Founder Stories: Michael Alonso, Immuno-Oncology Therapeutics

This is episode 29 of the Immunology Podcast, Immuno-Oncology Therapeutics with Dr. Michael Alonso. Hey everyone, this is Dr. Jason Goldsmith and Dr. Brenda Rao. Welcome back to Immunology Podcast. We have conversations with immunologists. Immunology Podcast is brought to you by Stem Cell Technologies, a global biotechnology company supporting life science research and fostering communication and collaboration in science. Today I have Dr. Michael Alonso on the podcast to talk about developing immunotherapies that activate myeloid cells. We've also got our usual roundup of recent highlights in immunology news coming up.

But first, if you enjoy the interviews on this podcast, we invite you to read more interviews on the Stem Cell Technologies website. In the Immunology Profiles series, immunologists tell their stories, discuss their research, and voice their thoughts and opinions on current topics in immunology. You can find these interviews at stemcell.com/immunoprofiles.

Hey, Brenda, how are you?

Hey, Jason, doing great. Doing great.

How are you holding up in conference season? I know we just got done with Immunology 2022.

It's nice to be updated with AAI 2022, but I was also somewhere else. You know, in Europe we also have some immunology-related meetings. So I was actually also attending a cancer immunotherapy conference last week. Pretty nice.

That was awesome. And I think you're going to talk about that here in a bit, right?

Yeah. Yeah. I have some, some highlights I just want to share with you. I know, I know you'll be interested.

Are they about enterocytes? No.

No. Sorry.

That's okay. I'm getting my enterocyte fill at the end of the week because I'm going to DDW, which is Digestive Disease Week, where I'll be sure to bring back some awesome mucosal immunology for all, all our listeners.

Can't wait. Wait, I'm looking forward. But regarding AAI 2022 meeting, so for all the listeners, remember we have 3 interviews with presenters at the conference, our 3 previous episodes. So if you want to go over those or you missed the meeting and want to get a sneak peek, make sure to check them out, right?

Absolutely. And we'll hope to see people next year in person at the meeting. Um, yes, maybe even have one of those big microphones where we can record people from the floor on the hall.

Wouldn't that be fun? Just walking around and asking people how they're doing and having a little conversation.

We could expand out our favorite immune cell question for sure.

We should carry out a proper survey. Uh, so many plans, so many things.

Absolutely. Get an IRB approval, get some p-values in there, you know. Really?

Yes. Yes, that's— that makes it all publishable.

Exactly. And you got to do right.

Indeed. Anyway, talking about publishing, why don't we talk about other people are publishing around the immunology, uh, you know, research sphere? They have something to bring to the table today.

I have a very interesting convergence of fields to get this one. It's called Fundamental Immuno— on Immune Oncogenicity Trade-offs Define Driver Mutation Fitness. First author is David Hoyos. Last author is Benjamin D. Greenbaum. It's in Nature. It came out the 11th of May. So this is a really interesting paper because it goes, okay, not proteins get mutated and those mutations then can lead to cancer. Okay. Some mutations though are so screwed up they kill a cell. So we know that's bad. So there's this mutational fitness where you want the mutation to promote cancer, but not kill the cell. So that's mutational fitness, right? How good is it at making it cancerous, survive the environment, do all those things. But we also know that mutated proteins are immunogenic and that can cause your immune system to detect them and then kill them. So what they did is using the p53 gene as a kind of core principle, they modeled out using, you know, real-world data from the Cancer Genome Atlas and then patient samples and then a bunch of cell studies that I'll get to. They modeled out and demonstrated that there's a fit curve to us, right? That maybe sometimes you have a really mutated cancer that's probably better at being cancer, a better protein that's better at making cancer, but it's going to die from the immune system. And so it's not what you see in most cancers. Simultaneously, you could have something that really evades immune checkpoint, right? Like it doesn't prevent, cause an antigen presentation that drives a lot of immunogenicity. But if the mutation is crap, it's also not going to cause cancer. And so they develop a fit model that can predict that combines these and then they demonstrate the fit model does indeed show the patterns of mutation most likely seen in people. And so looking at p53, right? So they use that as mutation distribution in a model system. And then they kind of used a couple of classic, a couple of well-known mutations. They found one mutation that wasn't super oncogenic, but it really evaded immune detection. And it was, you know, and they showed that if you took, you know, this protein and presented it to cells that the, you know, prime cells that the HLA presentation did not generate a strong response. And so they modeled that out in cell systems and similarly showed that also the self, this mutation, for instance, wasn't that oncogenic. So they really modeled it step by step for looking at the ability of these MHCs to present things.. And then so they made this fitness model and then let me hop in here. They applied it to a couple other genes, KRAS and PTEN, but they kind of paused there at one point and said, okay, cool. So we've shown this by principle. And then they showed they did what's called Pareto optimization, which is basically like at a certain point, further optimization on one curve doesn't get you anything and the cost optimization is worse than the gain.. So they, they, they show that it falls under something that's called Pareto optimal. The other example though, they showed there's these other mutations. So for, there's a couple of mutations. R175H is the most wild-like top, wild-like hotspot, right? So it's pretty similar to wild-type in its function, but has terrible MHC binding capacity, which is why it's a hotspot. Alternatively, R248Q and R248W mutations have complete, nearly completely loss of transcriptional function for p53. So they're really mutationed, right? They have a lot of cancer potential, but can generate neoantigens pretty well. And so the competitive proliferative management has to offset the immunogenicity. So they show these balances and they show that the different, it's a different calculus, right? So you can't weigh things equally. Cancer potential immunogenicity for KRAS, they do for p53. They also show that it's cell type specific. So they show, I think that, that a lot of in vitro work and then they hop in and look at patients and show that this comes up in patients with the same pattern. And I think they look in lung cancer cells and demonstrate that there in lung cancer cells, it's a little bit different balance, right? Depending on the type of cancer and how much immune cells they're exposed to at baseline. Right. So a cancer tissue that is immune privileged can afford to have a more mutation immunogenetic, immunogenic, or immunogenetic mutation in it because the cancer functions are outdrive it and then it's escaped. So they kind of go through this whole thing and map it out. And it's a pretty neat model that they can do. I thought it was interesting it made Nature because it kind of seems obvious. But then again, no one's ever actually demonstrated this pretty well, which is probably why it got there. But this trade-off was kind of intuitive, which I like. I like when my science makes sense to me when I think about it. They also showed that immune selective pressures matters a lot less in non-cancerous lesions than in tumor. And so they're suggesting immunotherapy for prophylaxis may be important in some genetic conditions where you have a lot of mutations pop up. So if you can get rid of it, then you're not— and then you need to drive the immune system to find those early, right? Versus the cancer where the immunogenicity is really predictive.

