Quick answer: Isogenic CRISPR modelling introduces a single genetic variant into an existing iPSC line while holding the rest of the genome unchanged. This enables any resulting phenotype to be linked directly to the variant of interest,  rather than to differences in the wider genetic background. pixCRISPR uses this approach to develop models of MASLD and a range of inherited and rare diseases, including alpha-1 antitrypsin deficiency, Wilson's disease, urea cycle disorders (UCD), familial hypercholesterolemia, and intrahepatic cholestasis. Knock-in, knockout, or gene-corrected clones are generated using the same differentiation and quality-control pipeline across pixlbio's wild-type cell portfolio.

Compare a familial hypercholesterolemia patient's liver sample to a healthy donor's. You are comparing two genomes, two lifestyles, two different ages, and often two different labs' worth of sample handling. Somewhere in that pile of differences sits the biology you wanted to study. Finding it is the hard part.

This is the challenge that isogenic modelling is designed to address, and one that CRISPR gene editing is particularly well suited to solving in iPSCs. Starting with a well-characterised cell line, a specific genetic change can be introduced while the wider genetic background remains unchanged. Differences between the edited line and its parental control can therefore be linked to the variant with much greater confidence, with less interference from background variability or other confounding factors. This is particularly valuable for studying rare diseases, where patient samples are often limited, providing a human-relevant and scalable model in which the effect of an individual variant can be investigated.

What Does 'Isogenic' Actually Buy You?

A widely cited review of iPSCs in disease modelling explains the key advantage clearly: parallel differentiation of an edited disease line alongside its gene-corrected, isogenic control allows a phenotype to be linked to a specific molecular lesion, rather than to other differences that may exist between whatever else happens to differ between unrelated cell sources (Rowe & Daley, Nature Reviews Genetics, 2019). This attribution is central to the approach. A patient-derived line carries a disease-causing mutation within the context of that patient's genome, while a healthy-donor control has a different genome altogether. An isogenic pair substantially reduces that ambiguity by providing a closely matched control.

The approach is also scalable. One of the largest demonstrations is iNDI, an initiative developing a library of isogenic CRISPR-edited iPSC lines recapitulating numerous neurodegenerative disease mutations on a single, shared genetic background. This allows researchers to compare the effects of different mutations using a common reference point, rather relying on whichever control line is available. Liver disease modelling is moving in a similar direction, although the filed currently has a shorter track record and fewer standardised resources.

What Does Isogenic CRISPR Modelling Look Like in Liver Disease?

Three of pixlbio's own disease model programmes are built on exactly this logic, and each has independent precedent in the published literature.

For alpha-1 antitrypsin deficiency (A1ATD), CRISPR-based technologies have been used to model the disease-causing Z allele of SERPINA1 (Glu342Lys) in iPSC-derived hepatocytes. One study generated syngeneic ZZ, MZ, and MM lines on a shared background, enabling the effects of zygosity – from heterozygous carriers to homozygous disease states – to be examined without the added variability of unrelated donor genomes. This comparison demonstrated that even a single Z allele can disrupt hepatocyte homeostasis, while the isogenic design allowed change in protein misfolding and metabolic function to be linked more confidently to the relevant genotype.  

Work conducted by pixlbio further demonstrates how this approach can be applied within a scalable drug-discovery platform. In its A1ATD study, pixlbio generated and differentiated CRISPR/Cas9-engineered wild-type and disease iPSC lines from multiple genetic backgrounds. The resulting hepatocytes reproduced key features of A1ATD, including intracellular accumulation of misfolded and polymeric A1AT, and were used to evaluate responses to both small molecules and RNA-based therapeutics. This highlights pixlbio’s ability to combine gene editing, iPSC differentiation, disease-relevant phenotypic assays and population-diverse models within a platform suitable for large-scale drug-efficacy screening.  

For Wilson's disease, a rare inherited disorder of copper metabolism, pixlbio has also direct experience in developing isogenic iPSC-based models. Starting from an in-house iPSC line, the team used CRISPR gene editing to generate clones carrying two prevalent pathogenic ATP7B mutations: p.H1069Q, commonly found in European populations, and p.R778L, which is more prevalent in East Asian populations. The edited iPSC lines were differentiated into hepatocytes using pixlbio's pixDiff process, and their copper-handling defects were characterised using high-content imaging.

