Trailhead Biosystems
Building More Reliable Human Cell Models

David Llewellyn, PhD, MBA, President and Chief Executive Officer

The most important lesson is that generating a cell with the right identity is not enough. The cell also needs to demonstrate the maturity, functional behavior and consistency required for its intended research application.

iPSCs provide an extraordinary starting point because they can theoretically produce almost any human cell type. That same biological potential makes differentiation difficult to control. Small differences in starting cells, timing, culture conditions or combinations of biological signals can materially change the output of a differentiation protocol. A protocol may produce encouraging results once, yet still fail to deliver the same quality across multiple runs or cell lines.

That has shaped how we develop products at Trailhead. We begin with the research application and define what the resulting cell must do, not simply which markers it should express. Identity, purity, maturity, functional performance, reproducibility and scalability all have to be considered together. The objective is not just to make a differentiated cell. It is to create a dependable biological model that researchers can use with confidence.

The Challenges of Achieving Reproducible and Scalable Cell Differentiation

One persistent challenge is that iPSC lines are not interchangeable. Genetic background, reprogramming history, culture conditions and the state of the starting population can all influence how a line responds to a differentiation protocol. A method that performs well in one line may produce a different mixture, yield or level of maturity in another.

Reproducibility is also difficult because differentiation is controlled by many interacting biological variables. Conventional optimization often isolates one variable at a time, even though changes in concentration, timing or culture conditions may alter the effects of several other inputs. This can result in protocols that work within a narrow set of conditions but are not robust when transferred across lines, operators or production scales.

Maturity creates another challenge. Cells can express relevant markers without achieving the functional behavior required for disease modeling or drug screening. Then the process must be scalable without losing the identity and performance established during product development. Solving these problems requires a deeper understanding of the biological process, clear product specifications and validation that confirms both marker expression and functional performance.

How Trailhead Evaluates the Functional Relevance of iPSC-Derived Cell Models

We start by understanding the intended use of the model and asking what a researcher must be able to measure with that cell type. Marker expression helps establish identity, but it does not establish physiological relevance on its own.

Our approach considers multiple factors, including morphology, positive and negative marker expression, purity, viability, maturity and functional behavior. The functional assays depend on the biology and application. For example, endothelial cells should demonstrate relevant vascular behavior, while neuronal models may require specific electrophysiological activity or responses to defined stimuli.

We also evaluate whether that performance can be reproduced across batches and maintained under the conditions a customer will actually use. A useful model must do more than resemble the intended cell type in a single experiment. Cells must provide relevant and interpretable readouts with enough consistency to support repeatable research.

Cell differentiation is governed by networks of interacting signals rather than a single dominant variable. Conventional optimization can identify whether one factor appears helpful under a particular set of conditions, but it can miss how the effect of that factor changes when other inputs also change.

HD-DoE® allows us to examine many biological factors and responses within a combined experimental and computational framework. The resulting data help reveal interactions, tradeoffs and regions of the biological design space that would be difficult to identify through sequential trial and error. Mathematical modeling then helps us understand which combinations are most strongly associated with the desired cell identity and performance. The predictions still have to be tested and biologically validated, but the platform gives our scientists a more systematic way to decide where to focus.

For product development, this means we are not simply looking for one condition that works. We aim to understand the process well enough to establish conditions that are robust and aligned with the product’s intended use. That understanding supports more consistent protocols and provides a stronger foundation for transferring them across production runs, scales and different iPSC lines.

Addressing Gaps in iPSC-Derived Models for Drug Discovery and Disease Research

Many commercially available iPSC-derived products are concentrated in cell types that are comparatively well established such as cardiomyocytes and neuronal cell types that are mixtures of subpopulations. Important gaps remain in cell models that require greater maturity, purity or functional specificity than existing products provide.

Blood, immune, nervous and vascular systems are areas of focus for Trailhead. Our development pipeline includes cell types such as A9 dopaminergic neurons, PV+ GABAergic interneurons, astrocytes and microglia, alongside our commercially available hematopoietic progenitor cells and endothelial cells. These models can support research in neurodegeneration, neurological disease, vascular biology, inflammation, drug discovery and toxicity testing.

We prioritize opportunities based on a combination of unmet scientific need, customer demand, biological feasibility and the ability of our platform to create a meaningful advantage. We are especially interested in applications where researchers cannot readily obtain the desired primary cells, or where existing iPSC models lack sufficient specificity or function. Our goal is not to build the largest catalog of familiar cell types. It is to expand access to physiologically relevant cell models that have historically been difficult or impractical to obtain.