So basically it's stating that those potential mutations that would be immunogenic are selected out of cancers that you actually can see and measure.

It depends. It's a, it's an advantage, right? So if you're really really oncogenic, right? And you may, yeah, on a growth, you may overwhelm the immune thing. And that was their example, right? You can have, you can have a mutation that's almost wild type, then it escapes immune system, or you can have one that's just crazy. The immune system will fight, but it doesn't matter cuz you're already screwed. Right. Basically. And it also depends on the tissue type and how much immune cells the tissue's exposed to.

I wonder, do they compare like cancers from people from, from different, like kind of backgrounds and they are presenting like they're characterized by different MHC molecules. For example, you know, Caucasians versus—

yeah, they did one of the more classic MHCs in here. I forget which one it was in their in vitro.

HLA-A2, for example, is very, very predominant in European people of European descent, but it's not as common in, for example, yeah, they use HLA-A2-1. Yeah, that's like Yeah, the most commonly studied.

Yeah, they just wanted to go off the base system. But yeah, presumably if different HLA is binding things differently, you'll have different outcomes in different people, which is kind of the point.

Okay.

Yeah, I guess what you say makes it sounds— it sounds, yeah, rational. I guess it doesn't feel like terribly extraordinary conclusions, but I think it's really nice to have it really looked at carefully and data to really support this idea. Otherwise it's just, well, a feeling. Of what should make sense.

Yep. All right, well, I think it's your turn.

Yeah. Um, so I want to start first, I want to talk about a paper, uh, that I really want to discuss today. I have to say it's a little bit— it's been published already some weeks ago, so it's a little bit not so hot off the press as usually we have, but I thought it was such an interesting publication and it had so much attention, so many people kind of discussing it that I thought I really wanted to get in that discussion as well. So I'm going to talk today about a paper called Reversal of the T cell immune system reveals the molecular basis for T cell lineage fate determination in the thymus, published in Nature Immunology on the 29th of April. It's from first author Miho Shinsawa, and from the lab of Alfred Singer at the NIH, who has been working with T-cell development for a while. And I thought it was super cool because basically what they did in this, in the study is they generate a mouse which they call the Flip-Flop mouse, in which they, they replaced inside the loci of the co-receptor CD4 and CD8, they replaced the actual gene being encoded, actual protein being encoded in each of these loci for each other. So they got mice that expressed the co-receptor CD4 under the influence of the locus of CD8. And they had a CD8 receptor that was— so it had a kind of a double gene placed into encoding for the two subunits of the CD8 receptor encoded inside the CD4 locus. And they looked at what happened. And I think the results were super interesting and really made a lot of people kind of take pause and rethink or like kind of think again, what is our understanding of how T cell in the thymus, development in the thymus happens? So, I just want to say very quickly a couple of kind of starters, things that are kind of facts that we need to keep in mind to understand this paper. So, within the thymus and thymic development of T cells, there's this stage in which they are expressing both CD4 and CD8 receptors.

They're called—

this is the double-positive stage. And in this stage, T cells are also having a rearranged TCR, and they are interacting with MHC molecules in the stroma of the thymus. And we know that from that, at some point you will end up having cells that have either a CD4 receptor or a CD8. And cells that have a CD4 receptor recognize antigens that are presented on MHC II, and cells that have a CD8 receptor in MHC I, because that's what the co-receptors combine to. Between these two stages, we do also know that at a first— in a first place, T cells receive stimulation but with both receptors. And then they tend to downregulate the CD8 receptor. And then in some cases, CD4 cells just conserve the CD4 receptor and continue their CD4 journey. Or if the TCR signaling is interrupted during the stage, then the CD8 the CD4 gets downregulated and the CD8 comes back again on the surface, and these cells become CD8 cells, MHC I restricted. So what the author showed is that if they do the opposite, if they have a CD4 molecule on a CD8 locus and the other way around, what they actually see is exactly the opposite. They see cells that first double positive, get stimulus, and then they downregulate CD4, the CD4 molecule, so that the protein gets downregulated. These cells are only expressing CD8. And then they either keep the CD8 or they downregulate the CD8 and upregulate the CD4. And you end up having CD8 cells that have a CD4 locus being expressed and the other way around. And this was very interesting. So what is— what's very interesting is that when they count the numbers within the thymus development, they have roughly the same amount of cells that you would expect from wild-type mice from either of the CD4, CD8 cells. But then in the periphery, they already saw that CD8 cells became a stark minority. They see very interestingly, they see that the transcription factors that characterize CD4 cells and CD8, so usually THPOC, CD40, and then a transcription factor, and then CD40L, GATA3. These molecules are related to CD4 identity. They are expressed in cells that are actually expressing a CD8 receptor. And the other way around, CD4 cells express RUNX3, perforin, IOMES, overall cytotoxic phenotype, and a transcriptome that is associated with cytotoxic CD8 cells in normal wild-type mice.

Various.

