Latest Articles Include:
- Postdoctoral training: Time for change
- Nat Cell Biol 13(7):735 (2011)
Article preview View full access options Nature Cell Biology | Editorial Postdoctoral training: Time for change Journal name:Nature Cell BiologyVolume: 13,Page:735Year published:(2011)DOI:doi:10.1038/ncb0711-735aPublished online01 July 2011 The increasingly pressing problems facing postdoctoral fellows in recent years call for a re-evaluation of the position of postdocs in academia and collaboration of involved parties to bring about positive change. Article preview Read the full article * FREE access with registration Register now * Already have a Nature.com account? Log in Additional access options: * Login via Athens * Login via your Institution * Purchase a site license * Use a document delivery service * Rent this article from DeepDyve * British Library Document Supply Centre * Infotrieve * Thompson ISI Document Delivery * You can also request this document from your local library through inter-library loan services. Additional data - Nature Publishing Group supports the ASCB's childcare awards
- Nat Cell Biol 13(7):735 (2011)
Article preview View full access options Nature Cell Biology | Editorial Nature Publishing Group supports the ASCB's childcare awards Journal name:Nature Cell BiologyVolume: 13,Page:735Year published:(2011)DOI:doi:10.1038/ncb0711-735bPublished online01 July 2011 Nature Cell Biology editorials have often highlighted the challenges of balancing the demands of an academic career with family commitments, and emphasized the importance of providing adequate institutional support in the form of progressive policies and relevant infrastructure. We are therefore pleased to announce that Nature Publishing Group will be supporting the American Society for Cell Biology's childcare awards for a period of five years, from 2011–2015. These awards are intended to meet the childcare needs of early-stage researchers who would otherwise find it difficult to participate in the ASCB's annual meeting. Article preview Read the full article * FREE access with registration Register now * Already have a Nature.com account? Log in Additional access options: * Login via Athens * Login via your Institution * Purchase a site license * Use a document delivery service * Rent this article from DeepDyve * British Library Document Supply Centre * Infotrieve * Thompson ISI Document Delivery * You can also request this document from your local library through inter-library loan services. Additional data - Might makes right: Using force to align the mitotic spindle
- Nat Cell Biol 13(7):736-738 (2011)
Article preview View full access options Nature Cell Biology | News and Views Might makes right: Using force to align the mitotic spindle * Oscar M. Lancaster1 * Buzz Baum1 * Affiliations * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:736–738Year published:(2011)DOI:doi:10.1038/ncb0711-736Published online01 July 2011 Both symmetric and asymmetric divisions rely on alignment of the mitotic spindle with cues from the environment. A study now shows that mitotic spindles find their position by reading the map of forces that load-bearing retraction fibres exert on the cell body. Figures at a glance * Figure 1: How do retraction fibres align the mitotic spindle? () Mitotic spindle alignment in tissue culture cells. The distribution of retraction fibres formed during mitotic cell rounding reflects the pattern of adhesion points in interphase cells. Retraction fibres are necessary for spindle alignment parallel to the extracellular growing surface, and on anisotropic patterns their cortical distribution guides spindle alignment in the x–y plane. () Spindles continually monitor the distribution of retraction fibres at the mitotic cell cortex. Laser ablation of retraction fibres along the long pattern axis induces the mitotic spindle to rotate and align with the remaining retraction fibres. No such realignment was seen in the same experiment for cells growing on control bar patterns. () Mitotic spindles monitor the distribution of forces exerted by retraction fibres on the cell body. To alter the distribution of forces (red arrows), mitotic cells were grown on flexible substrates that were subjected to an anisotropic stretch (blue arr! ows). Stretching an ellipse into a circle caused realignment of the spindle in the direction of the stretch. No spindle realignment was seen in the unstretched control. * Figure 2: Using force to align the mitotic spindle. On entry into mitosis, cells round up leaving retraction fibres tethering them to substrate adhesion sites. During this process, contraction of the actomyosin meshwork generates tensile forces in retraction fibres. In metaphase, the distribution of these tensile forces induces the polarization of a single dynamic cytoplasmic pool of actin filaments. Interactions between these actin structures and astral microtubules may generate transient pulling forces on the mitotic spindle that contribute to spindle alignment. Article preview Read the full article * Instant access to this article: US$18 Buy now * Subscribe to Nature Cell Biology for full access: Subscribe * Personal subscribers: Log in Additional access options: * Login via Athens * Login via your Institution * Purchase a site license * Use a document delivery service * Rent this article from DeepDyve * British Library Document Supply Centre * Infotrieve * Thompson ISI Document Delivery * You can also request this document from your local library through inter-library loan services. Author information Article tools * Print * Email * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg Affiliations * Oscar M. Lancaster and Buzz Baum are in the MRC Laboratory of Molecular Cell Biology, UCL, Gower Street, London, WC1E 6BT Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Buzz Baum Author Details * Oscar M. Lancaster Search for this author in: * NPG journals * PubMed * Google Scholar * Buzz Baum Contact Buzz Baum Search for this author in: * NPG journals * PubMed * Google Scholar Additional data - Multi-talented MCAK: Microtubule depolymerizer with a strong grip
- Nat Cell Biol 13(7):738-740 (2011)
Article preview View full access options Nature Cell Biology | News and Views Multi-talented MCAK: Microtubule depolymerizer with a strong grip * Stefan Diez1Journal name:Nature Cell BiologyVolume: 13,Pages:738–740Year published:(2011)DOI:doi:10.1038/ncb0711-738Published online01 July 2011 Microtubule-depolymerizing motor proteins regulate microtubule dynamics during chromosome segregation, but whether they can independently grip the ends of shrinking kinetochore microtubules has remained unresolved. MCAK, a member of the kinesin-13 motor protein family, is now shown to grip microtubules on its own and harness the forces of microtubule disassembly. Article preview Read the full article * Instant access to this article: US$18 Buy now * Subscribe to Nature Cell Biology for full access: Subscribe * Personal subscribers: Log in Additional access options: * Login via Athens * Login via your Institution * Purchase a site license * Use a document delivery service * Rent this article from DeepDyve * British Library Document Supply Centre * Infotrieve * Thompson ISI Document Delivery * You can also request this document from your local library through inter-library loan services. Author information Article tools * Print * Email * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg Affiliations * Stefan Diez is at B CUBE - Center for Molecular Bioengineering, Technische Universität Dresden and Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany Competing financial interests The author declares no competing financial interests. Corresponding author Correspondence to: * Stefan Diez Author Details * Stefan Diez Contact Stefan Diez Search for this author in: * NPG journals * PubMed * Google Scholar Additional data - Research highlights
- Nat Cell Biol 13(7):741 (2011)
Article preview View full access options Nature Cell Biology | Research Highlights Research highlights Journal name:Nature Cell BiologyVolume: 13,Page:741Year published:(2011)DOI:doi:10.1038/ncb0711-741Published online01 July 2011 The pole plasm, a cytoplasmic region containing maternal mRNAs and proteins at the posterior of Drosophila oocytes, is essential for germline and abdominal development. Pole plasm assembly is initiated by the microtubule-dependent transport of oskar (osk) mRNA to the posterior, where the Osk protein stimulates endocytic and actin-remodelling events essential for germ plasm functionality. Nakamura and colleagues (Development138, 2523–2532; 2011) have found that Mon2, a protein associated with Golgi and endosomes, acts downstream of Osk to remodel cortical actin and anchor the pole plasm. mon2 was identified in a screen for genes required for the localization of the pole plasm component Vasa. The authors found that Osk-induced formation of actin protrusions was perturbed in mon2 mutants, as seen previously in endosomal GTPase rab5 mutants. A mon2 mutation rescued actin defects in rab5 mutants, and Mon2 was found to interact with the actin nucleators Spire and Cappuccino, bot! h linked to Vasa localization in the screen. Cappucino and Spire were also required for Osk-induced formation of actin protrusions. Consistent with a role for Mon2 in mediating actin-related events, mon2 mutant oocytes did not exhibit posterior accumulation of the small GTPase Rho. Understanding exactly how Mon2 couples endocytic activity with cortical actin remodelling will be next on the agenda. NLB Article preview Read the full article * Instant access to this article: US$32 Buy now * Subscribe to Nature Cell Biology for full access: Subscribe * Personal subscribers: Log in Additional access options: * Login via Athens * Login via your Institution * Purchase a site license * Use a document delivery service * Rent this article from DeepDyve * British Library Document Supply Centre * Infotrieve * Thompson ISI Document Delivery * You can also request this document from your local library through inter-library loan services. Additional data - Origin and role of distal visceral endoderm, a group of cells that determines anterior–posterior polarity of the mouse embryo
- Nat Cell Biol 13(7):743-752 (2011)
Nature Cell Biology | Article Origin and role of distal visceral endoderm, a group of cells that determines anterior–posterior polarity of the mouse embryo * Katsuyoshi Takaoka * Masamichi Yamamoto * Hiroshi Hamada * Affiliations * Contributions * Corresponding authorsJournal name:Nature Cell BiologyVolume: 13,Pages:743–752Year published:(2011)DOI:doi:10.1038/ncb2251Received03 September 2010Accepted05 April 2011Published online29 May 2011 Abstract * Abstract * Author information * Supplementary information Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg Anterior–posterior polarity of the mouse embryo has been thought to be established when distal visceral endoderm (DVE) at embryonic day (E) 5.5 migrates toward the future anterior side to form anterior visceral endoderm (AVE). Lefty1, a marker of DVE and AVE, is asymmetrically expressed in implanting mouse embryos. We now show that Lefty1 is expressed first in a subset of epiblast progenitor cells and then in a subset of primitive endoderm progenitors. Genetic fate mapping indicated that the latter cells are destined to become DVE. In contrast to the accepted notion, however, AVE is not derived from DVE but is newly formed after E5.5 from Lefty1− visceral endoderm cells that move to the distal tip. Concomitant with DVE migration, all visceral endoderm cells in the embryonic region undergo global movement. In embryos subjected to genetic ablation of Lefty1-expressing DVE cells, AVE was formed de novo but the visceral endoderm including the newly formed AVE failed to migra! te, indicating that DVE guides the migration of AVE by initiating the global movement of visceral endoderm cells. Future anterior–posterior polarity is thus already determined by Lefty1+ blastomeres in the implanting blastocyst. View full text Figures at a glance * Figure 1: Two types of Lefty1+ cell in mouse embryos between E3.5 and E4.5. () Staining of an E3.5 mouse embryo harbouring the Lefty1-0.7-lacZ transgene, which recapitulates endogenous Lefty1 expression at pregastrulation stages5, with the LacZ substrate X-gal. Lefty1 expression was apparent in two cells of the inner cell mass. () Immunofluorescence microscopy of mouse embryos harbouring the Lefty1-0.7mVenus transgene at the indicated stages, stained with antibodies to GATA6. Nuclei were also stained with 4′,6-diamidino-2-phenylindole (DAPI, blue), and transgene expression was detected as membrane-localized Venus (mVenus) fluorescence. At E3.5, Lefty1+ cells (green) were negative for GATA6 (magenta, 7/9 embryos). At E4.0, cells expressing Lefty1 were either positive or negative for GATA6. At E4.5, Lefty1+ cells were positive for GATA6. () An E3.7 embryo harbouring the Lefty1(mVenus) transgene was cultured and monitored by time-lapse microscopy for 10 h. mVenus fluorescence (green; left) and phase-contrast microscopy images (middle) obtained at t! he indicated times are shown. The right panel of each set shows the region corresponding to the inner cell mass in the middle panel at higher magnification. The embryo was also subjected to immunofluorescence staining with antibodies to GATA6 (Supplementary Fig. S2a–c, magenta). Retrospective analysis of the time-lapse images enabled evaluation of Gata6 expression in most mVenus+ cells. Magenta arrowhead, GATA6+mVenus+ cell; yellow, white, and green arrowheads, GATA6−mVenus+ cells. Scale bars, 50 μm. * Figure 2: Fates of Lefty1-expressing cells in the blastocyst. (–) Lefty1(CreERT2) transgenic mice were crossed with Rosa26R mice, and embryos harbouring both the Lefty1(CreERT2) BAC and the Rosa26R allele were recovered at E4.5 (), E5.2 () or E6.2 (), treated with 4-OHTx for 7 h () or 6 h (,), and subjected to X-gal staining. A few cells positive for X-gal staining were apparent on the upper side of the primitive endoderm in . Cells positive for X-gal staining were specifically found in DVE in and in AVE in . () A transgenic embryo generated as in – was isolated at E3.5, treated with 4-OHTx in vitro for 3 h, transferred to the uterus of a pseudopregnant female and allowed to develop until E6.5 before staining with X-gal. Cells positive for X-gal staining were localized exclusively in the epiblast. (,) Transgenic embryos generated as in – were isolated between E3.75 and E4.0, cultured in hanging drops for 12–14 h (embryos develop to the stage equivalent to E4.5), exposed to 4-OHTx in vitro for 3 h and allowed to develo! p in utero until E5.5 () or E6.5 (). At E5.5, cells positive for X-gal staining were found exclusively in DVE () or in DVE plus visceral endoderm in the proximal region (arrowheads; , lower). At E6.5, cells positive for X-gal contributed to the visceral endoderm extending laterally from the embryonic–extraembryonic junction on the anterior side (); in some