Proceedings · Session S-315 · filed November 17, 2026

Translational ScienceSession paper

Preclinical Deals and Data: NIH's $2.5M and Ten Early-Stage Bets

NIH paid Sinopia $2.5M for a metabolomics model; Sapreme restored dystrophin above 55% in mice; Whitehead cut off-target editing 41-fold. November's preclinical claims, checked.

By Amara Osei4 min read806 words

Summary

  • NIH awarded Sinopia Biosciences $2.5 million to build a metabolomics foundation model for drug discovery.
  • Sapreme's SPM-2551 restored dystrophin to 35-55% in mice 56 days after single doses of 1-3 mg/kg.
  • Whitehead's VLP base editors cut off-target mitochondrial editing up to 41-fold versus plasmids.
  • Purespring's PS-003, first kidney-targeted gene therapy for Alport syndrome, showed COL4A5 expression in mouse GBM with an IND planned for next year.
  • ALLOX raised $10.5 million in seed funding for allosteric drug discovery.
Preclinical Pulse: The research catching our eye in November 2026
FigurePreclinical Pulse: The research catching our eye in November 2026 — tiburi / pixabay

By Amara Osei

The U.S. National Institutes of Health has paid Sinopia Biosciences $2.5 million to build a metabolomics foundation model, one of ten preclinical developments this month that R&D managers should weigh against their own pipeline math. That grant anchors a November selection spanning gene therapy, base editing and drug delivery, each claim examined below.

What did Sanofi buy from a four-month-old startup?

Sanofi has enlisted ScionxBio, a Cambridge, U.K. biotech launched in June 2026, to make its mRNA candidates persist longer in the body. ScionxBio's TERA platform, which the company describes as "powered by advanced deep learning architectures," deciphers "complex structure-activity relationships within high-throughput combinatorial screens" to engineer long-acting RNA. ScionxBio will run development to in vivo validation; Sanofi then takes on IND-enabling work, clinical studies, manufacturing and commercialization. Targets remain undisclosed beyond "conditions linked to protein expression" — a gap worth noting before crediting the deal with therapeutic scope. Financial terms were not revealed.

Sinopia's federally funded model will unify metabolomics data and feed its LEADS platform, which combines multi-omics, high-throughput screening and machine learning. LEADS has already produced SB-0110, a small molecule targeting PKA-II, a kinase downstream of the dopamine receptor, now in IND-enabling studies for Parkinson's disease and levodopa-induced dyskinesia. The platform's discovery record, so far, rests on one asset.

Which measured results stand up?

Sapreme Therapeutics presented data at the Oligonucleotide Therapeutics Society meeting in Colorado showing its exon skipper SPM-2551 restored dystrophin to more than 35% and 55% in mouse models 56 days after single doses of 1 and 3 mg/kg, respectively. That is roughly half to full restoration of a protein Duchenne patients lack. The single-stranded oligonucleotide achieved 5- to 10-fold higher exon skipping than four already-approved Duchenne skippers — though cross-trial comparisons in mice do not guarantee clinical superiority.

Whitehead Institute researchers reported a virus-like particle platform for mitochondrial base editors that cuts mtDNA-wide off-target editing by up to 41-fold versus plasmid delivery, while improving on-target editing. In mice, a single eye injection corrected the mutation behind Leber hereditary optic neuropathy; the editors installed a pathogenic mutation in 88% of mouse embryonic fibroblast mtDNA in testing. The work addresses a real liability of plasmid and AAV delivery, and the eye-injection data are measured, not projected.

Purespring Therapeutics presented the first kidney-targeted gene therapy, PS-003, at ASN Kidney Week 2026 in Colorado. In a 3D glomerulus model, PS-003 delivered COL4A5 — the gene mutated in about 80% of Alport syndrome cases — and protein was detected in the glomerular basement membrane-like matrix. A single dose expressed COL4A5 in the GBM of mice, with no treatment-related side effects observed at the preclinical level. An IND filing is expected next year; safety claims rest on animal data alone.

How solid are the earlier-stage claims?

Asgard Therapeutics published in Immunity that its off-the-shelf gene therapy AT-108 reprogrammed cancer cells into dendritic cells in mouse models of seven tumor types — melanoma, colon, bladder, breast, lung, pancreatic and glioblastoma — promoting antitumor T and B cell immunity via induced tertiary lymphoid structures. The Swedish company presented at the Society for Immunotherapy of Cancer conference in Arizona and expects clinical entry next year. Mouse breadth is encouraging; human translation of in vivo reprogramming remains unproven.

Propanc Biopharma reported its two-proenzyme formulation cut pancreatic cancer cell proliferation about 50% as monotherapy, and 75% and 66% in combination with galunisertib in cancer models and chemotherapy-resistant cells, respectively. It reduced cancer-associated fibroblast invasion by 85%, while its effect on normal fibroblasts — around 22% — was modest and not statistically significant. The company said the findings supported "investigating stromal selectivity as an important parameter when evaluating future combination strategies." Sample details went unspecified.

Link Biologics, a University of Manchester spinout, completed IND-enabling studies for LB001, a TSG-6-based eye drop for dry eye disease, with clinical entry planned for 2027. The company claims superiority over standards of care such as cyclosporine solutions and corticosteroids; no comparative figures were published.

ETH Zurich scientists built an "RNA archive" — engineered stress granules forming a protective film around mitochondria — that tracks several thousand RNA molecules over days. The method could reveal which cells survive chemotherapy and how cancers develop.

Spanish startup ALLOX closed a $10.5 million seed round to scale its allosteric-drug platform and AI models for identifying modulators from protein-protein interactions. The round funds discovery infrastructure, not yet a clinical candidate.

Most of these results come from corporate press releases or conference abstracts, and independent replication will be the test. Purespring's IND filing and Asgard's clinic entry, both slated for next year, are the nearest checkpoints.

via LABIOTECH.eu (Source)

Filed under

  • nih-funding
  • preclinical-research
  • gene-therapy
  • drug-discovery
  • biotech-partnerships
Share this article:

More from Amara Osei

Amara Osei

Show full bio

News editor covering money, funding and market shifts around research and development at Hypothesis Wire.

253 articles

References

  1. Sanofi Pays $1 Billion Upfront to Extend Regeneron Antibody Alliance
  2. Forbion raises $2.6 billion, its largest life sciences fund to date
  3. Bertis secures $2.23M state grant for AI drug-target platform
  4. NIH Funding Crisis Could Reshape Biotech for Years, Drug Discovery News Reports
  5. Pfizer to Deploy Edison's Kosmos AI Scientist Across R&D Pipeline

« Previous articleNext article »