So what they see is that when it comes to— so the signaling that is mediated by these two receptors is very interesting because normally CD4 cells, CD4 receptor is known for having a stronger signaling. And so there's always been this idea, this was for a long time this idea that Because the CD4 is a stronger receptor, this was the strength of this receptor that was kind of guiding the CD4 development in opposition to the CD8 development for cells that were MHC II preferential, so to say. But what they see is that when they express CD8— so this is— I just need a moment because this is very quite critical. When they have the mouse that has a transient CD8 expression, and then which results in an acute decline of CD4 protein. These cells result in a higher expression of a marker called CD5, which is related to kind of TCR signaling strength. But they also see that the actual LCK and the actual amount of CD8 proteins on the surface of these cells is lower. So they actually think that the signal is weaker. So CD5 expression, which was always kind of associated to the signaling strength, in this case, what they suggest is that it's not about the strength of the signal, but about the amount of time that happens when the cells downregulate the other locus, keep this locus, in this case is the locus of CD4, but coding CD8, And they show— and so they really show in a way that it's not about signaling strength, but it's about the signaling time. And it is the extension of the CD8 signaling in this case, which is in normal mice is meant to be CD4 signaling and not the actual strength of the receptor that dictates the diversion of the cell to a CD4, to a T helper phenotype because you end up having CD8 cells that are helper cells. They're expressing a lot of the hallmarks of CD4 cells. Not only that, within the CD8 positive cells, you have FOXP3 expressing T cells. They're not within the CD4 expressing cells, but within the CD8 compartment. And very interestingly, these cells can, well, this is that can mediate when in the cellular level, they can mediate, for example, graft rejection. So it suggests that this, and the cellular level, these cells can be cytotoxic, that they can act. But there are these CD8 helper cells are incapable of mediating, for example, B-cell responses. They cannot help B-cells. So these mice with this, mice with this Flip-flop mice have deficient humoral responses, B cells that don't class switch or have affinity maturation, and there's no germinal centers in these mice. They suggest that it's because as B cells cannot cross-present antigen through MHC I, which is what CD8 molecule can bind to, the helper, this kind of Flip-flop helper cells cannot help B cells because they cannot interact with each other through MHC I in the right way. Similarly, these mice cannot clear viral infection. And again, what I thought was very interesting is that they show that although CD8— one of the problems is that because CD4 MHC II-restricted cells are cytotoxic in these flip-flop mice, they seem to be restricting or reducing the immune response by actually killing antigen-presenting cells that are presenting the viral antigens through the MHC II, which in normal situations it would be presenting to helper cells and not cytotoxic cells. But in flip-flop mice, again, the roles are inverted. So I thought it was— this is a very cool paper. It took me a while to read it and kind of understand It's very complex. But I think what the authors really, really thought is that this model helped to answer the question, what is driving T cell differentiation, CD4 versus CD8? And they show that it's not about the receptor itself, but it's about the fact that one of the loci is designed to keep expressing for longer. And that is what really is driving the length of the stimulation for one of the loci is what's driving the differentiation in normal mice into CD4 cells that are helper. And then they're going to be expressing all these transcription factors related to helper cells and the other cells go into the CD8 kind of pathway.

So it sounds like the cell surface receptor is not the driver of differentiation. So even if you make that decision. Yeah, it's the other— it's the other epigenetic changes in the loci that are earlier than that. And that— but the CD4 and CD8 functions and what they can bind to for all the good immunological glory we know about, those are really important for that. And if you swap the top without the bottom, you get nothing that works.

Exactly. You have mice that cannot really mediate a response to any of the— to the pathogens or the situations that they expose them to. So it does seem that A cell that in order to have a functional T cell compartment, you need helper cells that recognize antigen in MHC class II and not MHC class I. So you need to make sure those two things coincide. And on the opposite side, you need to have CD8 cells that are recognizing antigen in MHC I that have a cytotoxic phenotype and not the other way around. So this is very critical. I know, again, sounds kind of obvious, but it's really nice to see it kind of really shown in this mouse and seen with so much detail. It's a very interesting read.

That is very good. So, all right. I don't have a good segue here, so I'll just guess I'll dive into it. It's a gut paper, though, so it's all good. It's actually an immunity, but it's about guts. All right. So it's the gut microbiota primes systemic antiviral immunity via the C. gas sting interferon-1 axis. First author is Saskia Ertman. Last author is Nelson O. Gekara. It's in Immunity. It came out the 10th of May. And I thought this is really interesting. So long and short, type 1 interferons protect against viruses, right? Yes. Cool. There's different pathways that can activate them. One is TLR. There's the RNA sensing pathway, and then there's the DNA sensing pathway of, you know, sea gas and sting. And it's been known that the microbiome needs to be present to have normal responses to viruses. So it's known that the microbiome, and this is also true in cancer, right? The microbiome is responsible for priming the immune system in such a way to maintain immune responses to bad things. Such as viruses and also to cancer and too much of it. Also, the microbiome can be primed to decide to go destroy your own intestinal tract. So there's a balance. What they show though is that if you do antibiotic ablation and followed by HSV infection, herpes simplex virus, the microbiome is really important for resisting herpes simplex virus infection. Cool. They show that, they establish that. And then they make, and Brenda is a postdoc, I think you will really appreciate this. They make, uh, I think 7 or 8 different knockout mice. Mighty88, TiCAM1, interferon beta with a luciferase trans, uh, they have a reporter luciferase, uh, for interferon beta. MAVE, MVS knockout, STING knockout, interferon. So they made like 3 or 4 mice and then mated them all with each other to have knockouts of everything along the pathway of the different pathways. So you can knock out the TLRs or the TLRs and STING or the RNA sensing gene, right?

They did it all. Oh my God, those breeding schemes must be a nightmare.

Although interestingly, if they started breeding them, whatever you got, since they wanted all of the flavors of ice cream, would be valid.

They just, they just got all the litters.