instances, extra X-gal+ cells are found in visceral endoderm in the extraembryonic region (arrowhead; , lower). (,) Mice pregnant with embryos generated as in – were fed with tamoxifen at E4.5, and embryos were isolated for X-gal staining at E5.5 (, upper, 3/6 embryos; lower, 3/6 embryos) or E6.5 (, upper, 6/11 embryos; lower, 5/11 embryos). The staining patterns were similar to those for the corresponding embryos shown in and (lower). () X-gal staining showing activation of the Rosa26R reporter allele by a Lefty1(Cre) BAC transgene at E6.5. Staining was apparent in a subset of epiblast cells (arrowheads) as well as i! n visceral endoderm including AVE. * Figure 3: Relation between DVE and descendants of Lefty1+ cells at E4.5. () Transgenic mice harbouring the Lefty1(CreERT2) and Lefty1(mVenus) BACs were crossed with Rosa26R transgenic mice, and pregnant females were treated with tamoxifen (Tx) at E4.5. Transgenic embryos were recovered at E5.5 and examined for Lefty1 expression (mVenus fluorescence, green; middle image). The same embryos were then stained with X-gal (blue; left image). A phase-contrast image (right) is also shown. Note that all X-gal-positive cells located at the distal side of the embryo (black arrowheads) express Lefty1 (green arrowheads). () Transgenic mice harbouring the Lefty1(CreERT2) and Cer1(mVenus) BACs were crossed with Rosa26R transgenic mice, and pregnant females were treated with tamoxifen (Tx) at E4.5. Transgenic embryos were recovered at E5.5 and examined for Cer1 expression (mVenus fluorescence, green; middle image). The same embryos were then stained with X-gal (blue; left image). A phase-contrast image (right) is also shown. Note that all X-gal-positive cells lo! cated at the distal side of the embryo (black arrowheads) express Cer1 (green arrowheads). * Figure 4: DVE-derived cells do not overlap with AVE at E6.5. () Lefty1(CreERT2) BAC transgenic mice were crossed with Rosa26R transgenic mice, pregnant females were treated with tamoxifen 3 h before E5.5 and transgenic embryos were recovered at E6.5 and stained with X-gal. () Alternatively, transgenic embryos harbouring the Lefty1(CreERT2) BAC and Rosa26R allele were recovered at E5.0, exposed to 4-OHTx in vitro for 6 h, cultured in the absence of the drug for 24 h and then stained with X-gal. Lateral (left) and anterior (right) views are shown for each embryo. In both embryos, X-gal+ cells were localized to the visceral endoderm region on the anterior–lateral side (arrowheads), forming a horseshoe shape in the anterior view. () Transgenic embryos harbouring the Lefty1(CreERT2) BAC and Rosa26R allele were recovered at E5.7, exposed to 4-OHTx in vitro for 8 h and then stained with X-gal. (,) Transgenic mice harbouring the Cer1(CreERT2) BAC were crossed with Rosa26R mice, and transgenic embryos recovered at E5.2 or E6.2 were e! xposed to 4 hydroxytamoxifen in vitro for 6 h before staining with X-gal. DVE and AVE cells were specifically positive for X-gal staining at E5.5 and E6.5, respectively, showing that the Cer1(CreERT2) BAC transgene recapitulates the endogenous Cer1 expression pattern. () Alternatively, pregnant females were treated with tamoxifen 3 h before E5.5, and the embryos were recovered at E6.5 and stained with X-gal. X-gal+ cells were detected in the most proximal portion of AVE and neighbouring visceral endoderm in the lateral region (arrowheads), a pattern similar to that in and . () Lefty1(CreERT2) BAC transgenic mice were crossed with Rosa26R transgenic mice, and pregnant females were treated with tamoxifen at E5.5. Transgenic embryos were recovered at E6.5 and stained with X-gal. Lateral (upper) and anterior (lower) views are shown. X-gal+ cells form a horseshoe shape in the anterior view and extend above the extraembryonic–embryonic boundary (dotted line). () Detection o! f Lefty1-expressing AVE cells by immunostaining of the E6.5 em! bryo with antibodies to Lefty (magenta). Nuclei were also stained with DAPI (blue). Note that Lefty1+ AVE cells are located below the extraembryonic–embryonic junction (dotted line). Magenta arrowheads indicate the proximal boundary of Lefty1+ AVE cells. * Figure 5: Distinct origins of DVE and AVE. () E5.5 transgenic embryos harbouring both Lefty1(mVenus) and Gata6(mTomato) BACs were monitored by time-lapse microscopy for 20 h. The mVenus and mTomato signals are shown in green and magenta, respectively (upper panels). Scale bars, 50 μm. The behaviour of Lefty1+ DVE-derived cells (green), AVE-fated cells (outlined in red, light blue, dark blue or yellow) and AVE cells (solid red, light blue, dark blue or yellow) is also illustrated schematically (lower panels). Lefty1-expressing AVE cells shown in solid light blue (12.5, 16.5 and 20 h) are derived from cells shown outlined in light blue (0 and 4.5 h) that do not express Lefty1 at earlier times. Similarly, Lefty1-expressing AVE cells shown in solid dark blue (16.5 and 20 h) are derived from cells outlined in dark blue (0, 4.5 and 12.5 h) that do not express Lefty1. Precursor cells for AVE (outlined cells), which do not express Lefty1, generate not only Lefty1+ AVE cells (solid colours) but also Lefty1− non! -AVE cells (outline colours). Brackets indicate AVE. () Immunostaining of an E6.5 transgenic embryo harbouring the Lefty1(mVenus) BAC with antibodies to Lefty. Lefty protein (magenta) was localized to AVE cells, whereas mVenus (green) was detected in DVE descendants and AVE cells. Green arrowheads, DVE-derived cells that are already formed by E5.5; magenta arrowheads, AVE cells that are newly formed at the distal region after E5.5. () Immunostaining of an E6.5 transgenic embryo harbouring the Cer1(mVenus) BAC with antibodies to Cer1. Cer1 protein (light blue) was localized to AVE cells (light-blue arrowheads), whereas mVenus (green) was detected in DVE descendants (green arrowheads) and AVE cells (light-blue arrowheads). * Figure 6: Global movement of visceral endoderm cells. () E5.2 transgenic embryos harbouring both Lefty1(mVenus) and Gata6(mTomato) BACs were monitored by time-lapse microscopy for 22.5 h. The mVenus and mTomato signals are shown in green and magenta, respectively (upper panels). Scale bar, 100 μm. The behaviour of Lefty1+ DVE and DVE-derived cells (light green), embryonic visceral endoderm cells (red, light blue, dark green and yellow) and extraembryonic visceral endoderm cells (grey and dark blue) is illustrated schematically (lower panels). DVE and embryonic visceral endoderm cells do not move for the first 6 h but show dynamic movement after this time. Extraembryonic visceral endoderm cells, on the other hand, do not move substantially. The bracket indicates the AVE. () Summary of the behaviour of five randomly selected visceral endoderm cells (three in the embryonic region and two in the extraembryonic region). The broken and solid lines indicate the behaviour of individual cells for the first 6 h and the remainin! g 16.5 h, respectively. A and P, anterior and posterior, respectively. (–) E5.0 transgenic embryos harbouring the indicated transgenes were treated with 4-OHTx for 6 h in vitro and then allowed to develop in culture for a further 26 h. The mTomato signal (magenta) was monitored. The behaviour of mTomato+ cells is illustrated schematically (lower panels). Scale bars, 100 μm. In control embryos harbouring Lefty1(CreERT2) and Gata6(mTomato) but lacking the Rosa26(DT-A) allele (,), mTomato-positive cells undergo global migration after 6 h. In embryos harbouring Lefty1(CreERT2), Gata6(mTomato) and Rosa26(DT-A) (,), mTomato-positive cells failed to move even after 32 h. The behaviour of five randomly selected visceral endoderm cells from corresponding embryos is summarized in and . * Figure 7: Role of DVE in AVE formation. (–) E5.0 transgenic embryos harbouring the indicated transgenes were exposed (–) or not (–) to 4-OHTx for 6 h in vitro and then allowed to develop in culture for up to 28 h before observation by phase-contrast and fluorescence microscopy. In embryos harbouring the Lefty1(CreERT2) BAC, Lefty1(Cherry) BAC and Rosa26(DT-A) allele (–), Lefty1 expression (Cherry fluorescence) was detected in DVE before () and immediately after () the 4-OHTx treatment. At 12 h (), Cherry-positive DVE cells had been ablated. At 18 h (), Cherry-positive cells newly appeared at the distal end of the embryo. At 28 h (), the Lefty1-expressing cells still remained at the distal tip of the embryo. In control embryos harbouring Lefty1(Cherry) and Lefty1(CreERT2) but lacking Rosa26(DT-A) (–), Cherry-positive cells had migrated normally at 18 h () and 28 h () after the onset of treatment with 4-OHTx. The presence of 4-OHTx per se did not affect formation or migration of DVE and AVE, ! given that the control embryos lacking the Rosa26(DT-A) allele developed similarly in the absence (–) or presence (–) of 4-OHTx treatment. Scale bars, 100 μm. (–) E5.0 transgenic embryos harbouring the Lefty1(CreERT2) BAC and Rosa26(DT-A) allele (,) or those harbouring Rosa26(DT-A) alone (,) were treated with 4-OHTx for 6 h and then allowed to develop in culture for a further 24 h. As a control, E5.0 transgenic embryos harbouring Lefty1(CreERT2) and Rosa26(DT-A) were incubated in the absence of 4-OHTx (,). All embryos were then subjected to in situ hybridization with probes specific for Cer1 and T (Brachyury) mRNAs. AVE is indicated by the black brackets. Anterior–posterior polarity (indicated by A and P, respectively) is indicated. * Figure 8: Origin and role of DVE in anterior–posterior patterning. Cells of DVE lineage are shown in blue, whereas AVE is shown in purple. Cells with solid colours express Lefty1, whereas cells with outline colours do not. In the wild-type embryo (upper panels), global movement of visceral endoderm cells is triggered by DVE migration at E5.5. While DVE cells migrate, AVE cells are newly formed at the distal end, migrate toward the proximal side following the DVE and finally occupy the entire region of AVE. When DVE is ablated (lower panels), AVE cells are newly formed at the distal end. However, AVE cells fail to migrate, because global movement of visceral endoderm cells is impaired. See the text for details. Author information * Abstract * Author information * Supplementary information Affiliations Contributions Project planning was carried out mainly by K.T. and partly by M.Y. and H.H.; all experiments involving mouse embryos were carried out by K.T.; the manuscript was written by K.T. and H.H. Competing financial interests The authors declare no competing financial interests. Corresponding authors Correspondence to: * Katsuyoshi Takaoka or * Hiroshi Hamada Author Details * Katsuyoshi Takaoka Contact Katsuyoshi Takaoka Search for this author in: * NPG journals * PubMed * Google Scholar * Masamichi Yamamoto Search for this author in: * NPG journals * PubMed * Google Scholar * Hiroshi Hamada Contact Hiroshi Hamada Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Abstract * Author information * Supplementary information Movies * Supplementary Information (160K) Supplementary Movie 1 * Supplementary Information (160K) Supplementary Movie 2 * Supplementary Information (200K) Supplementary Movie 3 * Supplementary Information (700) Supplementary Movie 4 * Supplementary Information (500K) Supplementary Movie 5 * Supplementary Information (500K) Supplementary Movie 6 * Supplementary Information (700K) Supplementary Movie 7 * Supplementary Information (740K) Supplementary Movie 8 * Supplementary Information (3M) Supplementary Movie 9 * Supplementary Information (600K) Supplementary Movie 10 * Supplementary Information (700K) Supplementary Movie 11 * Supplementary Information (600K) Supplementary Movie 12 * Supplementary Information (700K) Supplementary Movie 13 PDF files * Supplementary Information (3M) Supplementary Information Additional data - Differential requirement for the dual functions of β-catenin in embryonic stem cell self-renewal and germ layer formation
- Nat Cell Biol 13(7):753-761 (2011)
Nature Cell Biology | Article Differential requirement for the dual functions of β-catenin in embryonic stem cell self-renewal and germ layer formation * Natalia Lyashenko1 * Markus Winter1 * Domenico Migliorini1, 3 * Travis Biechele2 * Randall T. Moon2 * Christine Hartmann1 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:753–761Year published:(2011)DOI:doi:10.1038/ncb2260Received03 June 2010Accepted13 April 2011Published online19 June 2011 Abstract * Abstract * Author information * Supplementary information Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg Canonical Wnt signalling has been implicated in mouse and human embryonic stem cell (ESC) maintenance; however, its requirement is controversial. β-catenin is the key component in this highly conserved Wnt pathway, acting as a transcriptional transactivator. However, β-catenin has additional roles at the plasma membrane regulating cell–cell adhesion, complicating the analyses of cells/tissues lacking β-catenin. We report here the generation of a Ctnnb1 (β-catenin)-deficient mouse ESC (mESC) line and show that self-renewal is maintained in the absence of β-catenin. Cell adhesion is partially rescued by plakoglobin upregulation, but fails to be maintained during differentiation. When differentiated as aggregates, wild-type mESCs form descendants of all three germ layers, whereas mesendodermal germ layer formation and neuronal differentiation are defective in Ctnnb1-deficient mESCs. A Tcf/Lef-signalling-defective β-catenin variant, which re-establishes cadherin-mediated! cell adhesion, rescues definitive endoderm and neuroepithelial formation, indicating that the β-catenin cell-adhesion function is more important than its signalling function for these processes. View full text Figures at a glance * Figure 1: Characterization of β-catfl/fl and β-catΔ/Δ mESCs. () Genotyping PCR for β-catfl/fl (5E, 6E, 7G), β-catfl/Δ (7A, 3A, 10C) and β-catΔ/Δ (11A, 1B, 3B) individual mESC subclones. () Population doublings of β-catfl/fl and β-catΔ/Δ mESC clones grown for five days. Cells were collected and counted each day (carried out in triplicate). () Morphological appearance of β-catfl/fl and β-catΔ/Δ mESCs. Scale bars, 100 μm. () Confocal microscopy images of immunofluorescence staining for β-catenin, plakoglobin and E-cadherin of β-catfl/fl and β-catΔ/Δ mESC colonies. Scale bars, 20 μm. () Western blots for β-catenin, E-cadherin, plakoglobin and tubulin in mESCs. () Relative plakoglobin expression levels in β-catfl/fl and β-catΔ/Δ mESCs (n=1). () Western blot for E-cadherin, α-catenin, β-catenin and plakoglobin showing input and E-cadherin immunoprecipitates (IP) in β-catfl/fl and β-catΔ/Δ mESCs. () Histogram showing representative examples of BAR (TOPFlash) and fuBAR (FOPFlash) assays using stable β-catf! l/fl and β-catΔ/Δ mESC lines. () Histogram showing a representative example of a TOPFlash assay of stable BAR β-catfl/fl and β-catΔ/Δ mESCs stimulated with rWnt3a, Wnt3aCM or BIO. Error bars ± s.d in , and are calculated on the basis of three measurements. () Histogram