VDP-89 [Assessment]: ExcepGen Inc - RNA therapeutics for longevity

Proposals Active Proposals
1

VitaDAO Project Submission

One-liner: Optimizing RNA Delivery for Longevity Therapeutics

Project Name: ExcepGen Inc

Project lead name: Thomas Folliard

Project lead email address: Thomas.folliard@excepgen.com

Lead organization website: Excepgen.com

Key personnel
Thomas Folliard (Chief Executive Officer)
Barbara Mertins (Chief Technology Officer)
Katherine Barcay (Chief Operating Officer)
Imre Mager (Head of R&D)
Jason Wojcechowskyj (Director of Vaccines)
Pragya Mukherjee (Scientist)

Describe what you’ll produce in 50 characters or less:
RNA therapeutics for longevity/age-related disease

Please provide a fuller description of what you are doing and how it will impact longevity/healthspan/lifespan

Next-generation vaccines with better efficacy and broader protection are critical for increasing longevity. Severe viral infections have been associated immune dysfunction and chronic inflammation that match signatures of premature aging, an effect that can persist even after the pathogen has been cleared. Moreover, seasonal infections are a common cause of life shortening, frequently even in otherwise healthy individuals. The NIH’s National Institute on Aging has categorized flu as a very serious and life-threatening disease for elderly adults. The respiratory illness caused by viral influenza infection caused more than 12,000 deaths in older Americans over age 65 in 2018. Now in the endemic phase of Covid-19, the risk that aging populations face from viral infections has never been greater.

ExcepGen is developing a new generation of vaccines and therapeutics with its RNAx platform. ExcepGen’s RNAx is delivered alongside primary RNA cargo. RNAx is translated by the cell to produce a regulator protein that modulates signals between the nucleus and the cytoplasm of cells in a tunable manner. By doing this, we keep cells in a low-stress state improving the effectiveness of the RNA cargo to produce significantly more protein leading to increased protection.

We are applying this technology to produce RNA vaccines for challenging antigens, such as universal flu antigens.

While industry-standard messenger (mRNA) was successful in producing a COVID-19 vaccine, this was partly possible because the Spike antigen is highly immunogenic. Other RNA vaccines that target more challenging viral antigens may require longer antigen expression and better immune engagement to be effective.

Self-amplifying RNA (saRNA) is a promising modality for eliciting broader immunogenicity and making better RNA-based vaccines, but it comes with a catch-22. On one hand: saRNA allows for greatly increased antigen expression, potentially requiring lower doses, increasing valences (number of encoded antigens), and providing better adaptive immune responses. On the other hand: its delivery and replication cause inflammatory intermediates and harmful innate signaling that counteract the potential benefits.

With ExcepGen technology, saRNA translation is protected from these cellular triggers. We have demonstrated that

I. saRNAx produces a 400% boost in cargo expression in vivo in mice compared to standard saRNA

II. saRNAx produces up to 2600% improvement in cargo expression in innate-immune-competent human BJ cells

III. saRNAx majorly reduces, and in some cases completely prevents, deleterious inflammatory cytokine release

Based on these exciting platform data, we are pursuing the holy grail of vaccines - a fully differentiated universal pan-flu vaccine that we believe can only be unlocked by saRNAx technology. This product is our primary focus for the vaccine space with additional pipeline products planned for vaccines, protein replacement, and gene-editing.

How far along are you? How close to having intellectual property are you? For projects that you don’t think require IP, please explain.

We expect that we have already generated significant IP and we expect that we have strong IP coverage. We have a portfolio of patent applications submitted by Cooley LLP to protect our core technology and its applications.

What are your competitive advantages over other research and people in your space?

Core technology mechanism: We identified a counterintuitive, holistic way to make RNA perform better. Because our mechanism is counterintuitive, others in the space have not already been exploiting it.

Synergistic technology: Our technology works across many RNA types so we can work with our own RNA modalities or those generated by Pharma partners (including linear and self-amplifying, modified and unmodified, etc.)

IP Strategy: We have applied our technology broadly across RNA types and other nucleic acids to demonstrate our key benefits. We kept the company in stealth mode while developing our strong IP portfolio, which now gives us a strong foundation for technology partnerships in the RNA space.

Nimbleness: We are a small team and we can iterate quickly. Our research process enables rapid in vitro and in vivo testing and refinement.

Why did you pick this area to work on? Briefly discuss your domain expertise. How do you know people need specifically what you’re making?

Severe viral infections have been associated immune dysfunction and chronic inflammation that match signatures of premature aging, an effect that can persist even after the pathogen has been cleared.

The company founding team has deep technical expertise engineering reliability into biology for use in unreliable environments such as in microgravity and the rhizosphere. Educated at top universities, the co-founding team met at Oxford university and founded ExcepGen to engineer reliability into genetic medicines.