They just tested whatever they had. They used kind of what, like, we're going to breed them all together. But like, yeah, I would love to see the, uh, yeah, that breeding scheme. Yeah, that's other ones for RNAs, intracellular RIG-1-like receptors, RLRs. So, okay, holy God, they did, they did this. All right, so 7 years later when they had their mice— no, just joking. Um, but they basically showed that if you knocked out Sting or cGas, but starting with Sting, knocked out Sting, but none of the other ones, you lose the protection from HSV from the microbiota. But TLR, Mighty 88, MAVS, don't matter, doesn't care. TKAM1, only thing that matters was Sting. So it's saying it's a cGas-Sting pathway. So, uh, wasn't related to endogenous retrovirus signaling. Um, adjusting, and it works without cellular back. So, but they're like, how does the bacteria from the gut affect systemic HSV infection? So they looked at this and inoculated E. coli. In the blood or, you know, injected or the peritoneal, sorry, injections of E. coli or E. coli lacking an LPS that stimulates TLR4 signaling. So there's this mutant that doesn't activate TLR and show that, oh, the mutant doesn't make mice nearly as sick. But both the one that doesn't affect TLR, you know, where it's TLR signaling null and regular E. coli. Both equally protected against HSV signaling. So something about the bacteria that doesn't involve cellular, cellular contact is working. So enter, uh, MVs, which are, uh, membrane vesicles from bacteria are what's doing it. And they show this, uh, they can purify them, generate the effect. They can purify them from the weird E. coli, generate the same protection. It's all sting-mediated. They did, you know, cell experiments where they had the bacteria up top in a chamber and then the cells in the bottom and show the effect works. You don't have to have contact. They— I mean, they did all types of stuff. So they showed there's DNA inside of it. By just agarose gel. They show that if you precoat them, the MVs, with DNase and then RNase, right? You know, this, this pan-nucleus, and then do the experiment, it still works because the lipid protects the DNA on the inside, right? So no matter what you do, you have the effect. Um, and then they're like, you know, DNA bacteria can produce psychodinucleotides that may activate sting directly without C-gas. They show that it's sea gas dependent. They do that next, showing that, you know, if you don't knock out— if you knock out sea gas, you lose the effect.

Okay. So these vesicles are being released by this bacteria.

Then they transport into the circulation where they prime the immune system to have upregulated Sting and thus interfere on one signaling to protect you against viruses. And it works against RNA viruses too. So they say, well, it works against DNA viruses. It also primes you against RNA viruses because it generally increases your interferon-1 gamma tone, your interferon-1 tone, right? Yeah. Interferon-1 tone goes up, which means you have more viral immunity.

It's pretty cool because I guess, you know, there's also lately a lot of research on sting activation in cancer as well that seems to activate the immune system and might be beneficial for cancer responses. And so I wonder, yeah, if this is a common theme that having this double-stranded DNA being somehow delivered there and keeping the interferon response active has a kind of protective role.

It's very interesting. Yeah, super interesting. All right. So you have like a whole conference to present right now, right?

Yeah, well, I'm not going to present the whole conference. There was a lot of very interesting talks. So again, the Cancer Immunotherapy Conference, it was in Germany, in the heart of Europe. Very nice. I took the train, beautiful weather, and I just thought, do something different today. I just want to very quickly tell you about 3 stories that I thought were really cool. I tried to keep it short, but inspiring, interesting. And so I'm going to mention, first of all, Megan Berger was actually the first talk of the conference. It was 2.5 days. And I thought it was really cool because she had— she's at Koch Institute at the MIT, and she's opening her own lab very soon, location unknown at the moment, undisclosed. But she studied a model of lung adenocarcinoma in the mice. And that was already at her lab, had been used before for other studies. But this lung tumor model has basically mice that are KRAS mutated, and they also have floxed p53. And so by treating mice with viral vectors that are expressing the Cre recombinase, you can induce lung adenocarcinoma within 16 weeks in these mice. And this vector is interesting. It also contains two antigenic peptides included in the reading frame. So you have mice that have antigenically, kind of potently antigenic tumor cells that you can test because they have ovalbumin peptide and also a synthetic peptide that is also known to be very immunogenic. So what was— she looked at this model because here you have two different antigens being presented at kind of roughly the same— sorry, they're being expressed by this vector at the same concentration. But what's very interesting is that very quickly she realized that one of the peptides drives the dominant immune response. So in this case, it's the ovalbumin. So if you put these two peptides, basically mostly the ovalbumin generates T cells, specific T cells, and is the one that's driving the immune response. And so long story short, she shows that when you have this situation where you have multiple antigens being generated, that there is this antigenic dominance that has to do often with the fact that one of the peptides is presented more efficiently, and therefore it kind of dampens the response against the other peptide. But the bottom line of her— and sorry, and this subdominant antigen can actually drive suboptimal T cell responses, T cells that are characterized by the expression of TCR7 and TCF1. She names them TC17 because they also have some kind of Th17 properties. But what she suggests is that this subdominant response can actually be activated and harnessed with a vaccination, therapeutic vaccination. So in this mice, by adding therapeutic vaccination against this subdominant peptide, she can kind of bring back this response to bring it up to higher levels and break this suboptimal response to the subdominant peptide, suggesting that if you have many peptide, many potential antigens, you can kind of help out the underdog by using therapeutic vaccination. Really interesting story. And, and then there was another talk I really liked from Roby Meissner. He is assistant professor of pediatrics at Stanford University, and he works on immunotherapies for pediatric cancer, particularly CAR-T cells. And he had a very cool talk in which he went over the design of CARs for CAR-T cell therapy and the fact that we, all of the CARs that are used in clinics nowadays have basically the CD3 zeta, so what's the word, protein for signaling. So as part, which is part of the TCR signaling complex. And this, when these CARs were designed, they were thought, well, you know, that's CD37 is what's around the TCR, so this must be, you know, the best way of imitating T cell, TCR signaling. And what he does is he does a CRISPR knockout of all of the different members of the TCR signaling complex to identify which of those are really critical. And he shows that basically all of them except for Fyn are critical for the TCR signaling. And what he does is then he takes all of these different proteins, all of these different domains, and he makes CARs encoding, like, instead of CD3zeta, which is what all the other CARs have, he tries each of these different ones. And he finds that SAP70, which is one of the downstream kinases from the TCR/CD3, that is downstream from TCR/CD3, actually gives the strongest signal upon stimulation of the CAR, of the kind of peptide-binding part of the CAR on the outside. And he compares it with like a classical 4-1BB CD3 zeta CAR, and they show that they proliferate more. And I think what's very interesting, which is a huge, I think a huge issue when it comes to CAR-T cells, is that they have a lower baseline activation, so lower tonic signaling. And this really reduces the tendency of CARs to kind of become exhausted and start presenting exhaustion markers. So he thinks that this could be a better replacement for our good old CD3 theta. And in a different design, he goes even further down the signaling pathway and he focused on two molecules that are also downstream of SAB70. LAT, which is this large kind of— it's a molecule that acts kind of as an anchor for other peptides or other proteins to bind to it and kinases. And then SLP76, which is a kinase that is part of this signaling pathway. And actually, it's their proximity that is one of the downstream events from TCR signaling. So he suggests that you can have a heterologous CAR that has two different antibody kind of binding molecules. For example, again, CD19 and HER2, that's the example that he gives. And that when they bind to the molecules on the surface of cancer cells, they bring together these two elements. And that's what initiates the signaling downstream and the activation of the cell. Really cool because then it would allow you to be a bit more specific on which are the cells that you're targeting. I think it has a lot of, a lot of, it's a very interesting concept. So I really like his talk as well. So next generation CAR T-cell design, really cool. And following on the topic of CARs, what I thought was really cool and I And I thought was very interesting is not CARs in the real life, CAR T-cell therapy against solid tumors. So far, really disappointing results. So there's this clinical trial from BioNTech in which they have a Cloudin-6-specific CAR T-cell. Cloudin-6 is expressed on, not expressed on healthy islet tissue, but very highly, very commonly expressed on prostate cancer— sorry, testicular cancer, ovarian, and other types of cancer. And so what they did in this study is that they have the SCARs targeting claudin, but on top of that, being BioNTech, they treat the patients with therapeutic vaccinations in the form of mRNA of claudin-6 in lipoprotein, in liposomes. So they can kind of keep the activation of the Claudin-6 CAR-Ts and that keeps the cells for longer active. And long story short, they have seen really, really interesting results. So they have had response rates of 86% when it comes to kind of both, so the combination of partial responses and stable disease on patients that were previously progressing, which is one of the best results that we've seen so far for CARs for solid tumors. And these are the— they're the first ones to do this double therapy, CARs plus vaccination. And disclaimer, some of the patients were treated at my— by some of my colleagues. So that was really nice to see the data. But So this might be a new era on CAR-T cells for solid tumors.

