Mouse models and CRO selection critical to de

Mouse models and CRO selection critical to de-risking neurology drug development By Roohi Mariam Peter Add Labiotech as your Google Preferred Source 7 minutesmins July 28, 2026 7 minutesmins Share WhatsApp Twitter Linkedin Email Photo credits: GemPharmatech Add Labiotech as your Google Preferred Source Newsletter Signup - Under Article / In Page"*" indicates required fieldsFacebookThis field is for validation purposes and should be left unchanged.Subscribe to our newsletter to get the latest biotech news!By clicking this I agree to receive Labiotech's newsletter and understand that my personal data will be processed according to the Privacy Policy.*Company name*Job title*Business email* 2026 marks the third anniversary of Leqembi’s approval as the first disease-modifying Alzheimer’s therapy. While a landmark achievement, that it stands alone is indicative of the harsh reality that developers face when targeting the central nervous system (CNS). Clinical success is sparse across modalities and indications, yet CNS therapeutics fare even worse in clinical trials comparatively. GemPharmatech, a preclinical CRO service provider that specializes in mouse model creation, thinks upstream in vivo modeling decisions are at the core of these downstream failures.Table of contentsWhy use mouse models for CNS drug developmentFor successful drug development targeting the CNS, scientists must understand what happens in the brain before symptoms appear, how disease progresses at the cellular or circuit level, and more critically, how a therapy changes the course of a disease.Some of these questions for other indications like oncology are much easier to answer. But unlike a tumor, which can be serially biopsied, the brain cannot handle repeated sampling. This leaves researchers working from a single, isolated timepoint in the disease’s progression.“A lot of what we know about disease mechanisms comes from a snapshot, and a snapshot that is already very late,” explained Rikki Feng, GemPharmatech’s Neuroscience Pipeline Leader.Not to mention another challenge facing innovators in neuroscience: the human brain is remarkably well-protected by the blood-brain barrier (BBB), making it difficult to study as well as understand the efficacy of novel therapeutics. “The BBB is our body’s most elegant protective system. But for drug developers, it’s also the biggest problem. Large molecules like antibodies, enzymes, gene therapies, and vectors often cannot cross over, and that’s a huge problem.”Feng explained these are the main reasons mouse models offer a more “powerful window” into the brain, allowing researchers to ask questions and get answers they can’t get from the patients themselves. “We can trace a clear cause and effect line from molecular pathology all the way to behavior. That kind of readout is enormously difficult to get from human data alone. That’s the value of a good model.”But not all models are created equal, Feng pointed out, explaining how “traditional models” often fail to capture the complexity and late-onset progression characteristic of many neurological disorders.Gempharmatech’s mouse models mimic human disease pathologyAt GemPharmatech, the neuroscience team is focused on designing more precise, translationally relevant mouse models — genetically engineered platforms built to close the gap between preclinical promise and clinical outcome.To address the serious issue of the BBB, GemPharmatech looks to the transferrin receptor; the main shuttle for receptor-mediated transcytosis, a cellular transport mechanism used to transport molecules across the BBB. They create mice with a humanized transferrin receptor to test whether a drug that is designed to move past the BBB can truly do so. “This lets us de-risk brain delivery into a mouse model long before anyone runs an expensive human image or a cerebrospinal fluid (CSF) study. Without that step, many promising biologics simply never reach their targets at a meaningful level,” said Feng.Apart from models with humanized transferrin receptors, the company also develops models with humanized CD98HC and IGF-1R, the other keys to help unlock the BBB, Feng explained.As many drug candidates are optimized for human proteins, if they are only tested in mouse proteins, the pharmacological effect would be different from how it is in humans. That’s why humanized mouse models are essential: it bridges the species mismatch that occurs at the target level.Disease staging is another factor to keep in mind when designing these models.“Most animal studies start treatment very early, often before robust symptoms even appear. But human trials typically enroll patients who already have significant pathology. The timing difference can completely flatten an efficacy signal, so we need to provide model and study designs that match different stages of the disease, not just everything at the beginning,” said Feng.Moving past “traditional models” for Alzheimer’s and Parkinson’s diseaseGemPharmatech has developed the FAD3T mouse model to recapitulate the amyloid and tau pathology characteristic of familial Alzheimer’s disease (FAD), a form of Alzheimer’s disease caused by autosomal dominant mutations in APP and PSEN1. FAD3T carries humanized APP and PSEN1 familial AD mutations alongside a human MAPT/Tau transgene, which more fully reproduces the amyloid cascade and downstream tau pathology seen in human AD than APP/PSEN1 mutations alone.