We know people need specifically what we are making because of the number of major mRNA companies looking to evaluate our technology for licensing.

What is it about your approach or your findings that you think others don’t understand? What’s new about it and/or surprises people the most when they hear about it?

We encode RNA to produce a regulator protein that makes cells amenable to RNA instructions, resulting in majorly increased cargo expression and prevention of unwanted innate signaling and inflammatory responses. The regulator works by dampening transport of signals between the nucleus and the cytoplasm. Others think this sounds like it would be detrimental to RNA translation and cell function. Counterintuitively, by dampening that “information highway” for the cell, we actually prevent stress signaling and innate immune signaling. The expected MOA is that stress signals are activated when certain cell signaling pathways rise above “normal” signal thresholds. By slowing down the flow of information, ExcepGen technology makes cells amenable to executing the instructions they have been given. We believe that this technology will open up a whole new world of RNA drugs across vaccines and oncology.

Briefly explain how you will spend the money we would give you. This can be as simple as an itemized list of your costs.

ExcepGen is currently applying its technology to vaccines. An assessment from VitaDAO would fund in vitro research, scientific operations, and an in vivo immunogenicity study.

Any other project ideas or IP you’d like us to consider?
No

How did you hear about VitaDAO?
Pfizer - Mike Baran

  • Agree
  • Needs Revision
  • Disagree

0 voters

6 Likes
2

Some additional context: at Pfizer we have been talking with Tom and the ExcepGen team for 6 months+ and are intrigued by the technology. As most are all likely aware RNA therapeutics is one of the most promising modalities in therapeutic development. One of the engineering challenges to delivering RNA is maximizing its durability against the innate immune response which is actively trying to clear the foreign nucleic acid.

The ExcepGen technology potentially stabilizes LNP delivered RNA, leading to longer half-life and longer expression of payload. While the pilot project proposed for VitaDAO would be to explore in the context of a Universal Influenza Vaccine, the technology would be applicable beyond vaccines and of potential benefit to future drug discovery programs.

1 Like
3

Any more info to share on the specific ask?

3 Likes
4

There’s not a lot of data presented here to assess (or budget… is this a $100k ask, $250k ask, more?). Is the WG’s assessment forthcoming on this proposal? And are we voting on the proposal itself, or the WG’s assessment of it?

It sounds like the company is expressing either an NF-KB/IRF inhibitor or some kind of Ran GAP along with the saRNA. Once the patent filing becomes public, it would be help the assessment to have at least a ballpark idea of which pathways are being targeted. Are there any papers published showing this blockade does not lead to cell toxicity?

I can see how increasing mRNA survival/half-life will increase expression. Better expression means lower doses, means lower off-target effects, which is all good. But I thought the limiting factor is on-target effects-- engaging innate sensors like TLR7, NLRP3, RIG-I, MAVS, etc. Nor am I convinced that extending expression time would enhance vaccine outcomes.

My understanding is that the limiting inflammation for LNP delivered RNA is caused by activation of cytosolic innate sensors, and the LNP composition was as much a part of it as the RNA. Or did I read too much into Ira Mellman’s Nature Immunology paper from last year (PMID: 35332327)?

I also thought the limiting issue for flu vaccine efficacy in the elderly is the lack of an immune response. I am not convinced higher dosing is the long-term fix-- part of the immune system is broken in old people, and that’s where the longevity problem is.

It’s also not clear to me how this approach will generate a universal flu vaccine. Longer-lasting RNA doesn’t fix antigenic drift.

Also some concerns about model systems. If they’re going for a flu vaccine, mice are not a good model system. Fibroblasts like BJ cells might be a starting point, but I thought the main in vivo targets for an intra-muscular vaccine are muscle cells and any macrophages/DCs nearby. Macs and DCs have way more antiviral sensors than a fibroblast.

Why was “universal flu vaccine” considered the best target to pursue, even in the vaccine space? Seems like encoding an anti-HIV ribozyme, going after some other latent virus, or extending expression of a therapeutic you want expressed for a medium-ish time period (what is the t1/2 for expression vs AAV, regular RNA-LNP, or other delivery methods?) would be better choices.

Or in the longevity space, what about reprogramming sensencent cells?

2 Likes
5

Thanks Michael - I have asked Tom to supply their Non-conf presentation as well as a project plan with milestones/budget

1 Like
6

Great feedback Shrike - and agree there is no data shared to assess. Will add in a non-conf to give a better sense what they are doing + detailed project plan and budget.

High level the budget will be to support further in-vivo PoC for universal influence Vx + self amplifying RNA (saRNA) in an animal model. Success defined as showing prolonged expression and titer.

