relationship to where they work in the body. In the body, they’re jammed together. And cytokines don’t work like hormones, they don’t travel through serum. Cytokines work like neurotransmitters, they’re sole cells are right next to each other, and they’re secreting and receiving signal from an adjacent neighbor. That’s like my neighborhood in Maryland, where the houses next door are practically in my living room. But when the CAR-T guys came in, they actually rolled the dice in what is a kind of dangerous game. They built these totally synthetic molecules, and this is one of the things I applaud the most about them is they had the guts to discard biology essentially and build new things out of their imagination. And they have released the rest of us in a way. They’ve given the rest of us the license to stop building analogues of biological molecules, and really just go for it. I mean, we all knew that you could move domains around, and cell biologists have been doing this forever and be very cleverly, but these guys put it in a person jammed it through to clinical approval, and that’s really phenomenal. But along the way, the problem was, they changed the signal strength. That signal strength then gave you phenomenal replication of those cells. And of course, they had problems in cases where they had a high dose of modified cells and a very high tumor burden, they got a cytokine response. They’re overcoming these problems and working through it, this is a therapy improving every day. But it’s very interesting to look at data like this from one of their papers. Patients that had a successful treatment with CAR-T, they’re all about 0.1% or higher or those cells in the blood.

 

And the problem with their signal strengths are also because these types of molecules have avidity issues in addition to just binding and signaling through the molecule. I’m not sure anybody really appreciated the risk of avidity problems when they did it, that’s where a little bit of the off-target effect is coming from, but it’s just a matter of more engineering and more sophistication; but phenomenal and showed that we can do this, we can modify cells in the immune system, we can make them completely synthetic, divorce them from every regulation that they ever had, and it’s a successful therapeutic. And the proof is in the pudding, you see tumors resolving in patients.

 

So, that’s the numerology now, where you play with the signal strength, and you play with the repertoire complexity. In many patients, the CAR-T transduction commits 1-5% of total T cells to having a single receptor. Now, if you would have told Mel Cohn and his friends that tell you never can work because the only you survive an immune response is because it knows how to stop. We always used to say, it’s a waste of time to teach students how to activate T cells, you should tell them how to kill them because that’s the dangerous part. And so, these guys said, ‘we don’t think it’s that dangerous’, and they’re right. And that was pretty amazing.

 

The third lesson that I want to talk about just for a second is something that we’re doing in tumor immunology, and that is, HIV we have to rebuild with a complex set of T cells and count on the T cell receptor recognition. In CAR-T, they just threw it away and made something else, but they’re stuck with the problem, and they have to know exactly the molecule they’re after to build that receptor, and it’s not going to be possible in all cases. So, we said, what if we want to do something innovative about tumors, what should we do?

 

So, this is the third lesson in counting. Let’s pretend that we’re not polymorphic for antigen presenting molecule. Let’s pretend we’re monomorphic, and how does that work in the immune system? Well, let me just tell you a little bit about this.

 

These are innate immune cells, innate T cells. They’re surveillance for damaged cells. They really detect errors in metabolism and are stimulated to be activated and proliferate by recognizing cells that have errors in metabolism. And tumor cells virally infected cells, very often have a massive up regulation of membrane synthesis. And all the elements in that membrane synthesis pathway are screaming on, and intermediates in those pathways can activate subsets of gammadelta T cells, which are our predominant T cell subset in peripheral blood. Now, gammadeltas have complexity, they have a rearranged receptor. But they don’t see a polymorphic antigen presenting molecule. So, all of us in this room, we share a number of gammadelta T cell receptor sequences down to the nucleotide. Because we’re monomorphic, we’re the inbred mice. Congratulations, you finally devolved. And when you have a system like that, you can figure out how to introduce the metabolic insults by a viral vector. Now, you’re able to do something quite unique. You can do this!

