Longinus Therapeutics is developing the first targeted treatments for 15q11.2 BP1-BP2 microduplication — a genetic cause of autism affecting hundreds of thousands of families worldwide. The science exists. We are building the program.
Longinus Therapeutics was founded in Madrid by a parent of a child with 15q11.2 BP1-BP2 microduplication. After finding almost no targeted treatment options — despite a well-understood genetic mechanism — we decided to build the program ourselves.
The science exists. The CYFIP1 gene is well-characterized. The downstream pathways are druggable. A realistic ASO development timeline is 5–7 years to first patient dosed, 7–9 years to potential approval. What is missing is a focused, well-funded program to make it happen.
We fund the research, own the IP, and are building the data package that will bring pharma partners to the table. Our model is proven — it is exactly how Angelman syndrome went from no treatment to Phase 3 trials: patient foundation → preclinical data → pharma partnership.
The 15q11.2 BP1-BP2 microduplication places an extra copy of CYFIP1 in every cell. CYFIP1 overexpression disrupts two core pathways — NMDA receptor suppression and mTOR overdrive — both confirmed in human post-mortem brain tissue. CYFIP1 is dosage-sensitive in both directions: deletion and duplication each cause disease, which means the ASO therapeutic window must be carefully characterized.
Six interconnected biological pathways are perturbed by CYFIP1 overexpression — each offers a distinct therapeutic handle. The atomic-level AI design stack that targets these pathways is detailed in our AI Drug Design section.
Our program follows the proven rare disease pathway — fund the preclinical science, publish the data, attract a pharma partner for IND and trial. Pharma partners write meaningful checks at IND-ready or Phase 1 inflection — not at mouse efficacy. Our goal: reach that inflection with $2–4M, with a defined Month 24 Go/No-Go gate on the dose-response window.
⚠️ These are off-label uses requiring specialist neurologist supervision. Longinus does not prescribe treatments.
The curative CYFIP1 ASO is 5–7 years from first patient dosed. In the meantime, several FDA/EMA-approved drugs address the downstream consequences of CYFIP1 overexpression off-label. These require specialist neurologist supervision — but the biological rationale is strong, and the safety records are well-established. The right order matters: test first, then treat.
Before trialling any drug, three tests will tell you which treatments have the highest likelihood of benefit for your child specifically. These tests are non-invasive, available privately in Spain, and should be done regardless of whether drug treatment is pursued.
Test for folate receptor alpha antibodies (blood draw). If positive, leucovorin has the strongest clinical evidence for non-verbal autism. Should be done before any drug trials — a positive result changes the priority order entirely.
Measure pgACC glutamate levels (non-invasive MRI spectroscopy). Children with elevated glutamate on MRS respond 80% vs 20% to memantine — this is the biomarker that determines whether memantine is worth trying. Costs €300–800 privately in Madrid.
Review neurologist interpretation of the duplication, parental testing status, and whether Simons Searchlight enrollment is complete. Parental inheritance status affects prognosis framing and research eligibility.
Four approved drugs have mechanistic rationale against confirmed CYFIP1-duplication pathologies. All are Tier B — strong mechanistic fit with clinical evidence from adjacent synaptopathies (Fragile X, TSC). None have been tested in a 15q11.2 duplication cohort. All require a specialist.
Open-source, Nobel Prize–winning AI tools (RFdiffusion, AlphaFold3, ProteinMPNN) now make it possible to design atomic-precision binders against CYFIP1 at a fraction of historical cost. Alongside our ASO track, we are developing a custom peptide program that directly disrupts the CYFIP1–eIF4E protein–protein interaction — a bottleneck in the mTOR-dependent translation cascade that drives CYFIP1 overexpression pathology — and a complementary PROTAC track that degrades the excess CYFIP1 protein itself.
CYFIP1 directly binds eIF4E — the mRNA cap-binding protein that controls the rate of translation initiation. In 15q11.2 BP1-BP2 microduplication, excess CYFIP1 sequesters eIF4E in an abnormally stable repressor complex, dysregulating synaptic protein synthesis and impairing activity-dependent plasticity.