Will get Excepgen to answer your questions - but first pass from me below:

There’s not a lot of data presented here to assess (or budget… is this a $100k ask, $250k ask, more?). Is the WG’s assessment forthcoming on this proposal? And are we voting on the proposal itself, or the WG’s assessment of it?

Will post a non-conf deck with data. We are in process of pulling together a senior reviewer team. Internally at Pfizer this one has passed diligence review and we are looking to move forward with our own project. A few others pharma may be moving to MTA as well.

It sounds like the company is expressing either an NF-KB/IRF inhibitor or some kind of Ran GAP along with the saRNA. Once the patent filing becomes public, it would be help the assessment to have at least a ballpark idea of which pathways are being targeted. Are there any papers published showing this blockade does not lead to cell toxicity?

Target has not been made public. The ompany can provide more detail on MOA.

I can see how increasing mRNA survival/half-life will increase expression. Better expression means lower doses, means lower off-target effects, which is all good. But I thought the limiting factor is on-target effects-- engaging innate sensors like TLR7, NLRP3, RIG-I, MAVS, etc. Nor am I convinced that extending expression time would enhance vaccine outcomes.

You describe the approach well. This approach essentially blocks nuclear transport. So the innate sensors you describe would initiate transcription of the immune cascade in the nucleus to a lower extent & their transcripts would be trapped/or export minimized therefore lowering translation of immune machinery in the cytoplasm, which would then lead to longer RNA half life.

Our thinking at Pfizer is that this is potentially a way to overcome the challenges to date with self amplifying RNA. saRNA is believed to be one of the approaches to tackle the problem of RNA stability. (Others are excited by circular RNA). It is believed overcoming immunogenicity is one of the limiting factors to success with saRNA & this technology has potential to overcome that challenge.

My understanding is that the limiting inflammation for LNP delivered RNA is caused by activation of cytosolic innate sensors, and the LNP composition was as much a part of it as the RNA. Or did I read too much into Ira Mellman’s Nature Immunology paper from last year (PMID: 35332327)?

You are correct - the contribution of inflammation comes from both LNP & nucleic acid. Although nexgen lipid formulations are designed to minimize.

I also thought the limiting issue for flu vaccine efficacy in the elderly is the lack of an immune response. I am not convinced higher dosing is the long-term fix-- part of the immune system is broken in old people, and that’s where the longevity problem is.

This may be true. But also achieving high titer against a cocktail or potentially inclusive antigen should also have efficacy.

It’s also not clear to me how this approach will generate a universal flu vaccine. Longer-lasting RNA doesn’t fix antigenic drift.

Will get you the detail on this. The POC experiment will include a well published construct.

Also some concerns about model systems. If they’re going for a flu vaccine, mice are not a good model system. Fibroblasts like BJ cells might be a starting point, but I thought the main in vivo targets for an intra-muscular vaccine are muscle cells and any macrophages/DCs nearby. Macs and DCs have way more antiviral sensors than a fibroblast.

Will get you detail on this from the company.

Why was “universal flu vaccine” considered the best target to pursue, even in the vaccine space? Seems like encoding an anti-HIV ribozyme, going after some other latent virus, or extending expression of a therapeutic you want expressed for a medium-ish time period (what is the t1/2 for expression vs AAV, regular RNA-LNP, or other delivery methods?) would be better choices.

Will get an answer from company on this.

1 Like
7

I look forward to the company’s answers.

Improving RNA stability is a no-brainer for Pfizer and competitors if it looks like it might work. Lot of applications, which is why the current choice of target is somewhat mystifying.

In the meantime, two more concerns. The highest level view is that Excepgen is trying to reinvent RNA viruses. Since RNA viruses already exist, is Excepgen’s pathway already targeted by RNA viruses? If not, a big question would be why not? Do the viruses target a related part of the pathway, or did they give up on this mode of control?

Related to that, inflammasome activation doesn’t require nuclear access to destroy a cell, and necroptosis is the back-up cytoplasmic death mechanism characterized both for Casp8 failure, but also during flu infection. The cell studies need to be able to test if necroptosis, pyroptosis (and PANoptosis if we’re keeping with current fads), ferroptosis, or autophagy will be activated. Fibroblasts are inadequate to test most of those hypotheses.

Or will they need to encode Gasdermin and RIP kinase/mlkl inhibitors in as well? …and/or encapsulate something like MCC950?

1 Like
8

Answers below:

Are there any papers published showing this blockade does not lead to cell toxicity?

NCT inhibitors are used as potential anti-cancer drugs, due to their selective toxicity towards cancer cells, sparing non-transformed cells. Furthermore, the doses of saRNAx for vaccines remain low and given intramuscular injections, systemic toxicities are not expected. Also, in our in vivo study, using saRNAx intramuscularly did not induce body weight abnormalities or visible inflammatory reaction in the injection sites

But I thought the limiting factor is on-target effects-- engaging innate sensors like TLR7, NLRP3, RIG-I, MAVS, etc.