showing relative Axin2 expression level in β-catfl/fl and β-catΔ/Δ mESCs stimulated with Wnt3aCM and BIO. Data are the mean of two biological replicates. Uncropped images of blots are shown in Supplementary Fig. S8. * Figure 2: Self-renewal markers are not affected in β-catfl/fl and β-catΔ/Δ mESCs under different culture conditions. () Histogram showing relative expression levels of self-renewal marker genes Nanog, Oct4, Sox2 and Rex1 in β-catfl/fl and β-catΔ/Δ mESCs cultured in serum+LIF. () Western blot of β-catenin, Nanog, Oct3/4 and tubulin in β-catfl/fl and β-catΔ/Δ mESCs cultured in serum+LIF. () Western blots for Stat3 and p-(Tyr 705)-Stat3 levels in response to LIF stimulation. () Alkaline phosphatase staining on passage 25 β-catfl/fl and β-catΔ/Δ mESCs cultured in serum+LIF. () Morphology of β-catfl/fl and β-catΔ/Δ mESCs cultured in 2i+LIF without serum. () Histogram showing relative expression levels of self-renewal marker genes Nanog, Oct4, Sox2 and Rex1 in β-catfl/fl and β-catΔ/Δ mESCs cultured in 2i+LIF. Data in and are the mean of two biological replicates. Scale bars in and , 100 μm. Uncropped images of blots are shown in Supplementary Fig. S8. * Figure 3: Cell–cell adhesion defects in β-catfl/fl and β-catΔ/Δ mESCs in the presence and absence of Jup knockdown. () TEM micrographs of β-catfl/fl and β-catΔ/Δ mESC colonies at low magnification (top; scale bars, 10 μm) and high magnification of the outlined areas (bottom; scale bars, 2 μm), showing non-adherent spaces between adjacent β-catΔ/Δ mESCs (asterisks), compared with the tight cell–cell adhesion between β-catfl/fl mESCs. () Western blot for E-cadherin, plakoglobin and tubulin on mESC lysates 48 h after treatment with control or Jup RNAi. Quantified E-cadherin and plakoglobin protein levels normalized to tubulin are shown by the numbers below the corresponding lanes. () Phase contrast and confocal microscopy images of immunofluorescence staining (scale bars, 50 μm) for plakoglobin and PECAM1 on control-siRNA- and Jup-siRNA-treated mESCs 48 h after transfection. Cells stained with DAPI to visualize nuclei. () Confocal microscopy images (scale bars, 50 μm) of immunofluorescence staining for E-cadherin (bottom: merged images of E-cadherin and DAPI stainin! g) on control-siRNA- and Jup-siRNA-treated mESCs 48 h after transfection. Uncropped images of blots are shown in Supplementary Fig. S8. * Figure 4: Morphological appearance and plakoglobin levels of β-catfl/fl, β-catΔ/Δ, β-catrescWT and β-catrescΔC embryoid bodies during differentiation. () Phase-contrast images of embryoid bodies at days 3, 5 and 7 of differentiation. Scale bars, 200 μm. Right column, magnified view of single embryoid bodies of the different genotypes on day 7. () Western blot for β-catenin, plakoglobin and tubulin from mESC and embryoid body lysates analysed at indicated time points (d, day) during differentiation. Uncropped images of blots are shown in Supplementary Fig. S8. * Figure 5: Analyses of Tcf/Lef-mediated transcriptional activity. () Histogram showing representative examples of TOPFlash (BAR reporter) and FOPFlash (fuBAR reporter) activity in stable mESCs expressing rescue constructs from pGAGGS or the Rosa26 locus under unstimulated (PBS) and stimulated (3 μM CHIR99021) conditions. Error bars ± s.d. are calculated on the basis of three measurements. () Histogram showing relative expression of endogenous β-catenin/Tcf-regulated genes Axin2, brachyury (T) and Cdx1 on treatment with PBS or CHIR99021 (3 μM). Data are mean ± s.d. of three biological replicates. P values were calculated using a two-tailed Student t -test. * Figure 6: Neuroectodermal differentiation potential of β-catfl/fl, β-catΔ/Δ, β-catrescWT and β-catrescΔC embryoid bodies. () RT–PCR analyses of ectodermal marker genes Fgf5, Nes and Pax6 in mESCs and during the embryoid body differentiation time course of β-catfl/fl and β-catΔ/Δ mESCs. (,) Immunofluorescence staining for TuJ-positive neurons () in differentiated embryoid bodies at day 16, and in monolayer cultures of mESCs at day 14 (). Nuclei are stained for DAPI (blue). Scale bars, 50 μm. Uncropped images of blots are shown in Supplementary Fig. S8. * Figure 7: Mesendodermal differentiation potential of β-catfl/fl, β-catΔ/Δ, β-catrescWT and β-catrescΔC embryoid bodies. () Histograms showing relative expression levels of endodermal marker genes Gata6, Foxa2 and Mixl1 in mESCs and during embryoid body differentiation. Data are the mean of two biological replicates. () Immunofluorescence staining for β-catenin, Gata4/fibronectin (Fibro), E-cadherin and E-cadherin (E-cad)/Cxcr4 in differentiated embryoid bodies at day 8. All cells were also stained with DAPI. Scale bars, 100 μm. () Immunofluorescence staining for Foxa2 on plated embryoid bodies at day 16. Cells were counterstained with DAPI. Scale bars, 50 μm. Author information * Abstract * Author information * Supplementary information Affiliations * Research Institute of Molecular Pathology, Dr. Bohrgasse 7, 1030 Vienna, Austria * Natalia Lyashenko, * Markus Winter, * Domenico Migliorini & * Christine Hartmann * Institute for Stem Cell and Regenerative Medicine, Department of Pharmacology, Howard Hughes Medical Institute, University of Washington School of Medicine, Campus Box 358056, Seattle, Washington 98109, USA * Travis Biechele & * Randall T. Moon * Present address: Laboratory of Molecular Cancer Biology, University of Leuven, Herestraat 49, 3000 Leuven, Belgium * Domenico Migliorini Contributions N.L. carried out, analysed and interpreted experiments. M.W. and D.M. carried out pulldown studies. T.B. and R.T.M. generated and provided BAR/fuBAR lentiviruses, and commented on the paper. C.H. supervised the study and wrote the paper together with N.L. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Christine Hartmann Author Details * Natalia Lyashenko Search for this author in: * NPG journals * PubMed * Google Scholar * Markus Winter Search for this author in: * NPG journals * PubMed * Google Scholar * Domenico Migliorini Search for this author in: * NPG journals * PubMed * Google Scholar * Travis Biechele Search for this author in: * NPG journals * PubMed * Google Scholar * Randall T. Moon Search for this author in: * NPG journals * PubMed * Google Scholar * Christine Hartmann Contact Christine Hartmann Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Abstract * Author information * Supplementary information PDF files * Supplementary Information (900K) Supplementary Information Additional data - Opposing effects of Tcf3 and Tcf1 control Wnt stimulation of embryonic stem cell self-renewal
- Nat Cell Biol 13(7):762-770 (2011)
Nature Cell Biology | Article Opposing effects of Tcf3 and Tcf1 control Wnt stimulation of embryonic stem cell self-renewal * Fei Yi1 * Laura Pereira1 * Jackson A. Hoffman1 * Brian R. Shy1 * Courtney M. Yuen2 * David R. Liu2 * Bradley J. Merrill1 * Affiliations * Contributions * Corresponding authorsJournal name:Nature Cell BiologyVolume: 13,Pages:762–770Year published:(2011)DOI:doi:10.1038/ncb2283Received14 February 2011Accepted19 May 2011Published online19 June 2011 Abstract * Abstract * Accession codes * Author information * Supplementary information Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg The co-occupancy of Tcf3 with Oct4, Sox2 and Nanog on embryonic stem cell (ESC) chromatin indicated that Tcf3 has been suggested to play an integral role in a poorly understood mechanism underlying Wnt-dependent stimulation of mouse ESC self-renewal of mouse ESCs. Although the conventional view of Tcf proteins as the β-catenin-binding effectors of Wnt signalling suggested Tcf3–β-catenin activation of target genes would stimulate self-renewal, here we show that an antagonistic relationship between Wnt3a and Tcf3 on gene expression regulates ESC self-renewal. Genetic ablation of Tcf3 replaced the requirement for exogenous Wnt3a or GSK3 inhibition for ESC self-renewal, demonstrating that inhibition of Tcf3 repressor is the necessary downstream effect of Wnt signalling. Interestingly, both Tcf3–β-catenin and Tcf1–β-catenin interactions contributed to Wnt stimulation of self-renewal and gene expression, and the combination of Tcf3 and Tcf1 recruited Wnt-stabilized β-ca! tenin to Oct4 binding sites on ESC chromatin. This work elucidates the molecular link between the effects of Wnt and the regulation of the Oct4/Sox2/Nanog network. View full text Figures at a glance * Figure 1: Tcf3 regulates the Wnt stimulation and GSK3 inhibition of ESC self-renewal. (,) Top, images showing alkaline phosphatase staining of ESC colonies after four days in control or Wnt3a-conditioned media (top labels). Overexpression of full-length Tcf3 protein (Tcf3 OE) was induced at the onset of the experiment (bottom labels). Supplementary Fig. S2 shows western blot analysis of protein lysates taken from Tcf3-overexpressing cells. The percentage of AP+ ESC colonies (white number at the top of each image) represents the individual counts of biological duplicates. Bottom, flow cytometry analysis of cell cycle progression by propidium iodide staining (x axis) was carried out on separate cell cultures subjected to the same conditions for three days. The results in are from analyses carried out in media containing 1,000 U ml−1 of exogenous LIF, whereas those in are from analyses that did not contain exogenous LIF. () ESC colony formation from Tcf3+/+ or Tcf3+/+ cells (indicated on the left) subjected to serum-free culture conditions for four days an! d stained for alkaline phosphatase activity. Colonies were treated with N2B27 media without inhibitors (No Inh), N2B27 media+PD0325901 (ERK Inh), N2B27 media+CHIR99021 (GSK3 Inh) or N2B27+PD0325901 and CHIR99021 (2i). The percentage of AP+ ESC colonies (white number at the top of each image) represents the individual counts of biological duplicates. () Number of dye-excluding cells following each serial passage of Tcf3+/+ (green) and Tcf3−/− (red) ESCs in serum-free conditions as noted in . Each point represents an individual measurement of biological duplicates. * Figure 2: Wnt3a stimulates and Tcf3 inhibits expression of Oct4- and Nanog regulated genes. () PCA of gene expression in four conditions in ESCs: Tcf3+/+ (black); Tcf3−/− (green); Tcf3+/++Wnt3a (orange); Tcf3−/−+Wnt3a (blue). Each point represents one of three biological replicates for each condition. () Each point on a graph represents one of the 831 genes for which the expression level is significantly increased in any of the three comparisons (Tcf3−/− versus Tcf3+/+; Tcf3+/++Wnt3a versus Tcf3+/+; or Tcf3−/−+Wnt3a versus Tcf3−/−). Individual genes are listed in Supplementary Table S1. Group 1 (red) and group 2 (blue) genes are plotted according to the rank fold-effect of Wnt3a on Tcf3+/+ (x axis) and either effect of Wnt3a on Tcf3−/− (top) or effect of Tcf3 ablation (bottom). Pearson correlation coefficients (top of each graph) are shown for each comparison. Calculations for the formation of groups and the determination of correlation coefficients are presented in Supplementary Table S2. () Individual genes were plotted based on the rank f! old-effect of Oct4 shRNA treatment4 (x axis) and the effect of Tcf3 and/or Wnt3a (y axis). Group 1 and group 2 genes are the same as in . Pearson correlation coefficients (top of each graph) are shown for each comparison. Calculations for the formation of groups and the determination of correlation coefficients are presented in Supplementary Table S2. () Hierarchical clustering comparison of effects caused by ablation of Tcf3 and Wnt3a treatment. The heat map shows only genes for which the levels of expression were decreased by Oct4 and Nanog RNAi experiments4. Genes for which the levels of expression were increased in Tcf3−/− and further increased in Tcf3−/−+Wnt3a are highlighted with the green bar. Genes for which the levels of expression were increased by Wnt3a in a partially Tcf3-dependent manner are highlighted with the blue and orange bar. Expression values for the 311 genes are presented in Supplementary Table S3. * Figure 3: Nanog-independence of Wnt3a/Tcf3-mediated effects on gene expression. () Tcf3+/+ ESCs were transfected with Nanog siRNA or control (SCRM) siRNA and treated for 24 h with 50 ng ml−1 Wnt3a. Protein lysates were used for western blot analysis to detect Nanog and tubulin proteins. Identically treated cells were used for RNA isolation for the qPCR data shown in . () Tcf3−/− ESCs were stably transfected with either a Nanog shRNA or control pSuper vector. Protein lysates were used for western blot analysis to detect Nanog and tubulin proteins. Identically treated cells were used for RNA isolation for the qPCR data shown in . () Tcf3+/+ ESCs were stably transfected with a Nanog-overexpression (OE) plasmid. Protein lysates from cells grown in +LIF, self-renewal conditions were used for western blot analysis to detect Nanog and tubulin proteins. Tcf3+/+ and Tcf3−/− cells without the Nanog expression plasmid were included as controls. Identically treated cells were used for RNA isolation for the qPCR data shown in . () qPCR analysis measu! ring mRNA levels of 10 group 1 genes (Fig. 2). The level of each gene (rows) relative to GAPDH was used to determine the fold-change effects shown in the heatmap. Raw data are shown in Supplementary Table S4. Fold-changes are shown for each of the comparisons depicted at the top of each column, and the numbers within each rectangle equal fold-changes for each gene. Note that for most genes, knockdown of Nanog did not reduce the stimulation of gene expression caused by Wnt3a treatment or by Tcf3 ablation. In addition, overexpression of Nanog in Tcf3+/+ cells was not sufficient to activate expression of group 1 genes as well as either Wnt3a treatment or Tcf3 ablation. Uncropped images of blots are shown in Supplementary Fig. S7. * Figure 4: A combination of Tcf3–β-catenin-dependent and Tcf3–β-catenin-independent mechanisms mediates Wnt3a stimulation of self-renewal and target gene expression. () Alkaline phosphatase staining of Tcf3+/+, Tcf3−/− and Tcf3ΔN/ΔN ESC colonies after four days in serum-containing media without LIF, and without (Control; left) or with 50 ng ml recombinant mouse Wnt3a (right). The percentage of AP+ ESC colonies (white number at the top of each image) represents the individual counts of biological duplicates. () Luciferase reporter assays for β-catenin stimulation of SuperTOPFlash (top), Nanog promoter (middle) and Nr5a2 promoter (bottom). Increasing the amount of stable ΔNβ-catenin expression plasmid differentially activated reporter activities in Tcf3+/+ (white), Tcf3−/− (black) and Tcf3ΔN/ΔN (grey) ESCs. Data supporting the generation and primary characterization of Tcf3ΔN/ΔN ESCs are shown in Supplementary Fig. S4. The values represent the mean of biological duplicates, and the ends of the range bars correspond to the two individual measurements. * Figure 5: Endogenous Tcf1 stimulates Tcf3–β-catenin-independent activation of Wnt target genes. () Tcf cDNAs were measured by qPCR from RNA