The Needle Issue #10


Welcome to another issue of the Needle, a newsletter from Haystack Science helping you understand the latest translational research and the latest news from around the world on biotech startups.

Over the past few days, CAR-T cell therapies have been front and center in the news. Not your typical ex vivo CAR-T, but Capstan Therapeutics’ therapy that directly reprograms patient T cells in situ, which was the focus for a high-profile merger with AbbVie at the end of June. There continues to be a lot of innovation going on in CAR-T engineering in immune cells as illustrated by several papers captured in our survey of the translational literature. Elsewhere, several government and venture initiatives promise to boost funding available to early-stage startups. A bumper crop of financings and deals in recent days suggests that the early-stage space continues to tick over. Any that we missed, let us know (info@haystacksci.com).

Haystack chat

Just over a week ago, AbbVie paid $2.1 billion for Capstan Therapeutics’ in vivo anti-CD19 chimeric antigen receptor (CAR)-T cell therapy (CPTX2309) for B cell-mediated autoimmune disorders, which is currently in phase 1 testing. In the past few days, EsoBiotec (acquired by AstraZeneca earlier in the year) also published its first clinical data on a lentiviral-delivered anti-B-cell maturation antigen (BCMA) CAR-T approach (ESO-T01) for multiple myeloma, detailing responses in four patients, two of whom showed complete remission. With a host of other companies working on in vivo delivery into endogenous T cells—including Interius BioTherapeutics, Umoja Biopharma, and Orna Therapeutics, the field of in vivo delivered CAR-T cells appears poised at a tipping point.

Since transforming the face of cancer treatment in 2017, autologous CAR-T cell therapy has been dogged by logistical issues that have limited commercial rollout and increased costs—the need for leukapheresis, laborious cell harvesting, heterogeneous cell expansion, lengthy turnaround times, and inconsistency of batches—with access limited to just a few clinical centers. Extensive waiting lists can mean many patients die before even being treated, which has driven the search for ex vivo approaches that shorten manufacturing times using fully closed systems and/or miniaturization. Given these challenges, delivery of a CAR-encoding mRNA to a T cell in vivo could be a game-changing technology: No need for viral vectors; no leukapheresis/chemo; no ex vivo manipulation, no requirement for multiple patient hospital visits; no convoluted training of personnel; and no risk of second primary T-cell cancers due to insertional mutagenesis. This last issue has loomed over the field, with all CAR-T therapies carrying black box warnings, although at the end of June the FDA removed all requirements for Risk Evaluation and Mitigation Strategies (REMS).

Writing in Science, the founding team of Capstan Therapeutics, headed by Carl June and Bruce Levine at the University of Pennsylvania and Haig Aghajanian of Capstan, report proof of concept data that functional CAR T cells with antitumor activity can be produced in animal models without any ex vivo manipulation. A key breakthrough in their effort was the development of lipid nanoparticles (LNPs) specifically designed to target T cells and to overcome the propensity of LNPs to accumulate in the liver. To avoid this problem, the authors screened a set of ionizable lipids to identify L829, a lipid that incorporates a tertiary amine headgroup that reduces non-specific interactions with the hepatic system due to its pH-dependent protonation and neutral charge. Ester cleavage sites in the lipid also promote rapid breakdown in, and clearance from, hepatocytes. A final step was to decorate L829 LNPs with a mAb targeting CD5, a T-cell specific marker. The resulting LNP showed limited liver uptake in rodents and non-human primates compared with control LNPs.

To test the potential of L829-containing LNPs to generate functional CAR-T cells, the team engineered them to incorporate 1) mRNA encoding a CAR that binds CD19 on B cells and 2) an antibody targeting CD8+ T cells. These CD8-L829-CD19 targeted (t)LNPs successfully delivered the mRNA in vitro to CD8+ T cells from healthy subjects and from people with B cell-mediated autoimmune diseases. In vivo, these CAR T cells had anti-tumor activity in a humanized mouse model of B cell acute lymphoblastic leukemia.

In cynomolgus monkeys that received repeated doses of CD8-L829 tLNPs containing anti-CD20 CAR mRNA (instead of anti-CD19, which is not cross-reactive between human and monkey), sustained B-cell depletion was observed that lasted for one month. Importantly, reconstituted B cells were predominantly naïve, implying an immune reset — a key therapeutic goal in autoimmunity.

