RNA localization in confined cells depends on cellular mechanical activity and contributes to confined migration - PubMed (original) (raw)
RNA localization in confined cells depends on cellular mechanical activity and contributes to confined migration
Rebecca A Moriarty et al. iScience. 2022.
Abstract
Cancer cells experience mechanical confining forces during metastasis and, consequently, can alter their migratory mechanisms. Localization of numerous mRNAs to cell protrusions contributes to cell polarization and migration and is controlled by proteins that can bind RNA and/or cytoskeletal elements, such as the adenomatous polyposis coli (APC). Here, we demonstrate that peripheral localization of APC-dependent RNAs in cells within confined microchannels is cell type dependent. This varying phenotype is determined by the levels of a detyrosinated tubulin network. We show that this network is regulated by mechanoactivity and that cells with mechanosensitive ion channels and increased myosin II activity direct peripheral localization of the RAB13 APC-dependent RNA. Through specific mislocalization of the RAB13 RNA, we show that peripheral RNA localization contributes to confined cell migration. Our results indicate that a cell's mechanical activity determines its ability to peripherally target RNAs and utilize them for movement in confinement.
Keywords: Bioengineering; Biological sciences; Biophysics; Cell biology.
© 2022.
Conflict of interest statement
The authors declare no competing interests.
Figures
Graphical abstract
Figure 1
RNA localization patterns are altered in a cell type-dependent manner in confinement (A and B) Representative maximum intensity z-projected FISH images of RAB13, KIF1C, _RPL27_α, RPS20, and polyA RNAs in (A) MDA-MB-231 and (B) A375 cells in either 50-μm wide or 3-μm narrow microchannels. RNA signal is shown in white, nucleus in blue, and cell outline in red. Brackets above cells in 3 μm groups denote the microchannel width. The leading edge of the cells (determined by the side of chemoattractant addition) is toward the top of the page. (C) Schematic showing PDI index values of the indicated hypothetical RNA distributions. A more peripherally enriched RNA exhibits a PDI > 1, a diffuse RNA exhibits a PDI = 1, and a perinuclear RNA a PDI < 1. (D and E) PDI calculations of _RAB13_, _KIF1C_, _RPL27_α, _RPS20_, and _polyA_ RNAs for (D) MDA-MB-231 cells or (E) A375 cells. Dots on graphs in (D and E) represent individual cells pooled from at least three independent experiments (n > 25). Note that the PDI of polyA RNA, which is used as a proxy for the cytosolic volume, is slightly higher in cells in the 3-μm narrow microchannels. This is likely because in confined cells the cytosolic volume is more equally distributed throughout the cell body, whereas in unconfined cells most of the cytosol is found perinuclearly, leading to higher and lower PDIs, respectively. This difference is not sufficient to account for the more substantial PDI increase observed for r-protein mRNAs in the 3-μm narrow microchannels. (F) Schematic showing segmentation of the cell in the direction of migration, from the lagging edge through the nucleus to the leading edge. The intensity of the RNA signal was measured within ten evenly spaced sections starting and ending at the cell boundaries. (G and H) Quantification of the intensity of APC-dependent RNAs (RAB13, KIF1C) compared with polyA RNA in (G) MDA-MB-231 and (H) A375 cells along the length of cells in 3-μm narrow microchannels (n > 25). Bars represent mean ± SE. Scale bars: 10 μm. p Values represent ∗<0.05, ∗∗∗<0.001, ∗∗∗∗<0.0001; two-way ANOVA with Šídák’s test.
Figure 2
MDA-MB-231 cells have less detyrosinated (Glu) tubulin than A375 cells (A) Schematic representing the tubulin tyrosination and detyrosination cycle and the enzymes involved. The C-terminal tyrosine of α-tubulin can be removed by a tubulin carboxypeptidase (TCP), an enzyme that can be inhibited by parthenolide (PTL). The tyrosine can be added back through the action of tubulin tyrosine ligase (TTL), whose expression can be knocked down through targeted siRNA (si-TTL). (B and C) Representative immunofluorescence images of Glu-tubulin and tubulin in MDA-MB-231 or A375 cells in (B) 50-μm wide microchannels or (C) 3-μm narrow microchannels. Images are snapshots of 3D view renderings. White dashed lines indicate positions of nuclei. Scale bars: 5 μm. (D) Quantification of Glu-tubulin signal (diffuse background signal was thresholded out and normalized to total cell volume) in the indicated cells. Dots represent individual cells pooled from three independent experiments (n = 20–26); bars represent mean ± SE. p Values: ∗<0.05, ∗∗∗∗<0.0001; two-way ANOVA with Šídák's test.
Figure 3
Glu tubulin levels drive RAB13 RNA localization in cells in confinement (A and B) Representative maximum intensity z-projected RAB13 FISH images of (A) MDA-MB-231 cells and (B) A375 cells in 50-μm wide and 3-μm narrow microchannels after 3-h treatment with DMSO (vehicle control) or Parthenolide (PTL; 10 μM in (A) and 50 μM in (B)). RNA signal is shown in white, nucleus in blue, and cell outline in red. Brackets above cells in 3 μm groups denote the microchannel width. Scale bars: 10 μm for wide channel images and 5 μm for narrow channel images. (C and D) PDI calculations of RAB13 RNA distribution of (C) MDA-MB-231 cells and (D) A375 cells. (E) Representative RAB13 FISH images of MDA-MB-231 cells treated with control siRNA or TTL siRNA. (F) PDI calculations of RAB13 RNA distributions of MDA-MB-231 cells treated with control or TTL siRNA. Individual dots on graphs in (C, D, and F) represent individual cells pooled from three independent experiments (n = 14–32); bars display mean ± SE. p Values, ∗∗<0.01, ∗∗∗<0.001 by two-way ANOVA with Šídák’s test (C and D) or by unpaired t test assuming equal standard deviations (F).