isolated from Tcf3+/+, Tcf3−/− and Tcf3ΔN/ΔN ESCs. The values represent the mean of biological duplicates, and the ends of the range bars correspond to the two individual measurements. () Stable ΔNβ-catenin expression plasmid was co-transfected into ESCs with increasing concentrations of the indicated Tcf-expression plasmids. Relative SuperTOPFlash luciferase values represent the mean of biological duplicates, and the ends of the range bars correspond to the two individual measurements. () qPCR measurement of Tcf1 mRNA (top) from the Tcf1 shRNA and SCRM shRNA cell lines indicated at the bottom of the panel. The values represent the mean of biological duplicates, and the ends of the range bars correspond to the two individual measurements. Western blot (WB) analysis of protein lysates isolated from indicated ESC lines was carried out using the antibodies indicated on the right. Data supporting the effectiveness of shRNA knoc! kdown of Tcf1 are shown in Supplementary Fig. S5. () SuperTOPFlash luciferase reporter assay as described in using cell lines also used in . Uncropped images of blots are shown in Supplementary Fig. S7. * Figure 6: Combined effect of Tcf1 and Tcf3– β -catenin for the Wnt3a response of ESCs. () Alkaline phosphatase staining of colonies from ESC lines with different Tcf3 genotypes (indicated on the left) and expressing an shRNA (indicated at the bottom). Colonies were formed for four days in serum-containing media without LIF, and without (Control) or with Wnt3a. The percentage of AP+ ESC colonies (white number at the top of each image) represents the individual counts of biological duplicates. Wnt3a stimulated self-renewal at different levels in Tcf3+/+, Tcf3−/− and Tcf3ΔN/ΔN ESCs, whereas knockdown of Tcf1 significantly reduced stimulations in all three ESC lines. () Gene expression analysis in different Tcf3 ESC lines ±Tcf1 shRNA and ± recombinant Wnt3a. The relative level of mRNA for each gene is compared with untreated Tcf3+/+ cells set at 1.0. Numerical values ± standard deviations are shown in Supplementary Table S5. (–) Chromatin immunoprecipitated with antibodies against Tcf3 (), Oct4 () or β-catenin () was measured by qPCR using primers (Sup! plementary Table S5) specific for upstream regions of the ten genes labelled at the bottom of . A representative graph from three biological replicates is shown for each experiment. The values represent the mean of biological duplicates, and the ends of the range bars correspond to the two individual measurements. The assays in and were carried out using chromatin from Tcf3+/+ and Tcf3−/− ESCs. The histograms show the percentage of DNA immunoprecipitated relative to the input for each cell line. The assays in were carried out using chromatin from Tcf3 ESC lines ±Tcf1 shRNA and ±Wnt3a-conditioned media. The percentage of β-catenin immunoprecipitated was determined for control and +Wnt3a treatments, and the histogram shows the fold-increase caused by Wnt3a. Accession codes * Abstract * Accession codes * Author information * Supplementary information Referenced accessions Gene Expression Omnibus * GSE27455 Author information * Abstract * Accession codes * Author information * Supplementary information Primary authors * These authors contributed equally to this manuscript * Fei Yi & * Laura Pereira Affiliations * Department of Biochemistry and Molecular Genetics, University of Illinois at Chicago, 900 S Ashland Avenue, MBRB 2270, M/C 669, Chicago, Illinois 60607, USA * Fei Yi, * Laura Pereira, * Jackson A. Hoffman, * Brian R. Shy & * Bradley J. Merrill * Howard Hughes Medical Institute, Department of Chemistry and Chemical Biology, Harvard University, 12 Oxford Street, Cambridge, Massachusetts 02138, USA * Courtney M. Yuen & * David R. Liu Contributions All authors designed and analysed experiments. F.Y., L.P., J.A.H., B.R.S. and C.M.Y. carried out experiments. F.Y., L.P. and B.J.M. wrote the manuscript. B.J.M. and D.R.L. supervised the project. Competing financial interests The authors declare no competing financial interests. Corresponding authors Correspondence to: * Fei Yi or * Laura Pereira or * Bradley J. Merrill Author Details * Fei Yi Search for this author in: * NPG journals * PubMed * Google Scholar * Laura Pereira Search for this author in: * NPG journals * PubMed * Google Scholar * Jackson A. Hoffman Search for this author in: * NPG journals * PubMed * Google Scholar * Brian R. Shy Search for this author in: * NPG journals * PubMed * Google Scholar * Courtney M. Yuen Search for this author in: * NPG journals * PubMed * Google Scholar * David R. Liu Search for this author in: * NPG journals * PubMed * Google Scholar * Bradley J. Merrill Contact Bradley J. Merrill Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Abstract * Accession codes * Author information * Supplementary information Excel files * Supplementary Information (165K) Supplementary Table 1 * Supplementary Information (140K) Supplementary Table 2 * Supplementary Information (60K) Supplementary Table 3 * Supplementary Information (10K) Supplementary Table 4 * Supplementary Information (12K) Supplementary Table 5 * Supplementary Information (10K) Supplementary Table 6 PDF files * Supplementary Information (900K) Supplementary Information Additional data - External forces control mitotic spindle positioning
- Nat Cell Biol 13(7):771-778 (2011)
Nature Cell Biology | Article External forces control mitotic spindle positioning * Jenny Fink1 * Nicolas Carpi1 * Timo Betz2 * Angelique Bétard2 * Meriem Chebah1 * Ammar Azioune1 * Michel Bornens1 * Cecile Sykes2 * Luc Fetler2 * Damien Cuvelier2 * Matthieu Piel1 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:771–778Year published:(2011)DOI:doi:10.1038/ncb2269Received16 July 2010Accepted26 April 2011Published online12 June 2011 Abstract * Abstract * Author information * Supplementary information Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg The response of cells to forces is essential for tissue morphogenesis and homeostasis. This response has been extensively investigated in interphase cells, but it remains unclear how forces affect dividing cells. We used a combination of micro-manipulation tools on human dividing cells to address the role of physical parameters of the micro-environment in controlling the cell division axis, a key element of tissue morphogenesis. We found that forces applied on the cell body direct spindle orientation during mitosis. We further show that external constraints induce a polarization of dynamic subcortical actin structures that correlate with spindle movements. We propose that cells divide according to cues provided by their mechanical micro-environment, aligning daughter cells with the external force field. View full text Figures at a glance * Figure 1: Retraction fibre distribution dictates mitotic spindle orientation. () HeLa cell dividing on a bar-shaped fibronectin micropattern (left) and expressing MyrPalm–GFP to visualize the plasma membrane (middle and right). A maximum intensity projection (top view of the cell, middle) and three-dimensional reconstruction (side view, right) of 85 deconvoluted Z -stack planes at 0.3 μm intervals are presented. () Mitotic HeLa cells on bar- (top row) or cross-shaped (bottom row) micropatterns (outlined by the orange dashed line). Cells are expressing MyrPalm–GFP (green) and histone2B–mCherry (red) to visualize plasma membrane/retraction fibres and chromosomes. Retraction fibres orthogonal to the chromosome plate were laser ablated (white arrows at 0 min). To show both cell body and retraction fibres, two Z planes were superposed in the green channel for the first image in the top row. Other images are single confocal slices. Time is in min (′) and s (′′). () Observed chromosome plate orientation (red lines) on bar- and cross-shaped ! micropatterns (orange) before and after cutting. Retraction fibres/plasma membrane, green. () Total rotation (final−initial angle) of the metaphase plate in control cells and in cells after retraction fibre ablation. n=207 (bar control), 69 (cross control), 11 (bar ablation) and 10 (cross ablation). Error bars represent standard error. ***, P value <0.0001 (Student's t -test). Scale bars, 10 μm. * Figure 2: Retraction fibres exert strong forces on the mitotic cell body. () Mitotic HeLa cells on bar- and cross-shaped micropatterns (outlined by orange dashed lines) before and after ablation (white arrows) of retraction fibres at 0 min. Cells expressing MyrPalm–GFP (green) and histone2B–mCherry (red) to visualize membrane/retraction fibres and chromosomes. The first images show a superposition of two Z planes in the green channel. The other images are single confocal slices. The first row shows the same cell as in Fig. 1. Time is in min (′). () Chromosome plate angle (red) versus cell shape (width/length ratio, green) before and after retraction fibre ablation. The plots correspond to cells shown in (bar, dotted line; cross, solid line). () Forces in retraction fibres measured with optical tweezers. L, retraction fibre length; x, bead displacement; Fp, pulling force of optical tweezers; Ft, tension of retraction fibre. For , Ft=Fp×L/(4x) . Green, membrane; violet, bead. () Phase-contrast image showing a cell body and retraction fibres! with an attached bead. Optical tweezers pulling direction (arrow). () Measured tension in retraction fibres; n=116 on 20 fibres. () Shape factor of drug-treated cells dividing on bar-shaped micropatterns. n=30 (control, Ctrl), 54 (cytochalasin D, CytoD), 55 (blebbistatin, Bleb) and 43 (calyculin A, CalyA). Error bars represent standard error. ***, P value <0.0001 (Student's t -test). () Model for cell shape changes after laser ablation (black arrows). Intercellular contractility (blue arrows). Resisting retraction fibres (green arrows). Membrane/retraction fibres, green; chromosome plate, red; micropattern, orange. Scale bars, 10 μm. * Figure 3: Stretching retraction fibres induces spindle rotation. () Cell stretching set-up. Membrane, green; micropattern, orange. () Mitotic HeLa cell expressing MyrPalm–GFP to visualize membranes. Maximum intensity projections of deconvoluted Z stacks. The insets show a 1.5-fold magnification. The dashed lines represent the orientation of the retraction fibre before stretching. () Calculated force field before and after stretching. Black arrows and numbers represent total forces (nN) in X and Y axes. () RPE1 cells dividing on silicon substrates with oval ring micropatterns (orange). The pictures are a superposition of phase-contrast images and the Hoechst 33342 signal (red) to visualize the DNA/chromosomes. Horizontal stretch (white arrow) just after image acquisition at 0 min. () Chromosome plate orientation (red lines) in controls (upper row) and stretched cells (lower row). The initial orientation shows the orientation in the first recorded image and the final orientation shows the orientation at anaphase onset. Membrane, green; ! pattern, orange. () Scatter plot showing net rotations (α) of chromosome plates for controls and stretched cells. Negative values represent rotation towards 0° (final angle for controls on oval ring patterns); positive values represent rotation towards 90°. n=12 (circular ring), 13 (oval ring), 12 (oval ring to circular ring), 8 (oval to oval) and 14 (oval ring to circular ring, nocodazole). ***, P value <0.0001 (Student's t -test). Time is in min (′). Scale bars, 10 μm () and 20 μm (). * Figure 4: Adhesion geometry can bias dynamic subcortical actin structures in mitotic cells. () HeLa cells dividing on 12 μm fibronectin lines were fixed, stained for actin (phalloidin), microtubules and DNA and imaged using a spinning-disc confocal microscope. () Graph representing the distribution of subcortical actin structures in fixed HeLa cells dividing on 12 μm lines (n=70 from two independent experiments). 'Not aligned' means unipolar structure, but not aligned with the retraction fibre axis. Error bars represent standard error. () Mitotic HeLa cells on bar- (first row), cross- (second row) and disc-shaped (third row) micropatterns (dashed orange lines). Cells are stably expressing Lifeact–mCherry to visualize filamentous actin. The equatorial plane of cells was imaged during mitotic cell rounding. Images for Z projections, on the right, were acquired at 1 μm intervals, after complete cell rounding. Images for cells dividing on bar and disc shapes were mirrored to align with the cell on the cross shape. Scale bars, 10 μm. Time is in min! (′). * Figure 5: Quantification of subcortical actin polarization for different adhesion geometries. () Kymographs of 70 pixels in width were created from time-lapse movies (15 s) of cells expressing Lifeact–mCherry (top left picture). Cells were dividing on bar-, cross- and disc-shaped patterns (scheme: fibronectin pattern, orange; actin (retraction fibres, cortex), red; spindle, green; chromosome plate, black). Kymographs were generated in the main axis of the retraction fibres (bar and cross shape; 0°) or in the axis of the mitotic spindle (disc shape; 0°) and orthogonal to it (90°) as indicated by the arrows (top left picture). The second row shows a representative example of the resulting kymographs for a cell dividing on a bar-shaped pattern. (–) Residence time (), total intensity () and penetration depth () of actin structures measured using kymographs generated along 0° and 90°. The first row shows a twofold magnification of the white frame in . Black lines symbolize the measured height (residence time), area (intensity) and width (penetration depth) of t! he Lifeact–mCherry signal. Graphs represent values for cells dividing on bar- (n=14), cross- (n=13) and disc- (n=12) shaped patterns in 0° and 90° axes. Error bars represent standard error. NS, not significant. **, P=0.018 in ; ***, P<0.0001 (bar) and P=0.01 (cross) in (Student's t -test). * Figure 6: Polarization of dynamic subcortical actin structures persists when astral microtubules are depolymerized and the spindle is misoriented. () HeLa cell stably expressing Lifeact–mCherrry and EB3–GFP to visualize filamentous actin and spindle poles, respectively. The cell is dividing on a bar-shaped pattern as shown in the scheme (top, fibronectin pattern, orange; actin (retraction fibres, cortex), red; spindle, green; chromosome plate, black). As a result of treatment with low doses of nocodazole (20 nM), the spindle is misaligned during mitosis (middle). Kymographs were generated in the Lifeact–mCherry channel along the retraction fibre axis (0°) of the bar pattern and orthogonal to it (90°). Although the spindle poles are aligned with the 90° axis, polarization of subcortical actin structures remains in the axis of the retraction fibres (0°, bottom). (–) Graphs representing residence time (), total intensity () and penetration depth () of actin structures along 0° and 90° axes in HeLa cells dividing on a bar-shaped pattern (see Fig. 5 for measurement). Cells were treated with either dimethylsu! lphoxide (Ctrl, 0.2%, n=14) or nocodazole (NZ, 20 nM, n=15). Error