First published online 6 May 2008
doi: 10.1242/jcs.015842
Journal of Cell Science 121, 1803-1814 (2008)
Published by The Company of Biologists 2008
Myosin II regulates the shape of three-dimensional intestinal epithelial cysts
Andrei I. Ivanov1,*,
,
Ann M. Hopkins2,
G. Thomas Brown1,
Kirsten Gerner-Smidt1,
Brian A. Babbin1,
Charles A. Parkos1 and
Asma Nusrat1
1 Epithelial Pathobiology Research Unit, Department of Pathology and Laboratory Medicine, Emory University, Atlanta, GA 30322, USA
2 Department of Surgery, Royal College of Surgeons in Ireland, Dublin, Ireland

View larger version (48K):
[in this window]
[in a new window]
|
Fig. 1. Myosin II is enriched at the surface of 3D epithelial cysts. Caco-2 cells were embedded in Matrigel for 72 hours, fixed, and fluorescently double-labeled for F-actin (red) and either NMMIIA or NMMIIB (green). Cross-sections of Caco-2 cysts show that significant pools of NMMIIA and NMMIIB accumulate within F-actin bundles at the cyst surface (arrows). Scale bar: 20 µm.
|
|

View larger version (57K):
[in this window]
[in a new window]
|
Fig. 2. Inhibition of myosin II causes the formation of F-actin-rich protrusions at the surface of epithelial cysts. (A,B) Caco-2 (A) or SK-CO15 (B) cells were embedded for 72 hours into 3D Matrigel in the presence of either vehicle or the myosin II inhibitors blebbistatin (100 µM) or Y27632 (20 µM) and analyzed for cyst morphology by fluorescent labeling for F-actin. Vehicle-treated Caco-2 and SK-CO15 cells formed spherical cysts with smooth surfaces, whereas cysts grown in the presence of myosin II inhibitors developed radiating F-actin protrusions (arrows). (C,D) SK-CO15 cells were transfected with either control or NMMIIA-specific siRNA and, 72 hours later, were analyzed for NMMIIA expression (C) and morphology of 3D cysts (D). NMMIIA knockdown caused a significant decrease in the level of NMMIIA and induced the formation of numerous F-actin-rich surface spikes. *P<0.05.
|
|

View larger version (52K):
[in this window]
[in a new window]
|
Fig. 3. Formation of peripheral protrusions is mediated by actin polymerization and is not caused by stabilization of cortical F-actin. (A) Preformed Caco-2 cysts were treated for 12 hours with either blebbistatin alone (100 µM) or with a combination of blebbistatin and latrunculin B (1 µM) to inhibit actin polymerization. Inhibition of actin polymerization prevented the formation of peripheral protrusions in blebbistatin-treated Caco-2 cysts (*P<0.001). (B) Preformed Caco-2 cysts were treated for 12 hours with either vehicle, a cortical flow inhibitor (WGA; 500 µg/ml) or an F-actin-stabilizing drug (jasplakinolide; 0.5 µM). The formation of protrusions on the cyst surface was not induced by any treatment.
|
|

View larger version (49K):
[in this window]
[in a new window]
|
Fig. 4. Activity of Rac1 and Cdc42 is not required for the formation of blebbistatin-induced protrusions. (A) Caco-2 cysts were treated for 12 hours with either blebbistatin alone or in a combination with the Rac inhibitor NSC 23766 (100 µM) or the Cdc42 inhibitor secramine A (20 µM). Pharmacological inhibition of Rac and Cdc42 did not prevent the development of blebbistatin-induced spikes. (B) Caco-2 cells growing on Matrigel-coated coverslips were infected with adenoviruses bearing either control EGFP, or EGFP-tagged dominant-negative N17Rac1 or N17Cdc42 and treated for 12 hours with blebbistatin (100 µM). Cells infected either with control virus or with dominant-negative Rac1 or Cdc42 mutants developed peripheral protrusions upon inhibition of myosin II (arrows). (C) Caco-2 cells were treated for 12 hours with either blebbistatin (100 µM) or vehicle, and activation statuses of Rac1 and Cdc42 were determined in cell lysates using pull-down assays. No increase in the levels of activated Rac1 or Cdc42 were observed in blebbistatin-treated cells.
|
|

View larger version (24K):
[in this window]
[in a new window]
|
Fig. 5. PLC activity is required for the formation of peripheral protrusions in epithelial cysts. Caco-2 cysts were treated for 12 hours with either blebbistatin alone or in combination with the PLC inhibitors ET-18-OCH3 (10 µM) or U-73122 (1 µM). Both pharmacological inhibitors of PLC significantly attenuated the formation of blebbistatin-induced spikes (*P<0.001).
|
|

View larger version (50K):
[in this window]
[in a new window]
|
Fig. 8. Cofilin activity is required for the development of surface protrusions in blebbistatin-treated cysts. Caco-2 cells were embedded for 72 hours into 3D Matrigel in the presence of adenoviruses bearing either control EGFP (A) or an EGFP-tagged inactive S3E cofilin mutant (B), followed by 12 hours of treatment with blebbistatin. Inhibition of cofilin activity significantly attenuated the formation of blebbistatin-induced surface protrusions (D) (*P<0.01). (C) Caco-2 cells were grown in 3D Matrigel in the presence of adenoviruses bearing an EGFP-tagged constitutively active S3A cofilin mutant. Overexpression of active cofilin induced the formation of F-actin-rich processes on the cyst surface.
|
|

View larger version (42K):
[in this window]
[in a new window]
|
Fig. 9. Microtubules are involved in the formation of blebbistatin-induced protrusions. (A) Preformed Caco-2 cysts were treated for 12 hours with either blebbistatin alone or in combination with the microtubule-depolymerizing drugs nocodazole (30 µM) or vinblastin (10 µM). Microtubule depolymerization prevented the formation of blebbistatin-induced protrusions (*P<0.01). (B) Caco-2 cysts were treated for 12 hours with blebbistatin, fixed, and double-labeled for F-actin (green) and the microtubule marker -tubulin (red). Microtubules are enriched in the base of blebbistatin-induced spikes (arrows).
|
|

View larger version (24K):
[in this window]
[in a new window]
|
Fig. 11. Hypothetic mechanism for surface protrusion formation in myosin-II-inhibited intestinal epithelial cysts. The diagram depicts key signaling and morphological events involved in the development of peripheral spikes. For simplicity, these events are presented as a single cascade. Tentative and/or unproven connections between different steps in the cascade are marked by broken lines. See explanation in the Discussion.
|
|

CiteULike
Complore
Connotea
Del.icio.us
Digg
Reddit
Technorati
Twitter What's this?
© The Company of Biologists Ltd 2008