I was going to ask you if you could only list one favorite thing to go in one sentence or less, what it would be. But if you don't say what your coworkers did, you're going to be in trouble.

So I won't ask that. Yeah, it was always so good. Everything was so nice. You know what I liked? They have really good wine in that region of Germany. So it was really nice. Riesling?

Riesling. Exactly. All right, well, we're going to be speaking with Dr. Michael Alonso in just a moment, but before we get to that, explore scientific resources for your immunology research at the Stem Cells Technology Immunology Learning Center. Choose from different research areas and find expert interviews, technical tips, educational webinars, instructional videos, and much more.

Visit stemcell.com/immunology-research. Hi everyone, talking to us today is Dr. Michael Alonso. He is scientific co-founder and vice president of immunology and pharmacology at Bold Biotherapeutics, which is a company that develops cancer immunotherapeutics that activate myeloid cells. He is going to talk to us about his experience at the company and the science behind their products. Welcome to the Immunology Podcast, Michael.

It's a pleasure to be here. I love what you guys are doing with the podcast and also a big fan of stem cell technology. I've been using the Rosette set for the better part of the last 15 years. So really nice to be here.

Awesome. Thank you for coming. All right, Brenda, I'm going to dive in and start. I think you have an interesting story, Michael. Maybe you could start there, and that your grad school mentor is a co-founder of Bolt, and then you went off pretty early in your scientific career to industry and biotech, and you've been at the same place for most of that career then. So what, what has that been like? How did you What did you learn in lab or discover as that grad student slash maybe some postdoc time? I know how it gets nebulous. And then you hopped over and now you're, you know, in a clinical stage biotech.

Yeah, I'd say there was a lot of drinking through the fire hose, which has been extremely gratifying and invigorating at the same time. Certainly from our studies in the lab, right, this started off with a 2015 Nature publication. Where, you know, we have all these tumor models that are, you know, for the most part, you know, barren of T cells but have all of these myeloid cells present. And the big trick is, you know, how do you turn this franchise around and start to kill the tumor versus letting the tumor grow as more of a wound healing response? And it actually turns out that the, you know, that it was rather simple how to turn it around. You have to add two ingredients: anti-tumor antibody an immune stimulant. What does the anti-tumor antibody do? It essentially covers the tumor in eat me signals. The Fc is just a bunch of eat me signals and then your phagocytes or your Pac-Man of the immune response can come in and they can phagocytose the tumor. What happens when you phagocytose the tumor in the absence of immune stimulant? A little bit, not much. But if you add immune stimulant, now you've really got something to where you're upregulating co-stimulatory molecules. And you're chopping the tumor into itty-bitty pieces that you're going to present it on the surface of the peptide MHC that activates the T cells. And with that fundamental finding, we thought we had the keys to the kingdom. It was like, this is it. This is how we're going to cure cancer. I mean, your typical grad student, early founder days. This is it. Let's go. Let's go. So we went out there and we started looking for capital because, I mean, we had Ed Engelmann as the head of the lab. This guy has founded Dendrion. He's a venture capitalist. So we went out and the team universally got the same response from the investors and the clinicians. This is going to be very challenging to deliver because you have to deliver intratumorally and it's multiple components, right? So anybody thinking about getting into drug discovery or drug development, you really have to understand the PK/PD of what you're doing and how you're going to administer it clinically. And that turned out to be a big challenge for everybody. Nobody wanted to do intratumoral injections. And they definitely didn't want to do intratumoral injections of a multi-component mixture, essentially.

For the crowd who doesn't know what PK/PD means, could you— can you explain that for— we know everyone on the podcast knows cytokine and binding, but PK/PD may be newer for some people.