The Capstan in vivo mRNA-encoded CAR T platform eliminates the need for ex vivo manipulation and lymphodepleting conditioning. It avoids the risks often associated with the use of viral vectors that integrate into the genome. It also is transient, allowing dosages to be optimized and quickly stopped if patients suffer adverse events associated with neurotoxicity or cytokine-release syndrome. It will be interesting to see whether the approach is scalable and whether it can open up conditions where long-term CAR-T cell persistence might not be necessary, such as autoimmune disease.

Going forward, an important question will be to determine the potential immunogenicity of the tLNP formulation (especially as the mRNA treatment may be given multiple times), and whether tLNPs cause elevations of human liver enzymes like alanine transaminase or aspartate aminotransferase. Liver toxicity of a novel liposome formulation already caused a clinical hold for Verve Therapeutics’ base editing therapy last year. Future work will also need to define optimal dosing, durability, and long-term safety of this approach. But the work of June, Aghajanian and their colleagues is a compelling advance promising a new era of widely available adoptive T-cell therapies for B-cell driven hematological cancers and autoimmune conditions. A single dose of any of the seven currently approved commercial ex vivo CAR-T therapies costs ~$500,000. A vial of an in vivo treatment is likely to cost an order of magnitude less.

Papers: Best of the rest

Other engineered CAR immune cell therapies

An iPSC-derived CD19/BCMA CAR-natural killer cell therapy leads to month-long reduction in pathogenic B cells/plasma cells in a patient with systemic sclerosis | Cell

CD70-targeted CAR-T cells expressing IL-8 receptor and soluble LAIR2 block LAIR1-FXIII-A-collagen-mediated immunosuppressive tumor microenvironment in preclinical glioblastoma models | JCI

CAR-T cells engineered to secrete IL-2 or IL-12 under the control of tumor-restricted NR4A2 or RGS16 promoters show reduced toxicity and increase epitope spreading in preclinical breast cancer models | Nature

Target biology

Fibroblast lipid metabolism through ACSL4 regulates epithelial sensitivity to ferroptosis in IBD | Nature Metabolism

The prostacyclin receptor PTGIR is a NRF2-dependent regulator of CD8+ T-cell exhaustion | Nature Immunology

Proximity between LAG-3 and the T cell receptor guides suppression of T-cell activation and autoimmunity | Cell

GluN2B suppression restores phenylalanine-induced neuroplasticity and cognition impairments in a mouse model of phenylketonuria | JCI

Genetic disruption of F-box protein FBXW11 restores haploinsufficient neurofibromin and corrects neurobehavioral deficits in Nf1+/– neurofibromatosis model mice | JCI

Human cells transduced with AAV2 encoding CRISPR/Cas9 produce HLA-A∗2 restricted epitope that activates CD8+ lymphocytes, flagging broad immunogenicity risk | Mol Therapy Methods Clin Dev.

The synthetic lethal interaction between CDS1 and CDS2 is a vulnerability in uveal melanoma and across multiple tumor types | Nature Genetics

Platforms

Maturation of human cardiac organoids enables complex disease modeling and drug discovery | Nature Cardiovascular Research

UCSF startup Evertree publishes differentiation protocol for hiPSC-derived motor neuron like populations enriched for nitric oxide-production that engraft and improve gut motility defects in mice | Nature

Startup Arzanya publishes yeast display platform for affinity ranking of macrocyclic peptide libraries | Nature Communications

Functional assessment of all ATM single nucleotide variants using prime editing and deep learning | Cell

Roche/University of Basel CRISPRi screening collaboration shows knockout of AP1M1 delays transport of antisense oligonucleotides into lysosomes, increases endosome escape and enhances activity. | Nature Communications

Use of AlphaFold2, molecular dynamics, and quantum chemistry to refine derivatives of biochanin A inhibitor of metabolic enzyme 3βHSD1 improves affinity, oral availability and tumor-inhibitory activity in xenograft prostate cancer models | PNAS

Tools and lead compounds

Xyloglucan-conjugated dexamethasone prodrug activated in mouse gut by microbial glycosidases | Science

Development of ‘clinically viable’ non-muscle myosin II small molecule inhibitors | Cell