 

People have dreamed about viral vector therapy for cancer for a long time and direct vector administration. The problem was, you don’t hit very many cells, you hit a few. And so, everybody said, ‘Well, let’s put powerful cytokines on there, that will do it.’ It didn’t really do it. And twenty or so cytokines were tried clinical trials, they were vectored in by adenoviruses, adeno is a nice inflammatory inducing vector, and plus the cytokines, you got a whole immunological mess, but you didn’t really clean up tumors.

 

But in gammadelta cells, they have a very unique feature. Because of this, monomorphic presenting molecule. And that is, you have huge numbers of gammadelta T cells which recognize the same thing. It’s a five-carbon pyrophosphate from the mevalonate pathway. And you take that, it’s knucklehead immunology! You take the five-carbon pyrophosphate in a little IL-2 and PBMC, and you too can be a gammadelta T cells immunologist. It’s the easiest thing you can do in T cells. But the beauty of this is that they have a different set of receptors on the surface, which are NK receptors, they recognize the stimulating agent through the TCR, they recognized the tumor through the NKR. So, if you can just get a few cells programmed to stimulate their proliferation and activation, then they will kill all other cells in the tumor because they don’t use the T cell receptor, they don’t need the genetic modification to kill only to be stimulated. So far, this is working really well for us. We know these cells are enormously safe in the body, they never cause autoimmune disease. But this is another counting lesson of what happens when you have a system that thinks where monomorphic? It really changes the rules quite significantly and gives us a numerical advantage. All of you sitting here, about 1 in 40 of your memory T cells will react to IPP, this five-carbon pyrophosphate. It’s an enormous number, it’s not widely appreciated that this works. But this is your primary immune response in your body actually.

 

So, that’s really what I wanted to present was 3 aspects of problems and potential solutions for important clinical challenges. And I really want to pay homage back to those guys, Mel and all the other guys that were here in the 80’s, they were bunch buckets of them at Scripps and tons of them at UCSD. And there used to be these weekly immunological journal clubs, which usually started out nice and ended up by fights and throwing things. And it was really a barrel of laughs. And it was one of the most fun things about my whole scientific career was getting into those things and having really smart people using foul language. So, what’s going to happen next? We just scratched the surface of this, of taking this quantitative view of immunity and how we can manipulate these things for important next steps in the development of therapies. And I want to really emphasize my point I said before, no other tissue provides you a sample of itself all the time and continuously updates that sample. So, when you’re thinking of ‘just what am I going to do with my life,’ you might like to program T cells.

 

So, dosing, in all these regards, is a numbers game, but dosing depends also on signal strength. To an immunologist, activating cell proliferation and activating cell death are the same thing. We don’t think that’s different. We just think they’re varied outcomes of the same event. But if you start changing that balance by the way you stimulate, and the signal strength you provide, you could start doubling your weight every three or four days because of the enormous replication capacity of a T cell. A good old T cell going full speed will divide about every six to eight hours. So, it’ll take you over pretty quickly.

 

We have to really understand how to control the durability of these effector cells in places where they have off-target effects. And there’s a tremendous effort going on now about that, but there’s a lot of room for creativity there.

 

We have to know how to choose the specificity correctly. Are we going to engineer T cells and grind through this process and engineering, engineering, and engineering? We will find all the examples where engineering is going to work the best. And solid tumors will yield to these CAR-T type cells just as liquid tumors are yielding.

 

But sometimes, we may want to go to these innate systems because of enhanced safety and the fact that they don’t really make mistakes. Nobody’s engineering other white blood cell populations. I mean there’s a bunch of them, they might all be interesting. We kind of ignore them because, you know, starting in the 80’s, immunologists became fascinated with molecular immunology, and so the rearranging receptor became the de rigueur, and it still is today. But there’s a lot of cells in there that do really important functions, and we haven’t even started to dig around in there.

 

And how could we manipulate the bone marrow in new creative ways to drive the differentiation of these populations along the exact line we want? When are going to get to in vivo dosing?

 

So, these are my points, and I’m not going to do any of the rest of that. That’s something else.

And I’d like to just thank Jennifer for the kind invitation to speak. I’m sorry it was such a hurry, rushed, and patched together talk. But I had fun with it, and I hope you did too. Thank you very much.”