“We are not just looking at one piece of the puzzle. We are capturing the amyloid pathology, the tau pathology, the neuroinflammation, and the memory decline. That’s much closer to what we actually see in early onset of the Alzheimer’s disease patient,” said Feng.GemPharmatech uses that same principle — building a model that captures multiple, converging disease features rather than a single pathological marker — into their Parkinson’s disease model. They overexpressed human alpha-synuclein, the protein implicated in the neurodegenerative condition, to capture not just protein aggregation but the downstream motor and neuronal consequences that follow.“What’s special is that this model captures both the progress of the motor dysfunction and the alpha-synuclein aggregation. And crucially, it also shows the loss of the dopaminergic neurons over time; something that was often missing in our old models. The disease progress in the mouse really mirrors the human pathology sequence.”GemPharmatech builds these models to be translatable from the outset, designed to reflect the phenotypes, biomarkers, and endpoints clinicians rely on to evaluate patients. That alignment gives researchers a clearer read on whether functional deficits improve and whether biomarkers shift in directions consistent with what’s expected in human trials.“We pay very close attention to human cohort data,” said Feng. “That’s what makes the model a far more reliable platform for testing a drug.”Avoiding pitfalls in CNS drug development by partnering with a CROResearchers and drug developers need to know whether a therapeutic candidate lowered pathology markers, improved underlying biology, and is on a credible path toward the clinic. Those insights depend on translationally relevant models, which is exactly why selecting your partner for in vivo efficacy testing is such a consequential decision.“There’s a tendency, especially in smaller biotechs, to really want to run fast. They’re under a lot of pressure in terms of budget and timeline,” explained Brandy Wilkinson, North American CEO of GemPharmatech. She explained how this often results in scientists and companies choosing the most convenient model, rather than one that really replicates the disease. “That can be very dangerous,” she claimed, “particularly in something like neurology.”This race to produce promising data — whether to satisfy investors or claim first-to-market status — is understandable. But it also creates an opening for service providers to sell speed and convenience over rigor, offering a quick start to what is often expensive efficacy testing. Wilkinson positions GemPharmatech differently: as an expert partner working alongside developers from the outset, not a vendor for a single study.“It’s really critical for that consultative interaction to happen. The translational gap in neurology is pretty wide. If you’re not asking the right questions or the questions don’t get asked early enough, you can end up with running a lot of studies and, frankly, wasting a lot of money on models that may not really answer your question.”“It’s really critical for that consultative interaction to happen,” said Wilkinson. “The translational gap in neurology is pretty wide. If you’re not asking the right questions or the questions don’t get asked early enough, you can end up with running a lot of studies and, frankly, wasting a lot of money on models that may not really answer your question.”The earlier that GemPharmatech’s scientists can sit down with the client to understand the candidate’s therapeutic mechanisms and the disease it is targeting, the better, explained Wilkinson. “We want to see these drugs help patients as much as the developers themselves.”Feng concurs. While she was able to extensively elaborate on humanized models and endpoint alignment, the available behavioral services offered by GemPharmatech, and all the ways her team helps clients understand the microenvironment changes in the brain during treatment, she explained success ultimately comes down to their eagerness to partner with and provide their clients the best possible insights that will drive successful therapeutic development.“We actually treat every program not just as a single transaction, but a translational campaign,” said Feng. “We sit down with them and have a clear discussion about the therapeutic window, the target engagement evidence, and whether biomarker changes all point in the same direction. We want them to walk away with a comprehensive view of their drug profile and with much greater confidence going to the next stage.”Feng hopes that the line between preclinical and early clinical science becomes more connected.“If we can get this right, we might finally start delivering neurology drugs with a success rate that looks more like oncology. And that would be transformative for patients,” she said.To dive deeper into CNS drug development and the role of mouse models, listen to our recent podcast: Explore other topics: Alzheimer'sCDMONeurological disordersParkinson'sResearch