Cell clusters adopt a collective amoeboid mode of migration in confined non-adhesive environments.

Diane-Laure Pagès, Emmanuel Dornier, Jean De Seze, Li Wang, Rui Luan, Jérôme Cartry, Charlotte Canet-Jourdan, Joel Raingeaud, Raphael Voituriez, Mathieu Coppey, Matthieu Piel, Fanny Jaulin

Preprint posted on May 28, 2020

Squeezing from the back till the finishing line: cancer cell clusters migrate as a ‘collective amoeboid’ to invade their organ-specific targets.

Selected by Giuliana Clemente

Context and Background:

Decades ago, colorectal cancer (CRC) was a rather rare form of cancer but today represents the third most common cancer worldwide and sadly accounts for the death of 56% of patients diagnosed with it. To further complicate this clinical picture, 20% of patients with CRC quickly develop metastasis, long before the primary tumor can be diagnosed, and this accounts for the elevated tumor burden and poor prognosis. In an attempt to understand the mechanisms of CRC dissemination, Zajac et al. monitored ex-vivo a collection of CRC specimens from more than 50 patients. This study identified the frequent formation of large clusters of cells that disseminate efficiently as a collective. The clusters are rounded in appearance, display a marked epithelial organization and, even more surprisingly, are characterized by an inverted apico-basolateral polarity [1]. Because of these characteristics, the authors defined these clusters as tumor spheres with inverted polarity (TSIPs).

It is commonly accepted that cell clusters navigate the extracellular space by engaging with the surrounding environment via focal adhesion and generating traction forces that push the collective forward. The identification of cell clusters with inverted polarity therefore represents a huge paradox and poses the question of how such groups of cells can efficiently disseminate despite traction-based mode of migration being precluded to them.

In this preprint, using a combination of microchambers and channels, Pages et al. investigate whether amoeboid-like locomotion allows the spreading of metastatic cell spheres across the body.

Main findings:

Either TSIP clusters or clusters formed from HT29 or HT29-MTX cell lines were tested for their ability to migrate in PEG-coated, non-adhesive microchannels. The system allows to individually manipulate intrinsic and extrinsic (ECM composition) components of the cell adhesion machinery and test for cell autonomous and non-cell autonomous requirements for amoeboid-like locomotion of cell clusters. TSIP clusters plated in non-adhesive microchannels adopted a rounded morphology and migrated with high speed and directionality, favoring an adhesion-independent mode of migration. This is in line with the knowledge that TSIP clusters are characterized by inverted polarity and therefore by definition unable to adhere to the substrate. Strikingly, HT29 or HT29-MTX clusters displayed the same migratory behavior. These clusters were potentially capable of focal adhesion-dependent migration. Indeed, if plated in collagen-coated microchannels, they attempted to engage with the substrate as shown by the use of turquoise-paxillin and measured by quantification of maximum contact angle. However, in PEG-coated microchannels, they bypassed the requirement to adhere to the substrate and switched to an adhesion-free mode of migration.

As it happens for single cells displaying amoeboid locomotion, this adhesion-independent mode of migration was based on acto-myosin contractility at the rear of the cluster as suggested by the strong accumulation of cortical actin and myosin II at this location. The requirement for acto-myosin contractility was confirmed by treatment with the myosin inhibitor blebbistatin, which significantly inhibited cell movement, as well as by optogenetic-dependent activation of RhoA at the front of a migrating cluster, which was sufficient to stop and even revert direction of migration.


This work identifies a novel mode of migration named ‘collective amoeboid’, whereby a cluster of cells migrates as a coherent unit without adhering to the extracellular space and solely by using cortical acto-myosin contraction to move forward. The identification of this novel mode of migration solves the paradox of how TSIPs can still efficiently invade tissues and migrate despite their inverted apico-basolateral polarity preventing them from forming adhesions.

Questions to the authors:

  1. Which is the level of plasticity? Do you think that cluster of cells can adopt adhesive vs ameboid migration depending on characteristics of the environment? Have you tried to place HT29 or HT29-MTXtalin-RNAi clusters in collagen-coated channels and test whether they would still prefer to adopt ameboid-like mode of migration?
  2. How is acto-myosin contractility regulated within the cluster? What signal “strips away” myosin from the cortex of the cells at the front?
  3. What levels of therapeutic intervention can you envision for the treatment of TSIPs derived metastasis? Do you reckon one could force TSIPs to revert back their polarity?


  1. Zajac, O., et al., Tumour spheres with inverted polarity drive the formation of peritoneal metastases in patients with hypermethylated colorectal carcinomas. Nat Cell Biol, 2018. 20(3): p. 296-306.



Posted on: 8th June 2020 , updated on: 10th June 2020


Read preprint (No Ratings Yet)

Have your say

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Sign up to customise the site to your preferences and to receive alerts

Register here