Following copper challenge, the disease models displayed increased cellular toxicity and oxidative stress. These phenotypes could be rescued in a dose-dependent manner using a copper chelator, providing a clear pharmacological readout in models generated from a single genetic background without requiring recruitment of multiple patient cohorts. Independent published work on the same H1069Q mutation has reached a similar conclusion using CRISPR-corrected patient iPSCs: correcting a single ATP7B allele restores copper resistance, again isolating the mutation as the causal variable.  

For MASLD, the PNPLA3 I148M variant is among the strongest and most extensively studied genetic risk factors for hepatic fat accumulation and disease progression. A published isogenic hiPSC model used CRISPR/Cas9 to generate both a PNPLA3 knockout line and a line carrying the I148M variant on the same genetic background, differentiated both to hepatocytes, and characterised the resulting steatotic phenotype under free fatty acid challenge. The isogenic design is what let the authors attribute the phenotype specifically to the variant, not to background genetic risk the donor happened to carry.

Where Does pixCRISPR Fit In?

pixCRISPR is the gene-editing and differentiation service supporting pixlbio’s disease-modelling programmes, including models of MASLD and a growing range of inherited and rare diseases. The platform enables disease models to be developed as isogenic pairs or panels on a defined genetic background, providing a closely matched comparator for studying the effect of a specific variant.  

Using a licensed CRISPR-Cas9 workflow, pixCRISPR can introduce, disrupt, or correct selected variants in an existing, well-characterised iPSC line. The resulting clones are expanded, characterised, and taken through the same differentiation and quality-control pipeline across pixlbio's wild-type cell portfolio. This integrated process converts a defined genetic edit into a validated, disease-relevant cell model that can be supplied for mechanistic studies and drug-discovery applications.  

This consistency is fundamental to the value of a disease model. Researchers need confidence that an observed result reflects the biology of the variant under investigation, rather than an artefact introduced during gene editing, clonal selection, or differentiation. Therefore, every edited line undergoes karyotype and identity testing, confirmation of the intended edit by sequencing, and functional characterisation following differentiation. The resulting cells (hepatocytes, hepatic stellate cells, intestinal organoids, or macrophages) are then assesed using the same identity, quality, and functional markers across pixlbio’s wider portfolio.This consistent approach to model generation and characterisation also provides a strong foundation for deeper phenotypic analysis using pixCellPaint. The platform captures disease-associated changes across a multi-dimensional morphological profile, extending analysis beyond individual biomarkers or endpoints. This is exactly the setup a drug screen needs: a way to identify compounds that move a treated disease-line cell's phenotype back toward its healthy isogenic counterpart, giving a mechanistic, quantifiable readout of reversal rather than a binary disease/no-disease call.  

What Doesn't Isogenic Modelling Solve?

Isogenic modelling reduces variability from genetic background, but it does not eliminate every source of uncertainty.  CRISPR editing and clonal selection can introduce unintended changes, making sequencing-based off-target assessment, karyotyping and identity confirmation important components of model validation.  

In addition, and despite the power isogenic modelling provides in isolating the effect of a genetic variant, a single isogenic pair still represents only one donor background. The wider genome may contain modifier variants that strengthen, reduce, or otherwise alter the observed phenotype. Therefore, a result demonstrated in one isogenic pair establishes causality within that background, but not necessarily the causality across a wider patient population.

Addressing this limitation requires a complementary population-based approach. pixlbio can combine isogenic comparisons with panels of well-characterised wild-type iPSC lines from multiple donors, differentiated and assessed using the same processes. Where appropriate, the same disease-associated edit can also be introduced into more than one donor background. This combines the mechanistic clarity of isogenic modelling with the population diversity needed to assess phenotype penetrance, biological variability, and consistency of therapeutic response.

Finally, an isogenic edit does not automatically reproduce the complete disease environment. Some phenotypes depend on metabolic stress, drug exposure, interactions between different liver cell types or other external factors. pixlbio can incorporate relevant disease challenges, functional assays and complementary liver cell models into study design, helping translate a defined genetic change into a more biologically representative experimental system.

‍