bars represent standard error. NS, not significant. ***, P<0.0001 for 'Ctrl' and P=0.0060 for 'NZ' (Student's t -test). Scale bar, 10 μm. Time is in min (′). * Figure 7: Mitotic spindle rotation and movement strongly correlate with polarization of dynamic subcortical actin structures. () Ablation of retraction fibres (white arrows) in cells dividing on bar (top panel) and cross patterns (bottom panel). Retraction fibres are visualized by Lifeact–mCherry expression and are cut at 0 s. Patterns, dashed orange lines. () Evolution of subcortical actin structures (Lifeact–mCherry) and the mitotic spindle (EB3–GFP) after retraction fibre ablation of the cells shown in . Cell dividing on a bar pattern (top panel); cell dividing on a cross pattern (bottom panel). White arrows highlight regions with a higher density of subcortical actin structures. () Quantification of actin structure behaviour after ablation of retraction fibres. Actin repolarization means polarization of subcortical structures in the axis orthogonal to the ablated retraction fibres. Repolarization was taken into account only when lasting longer than 10 consecutive minutes. n=16 for bar shape; n=19 for cross shape. ***, P=2×10−5 (Chi-square test). () Graph representing the time when ac! tin starts to repolarize in the axis orthogonal to retraction fibre cutting (t1(repol)), time when the spindle starts to rotate (t2(rot)) and the delay between the start of actin repolarization and the start of spindle turning (t2–t1). Ablation is at t=0. () Quantification of spindle behaviour after ablation of retraction fibres. Total rotation means maximum spindle rotation towards the axis orthogonal to retraction fibre cutting. n=16 for bar shape; n=19 for cross shape. Error bars represent standard error. ***, P<0.0001 (Student's t -test). Scale bars, 10 μm. Time is in s (′′, ) and min (′, ). * Figure 8: Subcortical actin structures exert pulling forces on the mitotic spindle. () Time-lapse sequence of a HeLa cell stably expressing Lifeact–mCherrry and EB3–GFP. The white asterisk marks the position of the spindle pole at time zero. Scale bars, 10 μm (first image) and 2 μm (second image). Time is in s (′′). () A HeLa cell dividing on a bar pattern (as shown in the scheme: fibronectin pattern, orange; actin (retraction fibres, cortex), red; spindle, green; chromosome plate, black) and stably expressing Lifeact–mCherry and EB3–GFP to visualize actin filaments and spindle poles. Equatorial planes were imaged every 15 s using a spinning-disc confocal microscope. Kymographs of 70 pixel in width were created parallel to the spindle as indicated by the white bar and arrows in the picture. Scale bar, 10 μm. () Kymograph resulting from . The dashed white areas highlight the actin signal; the white lines highlight spindle boundaries. () Lifeact–mCherry temporal profile corresponding to the kymograph shown in (red signal) for the! right and the left side of the cell. () Spindle pole displacement corresponding to the kymograph shown in as determined by automated detection of the spindle boundary in the kymograph (green signal). Note that the spindle always changes direction with a delay, to eventually move towards the structures. () A HeLa cell dividing on a bar-shaped pattern as shown in . Kymographs were generated along the retraction fibre axis in the Lifeact–mCherry and EB3–GFP channel according to . Resulting kymographs of control cells (Ctrl, first row) and nocodazole-treated (NZ, 20 nM, second row) cells. Note that the spindle does not oscillate anymore in NZ-treated cells. () Quantification of spindle pole movements with respect to subcortical actin structures of control- (Ctrl, n=47 for 12 cells) and nocodazole- (NZ, 20 nM; n=42 for 11 cells) treated cells; correlated movement means spindle movement towards actin structures; anti-correlated means movement away from actin structures. ! **, P value is 0.014 (Chi-square test). Author information * Abstract * Author information * Supplementary information Affiliations * Institut Curie, CNRS UMR 144, 26 rue d'Ulm, 75248 Paris Cedex 05, France * Jenny Fink, * Nicolas Carpi, * Meriem Chebah, * Ammar Azioune, * Michel Bornens & * Matthieu Piel * Institut Curie, CNRS UMR 168, 26 rue d'Ulm, 75248 Paris Cedex 05, France * Timo Betz, * Angelique Bétard, * Cecile Sykes, * Luc Fetler & * Damien Cuvelier Contributions J.F. designed, carried out and analysed most experiments and wrote the article, N.C. carried out most cell stretching experiments as well as experiments shown in Fig. 2f and Supplementary Fig. S3, T.B. carried out and analysed optical trap experiments (Fig. 2c–e), A.B. carried out some cell stretching experiments, M.C. carried out the experiments shown in Fig. 2f and Supplementary Fig. S3, A.A. developed the method to produce micropatterns on stretchable substrates, M.B. contributed ideas, discussion and supervised part of the work of J.F., C.S. contributed ideas and discussion and supervised the work on optical trap experiments, L.F. carried out laser ablation experiments (Fig. 1), D.C. set up the cell stretching device, supervised the work of A.B. and contributed ideas and discussion, and M.P. supervised the work, carried out experiments and wrote the paper. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Matthieu Piel Author Details * Jenny Fink Search for this author in: * NPG journals * PubMed * Google Scholar * Nicolas Carpi Search for this author in: * NPG journals * PubMed * Google Scholar * Timo Betz Search for this author in: * NPG journals * PubMed * Google Scholar * Angelique Bétard Search for this author in: * NPG journals * PubMed * Google Scholar * Meriem Chebah Search for this author in: * NPG journals * PubMed * Google Scholar * Ammar Azioune Search for this author in: * NPG journals * PubMed * Google Scholar * Michel Bornens Search for this author in: * NPG journals * PubMed * Google Scholar * Cecile Sykes Search for this author in: * NPG journals * PubMed * Google Scholar * Luc Fetler Search for this author in: * NPG journals * PubMed * Google Scholar * Damien Cuvelier Search for this author in: * NPG journals * PubMed * Google Scholar * Matthieu Piel Contact Matthieu Piel Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Abstract * Author information * Supplementary information Movies * Supplementary Information (3M) Supplementary Movie 1 * Supplementary Information (5M) Supplementary Movie 2 * Supplementary Information (5M) Supplementary Movie 3 * Supplementary Information (7M) Supplementary Movie 4 * Supplementary Information (6M) Supplementary Movie 5 * Supplementary Information (5M) Supplementary Movie 6 * Supplementary Information (8M) Supplementary Movie 7 * Supplementary Information (8M) Supplementary Movie 8 PDF files * Supplementary Information (1200K) Supplementary Information Additional data - Myosin 1b promotes the formation of post-Golgi carriers by regulating actin assembly and membrane remodelling at the trans-Golgi network
- Nat Cell Biol 13(7):779-789 (2011)
Nature Cell Biology | Article Myosin 1b promotes the formation of post-Golgi carriers by regulating actin assembly and membrane remodelling at the trans-Golgi network * Claudia G. Almeida1, 2 * Ayako Yamada1, 3 * Danièle Tenza1, 4, 5 * Daniel Louvard1, 2 * Graça Raposo1, 4, 5 * Evelyne Coudrier1, 2 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:779–789Year published:(2011)DOI:doi:10.1038/ncb2262Received03 June 2010Accepted18 April 2011Published online12 June 2011 Abstract * Abstract * Author information * Supplementary information Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg The function of organelles is intimately associated with rapid changes in membrane shape. By exerting force on membranes, the cytoskeleton and its associated motors have an important role in membrane remodelling. Actin and myosin 1 have been implicated in the invagination of the plasma membrane during endocytosis. However, whether myosin 1 and actin contribute to the membrane deformation that gives rise to the formation of post-Golgi carriers is unknown. Here we report that myosin 1b regulates the actin-dependent post-Golgi traffic of cargo, generates force that controls the assembly of F-actin foci and, together with the actin cytoskeleton, promotes the formation of tubules at the TGN. Our results provide evidence that actin and myosin 1 regulate organelle shape and uncover an important function for myosin 1b in the initiation of post-Golgi carrier formation by regulating actin assembly and remodelling TGN membranes. View full text Figures at a glance * Figure 1: Distribution of Myo1b, MPR, F-actin and the Arp2/3 complex in the perinuclear region. () A representative spinning-disc confocal microscopy section throughout the nucleus after detection of Myo1b by immunofluorescence in HeLa cells. Note the concentration of Myo1b in the perinuclear region. The scale bar represents 10 μm. () A representative spinning-disc confocal microscopy section throughout the nucleus after detection of Myo1b (green), TGN46 (blue) and MPR (red) by immunofluorescence. Note the overlapping distribution of Myo1b with MPR and TGN46 (arrows). The white box indicates the enlarged TGN region. The scale bar represents 10 μm. () HeLa cells co-transfected with plasmids encoding GFP–Myo1b and LifeAct–Cherry were analysed by spinning-disc confocal microscopy. A representative spinning-disc confocal microscopy section through the nucleus shows the overlapping distribution of GFP–Myo1b (green) and F-actin foci detected with LifeAct–Cherry (red). The white box indicates the enlarged TGN region. Note the overlapping distribution of Myo1b wi! th F-actin foci in the TGN region (arrows). The scale bar represents 10 μm. () HeLa cells co-transfected with plasmids encoding GFP–Myo1b and p16–Cherry, a subunit of the Arp2/3 complex, were analysed by spinning-disc confocal microscopy. A representative spinning-disc confocal microscopy section through the nucleus shows the overlapping distribution of GFP–Myo1b (green) with p16–Cherry (red). The white box indicates the enlarged TGN region. Note the overlapping distribution of Myo1b with the Arp2/3 complex in the TGN region (arrows). The scale bar represents 10 μm. * Figure 2: Myo1b controls the steady-state distribution of MPR. () HeLa cells were transfected with Myo1bH siRNA or control siRNA and immunolabelled with anti-MPR antibody, and the distribution of MPR was analysed by confocal microscopy. Representative maximum projections are shown. The white boxes indicate the enlarged regions. The scale bar represents 10 μm. () The percentage of fluorescence corresponding to MPR in the perinuclear region was quantified as described in the Methods (n=6;N=56 control-siRNA-treated cells; N=66 Myo1b-siRNA-treated cells; *P=10−19, t -test, mean±s.e.m.). () HeLa cells were transfected with plasmids encoding GFP (green) or GFP–Myo1b (green) and immunolabelled with anti-MPR antibody (red), and the distribution of MPR was analysed by confocal microscopy. Representative maximum projections are shown. The scale bars represent 10 μm. () The number of cells exhibiting a dispersed distribution of MPR as shown for GFP–Myo1b-expressing cells or a normal distribution as shown for GFP-expressing cells was c! ounted and normalized to the total number of cells analysed in three independent experiments (n=3,N=97 GFP–Myo1b-expressing cells, N=187 GFP-expressing cells). * Figure 3: Myo1b knockdown impairs the TGN exit of cargo. (,) The internalized MPR in the perinuclear region was analysed by epifluorescence microscopy () and quantified (). HeLa cells were transfected with control siRNA or Myo1b siRNA and with a plasmid encoding chMPR (bovine murine MPR chimera). Cells were then labelled with anti-chMPR antibody for 5 min followed immediately by fixation (0 min), or incubated for 60 min and then labelled with antibody followed by fixation (60 min). The scale bar represents 10 μm (n=2,N=8–10 control-siRNA-treated cells, N=7–12 Myo1b-siRNA-treated cells, mean). () The kinetics of the exit of GFP–MPR from the TGN was monitored by time-lapse imaging using spinning-disc confocal microscopy in control-siRNA- and Myo1b-siRNA-treated HeLa cells after temperature block (see Supplementary Movie S1). Representative maximum intensity projections at 0 and 120 min are shown. The scale bar represents 10 μm. () The amount of GFP–MPR fluorescence was quantified in the TGN region as a functi! on of time and as a percentage of the fluorescence detected at 0 min (n=3,N=16 control-siRNA-treated cells, N=21 Myo1b-siRNA-treated cells; mean±s.e.m.). () HeLa cells were transfected with Myo1b siRNA or control siRNA, immunolabelled with anti-cathepsin D and TGN46 antibodies and analysed by epifluorescence microscopy. The scale bar represents 10 μm. () The amount of cathepsin D in the TGN region was quantified (n=3,N=47 control-siRNA-treated cells, N=67 Myo1b-siRNA-treated cells; *P=10−14, t -test, mean±s.e.m.). () β -hexosaminidase activity was quantified in conditioned media of control-siRNA- and Myo1b-siRNA-treated HeLa cells (n=2, carried out in duplicate; mean). () The kinetics of exit of p75–GFP from the TGN (see Supplementary Movie S2) was analysed as described in , for GFP–MPR exit (n=4,N=77 control-siRNA-treated cells, N=58 Myo1b-siRNA-treated cells; mean±s.e.m.). () Arrival of p75–GFP at the plasma membrane in control-siRNA- and Myo1b-siRNA-trea! ted HeLa cells was quantified as a function of time (n=3, N=47! –61 control-siRNA-treated cells, N=38–75 cells Myo1b-siRNA-treated cells; mean±s.e.m.). () The kinetics of exit of GPI–GFP from the TGN (see Supplementary Movie S3) was analysed as described in , for GFP–MPR (n=3,N=72 control-siRNA-treated cells, N=68 cells Myo1b-siRNA-treated cells; mean±s.e.m.). () HeLa cells transfected with control siRNA or Myo1b siRNA were analysed by conventional electron microscopy. On Myo1b depletion, Golgi stacks are more compact (low magnification; left panel), shorter and often more dilated at the rims (high magnification, arrows; right panels). The scale bars represent 250 nm. * Figure 4: Myo1b controls the formation of tubular carriers at the TGN. () HeLa cells were transfected with Myo1b siRNA or control siRNA, and with a plasmid encoding GFP–MPR. GFP–MPR carriers were monitored at 37 °C by time-lapse imaging using spinning-disc confocal microscopy (see Supplementary Movie S4). The first frames of representative movies and representative kymographs (that cover 26 s) revealing the sequence of events of tubule formation39 and scission (*) are shown. The arrow indicates the beginning of tubule formation. The scale bars represent 10 μm. () The average number of cytoplasmic carriers observed per frame was quantified (n=3,N=24 control-siRNA-treated cells, N=19 Myo1b-siRNA-treated cells; *P=10−5, t -test, mean±s.e.m.). () The number of tubules that formed and underwent scission per minute was quantified (n=3,N=247 tubules in 21 control-siRNA-treated cells, N=103 tubules in 25 Myo1b-siRNA-treated cells; *P=10−7 t -test, mean±s.e.m.). (,) HeLa cells were