Our proposal focuses on the research and development of a self-amplifying RNA (saRNA) universal influenza vaccine. The central problems facing saRNA are suboptimal antigen expression and uncontrolled innate signaling. Our technology is uniquely suited to enable saRNA as a platform by increasing cargo gene expression both in vitro and in vivo and by decreasing innate signaling, which we have shown in vitro. There are different and unique shortcomings for mRNA-based therapeutics which ExcepGen is attempting to address, but those are not a focus of this proposal.

Nor am I convinced that extending expression time would enhance vaccine outcomes.

One of the central tenets of vaccinology is that increasing the dose of antigen without overwhelming the body with innate signaling increases the potency of a vaccine. This issue is particularly acute for current influenza vaccines. The dose of HA given to individuals is rate limiting and unfortunately, current vaccines have suboptimal potency. This is of urgent concern for increasing longevity in people. The elderly must be given 3x the dose of HA in a vaccine, and if feasible, would be given more. Right now, the inability to deliver a sufficient dose of HA for an influenza vaccine is hindering its potency and overall effectiveness.

While it is generally known that increasing the dose of antigen improves potency, there is a growing body of literature from multiple investigators across many vaccine antigens that the extended presence of antigen in the lymph nodes increases the magnitude and quality of the antibody response (PMIDs 36131022, 31080066, 31358641, 27702895). One of the most exciting findings from these studies is that expressing a poorly immunogenic antigen for an extended duration dramatically increased the neutralization potential of the antibody response.

My understanding is that the limiting inflammation for LNP delivered RNA is caused by activation of cytosolic innate sensors, and the LNP composition was as much a part of it as the RNA. Or did I read too much into Ira Mellman’s Nature Immunology paper from last year (PMID: 35332327)?

There are many potential sources of innate signaling from an LNP-delivered RNA, e.g. as you point out from nucleic acid sensors, lipid components of the LNP, but also route of delivery. Ira’s paper does a wonderful job of illustrating how both the composition of the nucleic acid and lipid composition drives inflammation during IV administration. Our technology is unique because of how dramatically it suppresses saRNA-derived sources of innate signaling, e.g. double stranded RNA, thereby maximizing cargo gene expression.

I also thought the limiting issue for flu vaccine efficacy in the elderly is the lack of an immune response. I am not convinced higher dosing is the long-term fix-- part of the immune system is broken in old people, and that’s where the longevity problem is?

The question of why the elderly do not mount as effective immune response to vaccines in general is one that continues to vex the scientific community. To improve their immune response to the influenza vaccine, the elderly are given triple to dose of Influenza HA. If we can achieve a higher dose from a dose sparing effect, or simply increase immunogenicity potency from our technology, then we expect to further improve vaccine efficacy in the elderly.

It’s also not clear to me how this approach will generate a universal flu vaccine. Longer-lasting RNA doesn’t fix antigenic drift.

You rightly point out that antigen drift is the central issue facing an influenza vaccine. A universal vaccine must be able to target multiple strains from one formulation. There are 2 general strategies to achieve universality: 1) include multiple diverse strains (multi-valency) or 2) use a highly conserved antigen. Each of these approaches is currently limited by suboptimal vaccine potency, which we argue will be addressed by higher and longer antigen expression (see refs above regarding extended Ag in the lymph nodes).

The ability to maximize valency of a vaccine product is driven by the potency of each component. In other words, the more potent and dose sparing your product, the more you can administer in 1 shot.

Highly conserved antigens , eg HA stem, are the targets of broadly neutralizing antibodies which protect from diverse influenza strains. However, they are also intrinsically poorly immunogenic. The central thesis of our technology is the ability to maximize the immunogenicity potency from saRNA, thereby enabling previously poorly immunogenic (and highly conserved) antigens.

Given this, we strongly believe that a higher magnitude and duration of antigen expression from saRNA encoding our technology will in fact enable a universal influenza vaccine. Higher and longer term antigen expression leads to a higher quality of the antibody response, which is in turn needed to enable more dose sparing (multivalency) and/or higher potency (HA stem).

Why was “universal flu vaccine” considered the best target to pursue?

We believe a universal influenza vaccine aligns well with a critical unmet need for improving longevity while also being the ideal biological test case for how a more potent vaccine will enable an entirely new product class via saRNA. Because saRNA has yet to be successfully deployed in the clinic, we believe that enabling saRNA, which our technology is uniquely suited to do, will open the door for many more targets and diseases. Further, because of ExcepGen’s unique ability to majorly boost antigen expression, we believe that we can uniquely improve the immune response against already known universal flu antigens that have been historically poorly immunogenic.