Custom peptides are designed to mimic or block the CYFIP1 binding surface on eIF4E — either competing with the repressor complex or stabilizing the release of eIF4E at the right moment during synaptic activation.
We use the same Nobel Prize–winning AI stack that underpins our PROTAC program to design peptide binders from scratch:
A therapeutic peptide must cross both the blood–brain barrier and the neuronal plasma membrane to reach its intracellular target. We are evaluating two delivery strategies:
Every peptide lead is validated in cortical neurons derived from iPSCs reprogrammed from patients carrying the 15q11.2 BP1-BP2 microduplication — the same human model system used across all Longinus drug programs.
Readouts in patient neurons include: eIF4E co-immunoprecipitation (target engagement), polysome profiling (translation correction), dendritic spine morphometry (structural rescue), and synaptic protein levels (functional correction). This is the same validation funnel used to qualify ASO and PROTAC leads before animal studies.
We follow the proven rare disease model: fund the preclinical science, generate the data package, attract pharma, run the trial.
Derive iPSC-derived cortical neurons from patients with the 15q11.2 BP1-BP2 duplication. These are the human validation model for every drug and gene therapy candidate — confirming what works in patient-specific cells, not just a mouse.
Develop and validate a CYFIP1-targeting intrathecal ASO. The central scientific question is the dose-response window: CYFIP1 is dosage-sensitive in both directions (deletion also causes disease). Full dose-response curve with haploinsufficiency comparison is Year 1 priority. ASO is first choice for reversibility — critical given the U-shaped risk profile. ION582 (Ionis Angelman, Phase 3) and ATTUNE (ION440, MECP2 Dup, Phase 1/2) are the direct platform benchmarks.
Parallel small molecule program using AlphaFold3, RFdiffusion, and Chai-1 to design PROTAC protein degraders targeting CYFIP1. A PROTAC could degrade exactly the excess 33% of CYFIP1 in a dose-titratable way — potentially solving the therapeutic window problem. Oral delivery vs lumbar puncture. Timeline: lead compound validated in iPSC neurons within 18 months. Cost to lead: €200-600K.
AAV-delivered RNAi is the Plan B modality — reserved for after the dose-response window is well-characterized in the ASO program. Single-administration advantage is compelling, but irreversibility is disqualifying until the safe window is defined. CRISPRi epigenetic silencing of the CYFIP1 promoter remains a longer-term option (10+ years). CYFIP1 is not imprinted, removing the elegant Angelman-style epigenetic handle.
Fund a longitudinal natural history study in collaboration with Simons Searchlight. Without this, FDA will not approve a treatment. Documents disease trajectory, cognitive and language endpoints, biomarkers — the data package that makes a clinical trial designable.
CYFIP1 does not yet have a clinical program — but every neighboring ASO trial that reads out changes the de-risking calculus for ours. These are the developments we are tracking most closely.
n-Lorem and collaborators reported the first individualized ASO treatments for KCNT1-related epilepsy in Nature Medicine (PMID 41981306). The cases extend the bespoke-ASO precedent (milasen, atipeksen) into a new gain-of-function neurodevelopmental gene — the same class of genetic mechanism as CYFIP1 overexpression. This is a direct platform read-through for our program.
Read on PubMed →Jagasia et al. (Hoener lab, Roche) reported in Nucleic Acids Research the full preclinical package for rugonersen, a clinical-stage ASO targeting UBE3A-ATS to reactivate the silent paternal UBE3A allele in Angelman syndrome (PMID 40884397). A single intrathecal dose produced sustained UBE3A re-expression in cynomolgus monkeys for months. This is a second industry-grade ASO program (alongside Ionis ION582) validating the intrathecal ASO modality for 15q11-q13 region neurodevelopmental disorders — the same chromosomal neighborhood as CYFIP1.
Read on PubMed →Smeenk et al. from the Elgersma group at Erasmus MC published a xenotransplantation model in Scientific Reports enabling preclinical testing of human-specific ASOs in vivo — using human iPSC-derived neurons grafted into mouse brain (PMID 41764319). This addresses a core translational gap: human ASOs cannot be tested in standard rodent efficacy models. The platform is directly applicable to a human-specific CYFIP1 ASO.