Figure 4
Loss of myosin activity is not sufficient to disrupt peripheral RAB13 RNA localization in cells in confinement (A) Representative maximum intensity z-projected images of phospho-myosin light chain (Ser19) (pMLC) staining in MDA-MB-231 and A375 cells in 50-μm wide and 3-μm narrow microchannels. Brackets above cells in 3 μm groups denote the microchannel width. (B) Quantification of pMLC signal (after thresholding out diffuse background signal) of MDA-MB-231 and A375 cells in 50-μm wide and 3-μm narrow microchannels (n = 22–37). (C) Representative maximum intensity z-projected RAB13 FISH images of A375 cells treated with DMSO (vehicle control) or 50 μM ML-7 in 50-μm wide and 3-μm narrow microchannels. RNA signal is shown in white, nucleus in blue, and cell outline in red. (D) Quantification of RAB13 PDI of DMSO control or 50 μM ML-7 treated cells. Individual dots on graphs represent individual cells pooled from three independent experiments (n = 16–43); bars display mean ± SE. p Values, ∗<0.05, ∗∗<0.01, ∗∗∗<0.001 by two-way ANOVA with Šídák’s test. Scale bar: 10 μm.
Figure 5
The Piezo1 channel and myosin activity function redundantly to regulate peripheral RAB13 RNA localization in cells in confinement (A) Representative maximum intensity z-projected FISH images of RAB13 RNA (in white dots, nucleus in blue, cell outline in red) in A375 cells in 50-μm wide microchannels treated with DMSO (vehicle control) or 50 μM ML-7 along with the indicated siRNAs. (B) Quantification of RAB13 PDI of A375 cells in 50-μm wide microchannels treated with DMSO or ML-7 and Piezo1 siRNAs (n = 38–47). (C) Representative maximum intensity z-projected FISH images of RAB13 RNA (in white dots, nucleus in blue, cell outline in red) in A375 cells in 3-μm narrow microchannels. Cells were treated with DMSO (vehicle control) or 50 μM ML-7, along with the indicated siRNAs. (D) Quantification of RAB13 PDI of cells in 3-μm narrow microchannels (n = 28–55). Individual dots on graphs represent individual cells pooled from three independent experiments; bars display mean ± SE. Scale bars: 10 μm (A), 5 μm (C). p Values, ∗<0.05, ∗∗<0.01, ∗∗∗<0.001, ∗∗∗∗<0.0001 by two-way ANOVA with Šídák’s test.
Figure 6
The Piezo1 channel and myosin activity function redundantly to regulate the detyrosinated tubulin network in cells in confinement (A) Representative western blot showing effect of Piezo1 knockdown and ML-7 co-treatment (DMSO is used as the vehicle control) on Glu-tubulin protein levels for cells on a 2D surface. (B) Quantification of corresponding Glu-tubulin to tubulin protein levels from (A); individual dots represent one independent experiment. (C) Representative immunofluorescence images of Glu-tubulin in A375 cells in 3-μm wide microchannels after treatment with Piezo1 siRNA and/or ML-7. Images are snapshots of 3D view renderings. Dashed white circle represents position of nucleus. (D) Quantification of mean intensity of Glu-tubulin signal per cell (n = 35–48). Dots represent individual cells pooled from three independent experiments. Bars reflect mean ± SE; scale bars: 5 μm. p values, ∗<0.05, by two-way ANOVA with Šídák’s test.
Figure 7
Peripheral RNA localization functionally contributes to A375 cell migration in wide and narrow channels (A and B) Cell speed (A) and persistence (B) of MDA-MB-231 cells in 50-μm wide and 3-μm narrow microchannels were quantified after treatment with control morpholino oligos or oligos directed against localization sequences in the 3′ UTR of RAB13 RNA. (C and D) Cell speed (C) and persistence (D) of A375 cells in 50-μm wide and 3-μm narrow microchannels were quantified after treatment with control morpholino oligos or oligos directed against the 3′ UTR of RAB13 RNA. Dots represent individual cells pooled from three independent experiments (n > 60). Bars represent mean ± SE. p values, ∗<0.05, ∗∗∗∗<0.0001 by two-way ANOVA with Šídák’s test.
Figure 8
Proposed model Flow chart depicting how the mechanical properties of cells determine RNA localization in confined cells. Peripheral RNA targeting can promote migration through narrow spaces. We note that our data suggest that formation of a robust detyrosinated tubulin network is necessary to promote peripheral RAB13 RNA localization. It is possible that mechanical activity might act through additional ways to affect cytoplasmic RNA distributions.
References
- Alexander S., Koehl G.E., Hirschberg M., Geissler E.K., Friedl P. Dynamic imaging of cancer growth and invasion: a modified skin-fold chamber model. Histochem. Cell Biol. 2008;130:1147–1154. - PubMed
- Alexander S., Weigelin B., Winkler F., Friedl P. Preclinical intravital microscopy of the tumour-stroma interface: invasion, metastasis, and therapy response. Curr. Opin. Cell Biol. 2013;25:659–671. - PubMed
- Andreassi C., Riccio A. To localize or not to localize: mRNA fate is in 3′UTR ends. Trends Cell Biol. 2009;19:465–474. - PubMed
LinkOut - more resources
Full Text Sources