Since RNA viruses already exist, is Excepgen’s pathway already targeted by RNA viruses? If not, a big question would be why not? Do the viruses target a related part of the pathway, or did they give up on this mode of control?

Our choice of innate signaling modulator was chosen precisely because many diverse viruses also target the same cellular pathway. Viruses have evolved to do so for the same reason we would like to improve vaccines: express cargo/antigen at a higher level for a longer period of time. Our platform of choice is saRNA, which in its current form lacks a means of regulating innate signaling, which many others have shown is needed to maximize its potential.

The cell studies need to be able to test if necroptosis, pyroptosis (and PANoptosis if we’re keeping with current fads), ferroptosis, or autophagy will be activated. Fibroblasts are inadequate to test most of those hypotheses.

Control of innate signaling is a key driver of maximizing cargo gene expression and immunogenicity from saRNA. We have shown that our technology leads to dramatic increases in cargo gene expression from saRNA in vitro and in vivo concomitant with a reduction in innate signaling.

Or will they need to encode Gasdermin and RIP kinase/mlkl inhibitors in as well? …and/or encapsulate something like MCC950?

Given the unique challenges to mRNA, this is a clever idea to address those. Unfortunately this approach is outside of the scope of our current work given that we are focusing on saRNA.

1 Like
9

Hi All - Below is a link to ppt from Excepgen.

1 Like
10

I notice that they are requesting $200k from VitaDAO. I thought Assessments were not directly funded by VitaDAO, but just putting the WG assessment “on-chain”. Should this be a proposal instead of an assessment, with the $200k ask added to the main proposal body?

3 Likes
11

IDK still learning myself - @timrpeterson do I need to update the title since there is a $ ask?

1 Like
12

Something feels ‘off’ to me about this entire proposal.

On one hand, it sounds to me that Pfizer is supportive of this program overall and has access to the non-public data. Improving RNA expression is an obvious fit there regardless of other applications, so that part doesn’t surprise me.

The part that concerns me is that the company brushed off all questions about innate immune activation with zero data. For example,

No convincing public data have been shown, even with the slide deck posted. Innate signaling can happen without nuclear invovlement. And based on the cancer cell toxicity, at least one of those pathways could be activated.

No data for this one either. When flu fails to inhibit necroptosis, that cytoplasmic death pathway eliminates infected cells.

Control of innate signaling not been shown publicly. Only thing shown is that Excepgen can boost luciferase production a few days after transfection. It’s a nice boost, and would be helpful for lowering the dose on existing vaccines, but not enough to show the main assertions–increased duration and lack of innate immune signaling.

Without data showing “control of innate signaling”, I flat up do not believe the claims. Answering questions with assertions instead of data sounds to me like the data don’t exist, or they are poor. This leaves aside the issue of if the company’s secret Ran GAP is immunogenic itself (presumably it is a neutered version of the VEEV capsid/other viral protein already optimized to lack immunogenicity?)

Maybe these data exist. But using a fibroblast line that I don’t expect has key innate immune signaling mechanisms does not inspire confidence. There are a lot of easy in vitro assays to do that would address these concerns, which if true would be lethal to the whole program. That they are brushed off instead is …concerning.

No data showing longer duration of expression has been made public. Do these data exist? How do they compare to AAV, or other delivery methods?

At a high level, one challenge to Shane Crotty’s ‘let’s dose more like real viruses do’ approach for flu is that weakly immunogenic “universal” flu antibodies don’t seem to be raised during longer, natural infections. Otherwise, antigenic drift wouldn’t be a problem. So longer duration may not overcome poor immunogenicity. Also, if duration is the name of the game, why not AAV?

I do not agree with a ‘dose them higher’ approach to try to improve efficacy. Seems like a minefield between more side effects and increased innate challenges on one side and switching to Treg polarization on the other.

The timeline and plan in the slide deck are both vague and unrealistic. Three months will not be long enough to do the experiments in mice. What ‘stem antigen optimization’ is needed? Does expression testing include testing in DCs, macrophages and muscle cells? For the mice, will IL-1RA KO mice be used? Will an aged mouse model or other attempt to model elderly immune deficiency be used? What route of delivery? If it’s just immunogenicity testing, there’s no plan to challenge with a lethal flu dose to show efficacy? And how will you measure durability of response when the endpoint is only a month (or less!) after a standard immunization? Or is function planned for ferrets?

The lack of apparent consideration of these issues does not inspire confidence in the team. This is also the part that confuses me the most. How did it pass diligence from the company’s own SAB (which has 2 immunologists?) and one or more companies without addressing some of these key issues? do the data exist, and they’re just not public? But if so, why evasive non-answers to the questions about innate signaling? Am I missing something here?