Yeah, absolutely. So, you know, PK is pharmacokinetics and PK is pharmacodynamics, right? So what does pharmacokinetics mean? How long is your drug present in the circulation or in the tumor? Pharmacodynamics, what is it doing during the time that it's there? Is it fast-acting? Is it long-acting, etc.? In the case of our immune stimulant, what we used in the Nature paper was interferon gamma or TNF-alpha, which are cytokines, and then CD40 agonists, which was either a ligand or an antibody. Then obviously the anti-tumor antibodies. Each one of those components would have a different half-life, a different PK/PD profile. If you want to treat patients, that just gets complicated fairly quickly. We struggled with that for a bit and then we had our aha moment. It's like, aha, well, we'll just tie the two of these together or physically conjugate them and then be able to administer it systemically. And this is another thing to think about when you're thinking about drug development. Like, can you actually safely administer immune stimulants systemically via IP or IV administration? IP if we're in the preclinical models and IV if we're in the clinical models. And there was a lot of design process that went into thinking about how to actually administer an immune agonist systemically, right? So obviously, if you're thinking about interferon gamma, It's very challenging to do because the surface receptor for interferon gamma is on the cell surface. So what we decided to do is we decided to utilize a TLR7 and TLR8 agonist, right? These receptors are located inside of cells in the phagolysosomes. And more specifically, they're located in your myeloid antigen-presenting cells and your plasma cytoid antigen-presenting cells, not in the tumor or in many of these other healthy cells. With that, we felt that we could actually systemically administer these agents and we had a real focus on safety. Can we do it safely such that you can control the activation not in the periphery, but in the tumor microenvironment? Obviously, a lot went into that and I'm trying not to get too long-winded in my answers. But once you design that process, our goal was not to break the immune stimulant. And what we found out, and I'll never forget this day, was Shelley Ackerman and I, because she was a very talented graduate student in the lab who knew how to conjugate immune stimulants to antibodies, and we put our ISAC, or immune stimulating antibody conjugate, into the, you know, a co-culture of monocytes. You know, these are your precursor cells with CD20 expressing tumor cells. And we came back the next day We're shocked we didn't find monocytes in the tissue culture dish. We found dendritic cells. Anybody who works with monocyte-derived dendritic cells knows that this takes 3-5 days to differentiate. We were seeing them in 18 hours. Right away, it's like this is something different and we definitely didn't break the immune stimulant, we made it better. Jason, with that finding, We characterized it and we said, okay, what's really going on is we found a more efficient way to deliver immune stimulant to dendritic cells. It's through the Fc receptor with phagocytosis, and now you've concentrated that immune stimulant at the exact place where it needs to be. The phagolysosome with your TLR-containing compartments. We're no longer relying on passive diffusion of your TLR agonist. It made a lot of sense and with that, we demonstrated preclinically that we could elicit complete tumor regression, etc. And we were off to raise our Series A funding and to start building the team.

Yeah, on that note, so I see, so for example, these immune stimulating antibody conjugates, or ISACs as you call them, and some of them are, for example, targeted against HER2. I think it's your first product. And I have a question, maybe, maybe it's a little bit naive, but you published this information in Nature Cancer in 2020. But then the company was founded much earlier. So how did it work? I guess you already had, you were developing the IP and you had to wait, right, until you could publish this. So maybe can you talk about the experience of this particular target and what the story of this target was.

Yeah, I mean, the story of HER2, right? HER2 is just one of those fantastic tumor antigens because it's expressed at very high levels on several tumor types. And if you think about a phagocyte that eats based on the presence of eat me signals, if it's HER2, there's a whole lot of eat me signals on the surface of those cells. And it's been clinically validated as a very nice therapeutic in Herceptin. So we knew it was safe as an antibody. We knew it had low immunogenicity, and it made sense of it as an Isaac. And you're absolutely right. We spent a lot of time sitting on a bunch of preclinical data because we weren't ready yet to release it to the world, which was appropriate as a young startup. Who, you know, really feels that we're on to something that, you know, could be incredibly impactful. So we absolutely timed the publication, you know, as close to as possible with the initiation of the clinical trial while releasing a little bit of data through posters over time. But, you know, it certainly drove the investors nuts at times with how secretive we were, but I think it worked out for the better with, you know, when we were able to release information.

That seems to be very, very common. No, I think it's a very huge difference between, you know, the industry and especially starting biotechs that really need to protect their IP, and that's basically all they have. And I have also, you know, we also talk to other people that have companies, and this is— can be really difficult because as an academic, you are very prone to sharing and to kind of showing the world what you're doing, but then you need to be very careful when you are— this is not the basis of your company.

Yeah, like 100%, Brenda. I think the equation changes a little bit for academic or somebody in biotech when it moves from, you know, sharing the world with the technology versus trying to share with a product and be able to take something from bench to bedside. In all honesty, the publication was not high on our minds.

And one more thing.

You—

so do you— this is— this company associated with your university? I see that your publications, for example, is still associated with Stanford. So my question is, how are— what is the role of universities in maybe helping or incubating this type of companies? Was that the case for you guys?

Yeah, absolutely. I mean, Stanford is such a fantastic place. For developing cutting-edge technologies and then for licensing that technology to biotech companies, particularly not just like large biotech, but they really try to license this technology to the founders if they are interested in pursuing it. So from Stanford standpoint, absolutely the technology was developed in Stanford labs, right, with Ed Ingelman. And then we filed the IP and then started to, you know, bring that technology out, and Bolt Biotherapeutics licensed it. Now, in our case, it was actually very unique because Bolt maintained a very close collaboration with Stanford during the first year or two years of our existence, because as I mentioned, a graduate student, Shelley Ackerman, remained a graduate student at Stanford, and her thesis became these, these bold body Isaacs. So we collaborated very closely with her and with a postdoc in the lab, Justin Kenkel, who was developing the technology for our TAM agonist, or tumor-associated macrophage, that DECTIN2 program. So, you know, we have a lot of very close ties with Stanford.

So I've asked this next question of a couple other biotech guests I wanted to ask you as well, because it's always good to hear multiple opinions. What do you think is the hardest phase shift from academia where, you know, hey, you found something, you're ready to cure cancer, and then biotech/pharma, right? Every biotech, you know, you're on the biotech path that morphs, differentiates into pharma over time, ideally, if you want to use that type of terminology. Where do you, where do you think for people interested in that, what are the skills they need to get while in grad school or postdoc? Where is the big surprises you didn't see coming that you're like, oh man, in hindsight, man, I wish I had known that, or that was a big phase shift.