co-transfected with plasmids encoding GFP–MPR and Li! feAct–Cherry () or GFP–MPR and Cherry–Myo1b (). The first frames of representative movies and representative kymographs of cells expressing low levels of the recombinant proteins (that cover 6 s) revealed the presence of Cherry–Myo1b (red; see Supplementary Movie S5) and LifeAct–Cherry (red) at the base of nascent GFP–MPR (green) tubules emanating from the TGN (arrows; ref. 39). The scale bars represent 10 μm. A schematic representation of the distribution of Cherry–Myo1b (red) and LifeAct–Cherry (red) at the base of the nascent tubes (green) is also shown. () HeLa cells were co-transfected with plasmids encoding GFP–MPR (green) and Cherry–Myo1b (red) or with GFP–MPR alone (Mock). GFP–MPR carriers were monitored at 37 °C by time-lapse imaging using spinning-disc confocal microscopy (see Supplementary Movie S6). The first frames of representative movies and representative kymographs (that cover 26 s) revealing the sequence of tubule formatio! n39 and scission (*) are shown. Note that the tubule did not u! ndergo scission in the Cherry–Myo1b-expressing cell. The arrow indicates the beginning of tubule formation. The scale bar represents 10 μm. () The average number of cytoplasmic carriers observed per frame was quantified (n=3, N=28 Cherry–Myo1b-expressing cells, N=24 mock cells; P=0.87, t -test mean±s.e.m.). () The number of GFP–MPR stable tubules was quantified (n=4,N=49 Cherry–Myo1b-expressing cells, N=40 mock cells; *P=0.0014 t -test, mean±s.e.m.). * Figure 5: Myo1b depletion reduces the number of F-actin foci and of Arp2/3 complex structures in the vicinity of the TGN. () HeLa cells transfected with Myo1b siRNA or control siRNA, and with plasmids encoding LifeAct–Cherry and GFP–MPR, were monitored by time-lapse imaging using spinning-disc confocal microscopy (see Supplementary Movie S7). The first frames of representative movies are shown. The white boxes indicate the enlarged TGN region shown in . The scale bar represents 10 μm. () The TGN pool of GFP–MPR (green) and LifeAct–Cherry (red) alone or merged and overlapping pixels are shown at high magnification. The arrows indicate sites of overlap. The scale bar represents 10 μm. (,) The number of LifeAct–Cherry puncta (*P=0.00005, t -test, ) and the number of interactions between GFP–MPR and LifeAct–Cherry (*P=0.0003, t -test, ) in the TGN region were quantified as described in the Methods (n=4,N=35 control-siRNA-treated cells, N=32 Myo1b-siRNA-treated cells). () HeLa cells transfected with control siRNA or Myo1b siRNA were labelled with fluorescent phalloidin, immunolab! elled with anti-GM130 antibody and analysed by three-dimensional deconvolution microscopy. Representative maximum-intensity projections of merged F-actin (red) and GM130 (green) are shown at low magnification. The white boxes indicate the enlarged Golgi regions. A single focal plane of F-actin in the Golgi region is shown at high magnification. The scale bar represents 10 μm. () The number of F-actin foci in the Golgi region was quantified as described in the Methods (n=2,N=12 control-siRNA-treated cells, N=18 Myo1b-siRNA-treated cells). () HeLa cells were transfected with plasmids encoding Cherry or Cherry–Myo1b, immunolabelled with anti-p34 and GM130 antibodies and analysed by three-dimensional deconvolution microscopy. Representative maximum-intensity projections of Cherry (red), Cherry–Myo1b (red) and merged p34 (green) and GM130 (red) are shown at low magnification. The white boxes indicate the enlarged TGN region. Single focal planes of p34 in the Golgi region ! are shown at high magnification. The scale bar represents 10�! �μm. Note the increase of p34 puncta in Cherry–Myo1b-expressing cells in the Golgi region. (,) The number of p34 puncta in the volume occupied by GM130 in the Golgi region (*P=0.0032, t -test, ) and at the ventral plasma membrane (P=0.26, t -test, ) was quantified as described in the Methods (n=3,N=13 Cherry-expressing cells, N=21 Cherry–Myo1b-expressing cells). * Figure 6: Functional Myo1b is required for organizing F-actin–Arp2/3 foci. () HeLa cells were transfected with control siRNA or Myo1b siRNA and with plasmids encoding FlagHA–Myo1b-5M, FlagHA–Myo1b-5MR or FlagHA–Myo1b-5ME, immunolabelled with anti-MPR (green), anti-TGN46 (red) and anti-HA antibodies and analysed by epifluorescence microscopy. The scale bar represents 10 μm. () The percentage of fluorescence corresponding to MPR in the TGN region was quantified as described in the Methods (n=3,N=55 control-siRNA-treated cells, N=54 Myo1b-siRNA-treated cells, N=41 Myo1b siRNA + FlagHA–Myo1b-5M-treated cells, N=46 Myo1b siRNA + FlagHA–Myo1b-5MR treated cells, N=56 Myo1b siRNA + FlagHA–Myo1b-5ME treated cells, *P<0.05 versus control-siRNA-treated cells, analysis of variance (ANOVA), mean±s.e.m.). () HeLa cells were transfected with control siRNA or Myo1b siRNA and with plasmids encoding FlagHA–Myo1b-5M, FlagHA–Myo1b-5MR or FlagHA–Myo1b-5ME, immunolabelled with anti-p34 (green), HA and MPR (red) antibodies and analysed by three-dime! nsional deconvolution microscopy. Representative maximum-intensity projections at low magnification of HA and merged MPR (red) and p34 (green) are shown. Single focal planes at high magnification of merged MPR and p34, and p34 alone, are shown. The white boxes indicate the enlarged TGN region. The scale bar represents 10 μm. (,) The number of p34 puncta in the volume occupied by MPR in the TGN region (*P<0.05, versus control-siRNA-treated cells, ANOVA, mean±s.e.m., ) and at the ventral plasma membrane (P>0.05, versus control-siRNA-treated cells, ANOVA, mean±s.e.m., ) was quantified as described in the Methods (n=3,N=18 control-siRNA-treated cells, N=17 Myo1b-siRNA-treated cells, N=22 Myo1b siRNA + FlagHA–Myo1b-5M-treated cells, N=22 Myo1b siRNA + FlagHA–Myo1b-5MR treated cells, N=19 Myo1b siRNA + FlagHA–Myo1b-5ME treated cells). * Figure 7: Depletion of the Arp2/3 complex induces MPR accumulation in TGN and inhibits post-Golgi carrier formation. () HeLa cells were transfected with p34-A siRNA or control siRNA, immunolabelled with anti-p34 and GM130 antibodies, labelled with fluorescent phalloidin and analysed by three-dimensional deconvolution microscopy. Representative maximum-intensity projections of merged p34 (red), F-actin (green) and GM130 (blue) are shown at low magnification. The white boxes indicate the enlarged Golgi regions. Single focal planes of p34 and F-actin in the Golgi region are shown at high magnification. The scale bar represents 10 μm. () HeLa cells were transfected with p34-A siRNA, Myo1b siRNA or control siRNA, immunolabelled with anti-MPR antibody and analysed for the distribution of MPR by epifluorescence microscopy. The scale bar represents 10 μm. () The percentage of MPR fluorescence in the TGN region was quantified as described in the Methods (n=3,N=62 control-siRNA-treated cells, N=45 p34-A-siRNA-treated cells, N=68 Myo1b-siRNA-treated cells; *P<0.001, ANOVA, mean±s.e.m.). () HeL! a cells were transfected with p34-A siRNA or control siRNA, and with a plasmid encoding GFP–MPR. GFP–MPR carriers were monitored at 37 °C by time-lapse imaging using spinning-disc confocal microscopy (see Supplementary Movie S8). The first frames of representative movies are shown at low magnification. The white boxes indicate the enlarged TGN regions. Note the tubules containing GFP–MPR (arrows). The scale bar represents 10 μm. () The average number of cytoplasmic carriers per frame was quantified as described in the Methods (n=2,N=18 control-siRNA-treated cells, N=18 p34-A-siRNA-treated cells, mean). () The number of tubules that formed and underwent scission per minute was quantified (n=2,N=18 control-siRNA-treated cells, N=18 p34-A-siRNA-treated cells, mean). * Figure 8: Model for the role of Myo1b in the formation of tubular carriers at the TGN. () By actively tethering and orienting the polymerizing F-actin to the TGN membrane Myo1b may increase its stability and thus induce the formation of F-actin foci. In addition, active Myo1b would generate a force to deform the TGN membrane. Myo1b is represented before (dashed blue line) and after (solid blue) the power stroke, during which it pulls the actin filament (initial position, dashed grey line; final position, solid grey line) towards the cytosol (arrows indicate the directions of F-actin and Myo1b movements) (1). By deforming the membrane Myo1b would facilitate the kinesin function to extend membrane tubular precursors along microtubules (MT, 2). Finally, the scission machinery, which includes dynamin and/or myosin 2 and F-actin, will lead to the formation of post-Golgi tubular carriers (3). () The absence of Myo1b or the inhibition of its motor activity would not properly orient polymerizing actin at the TGN membrane and would render F-actin unstable, leading to a! reduction of F-actin foci. Without Myo1b activity and F-actin foci, the formation of post-Golgi carriers would be inhibited and the exit of cargo and TGN morphology would be impaired. () The excess of Myo1b would increase the number of sites where actin polymerization is stabilized to form more F-actin foci. Consequently, sites where Myo1b induces membrane deformation would increase and extra tubular precursors pulled by kinesins would be detected due to the rate-limiting scission machinery. Note that the scale of the different panels of the model is not the same. Author information * Abstract * Author information * Supplementary information Affiliations * Institut Curie, Centre de Recherche, Paris, F-75248, France * Claudia G. Almeida, * Ayako Yamada, * Danièle Tenza, * Daniel Louvard, * Graça Raposo & * Evelyne Coudrier * Morphogenesis and Cell Signalization CNRS, UMR144, Paris, F-75248, France * Claudia G. Almeida, * Daniel Louvard & * Evelyne Coudrier * Cell and Tissue Imaging Facility, CNRS UMR 144, Paris F-75248, France * Ayako Yamada * Membrane and Cell Functions, CNRS UMR 168, Paris F-75248, France * Danièle Tenza & * Graça Raposo * Structure and Membrane Compartments CNRS, UMR144, Paris, F-75248, France * Danièle Tenza & * Graça Raposo Contributions C.G.A. and E.C. conceived the project and wrote the manuscript; C.G.A. generated and analysed most of the data; A.Y. carried out Myo1b and Myo1b mutant purification as well as their characterization in vitro; D.T. and G.R. generated and analysed the electron microscopy data; D.L. revised the manuscript. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Evelyne Coudrier Author Details * Claudia G. Almeida Search for this author in: * NPG journals * PubMed * Google Scholar * Ayako Yamada Search for this author in: * NPG journals * PubMed * Google Scholar * Danièle Tenza Search for this author in: * NPG journals * PubMed * Google Scholar * Daniel Louvard Search for this author in: * NPG journals * PubMed * Google Scholar * Graça Raposo Search for this author in: * NPG journals * PubMed * Google Scholar * Evelyne Coudrier Contact Evelyne Coudrier Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Abstract * Author information * Supplementary information Movies * Supplementary Information (300K) Supplementary Movie 1 * Supplementary Information (200K) Supplementary Movie 2 * Supplementary Information (250K) Supplementary Movie 3 * Supplementary Information (4M) Supplementary Movie 4 * Supplementary Information (400K) Supplementary Movie 5 * Supplementary Information (4M) Supplementary Movie 6 * Supplementary Information (1600K) Supplementary Movie 7 * Supplementary Information (700K) Supplementary Movie 8 PDF files * Supplementary Information (2M) Supplementary Information Additional data - Intraflagellar transport delivers tubulin isotypes to sensory cilium middle and distal segments
- Nat Cell Biol 13(7):790-798 (2011)
Nature Cell Biology | Article Intraflagellar transport delivers tubulin isotypes to sensory cilium middle and distal segments * Limin Hao1 * Melanie Thein1 * Ingrid Brust-Mascher1 * Gul Civelekoglu-Scholey1 * Yun Lu2 * Seyda Acar1 * Bram Prevo1, 3 * Shai Shaham2 * Jonathan M. Scholey1 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:790–798Year published:(2011)DOI:doi:10.1038/ncb2268Received10 August 2010Accepted21 April 2011Published online05 June 2011 Abstract * Abstract * Author information * Supplementary information Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg Sensory cilia are assembled and maintained by kinesin-2-dependent intraflagellar transport (IFT). We investigated whether two Caenorhabditis elegansα - and β-tubulin isotypes, identified through mutants that lack their cilium distal segments, are delivered to their assembly sites by IFT. Mutations in conserved residues in both tubulins destabilize distal singlet microtubules. One isotype, TBB-4, assembles into microtubules at the tips of the axoneme core and distal segments, where the microtubule tip tracker EB1 is found, and localizes all along the cilium, whereas the other, TBA-5, concentrates in distal singlets. IFT assays, fluorescence recovery after photobleaching analysis and modelling indicate that the continual transport of sub-stoichiometric numbers of these tubulin subunits by the IFT machinery can maintain sensory cilia at their steady-state length. View full text Figures at a glance * Figure 1: Characterization of the dyf-6, ift-81, ift-74, tba-5 and tbb-4 mutants. () Schematic representation of the structure of the phasmid endings. (,) The DYF-1::GFP marker was used to visualize the phasmid ciliary morphology of the wild type, three IFT-B mutants () and two tubulin mutants (). () The DYF-1::GFP marker was used to visualize the phasmid ciliary morphology of the double mutants, dyf-6;klp-11, ift-81;klp-11, ift-74;klp-11, tba-5;klp-11 and tbb-4;klp-11, demonstrating that dyf-6, ift-81 and ift-74 are distinct from tba-5(qj14) and tbb-4(sa127). () Four different markers were used to visualize the phasmid ciliary morphology in the tubulin mutants. In –, scale bars, 5 μm. Arrows point to transition zones with cilia oriented upward. See also Supplementary Fig. S3, which shows the ciliary morphology of these mutants in the phasmid and amphids using other markers. () Electron micrographs of amphid middle segments in wild-type (left), tba-5(qj14) (middle) and tbb-4(sa127) (right) adult animals. Arrows in the wild-type section point to singl! et microtubules that occur less frequently in tbb-4(sa127) mutants. () Same as in , except the amphid distal segment is shown. White arrows point to the empty distal channel. In and , scale bars, 200 nm. * Figure 2: Expression and localization of two axonemal tubulins, TBA-5 and TBB-4, and characterization of their missense mutations. (,) Models of the tba-5 and tbb-4 gene. Two tba-5 missense mutations, qj14 and dyf-10, and a deletion mutation, tm4200 (), and a tbb-4 missense mutation, dyf-12(sa127), and a deletion mutation, tbb-4(OK1461) (), are shown. (,) Inner () and outer () views of the structure of the predicted TBA-5 and TBB-4 heterodimer based on the porcine brain tubulin dimer structure, 1JFF. The three point-mutation sites (P360, A19 and L253) and the loop that contains P360 are shown in green. (–) Dyf assays (dye-filling assays) on dpy-6 (,) and dpy-6;tba-5(tm4200);tbb-4(OK1461) (,) worms. There are no obvious defects in ciliary structure in the mutants. Scale bars, 10 μm. (,) A transgene tba-5p::tba-5::GFP was expressed in amphid neurons () and phasmid neurons () in tba-5(qj14) worms. Scale bars, 10 μm. (–) Cilium formation was rescued in amphids and phasmids of tba-5(qj14) worms by expression of the transgene, tba-5p::tba-5::GFP. Gene expression indicated in green in cilia and dendr! ites (,) and intact cilia shown by dyf assays (,; dye distribution shown in red). Scale bars, 5 μm. (–) TBB-4::YFP restored the ciliary length of amphids () and phasmids () in tbb-4(sa127) and localized to the entire cilia nearly homogeneously; TBA-5::GFP restores the lengths of amphid () and phasmid () cilia in tba-5(qj14) and extended from the distal regions of middle segments to the distal tips of distal segments. Scale bars, 5 μm. Right: schematic representations of the structure of the cilia and dendrites in amphids and phasmids. Arrows point to transition zones with cilia oriented upward. A, axon; C, cilia; CB, cell body; D, dendrite. * Figure 3: Tubulin point mutants are temperature sensitive. () The wild type and the two deletion mutants, tbb-4(OK1461) and tba-5(tm4200), were nearly 100% stained in the amphid and phasmid neurons at 15 °C, 20 °C and 25 °C. tbb-4(sa127) and tba-5(qj14) worms were not stained when grown at 15 °C. However, at 25 °C, around 50% of the amphids and 20% phasmids are stained, whereas tba-5(dyf-10) has very little temperature effect. n indicates the number of amphids or phasmids. () Visualized with a TBB-4::YFP tubulin marker, tba-5(qj14) possessed only the middle segment of the amphid and phasmid cilia at 15 °C, but full-length cilia could be seen in tba-5(qj14) at 25 °C. Scale bar, 5 μm. Arrows point to transition zones with cilia oriented upward. () Visualized with an OSM-6::GFP marker, tbb-4(sa127) possessed only the middle segment of the amphid and phasmid cilia at 15 °C, but full-length cilia could be seen in tbb-4(sa127) at 25 °C. Scale bar, 5 μm. Arrows point to transition zones with cilia oriented ! upward. The images of the TBB-4::YFP marker in wild-type cilia in and is the same as used in Fig. 1e and Supplementary Fig. S3b. * Figure 4: Dynamics of axonemal microtubules at the middle segment and distal segment tips. (–) Cilia expressing TBB-4::YFP were photobleached in different regions and recovery was recorded for entire cilia (), tips of middle segments () and distal segments () in phasmids. In each case, images are shown before (at 0 s) and after photobleaching. The arrows point to the recovery regions. The schematic at the upper left for each set illustrates the region of cilia that was analysed; the photobleached region is shown by a black rectangle and the region used for recovery analysis is shown by a red rectangle. Scale bars, 5 μm. (,) The kinetics of FRAP recovery at the tips of middle segments () and the distal segments () were fitted with a single exponential equation (pink line). The fluorescence intensity is normalized to the prebleach. () EBP-2::GFP proteins are more concentrated at the tips of middle segments and distal segments. Scale bar, 5 μm. () A line scan along the cilia in . () Dynamics of EBP-2::GFP in dendrites where the EB1 homologue tracks the tips! of the microtubules. The arrows point to the comets. Scale bar, 5 μm. () A kymograph of EBP-2::GFP comets from . Horizontal scale bar, 10 μm; vertical scale bar, 10 s. BB, basal body (equivalent to TZ); CB, cell body; D, dendrite; DS, distal segment; MS, middle segment; TZ, transition zone. * Figure 5: Analysis of TBB-4::YFP transport rate in cilia. Kymographs of DYF-1::GFP and TBB-4::YFP in IFT assays under exactly the same conditions, except that TBB-4::YFP was photobleached with a mercury lamp before recording to reduce the background. () DYF-1::GFP represents the IFT transport in cilia, and the IFT tracks are clear and thick in the kymograph. () The tracks of TBB-4::YFP in cilia are faint and thin when compared with IFT tracks, for example in . () For comparison, OSM-9::GFP, which is proposed to be transported by IFT, was used as a control. All of the recorded movies were processed using the basic filters (Sharpen High and Low pass) before creating kymographs. K is the kymograph that was created along the cilia and K′ is a drawing of the kymograph lines in K. In –, horizontal scale bars 2.5 μm; vertical scale bars, 5 s. (–) Modelling of microtubule dynamics in a cilium. Dynamic instability and the delivery of tubulin subunits through IFT can constrain the length fluctuations of microtubules in both the mi! ddle (blue) and the distal (black) segments to a narrow range (); in silico FRAP of the cilium shown in for both the middle () and distal () segments indicates similar recovery curves to the experimental results (Fig. 4d,e). The fluorescence intensity is normalized to the prebleach. DS, distal segment; MS, middle segment. Author information * Abstract * Author information * Supplementary information Affiliations * Department of Molecular and Cellular Biology, University of California at Davis, One Shields Avenue, 145-B Briggs Hall, Davis, California 95616, USA * Limin Hao, * Melanie Thein, * Ingrid Brust-Mascher, * Gul Civelekoglu-Scholey, * Seyda Acar, * Bram Prevo & * Jonathan M. Scholey * Laboratory of Developmental Genetics, The Rockefeller University, 1230 York Avenue, New York, New York 10065, USA * Yun Lu & * Shai Shaham * Department of Physics and Astronomy and Laser Centre, VU University, De Boelelaan 1081, 1081 HV Amsterdam, The Netherlands * Bram Prevo Contributions J.M.S is the principal investigator of the grant and laboratory that support this IFT project. L.H. and J.M.S. designed the experiments and drafted the manuscript. J.M.S. wrote the manuscript. L.H. carried out most of the experiments. M.T. characterized the qj14 mutant by crossing all of the used IFT markers into it, made the TBB-4::YFP transgenic worms and carried out the EBP-2 experiments. I.B-M. implemented the stochastic tubulin transport and dynamics model and the in silico FRAP model and analysed the results, and helped with the FRAP experiment and transport assay of TBB-4::YFP and analysis of the results. G.C-S. designed and wrote the stochastic tubulin transport and dynamics model and the in silico FRAP model scripts. Y.L. and S.S. carried out the electron microscopy studies and analysed the results. S.A. and B.P. carried out the Y2H assays. All of the authors read the manuscript. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Jonathan M. Scholey Author Details * Limin Hao Search for this author in: * NPG journals * PubMed * Google Scholar * Melanie Thein Search for this author in: * NPG journals * PubMed * Google Scholar * Ingrid Brust-Mascher Search for this author in: * NPG journals * PubMed * Google Scholar * Gul Civelekoglu-Scholey Search for this author in: * NPG journals * PubMed * Google Scholar * Yun Lu Search for this author in: * NPG journals * PubMed * Google Scholar * Seyda Acar Search for this author in: * NPG journals * PubMed * Google Scholar * Bram Prevo Search for this author in: * NPG journals * PubMed * Google Scholar * Shai Shaham Search for this author in: * NPG journals * PubMed * Google Scholar * Jonathan M. Scholey Contact Jonathan M. Scholey Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Abstract * Author information * Supplementary information Excel files * Supplementary Information (13K) Supplementary Table 1 * Supplementary Information (15K) Supplementary Table 2 * Supplementary Information (15K) Supplementary Table 3 * Supplementary Information (10K) Supplementary Table 4 PDF files * Supplementary Information (800K) Supplementary Information Additional data - Heterochromatin boundaries are hotspots for de novo kinetochore formation
- Nat Cell Biol 13(7):799-808 (2011)
Nature Cell Biology | Article Heterochromatin boundaries are hotspots for de novo kinetochore formation * Agata M. Olszak1 * Dominic van Essen1 * António J. Pereira2 * Sarah Diehl1 * Thomas Manke1 * Helder Maiato2, 3 * Simona Saccani1 * Patrick Heun1, 4 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:799–808Year published:(2011)DOI:doi:10.1038/ncb2272Received28 February 2011Accepted28 April 2011Published online19 June 2011 Abstract * Abstract * Author information * Supplementary information Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg The centromere-specific histone H3 variant CENH3 (also known as CENP-A) is considered to be an epigenetic mark for establishment and propagation of centromere identity. Pulse induction of CENH3 (Drosophila CID) in Schneider S2 cells leads to its incorporation into non-centromeric regions and generates CID islands that resist clearing from chromosome arms for multiple cell generations. We demonstrate that CID islands represent functional ectopic kinetochores, which are non-randomly distributed on the chromosome and show a preferential localization near telomeres and pericentric heterochromatin in transcriptionally silent, intergenic chromatin domains. Although overexpression of heterochromatin protein 1 (HP1) or increasing histone acetylation interferes with CID island formation on a global scale, induction of a locally defined region of synthetic heterochromatin by targeting HP1–LacI fusions to stably integrated Lac operator arrays produces a proximal hotspot for CID depos! ition. These data indicate that the characteristics of regions bordering heterochromatin promote de novo kinetochore assembly and thereby contribute to centromere identity. View full text Figures at a glance * Figure 1: Distinct islands of CID deposition remain after pulse overexpression of CID. () Drosophila Schneider S2 cells stably transfected with inducible CID–GFP were pulse induced for CID–GFP overexpression for 48 h, sorted by fluorescent activated cell sorting (FACS) for high GFP (>75% of maximum intensity; day 0, correlating with >90% of cells showing mitotic defects25) and released from the induction into the chase (>day 0). Mitotic chromosome spreads were monitored for clearing of CID and regions with stable CID incorporation at ectopic sites (white arrows). DAPI, 4,6-diamidino-2-phenylindole. () Quantification of the percentage of cells with global mislocalization of CID (similar to day 0) or cells containing CID islands (similar to days 3–6) in a 14 day period. () CID islands recruit inner- and outer-kinetochore proteins. CID–GFP was simultaneously localized with the inner-kinetochore protein CENP-C and the outer-kinetochore proteins NDC80, ROD and POLO by indirect immunofluorescence microscopy on fixed mitotic chromosomes at day 4 after pul! se induction. Chromosomes with CID islands show co-localization with all kinetochore proteins tested in addition to the endogenous centromeres (green arrows). Scale bars=3 μm. () Schematic representation of the inner- and outer-kinetochore organization including microtubules. () Quantification of the percentage of centromeres (dark-grey bars) or CID islands (light-grey bars) co-localizing with inner- and outer-kinetochore proteins (representative images shown in : nCENPC=189, nNDC80=179, nROD=109, nPOLO=109). Data are mean ±s.e.m. ( Mitotic defects were assayed on the basis of the presence of anaphase bridges, stretched or lagging chromosomes in fixed preparations of stably transfected S2 cells at day 0 (after 48 h of CID–GFP induction) or day 4 of the chase (examples shown in Supplementary Fig. S1b). Induced cells at both times show a significant enrichment of mitotic defects as compared with untransfected and uninduced stably transfected control cells (Student's! t-test P<0.05). Data are mean±s.e.m.; cells analysed: nCONT=! 87, nUNIND=98, nday0=97, nday4=66). * Figure 2: CID islands are functional ectopic kinetochores that can mediate poleward movement of acentric chromosome fragments and self-propagate through multiple rounds of the cell cycle. () Uninduced (left column) and pulse-induced CID–GFP cells with CID–GFP islands (right column) were subjected to laser microsurgery to cut chromosome arms at day 4 in mitosis. Time-lapse imaging with 1 frame min−1 was used to monitor movements of induced chromosome fragments as the cells enter anaphase. Quick-projected images are shown. Open arrow heads indicate chromosome fragments; filled green arrow heads indicate sites of laser surgery. The dashed white line represents the metaphase plate. Shown is one of seven examples for uninduced and one of three examples for pulse-induced cells. Green=CID–GFP, red=histone H2B–mRFP (monomeric red fluorescent protein). () S2 cells stably transfected with inducible CID–GFP and low-constitutive CID–HA were subjected to 48 h of induction of CID–GFP (day 0—start chase; left column) and chased for 4 days (right column). Note the constitutive presence of CID–HA at the centromeres relative to its absence at ectopic ! sites at day 0 versus its accumulation on 84% of the CID islands on day 4. Green arrows, endogenous centromeres; white arrows, CID islands–ectopic kinetochores. Scale bar, 3 μm. () Quantification of levels of CID–GFP and CID–HA at endogenous centromeres and CID islands. Data are mean±s.e.m., nendog.CEN=40, nCID island=40. * Figure 3: Telomeres and pc-het are enriched for CID islands. () Fixed mitotic chromosomes with CID islands were stained for CID, HP1 and H4ac to help identify the heterochromatin–euchromatin boundary. Examples are shown for three categories of CID-island localization: proximal to pc-het, the telomere (tel) and in between (arm). Brackets indicate the spatial extension for each category. () Quantification of the localization of CID islands within the three spatial categories: pc-het, arm, tel. Proximity to pc-het and tel is scored, when CID islands are found near these regions within a distance of 10% of the average arm length (equivalent to expected random distribution). CID islands analysed: nCID pulse=391, nCID pulse+HP1 RNAi=70. The difference in distribution of CID islands in either experiment is highly significant when compared with the expected random distribution and with each other (χ2 test P<0.01). Data are mean±s.e.m. () CID islands combined with HP1 or H4ac staining only show limited overlap on fixed mitotic chromo! somes. HP1 staining was applied to visualize pc-het and tel. Regions marked by an asterisk are shown as insets amplified twofold and shown on the right. () Example of a structurally less preserved mitotic autosome with a telomeric CID island to reveal the pronounced separation of the CID island from the telomeric HP1 focus (see the inset). () Chromosomes stained for H4ac to visualize euchromatin. Inset: a X chromosome. () Staining as in , with an X chromosome containing two CID islands separated by a region of strong H4ac staining. () Example of two mitotic autosomes with CID islands in the middle of the arm. Note the presence of HP1 and H4ac staining in this