I think a pay-by-milestone-based approach would be better risk managment for VitaDAO. First milestone would be show increased duration of flu antigen expression in a muscle cell line and primary macrophages, and DCs, while also measuring amount of antigen, nSP1-4, Ran GAP expression, cytokine protein production by the macs/DCs (TNFa, IL6, IL-1a, IL-1b, IL-18, type I IFN), nuclear translocation (HMGB1 does double duty for alarmin and nuclear transport) and amount/type of cell death. Compare to control, AAV, mRNA/LNP, saRNA/LNP.

If it fails, only $20-$30k spent exploring a potentially interesting idea.

If that one is hit, then in vivo testing would be better justified. But here, along with AAV, mRNA/LNP, saRNA/LNP, test vs the current (more antigenic) parts of HA using standard immunization methods. Also need to test for antigenicity of nSP1-4, the secret Ran GAP, and measure durability of the response/expression time. At least one cohort should use IL-1RA KO mice to address potential for “unexpected” side effects in human.

2 Likes
13

Will look to get more data from the company. Also we will schedule Tom to come present on a Friday with live Q&A.

1 Like
15
16

A few more responses from the company. And Tom will be coming to present at this Friday’s deal flow call - please join if interested to dig in more.

-Mike

We thank the commenters for their questions. We have prepared a slide deck with key findings that support our central conclusions: Our tech improves cargo gene expression from saRNA while decreasing inflammatory cytokine release.

Link to slide: https://drive.google.com/file/d/16ryZHppiODo8IXIdwuik-KaicpBGzrko/view?usp=share_link

This proposal is focused on evaluating how saRNAx may improve immunogenicity of a universal influenza antigen in mice. Many variations and additional comparisons, while interesting, are outside of the scope of this proposal.

Please see below for answers to specific questions. We apologize that at least one question was missed during the last round of comments. We include it here.

Also some concerns about model systems. If they’re going for a flu vaccine, mice are not a good model system.

While we strongly share the sentiment that mice are not sufficient to demonstrate the efficacy of an influenza vaccine in total, we respectfully disagree with the implication here that mice are not worth testing first. Mice are a gold standard model for early stage preclinical testing of influenza vaccines. There are no perfect preclinical models for any vaccine, but practical concerns limit our options. Future studies will test ferrets and/or nonhuman primates.

Fibroblasts like BJ cells might be a starting point, but I thought the main in vivo targets for an intra-muscular vaccine are muscle cells and any macrophages/DCs nearby. Macs and DCs have way more antiviral sensors than a fibroblast.

We chose normal human fibroblasts (BJ cells, MRC-5 cells) for a number of important reasons. These cells have intact innate signaling, primary cell-like nature, are not transformed, and are relatively easy to use. Leaders in the field of RNA vaccines also agree with their utility, e.g. teams at Moderna (PMID 32637598, 36357718) use BJ cells and the team at the Imperial College London (PMID 33352107) utilize MRC-5 cells. You are correct that macrophages and dendritic cells have more innate sensors than fibroblasts. However, it is not known the specific role these cells play with RNA vaccines. While engagement of antigen with antigen presenting cells such as this are essential for a robust immune response, it is not known whether or not these cells are or must be directly transduced by (sa)RNA-LNPs, or whether the saRNA must express in these specific cells to induce a potent immune response. There are many paths that antigen can take to these cells and studying how saRNAx may or may not interface directly with these cells would certainly be an interesting separate study. It is not yet technically possible to model the priming of an immune response in vitro, so we have not prioritized these studies before deploying saRNAx in vivo in a range of contexts.

No convincing public data have been shown, even with the slide deck posted. Innate signaling can happen without nuclear involvement. And based on the cancer cell toxicity, at least one of those pathways could be activated.

We have claimed (and are now presenting data) that “ saRNAx majorly reduces, and in some cases completely prevents, deleterious inflammatory cytokine release.” From saRNA, we show a reduction in IFN-beta and IP-10 (CXCL10). From an inflammatory mRNA, we observed a decrease in IFN-beta, IFN lambda 1, IFN lambda 2/3, and IP-10 (the only cytokines in the panel tested that showed an increase in the absence of our tech). A comprehensive understanding of how the tech affects all forms of cell stress, inflammation, and death across cell types is certainly very interesting, but not in this proposal for VitaDAO.

“Our technology is unique because of how dramatically it suppresses saRNA-derived sources of innate signaling, e.g. double stranded RNA, thereby maximizing cargo gene expression”

No data for this one either. When flu fails to inhibit necroptosis, that cytoplasmic death pathway eliminates infected cells.

We have now included experiments where our tech protects cargo gene expression from mRNA in the presence of saRNA-derived sources of innate signaling (dsRNA, IFN-beta).

