Thread by @LunedBadder

Thrilled that our work on multiparametric imaging of #organoids is now out (and was perfectly timed for my birthday!) @TrevorCDale @DrPaulShaw @AJ_Hollins @DrMairianThomas journals.plos.org/plosone/articl…

Patient-derived tumour organoids have been shown to closely model patient disease and predict drug responses, and are gaining increasing attention as 3D in vitro models.

They provide a way to study complex cell-cell interactions, which have been somewhat limited in 2D cell cultures. One potential application is to study the effects of cancer stem cell targeting agents (e.g. Wnt signaling inhibitors) on tumour cell populations.

Complex 3D model systems also present their own challenges - one of which is to be comprehensively characterised against compounds that have differential effects on tumour cell subpopulations, which may be beyond the scope of classical viability readouts.

We generated a set of tumour organoids from colorectal cancer patients and tested their responses to novel inhibitors of Tankyrase (TNKSi) which are known to modulate Wnt signalling.

Using single-metric viability assays we were able to see partial responses to TNKSi in organoids. However, we found that our ability to measure responses to treatment were greatly enhanced by 3D image-based analysis.

We looked at ~600 different morphological readouts from individual organoids following treatment with TNKSi and found multiple features that were altered as a result of drug-specific changes, such as changes in organoid shape and branching, as well as nuclei shape and size.

By analysing multiple morphological measurements, we were able to detect clear responses to TNKSi over a range of potencies. Whilst these responses followed a similar trend to viability assays, subtle alterations of tumour cells were detected far more effectively.

We then analysed downstream Wnt signaling pathway targets following TNKSi treatment. Overall, TNKSi sensitivity was related to drug- induced suppression of stem cell markers. However, the suppression of stem cell markers did not always lead to a functional response.

Our findings highlight that complex 3D models require comprehensive analyses, particularly for testing drugs that act on a crucial subset of cell populations only. Combining biochemical and image-based analyses could effectively exploit organoids for predictive drug testing.

Whilst our work has focused on the power of image based analyses for cancer stem cell therapies, we believe this could be applicable to other compounds to investigate gene-drug interactions, and could be used to rationalise drug combinations that act on different cell types.

This has been a huge collaborative effort - with surgical, histopathology teams at UHW @PM40Doc, aswell as team members at @walescancerbank @WalesGenePark @OcellO_3D @Cellesce, to name a few.

Particularly grateful also to our funders at @cancer_wales ECMC and CRUK and to any other contributors that I've not listed here!