Read on PubMed →The first Phase 1/2 ASO trial for a chromosomal duplication syndrome — MECP2 duplication — is enrolling boys aged 2–17. ATTUNE is the closest analog to a future CYFIP1 ASO program: intrathecal delivery, overexpression reduction, neurodevelopmental endpoints. Top-line readout expected 2027. A positive result directly de-risks our mechanism class.
ClinicalTrials.gov →Servier's ASO program targeting KCNT1 (a gain-of-function channelopathy) entered clinical development. With ATTUNE (Ionis) and KANDLE (Servier) both active, the regulatory and manufacturing playbook for intrathecal ASOs targeting overactive neuronal genes is consolidating — exactly the path a CYFIP1 ASO would follow.
ClinicalTrials.gov →Ionis' Angelman ASO is now the most advanced neurodevelopmental ASO in pediatric autism-spectrum indications. Phase 1/2 data showed gains in communication, cognition, and motor function. This is the platform proof-of-concept that intrathecal ASOs can drive functional improvement in a genetically defined pediatric neurodevelopmental population — the exact thesis of CYFIP1.
Ionis pipeline →The Ming/Song lab (Johns Hopkins) showed in Biological Psychiatry that CYFIP1 overexpression and deletion produce opposite behavioral and molecular phenotypes via diametric control of NMDA receptor complex translation (PMID 34247782). This is the mechanistic anchor for an ASO that lowers CYFIP1 back toward wild-type dosage.
Read on PubMed →The Rare Pediatric Disease Priority Review Voucher program is currently authorized through September 2029. Recent PRV transactions: Zevra $150M, Abeona $155M, Jazz $200M. CYFIP1 ASO qualifies as a rare pediatric serious disease treatment — securing the designation before sunset (or under any reauthorization) is a meaningful component of program value.
Investor inquiries →A CYFIP1 ASO program needs deep expertise across mGluR/mTOR synaptic biology, neurodevelopmental clinical trial design, ASO chemistry, and 15q11.2 patient phenotyping. We are assembling that team now.
One of the world's leading authorities on mTOR-pathway neurodevelopmental disorders, including tuberous sclerosis, Angelman syndrome, and MEF2C haploinsufficiency. His lab co-developed the preclinical groundwork that enabled the Angelman ASO programs now in clinic, and his group has published extensively on translating mouse synaptic-plasticity findings into therapeutic targets.
Chief of Neurology at CHLA and a leading clinical investigator in neurogenetic syndromes, including Dup15q syndrome, Angelman syndrome, and tuberous sclerosis complex. Co-author of the natural history studies and EEG biomarker work that underpins current trial-readiness for several rare CNV-driven conditions.
We are seeking a senior advisor with hands-on experience taking an intrathecal ASO from guide-RNA screening through IND-enabling tox (NHP), ideally with prior involvement in MOE/cEt gapmer programs. Background on the ION440 (ATTUNE) or nusinersen platforms is ideal.
We will reserve a permanent advisory seat for a parent or adult patient with a 15q11.2 BP1-BP2 microduplication diagnosis. Drug programs that ignore lived experience design the wrong endpoints and miss the symptoms families actually want fixed.
Whether you are a family with a diagnosis, a researcher, an investor, or a potential pharma partner — there is a concrete role for you.
Register in Simons Searchlight, bank your child's cells for iPSC research, and connect with our patient community. Your data and your cells are the foundation of any future treatment.
Connect with usWe are seeking academic collaborators for iPSC neuron experiments, ASO guide RNA screening, mouse model work, and CSF biomarker development. Funded sponsored research agreements available.
CollaborateThe program reaches a defensible Go/No-Go decision for $2–4M in 24 months. PRV on approval is worth $150–200M (recent 2025–26 transactions: Zevra $150M, Abeona $155M, Jazz $200M) — subject to program reauthorization (current sunset Sept 2029).
Investor inquiriesSubmitting opens your email client with the message pre-filled to marcio@longinus.bio.