Yeah, honestly, there's many phase shifts that happen. We touched on one of them, right? PK/PD relationships. As a grad student, it's just not something that we were concerned with. You dose your preclinical models until you see responses, then you can just dose them much more frequently. The biggest surprise that we had, and I still tell young founders or early founders this because it shocked me, we had raised our Series A round and had millions of dollars in the bank. I've never seen this many zeros in a bank account. I remember trying to buy reagents and all of the vendors told us no, we don't have sufficient credit to buy my rosette set from Stem Cell Technologies or to buy antibody. I'm like, this is crazy. Look at how many zeros are in our bank accounts. But you have to establish credit as a company. And in the early days, we were absolutely— we were using our personal credit cards to buy pipettes and antibodies. So that was a very early shocker to me as an early founder. But a second big phase shift from academia to industry is building the team, right? In academia, you come in, you've got a team of brilliant scientists who are well-published, or young graduate students who are extremely eager. And now you're at a biotech company and it's a very small, close-knit group, and you've got to expand the team. And you have the choice, do I bring in the best scientist? Do I bring in the best people, or a combination of the two? And early on at Bold, I found it extremely useful to, to bring in the best people or the best athletes, as you will, somebody with diverse skill sets, because you are going to be wearing multiple hats and you are going to be in the trenches with your colleagues and your teammates. So it's very important to develop that camaraderie and, you know, hire in people or bring in people to the team that want to be there, that share your drive and are ready for the ups and downs. Of startup life. And obviously, Jason, as you transition to bigger pharma, that equation changes a bit.

So having now established credit, you've gone through some series of funding, you have some, you know, things in phase— in a phase 1, 2 trial. You know, where you've done grad school, postdoc, where do you think your next step of growth challenge is besides get the drug all the way through? Like, put that aside.

Yeah.

That's of course every, every, every company's gonna do that. But where, where do you see the next step of your journey as a scientist slash now biotech entrepreneur, head of a lot of science for a company?

Yeah, look, look, absolutely. I, I can't set that aside, right? I mean, the, the biggest personal goal that I have is to impact patient lives and ensure that, you know, we change them from patients to survivors. I'm very bullish that BDC-1001 can do that and that the platform will be extremely useful for clinicians and for our patients. My next steps professionally is to build out the pipeline to understand the technology as we knew it when we developed BDC-1001, the technology as we know it today, and how we can apply our learnings for future programs, both as immune-stimulating antibody conjugates and as myeloid agonists, as is the case with our ADACNT2 program.

In the case of the myelin antagonist, what are the types of tumors that you're targeting? Because HER2, I think, was it breast cancer probably? And I mean, there's other tumors, but I think this is the big one probably. And what is your scope in that sense?

With the HER2-targeted agent BDC-1001, breast and gastric cancers are the big ones. But surprisingly, there's high levels of HER2 on many other tumor types. With the advent of some of the sequencing technology, those patients are now being able to take advantage of some of these HER2-targeted therapies. We're certainly very interested in those indications. In the standpoint of the Dectin-2 agonist, as I mentioned earlier, a lot of these tumors, be they hot or be they cold, they have high levels of myeloid cells in them. Certainly if you can activate the myeloid cells, we think there's tremendous progress for then engaging the adaptive immune response, so bridging this innate and adaptive immune response. And in the past, you know, people have really focused on depleting the myeloid cell. Let's just, you know, get rid of them and hope that, you know, the myeloid cells that infiltrate are not suppressive anymore. Whereas we're very much taking the opposite approach. We're trying to, you know, Ed used to say we're trying to redeem them or revitalize them again, convert them from these tumor suppressive cells to these, you know, immune cell fighters to the cancer fighters that we know they can be as we see them in the context of infectious disease.

Can you go a little bit more into— because maybe I'm— yeah, on the mechanism. So it is easier for me to see how TLR agonists would activate a myeloid cell. How is it— what is the difference between using that and, for example, DECTIN-2?

Yeah, so Dectin-2 would be a direct agonist play, and I can't go in— I'm not gonna— I'm gonna be more careful with my words as it relates to this based on, you know, the limited published information. But that would be a direct agonist play, whereas, you know, with BDC-1001 and some of these other HER2 target agonists, you're utilizing other pathways to get NF-kappaB-induced signaling, IL-12 secretion, TNF-alpha, and some of these other very pro-inflammatory cytokines and co-stimulatory molecules.

All right.

Okay.

Then I guess we'll have to wait for more information to come to the public domain. But it's, I mean, it's very exciting. And I think the point that you make of instead of seeing myeloid cells as a kind of a hindrance to tumor immunity, try to actually get them on your side because, yeah, they're there and they're hard to get rid of, so you might as well recruit them properly. So I think trying to really model or influence tumor microenvironment is a lot of— a lot of different approaches are trying to do this, and it's extremely important. And so when it comes to the— to your kind of already published data and your preclinical models, what are— what is the main difference that you see when you do this activation? What are the main markers or the main populations that you see coming up when you use, for example, your ISAC, your ISAC well-balled bodies?

Yeah. I mean, that's a fantastic question because really what we were doing, we were bridging the innate and adaptive immune system in a very productive manner against cancer. So what happens 24 hours after Isaac treatment within the tumors, you see a massive infiltration of myeloid cells, right? These cells with all this phagocytic potential that can then chew the tumor up via phagocytosis and get ready for antigen presentation. Then 7 or so days later, at least as we published in the Nature Cancer manuscript, you now start to see more increase of these myeloid cells. You start to see the T cells coming in. Then you've really bridged this and the biology of the recruitment fits with what cytokines are released at which times. Again, this gets back to this PK/PD relationships. You get this massive release of chemokines and cytokines that prep the myeloid cells, and then the T cells will come in and start to mediate immunologic memory, not just against the HER2 antigen. That's really part of the beauty of the approach. Once you get phagocytosis, again, you cut the tumor up in these itty-bitty pieces and you present them on the peptide MHC such that you're exposing the T cells to the wealth of tumor neoantigens. Then we get a very diverse T cell population that can mediate hopefully long-term responses, and it certainly did in our preclinical models.

Very exciting. I always like when the T cells come to the rescue. That's what I was going to. I was like, are the T cells finding their way eventually? Jason says no, stop talking about T cells.