“Control of innate signaling is a key driver of maximizing cargo gene expression and immunogenicity from saRNA. We have shown that our technology leads to dramatic increases in cargo gene expression from saRNA in vitro and in vivo concomitant with a reduction in innate signaling.”

Control of innate signaling not been shown publicly. Only thing shown is that Excepgen can boost luciferase production a few days after transfection. It’s a nice boost, and would be helpful for lowering the dose on existing vaccines, but not enough to show the main assertions–increased duration and lack of innate immune signaling.

Without data showing “control of innate signaling”, I flat up do not believe the claims. Answering questions with assertions instead of data sounds to me like the data don’t exist, or they are poor. This leaves aside the issue of if the company’s secret Ran GAP is immunogenic itself (presumably it is a neutered version of the VEEV capsid/other viral protein already optimized to lack immunogenicity?)

See included data showing a reduction in saRNA-derived cytokine release via our tech. We can also reduce the production of cytokines from an artificially inflammatory mRNA that co-expresses our tech. Combined with the observation that our tech can ‘protect’ modified mRNA from exogenous dsRNA or IFN-beta, we are confident in the functional link between the enhancement of cargo expression from saRNA and innate signaling.

Maybe these data exist. But using a fibroblast line that I don’t expect has key innate immune signaling mechanisms does not inspire confidence. There are a lot of easy in vitro assays to do that would address these concerns, which if true would be lethal to the whole program. That they are brushed off instead is …concerning.

Normal human fibroblasts such as BJ cells (and MRC-5 cells) are well known to have intact innate signaling. Teams at Moderna (PMID 32637598, 36357718) and the Imperial College London (PMID 33352107), both of which specialize in (sa)RNA-based vaccines, also utilize these cells for understanding how RNA-based vaccines intersect with innate immunity. We have also included data in A549 cells and C2C12 cells (mouse muscle).

“Higher and longer term antigen expression leads to a higher quality of the antibody response, which is in turn needed to enable more dose sparing (multivalency) and/or higher potency (HA stem).”

No data showing longer duration of expression has been made public. Do these data exist? How do they compare to AAV, or other delivery methods?

We have included data showing that saRNAx enhances cargo gene expression overtime. Since we are focused on improving saRNA, We are not suggesting that AAV or other modalities are compared as this would significantly increase the timeline and budget beyond the scope of this proposal.

At a high level, one challenge to Shane Crotty’s ‘let’s dose more like real viruses do’ approach for flu is that weakly immunogenic “universal” flu antibodies don’t seem to be raised during longer, natural infections. Otherwise, antigenic drift wouldn’t be a problem. So longer duration may not overcome poor immunogenicity. Also, if duration is the name of the game, why not AAV?

Antigenic drift is a problem because the immunodominant domains of flu HA are highly variable. Epitopes that give rise to universal antibodies are subdominant. So regardless of how long a full length flu HA is expressed, overcoming this immunodominance is a challenge. Universal flu antibodies are in fact elicited by natural infection and vaccines, however the levels induced are not of sufficient quantity for broad protection

As interesting as AAV is in an academic setting, it is unsuitable for infectious disease vaccines.

“Inability to deliver a sufficient dose of HA for an influenza vaccine is hindering its potency and overall effectiveness”

I do not agree with a ‘dose them higher’ approach to try to improve efficacy. Seems like a minefield between more side effects and increased innate challenges on one side and switching to Treg polarization on the other.

The dose of HA units delivered to elderly people is rate limiting for their immune response. This is why the elderly are giving a 3x dose of the current vaccine. Each vaccine platform has its own limit of actual dose of drug substance that can be given, so we agree that some vaccines cannot just be administered at a ‘higher’ (depending on what is baseline) dose. For saRNA, we and many industry leaders all agree with the hypothesis that increasing saRNA cargo gene expression will improve immunogenicity. We have built a novel platform to do so that also decreases inflammatory cytokine release, which the same industry leaders also hypothesize will have a beneficial impact on saRNA-based vaccines.

The timeline and plan in the slide deck are both vague and unrealistic. Three months will not be long enough to do the experiments in mice.

We respectfully disagree. It takes a few weeks to generate LNPs and a standard immunogenicity study lasts 42 days. T cell analysis is not prolonged and ex vivo testing (ELISA, neutralization) are typically conducted in real time during the study.

What ‘stem antigen optimization’ is needed?

The precise sequence and antigen construct for delivering the HA stem needs some engineering. For IP reasons we cannot disclose details.

Does expression testing include testing in DCs, macrophages and muscle cells?

We plan on testing in BJ and 293T cells to confirm expression of antigen.

For the mice, will IL-1RA KO mice be used? Will an aged mouse model or other attempt to model elderly immune deficiency be used?