Small-molecule proline/phenylalanine mimetics against prostatic acid phosphatase identified via DNA-encoded library screening improve survival in preclinical prostate cancer models when conjugated to radioligands or cytotoxic agents | Nature Biomedical Engineering

Direct evidence that IgG Fab domains interact with human FcRn upon IgG-FcRn complex formation | Communications Biology

Trefoil factor 2-mouse serum albumin fusion peptide synergizes with anti-PD-1 to inhibit tumor growth/metastasis and prolong survival in gastric cancer mouse models | Cancer Cell

FcγR1-serum albumin-poly(l-lactide) fusion ‘nano-adaptor’ carriers for bi-, tri- or tetra-specific antibody mixtures in preclinical breast cancer, melanoma and pancreatic cancer models | Nature Biomedical Engineering

Patient PPP2R1A mutations predict improved survival after checkpoint blockade, with genetic or pharmacological blockade of PPP2R1A potentiating PD1 inhibitor activity in preclinical ovarian clear cell carcinoma models | Nature

Oligonucleotide mimics of NDUFA4L3/miR-147 inhibit cytochrome c oxidase subunit NDUFA4 to increase mitochondrial DNA release/STING activation in tumor-associated macrophages in melanoma xenograft models | Immunity

LNP-delivered modRNAs encoding mitochondrial DdC base editor fix mtDNA mutations in primary patient cells derived from patients with Gitelman-like syndrome | PLoS Biology

Startup news

Last call, if you're working on a novel solutions to address non-T helper 2 pathways in respiratory disease: J&J Innovation’s deadline for applications is July 11.

Good news for those in search of funding. Several initiatives proposing funding for early-stage biotech startups and translational investigators:

UK government to invest £380 million in ‘engineering biology’ with initial £196 million for R&D and £184m for pilot and scale-up infrastructure.

Japanese venture fund AN Venture Partners closes $200 million fund investment strategy includes pre-proof-of-concept startups using any therapeutic modalities in any disease

New York-based Catalio Capital Management fund raises $400 million fund IV to target early-stage startup formation and assets from distressed biotechs

$200 million Chugai Venture Fund launches to invest in startups with innovative drug programs

Aditxt ‘social innovation platform’ announces bitcoin-based strategy to accelerate biotech commercialization, with lingering questions around liquidity

Meanwhile, choppy waters for an established NYC life science venture fund ATP:

Legal case pits ATP partners against its lone investor, the Rigmora Holdings family office of Russian billionaire Dmitry Rybolovlev, which is withholding funds for portfolio company capital calls, with many facing “imminent collapse”

And another story following up on Boston Globe’s April story about the frosty fundraising environment sending Boston’s biotech cluster from boom to bust:

Party is over for Boston biotech sector

​Industry joins academia’s strong rejection of proposed cuts to NIH and STTR/SBIR grants, while China is set to take advantage of the current climate in the US:

Open letter to US Congress from 110 biomedical, health sciences industry leaders tell Trump administration’s proposed cuts could have “catastrophic effect”

The gap in NIH extramural grant disbursements now lags annual 2016-2024 average by whopping $4.7 billion

China set to overtake the US on R&D spending this year:

Preclinical funding

Biotech Funding Table
Date Company (location) Amount (millions) Funding type (lead investors) Therapeutic (lead) focus
June 10, 2025 Mosanna Therapeutics (Basel, Switzerland/Redwood City, CA) $80 Series A (Pivotal bioVenture Partners, EQT Life Sciences, Forbion, Broadview Ventures, and Norwest) Nasally administered small molecule pan-K+ channel inhibitor to increase upper airway muscle tone and reducing apnea and hypopnea episodes
June 18, 2025 Infinity Bio (Baltimore, MD) $8 Series A (Illumina Ventures) ORFeome-scale libraries of proteins covalently coupled to unique DNA barcodes for analysis of autoantibodies in plasma by sequencing
June 19, 2025 Anoat Therapeutics (Paris, France) €2 Seed (AdBio Partners) Antisense and siRNA oligonucleotides to address ENaC for cystic fibrosis patients unresponsive to small-molecule CFTR modulators
June 27, 2025 Astraveus (Paris, France) €16.5 Seed (Adbio Partners) Single-use, miniaturized microfluidic devices for manufacturing CAR-T and other cell therapies
June 30, 2025 Laverock Therapeutics (London, UK) £6.5 Seed extension (Calculus Capital) iPSC-derived monocyte/macrophage-like cells in which myeloid microRNA genes are edited to promote myeloid cell differentiation for use in hematological and solid tumors
June 30, 2025 Pelthos Therapeutics (Freehold, NJ) $50 Private placement (Murchinson and Ligand Pharmaceuticals) Small molecule nitric oxide-releasing topical treatment for poxvirus-mediated molluscum contagiosum and acne
July 1, 2025 Syntis Bio (Boston, MA) $38 Series A (Cerebrus Ventures) 24 h polydopamine coating that masks duodenum wall for diabetes or delivers methionine gamma lyase or leucine decarboxylase for homocystinuria and maple syrup disease, respectively
July 1, 2025 Quiver Bioscience (Cambridge, MA) $2.5 Grant (NINDS SBIR) Use of human cells in vitro to train machine learning models developing safety and tolerability of CNS-targeted ASOs targeting SCN9A and KCNT1 for neuropathic pain and epilepsy
July 1, 2025 FELIQS (Fukuoka, Japan) $9 Series A (Unnamed pharma and Beyond Next Ventures) Once-daily oral small-molecule binder of lipid hydroperoxides and modulator of ferroptosis/lipid peroxidation enzyme pathways for retinopathy of prematurity and for other rare pediatric retinal diseases
July 2, 2025 Argon AI (New York, NY) $5.5 Seed (Crosslink Capital and Wireframe Ventures) ‘AI-native workspace’ database covering assets, diseases, patient journeys, and clinical trials for competitive intelligence

Preclinical deals

Biotech Deal Table
Date Type Payer (location) Payee (location) Upfront payment (millions) Milestones amount (millions) Total (up to millions) Therapeutic Focus
June 24, 2025 Strategic collaboration Plectonic Biotech (Munich, Germany) Sonrai Analytics (Belfast, UK) ND ND ND Plectonic’s T-cell engager F(ab) and scFv mAbs on DNA origami nanocarriers to be targeted against epitopes identified by Sonrai’s multi-modal models trained on single-cell RNA sequence and proteomic mass spectrometry data for predicting antigen co-expression patterns in cancer
June 26, 2025 License Percheron Therapeutics (Melbourne, Australia) Hummingbird Bioscience (Singapore) ND ND $290 VISTA-targeting IgG4 mAb that blocks VSIG-3 and PSGL1 engagement
June 27, 2025 License Calico (S. San Francisco, CA) Mabwell (Shanghai, China) $25 $571 $596 Global rights (except mainland China and territories) to mAbs targeting interleukin 11, including lead asset (9MW3811)
July 1, 2025 Strategic collaboration Argenx (Ghent, Belgium) Unnatural Products (Santa Cruz, CA) “Double-digit $ millions” upfront and “un-disclosed equity” payment $1,500 $1,500 AI-enhanced oral synthetic macrocyclic peptide-discovery platform for use against undisclosed targets decided by Argenx
July 1, 2025 Asset sale Ho’ola Therapeutics (Aiea, HI) Oncternal Therapeutics (San Diego, CA) $3 $65 $68 ROR1-targeting zilovertamab and anti-ROR1 CAR-T cell therapy for mantle cell lymphoma/chronic lymphocytic leukemia and relapsed B-cell lymphoma
July 1, 2025 License HanchorBio (Taipei, Taiwan) Henlius (Shanghai, China) $10 $182 $192 Engineered SIRPα-IgG4 Fc fusion
July 1, 2025 License Beacon Therapeutics (Alachua, FL) Abeona Therapeutics (Cleveland, OH) ND ND ND AAV-204 vector derived via directed evolution of library of comprehensive AAV serotypes for intravitreal injection for eye diseases

Stay in touch

We hope you enjoyed this issue of The Needle and hit the button below to receive forthcoming issues into your inbox

If you’re interested in commercializing your science, get in touch. We can help you figure out the next steps for your startup’s translational research program and connect you with the right investor. Follow us on X, BlueSky and LinkedIn. Please send feedback; we’d love to hear from you (info@haystacksci.com).

Until next week,

Juan Carlos and Andy

Subscribe to The Needle