Page 318 - Read Online

Mokhamatam et al. J Cancer Metastasis Treat 2020;6:28 I http://dx.doi.org/10.20517/2394-4722.2020.38 Page 5 of 9

As the attachment of cells to the surface is too strong in 2D cultures using polystyrene plates, efforts were
made to minimize the cell to surface attachment and to increase cell to cell contacts. One such effort was
done by the Fujibayashi group, who designed nano-culture plates (NCP) using inorganic nanomaterial
scaffolds and nanoimprinting technology. Increased cell migration and spheroid formation at different
[61]
[62]
locations on the plates was demonstrated, unlike in 2D plates . Arai et al. used this NCP-based gel and
soft-agar free 3D-HTS system for screening 1,330 compounds for spheroid EMT inhibitory (SEMTIN)
activity. They found 9 compounds with significant activity. SB-525334, a TGF-βR1 inhibitor, and SU9516,
a CDK2 inhibitor, were shown to have SEMTIN activity. This is also an example of the successful culture
[63]
of A549 lung cancer cells on the NCP platform and its EMT features were established clearly. Aref et al.
developed a 3D microfluidic assay that provides a 3D microenvironment for cells to mimic EMT using
HUVEC cells, and the assay is quantifiable as well. They co-cultured A549 lung cancer cells with HUVEC
cells in this system and compared several metastatic inhibiting drugs between 2D and 3D models. In their
system, A83-01, which is a TGF-βR inhibitor, significantly lost potency in the 3D system (5 nmol/L vs.
[63]
2.5 µmol/L) . 3D models of HNSCC expressed EMT markers better than 2D models. They also showed
a decreased sensitivity to cisplatin and cetuximab, unlike in 2D models. This suggests that 3D models can
provide better simulations of drug activity . These reports suggest that 3D models are far superior to 2D
[64]
models in mimicking EMT phenomena and predicting the potency of the drugs.

NEXT-GENERATION DISCOVERY ATTEMPTS FOR CONTROLLING EMT
Though 3D assays are better than the 2D assays for efficient drug screening, they are limited by the ease
of the experiment analysis for doing primary and secondary assays. Most of the 3D assays utilize artificial
substrates or gel components, which in turn make those assays different from in vivo or clinical models.
Although they can form 3D architects, they might be different from the original ones and might have been
induced by the nature of the materials that were used. Tumorspheresare also no longer considered as exact
replicas of the original tumour. They are only rich in the CSC population and lack the heterogeneity of the
[65]
tumour in vivo . Maintenance of tumorspheres for long term with the same properties and composition is
[66]
also challenging .
Owing to all these problems with traditional 3D-HTS systems, organoid cultures are becoming more
popular, as they are small pieces of original patient-derived tumours that were grown in laboratory
conditions. They form a full tumour and express the markers and mimic the organ properties from which
[67]
they were derived . Organoids use basement membrane components like Matrigel, which arecloser to
the natural system compared to ultralow attachment surfaces for tumorspheres which cannot be found in
[68]
nature . Though the term organoid has been used in many different contexts, its actual popularity started
when intestinal organoids were developed by the Hans Clevers group, using Lgr5 stem cells . LGR5 was
[69]
+
found to activate EMT in glioma stem cells and is a better therapeutic target for EMT control. It functions
through the WNT/β-catenin pathway . Because of the importance of organoids in many fields, it was
[70]
[71]
[72]
considered as the “Method of 2017” . Fan et al. extensively reviewed the organoid models that were
developed for different types of cancers.
Patient-derived tumour organoids are cheaper, faster, and easier to handle compared to patient-derived
tumour xenografts, which were traditionally used. These can also be used for high throughput screening
of drugs, that can be administered immediately to the corresponding patients [72,73] (personalised cancer
medicine). Despite the current lack of studies reporting on EMT using organoids, many more are expected
soon. Hypoxic gradients that are much needed for EMT induction cannot be modelled by traditional 3D
[75]
[74]
assays but can be done by using organoid models . Hahn et al. developedan intestinal organoid-based
EMT model by inducing intestinal organoids with TGF-β1 and TNF-α. This suggests that it is possible
to get EMT models for different types of cancers, which can be further used for high throughput drug
screening. Drugs that can come through the organoid screen are more likely to work better in patients and