We believe that it is outside the scope and budget of this proposal to include an influenza challenge model or to duplicate the study design to include IL-1RA KO or aged mice. These are all attractive next steps following promising results in WT mice. We believe a straightforward immunogenicity study in a well established mouse model is the most prudent next step.

What route of delivery?

intramuscular

If it’s just immunogenicity testing, there’s no plan to challenge with a lethal flu dose to show efficacy?

That is correct. Homologous and heterologous Influenza challenge will proceed in a follow up study.

And how will you measure durability of response when the endpoint is only a month (or less!) after a standard immunization? Or is function planned for ferrets?

We certainly have the option of extending the mouse study to gather antibody response data beyond 42 days. However, mice may not be ideal models for durability. To balance time and costs, we are focusing this study on quantifying the potency (ELISA) and quality (neutralization) of the antibody response from saRNAx. Ferrets and/or nonhuman primates will be a natural next step following a challenge study and will glean more relevant insights into durability.

The lack of apparent consideration of these issues does not inspire confidence in the team. This is also the part that confuses me the most. How did it pass diligence from the company’s own SAB (which has 2 immunologists?) and one or more companies without addressing some of these key issues? do the data exist, and they’re just not public? But if so, why evasive non-answers to the questions about innate signaling? Am I missing something here?

We appreciate your understanding that we cannot share most of our data at this time without confidentiality agreements in place.

1 Like
17

When on Friday is the call? I have a final exam to deal with, so it may not work out for me.

For me, the weaknesses still outweigh the strengths.

One big concern is that I’m not convinced the proposed experiments will catch disqualifying problems that could sink the project later in the pipeline. I can’t tell if I’m missing something, or if this approach is just the biotech/VC funding game. Personally, I’d rather find the lethal problems sooner rather than later. Did the SAB consider these problems, and see nonpublic data that mitigated these problems, did they judge the problems non-lethal, or not consider these problems?

I don’t like the mouse model system for flu challenge because either way the results go, an optimist or a pessimist will insist they could still be right. Targeting studies and immunogenicity in mice seem reasonable starts, but seems like ferrets would be better sooner rather than later.

The need to engineer the antigen is also concerning. It’s not clear if this is the endless ‘let’s tinker and make it a little better’ or ‘we must fix this for it to work’. If the latter, you have a lethal pitfall for the project.

The fibroblasts are another key weakness. Need to use a more relevant cell line and/or test additional cells. Specifically, you need to look at necroptosis and pyroptosis, which are the obvious limiting factors. If this is widely ignored by “industry leaders”, maybe that’s why some of these things have had unexpectedly potent inflammatory side-effects. HT29 cells are standard for necroptosis. Primary macs will be good for pyroptosis, especially if you prime them first.

Also ’ it is not known the specific role [macs and DCs] play with RNA vaccines.'… I thought Pfizer and others had extensive data on the cell types that take up LNPs by formulation, and how targeting changes that. But without the data on hand, the obvious hypotheses in the absence of any targeting are the LNPs get taken up by muscle cells and phagocytes.

The slide deck lacks controls. No controls for extent of cell death throughout.

Slide 3: why are “untreated” fibroblasts making IFNb? Presumably they were treated with something. Which means a negative control is missing. Also, measured at day 6, not at day 20.

Slide 4: Is this normalized to cell counts? What were cell counts on each day? Where are your controls of saRNA alone and mRNA-LNP?

Slides 5 and 6: this is measured 5 min after addition of IFNb or polyI:C? I am surprised that you see much effect that quickly, especially at those doses. How do the cells look if you take the assay out 24-48 h? What controls show that IFNb and polyI:C are acting via expected pathways?

Additional data hinted at: A549 are epithelial-ish cells, and not relevant to an intramuscular delivery model. For C2C12, were these done as myoblasts or were they differentiated into myotubes first? Which data do you have in those? If you’re testing expression in BJ, why use 293T? They express everything. Why not a hard-to-deliver cell?

Your cytokine screen showed only a few innate cytokines upregulated. But if this is differential expression, which ones were elevated in both? This seems key to figuring out if your saRNAx tech will hit a different road block. And I don’t trust fibroblasts as a good screen because they will miss other innate effectors. Why not C2C12 myotubes and bone-marrow derived macrophages?

Also, no mechanism to show that the system is working like it should be. Is nuclear transport inhibited in these cells? If you transfect with STING ligands, how well does the saRNAx protect the cells? If you use IFNAR and TLR3 knockout cells, does that block the IFN and polyI:C effects? While I understand IP protection limits the data that can be shared publicly, do these data exist, and how robustly do they support the proposed mechanism?

2 Likes
18

Any update on the senior reviews? @scienceman

19

The senior reviews have been carried out, I am expecting replies by ExcepGen next week

2 Likes