Brenda's favorite immune cell is a T cell.

But for me, it is not. Just kidding. I, I appreciate all immune cells equally.

So what I wanted to ask, and I know— so I'm going to try to phrase this in a way based on what you can say from your, your IND package and everything else. Does your approach with these immune-stimulating antibody conjugates help get rid of some of the concern for cytokine storm, cytokine release syndrome? Because of the internalization? Is that one of the— what you guys are believing is a competitive advantage or have some preclinical data, whatever you can say?

Yeah. Yeah. So it's a brilliant question because it gets back to the, you know, the meat of it. How do you deliver an immune agonist systemically? Right. So, so as I indicated, right, we chose TLR7 and 8 because it would require internalization. But we also, you know, took great care to ensure that, you know, if the immune agonist did cleave, because we were using non-cleavable linkers to make sure it's on there, but even if it did cleave, it was going to be cell membrane impermeable. So again, to try to control some of the potential for, you know, off-target immune activation and cytokine release syndrome. And certainly in our preclinical models, we did a good job of of minimizing that potential. And thus far, you know, in the clinic, we've done a great job of minimizing that potential. And, you know, we— our Isaacs require what we call a 3-factor authentication. And in the Nature Cancer paper, you kind of see us go through each step. You have to get tumor binding first and foremost to get any type of, you know, reasonable activation in vivo. Once you get that, you need the Fc, right? Remember I said the Fc is that giant eat me signals. If you ablate the Fc and do an Fc-null ISAC, nothing happens to the tumor. So you need that phagocytosis. And then the same thing, if I inactivate the TLR agonist, which I think we call the TLR-null in the paper, we call it deadbolt in our company, you don't see any of these responses. So it really is, you need the tumor-antigen engagement, Fc-dependent effector functions, and the TLR agonism. Which again, Jason, we took great care preclinically to design around this because it matters to us greatly and it matters to our patients, right? Because a lot of these treatments that are out there have pretty severe side effects. So it was important for us to hopefully extend their survival but maintain as high a quality of life as possible.

So important, good design and is critical, and this kind of several layers of control is critical because once you, of course, once you inject that into the patient, that's, there's nothing you can do anymore. So very, I think it's very, very interesting, and I hope that you guys will have a good first clinical trial. So you say you are now in phase 1-2 at the moment? So let's hope that the results are encouraging and actually you can show that you provide a nice therapeutic option for patients. And also, as I think what we mentioned with Jason is your career story has been also very, very interesting. And I think for many of our listeners, a nice role model to see how you have follow this, this, this research expertise and joined a company that does such nice things. And now you have a very, you know, a top position within this company. And I guess this brings us to a question that we would like to ask you at the end of our conversations. We like to ask our guests something not so much maybe related to science, But in your case, what we'd like to ask you is, what is the best piece of advice that you've ever been given? And it could be professional or not.

Yeah. So, so I'm going to tie this back to Bolt again because this was during our Series A days. And Jason, it gets back to your question, you know, how do you, how do you design these things to make sure that they're, they're safe and you could administer them? With our early constructs, we administered them in our preclinical models, and we found that we encountered extremely high levels of anti-drug antibody, right? Because again, what did we, what did we actually do, right? What's an ISAC? We tied an immune stimulant to an antibody. So obviously, you know, if that antibody has immunogenicity, you run, you know, a real risk of amplifying that response. And, you know, that's what we found in our early preclinical days. And I remember Richard Miller, who's the co-founder of Corvus with Pharmacyclics, huge in the development of rituximab and a brilliant mind. He looked at us, it was myself, it was Grant Yonahiro, who was our chief business officer. And he looks at us, he's like, look, that one's not going to hunt, go get a new one. And that piece of advice, really, we took it as embrace your challenges, embrace bad data, because that's what that was. That was devastating data to a company with Series A funding. We took that challenge and the next several design cycles, we were laser-focused on how to reduce the potential for immunogenicity versus just saying, hey, they're preclinical models, they're not predictive. We really took the bull by the horns and tried to design around that. And I'm very proud to say that in the clinic, we have encountered very low levels of immunogenicity. And that advice of embracing the bad data and then finding a way to solve that problem is just key to anybody in academia or in industry or in life in general, right? Like, tackle the hard problems because our time here is limited. So get the work done.

My boss actually has given the advice that bad news early is good news. It's both a way— don't hide problems, but also under this concept, I think that's a very good truism, as you put it. Embrace the problem and solve it. Don't put it off.

Absolutely. Does that take some— does take some courage, though, to accept, and then especially when you're feeling the pressure of— making what you have has to work. It has to work. We're so far, but it has to work. And then you realize that in the end, everything will— like, you think you're so far, but actually, most likely, it's just the beginning. So you might as well, you might as well start over anyway.

Yeah. And Brenda, honestly, that's why towards the beginning of this, I focused on, on the team. The team that you have around you have to be ready to tackle the tough problems and really be committed to developing good drugs. To make sure that you don't run from the problems or sweep them under the rug, that you attack them head on. So absolutely agree with you.

It's good to have a good team to support you in those times. And then, well, eventually, hopefully, you'll succeed, and then you can go and have a drink and celebrate.

Well, thank you very much for being here. If we want to find you, I think your company Twitter is @BoltBio. Right? So we can check you out there. And you're also on the internet and LinkedIn and everywhere else. I know you guys have a bunch of job openings, so they should check— people should check out your career page as well, right?

Yeah, absolutely. We can't hire fast enough. And it's a great place to work. We are scientifically voracious. So if you share that spirit and that culture, please do come and apply.

There you go, listeners. In the West Coast, you go and check out the open positions at Bold. Good luck with your ongoing studies and your research, and I will— we'll see what comes from that.

Thank you so much. It was a real pleasure.

That brings us to the end of our show. Don't forget, if you want to subscribe to our newsletter at immunologypodcast.com, you can get there show notes, also episode summaries, links to all the interview around the papers. And if you feel like reaching out to us, you can do that on Twitter @immunopodcast or send us an email at info@immunologypodcast.com if you have any feedback or if you have a guest you would like to suggest. See you next time.

Mm.