chromosome region next to CID islands showing little overlap. Scale bars, 3 μm. In ,–, green arrows represent endogenous centromeres. In ,, white arrows represent CID islands–ectopic kinetochores. * Figure 4: CID hotspots cover extended areas of 100–200 kb and correlate with transcriptionally silent, intergenic domains. () A genome-wide karyotype view of ChIP-seq data representing the ratio of CID–GFP ChIP versus chromatin input on chromosome arms. Genomic regions are binned into 100 kb and shown as single bars. Values >0.2=blue, <0.2=orange (natural log scale). T symbols represent telomeres; circles represent centromeres. Density plots in grey indicate binding of HP1 (refs 33, 34, 35). Open triangles point to a subset of hotspots visualized in (blue bars) and filled triangles to top-ranking hotspots visualized in . () Regions of 200 kb bins were ranked according to the significance of the CID–GFP ChIP/input ratio and plotted 1/rank across the genome. Grey bars indicate pc-het extension. The top ranks for each chromosome arm (excluding the fourth) show a highly significant overlap with cytologically defined hotspots within 0–200 kb of the telomere (Fisher's exact test, P=0.0027). () qPCR was carried out to validate the ChIP-seq results using specific primer pairs for all candi! date hotspots (all triangles in ) and control regions with depleted CID–GFP–LacI. Values were normalized to a control region (Xdepl1; 100%) at day 0. Horizontal bars show the mean of the PCR amplicons. Differences of means (depleted versus hotspot regions) are highly significant for induced (Pday 0=1.8×10−6, Pday 5=1.3×10−8, Pday 6=5.4×10−8) but not non-induced cells (Pnon−ind=0.5), excluding the grey plus symbol, which probably represents an endogenous centromere sequence. U=unordered, unoriented chromosome, rep=unmapped hotspot (CTCTT-repeat associated). () CID–GFP ChIP/chromatin input ratio for the pc-het of chromosome 2R. Green brackets indicate the actual extent of CID hotspots identified in . Below, density plots of markers that correlate with transcriptionally active chromatin or silent chromatin. The multicolour in-between plot represents the nine-state model of prevalent chromatin states34 (light grey = state 9; all other states are detailed ! in Supplementary Fig. S4). () Similar to , a 1 Mb region at ! the left telomere of chromosome X is shown. * Figure 5: The heterochromatin–euchromatin transition zone is a hotspot for ectopic kinetochore formation and correlates with NDC80 binding before CID clearing. () A mitotic X chromosome stained for HP1, CID and H4ac is shown. The dashed line indicates the position of the pc-het–euchromatin boundary. () Example of a line profile (Metamorph) to show staining intensity along the length axis of the chromosome shown in . () Two mitotic X chromosomes are shown for day 4 (DAPI, CID) and day 0 (DAPI, CID, NDC80). () Line scans of X chromosomes as shown in were quantified for CID-island localization within similarly sized chromosome regions from 1 (left telomere) to 11 (endogenous centromere) in 20 cells. The fraction of cells with CID islands in a particular region was plotted as a frequency relative to the region containing the endogenous centromere (100%). The patterns of CID islands in the euchromatin arm (regions 1–6) at day 4 and NDC80 foci at day 0 are almost indistinguishable (χ2 test P>0.9), whereas their differences when compared with the NDC80 staining in uninduced cells are highly significant (χ2 test P<0.001; number of ch! romosomes: nCID islands D4=20, nNDC80 D0=80, nNDC80 unind=53). The green double arrow points to the position of the endogenous centromere. * Figure 6: Increased acetylation and high levels of HP1 inhibit whereas reduction of HP1 promotes the formation of CID islands. () Mitotic chromosome spreads of CID–GFP cells that were pulse induced in the presence of HDAC inhibitor (TSA) show increased acetylation levels. () Quantification of the percentage of cells with global mislocalization of CID and CID islands in the absence (untreated) and presence of TSA. Data are mean±s.e.m., cells analysed (untreated, +TSA): nmisloc=42, 30, nCID islands=65, 25. () Stably transfected S2 cells were pulse induced for CID–GFP alone (untreated for HP1), in combination with pulse-induced HP1–V5 or with reduced levels of HP1 (RNAi). Representative images of fixed mitotic chromosome spreads are shown, stained for DNA, CID and HP1. Note the remaining HP1 staining at telomeres (red arrows) and pc-het in the HP1 RNAi cells at day 0. Green arrows in the HP1 RNAi cells at day 4 indicate CID islands. Scale bar, 3 μm. () Western blot using cells induced for pulse overexpression of CID and HP1 probed with anti-HP1 antibody. The HP1–V5 levels at day 0 are est! imated to be fivefold overexpressed relative to wild-type HP1 levels (loaded left of the dashed line). α-tubulin serves as a loading control. () Western blot showing the reduction of HP1 levels achieved by RNAi (started day −7 relative to chase) from day −5 to day 8. The reduction is estimated to be 0.1- to 0.2-fold of wild-type HP1 levels. (,) Quantification of the percentage of cells with global mislocalization of CID () or cells containing CID islands () after CID pulse overexpression alone (control) or combined with HP1 RNAi or with HP1–V5 pulse co-expression. Data are mean±s.e.m., cells analysed (days 0, 4, 8): nCID pulse=23,52, 56, nCID pulse+HP1 RNAi=45, 51, 35, nCID+HP1 pulse=45,90,40. Uncropped images of blots are shown in Supplementary Fig. S6. * Figure 7: Local targeting of HP1–LacI induces a new hotspot for CID-island formation. (,) Mitotic chromosome spreads of S2 cells containing stably integrated LacOp repeats are subjected to pulse expression of CID–GFP alone (control; ) or in the presence of constitutively expressed HP1–LacI (). The red arrow points to the position of the LacOp sequences, as visualized by LacOp–FISH (red; ), or antibodies against the V5 epitope for HP1–LacI–V5 (red; ). Values represent the mean percentage of LacOp sites associated with CID islands±s.e.m; LacOp sites: ncontrol=63, nLacOp=37. Scale bar=3 μm. () The presence of HP1–LacI targeted to LacOp sequences correlates with increased frequency of double-stranded breaks at the LacOp indicative of mitotic defects that are significantly different from both control cell lines carrying only LacOp sequences and combined with GFP–LacI expression (χ2 test P<0.001). Red arrows: LacOp sequences. Scale bar=3 μm. () Quantification of chromosomal breakage at the LacOp site. Data are mean±s.e.m., cells analysed: nco! ntrol=60, nGFP−LacI=64, nHP1−LacI=16. () Schematic model summarizing that ectopic kinetochores are formed preferentially on transcriptionally silent, intergenic domains (light grey) at heterochromatin boundaries, lacking marks of both pc-het (dark blue) and transcriptionally active chromatin (red). We propose a dynamic boundary of heterochromatin on the basis of the cooperative binding of HP1 in repeat dense regions versus regions of higher complexity59. Individual cellular levels of heterochromatin proteins such as HP1 could lead to the heterochromatin boundary either spreading towards complex DNA (refractory for CID islands) or retreating towards repeat dense regions (permissive for CID islands), enabling ectopic kinetochore assembly. Increasing experimentally the levels of HP1 or of the active histone acetylation mark thereby 'seals' the silent domains, whereas reducing HP1 levels using RNAi uncovers these regions, thus promoting the formation of CID hotspots. Author information * Abstract * Author information * Supplementary information Affiliations * Max Planck Institute of Immunobiology and Epigenetics, Stübeweg 51, 79108 Freiburg, Germany * Agata M. Olszak, * Dominic van Essen, * Sarah Diehl, * Thomas Manke, * Simona Saccani & * Patrick Heun * Department of Experimental Biology, IBMC-Instituto de Biologia Molecular e Celular, Universidade do Porto, Porto, Portugal * António J. Pereira & * Helder Maiato * Department of Experimental Biology, Faculdade de Medicina, Universidade do Porto, Porto, Portugal * Helder Maiato * BIOSS Centre for Biological Signalling Studies, Albert-Ludwigs-Universität Freiburg, Freiburg, Germany * Patrick Heun Contributions A.M.O. carried out the cytological characterization of CID islands. A.J.P., A.M.O. and H.M. carried out laser-microsurgery experiments. D.v.E. and S.S. conducted the ChIP-seq and qPCR. T.M. and S.D. carried out bioinformatic analysis of ChIP-seq. A.M.O. and P.H. did the project planning and data analysis. P.H. wrote the manuscript. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Patrick Heun Author Details * Agata M. Olszak Search for this author in: * NPG journals * PubMed * Google Scholar * Dominic van Essen Search for this author in: * NPG journals * PubMed * Google Scholar * António J. Pereira Search for this author in: * NPG journals * PubMed * Google Scholar * Sarah Diehl Search for this author in: * NPG journals * PubMed * Google Scholar * Thomas Manke Search for this author in: * NPG journals * PubMed * Google Scholar * Helder Maiato Search for this author in: * NPG journals * PubMed * Google Scholar * Simona Saccani Search for this author in: * NPG journals * PubMed * Google Scholar * Patrick Heun Contact Patrick Heun Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Abstract * Author information * Supplementary information PDF files * Supplementary Information (1800K) Supplementary Information Additional data - A systematic RNAi synthetic interaction screen reveals a link between p53 and snoRNP assembly
- Nat Cell Biol 13(7):809-818 (2011)
Nature Cell Biology | Article A systematic RNAi synthetic interaction screen reveals a link between p53 and snoRNP assembly * Dragomir B. Krastev1, 2 * Mikolaj Slabicki2 * Maciej Paszkowski-Rogacz1, 2 * Nina C. Hubner3, 4 * Magno Junqueira2, 5 * Andrej Shevchenko2 * Matthias Mann3 * Karla M. Neugebauer2 * Frank Buchholz1, 2 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:809–818Year published:(2011)DOI:doi:10.1038/ncb2264Received26 October 2010Accepted20 April 2011Published online05 June 2011 Abstract * Abstract * Author information * Supplementary information Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg TP53(tumour protein 53) is one of the most frequently mutated genes in human cancer and its role during cellular transformation has been studied extensively. However, the homeostatic functions of p53 are less well understood. Here, we explore the molecular dependency network of TP53 through an RNAi-mediated synthetic interaction screen employing two HCT116 isogenic cell lines and a genome-scale endoribonuclease-prepared short interfering RNA library. We identify a variety of TP53 synthetic interactions unmasking the complex connections of p53 to cellular physiology and growth control. Molecular dissection of the TP53 synthetic interaction with UNRIP indicates an enhanced dependency of TP53-negative cells on small nucleolar ribonucleoprotein (snoRNP) assembly. This dependency is mediated by the snoRNP chaperone gene NOLC1 (also known as NOPP140), which we identify as a physiological p53 target gene. This unanticipated function of TP53 in snoRNP assembly highlights the potenti! al of RNAi-mediated synthetic interaction screens to dissect molecular pathways of tumour suppressor genes. View full text Figures at a glance * Figure 1: Genome-scale TP53 synthetic interaction screen. () Schematic representation of the screen assay. Equal numbers of HCT116 wild-type (red) and knockout (green) cells were reverse transfected with esiRNAs. Possible scenarios and the number of knockdowns causing indicated phenotypes are shown. Screen controls are shown in Supplementary Fig. S1a–c. () Representative images from transfections with esiRNAs of the non-targeting luciferase control (Luc), the kinesin family member 11 (KIF11), the p53 regulator HDMX and the inner centromere protein antigens 135/155 kDa (INCENP) exemplifying knockdown phenotypes (scale bars, 100 μm). () Statistical evaluation of the screening results. A histogram of the log2(wildtype/knockout) calculated in a plate-wise manner is shown. The Gaussian fit to the distribution (red line) underlines the need to use an asymmetry- and outlier-insensitive quartile (Q)-based threshold (dashed lines, see Methods) with a targeted error rate of 0.05 to determine the primary hits. * Figure 2: Depletion phenotypes of knockdowns increasing the wild-type/knockout ratio. () RNAi phenotype of the three genes decreasing the number of knockout cells—UNRIP, MASTL and KIAA1344. The cell number for each cell type is presented as a percentage of the total cells counted per well (left y axis) and the corresponding z score (right y axis, calculated versus Luc knockdown, see Methods). The knockdown of HDMX is shown as a positive control. Two esiRNAs (1 and 2) are shown for each hit gene. () Time-lapse analysis of the RNAi phenotypes. Wild-type and knockout cells transfected with the indicated esiRNAs were imaged for a period of 72 h, starting at 24 h post-transfection. The doubling time was calculated from an exponential fit of the growth curves (values next to the curves). (,) Quantification of UNRIP knockdown. mRNA and protein depletion levels normalized to luciferase-depleted cells. GAPDH or α -tubulin served as an internal loading control (mean±s.d., n=3). () Long-term survival (colony formation assay) of HCT116 and RKO cells on UNRIP depl! etion. The quantification presents the number of colonies relative to control transfection. The significance is determined with Student's two-tailed t -test, * P values<0.05, ** P values<0.01 (mean±s.d., n=3). The depletion phenotypes are further characterized in Supplementary Fig. S2. In all panels, red and green represent wild-type and knockout cells, respectively. Uncropped images of blots are shown in Supplementary Fig. S7a. * Figure 3: SMN complex localization on UNRIP depletion. () Immunofluorescence microscopy analysis of the cellular localization of SMN1, Gemin2 and COIL in wild-type and knockout cells on UNRIP depletion. Cells were stained with DAPI (4,6-diamidino-2-phenylindole; blue), α -tubulin antibody (green) and antibodies against the indicated proteins (red) for illustrated cells and knockdowns. Note that SMN1 and Gemin2 Cajal body (white arrows) localization is affected by UNRIP depletion only in knockout cells. Scale bars, 5 μm. () Quantification of the number of SMN1-positive Cajal bodies per nucleus. The ball size corresponds to the percentage of cells with a particular number of foci (numerically shown on the right of each ball). The distributions of indicated cells and knockdowns were compared with Student's two-tailed t -test, * P values<0.01, ** P values<10−5, *** P values<10−10, 200–300 nuclei counted per condition in three independent experiments. Quantification of Gemin2 and COIL foci is provided in Supplementary Fig! . S3a,b. () SMN1 nucleo-cytoplasmic distribution on UNRIP depletion. Western blot analyses of indicated proteins in cytoplasmic and nuclear extracts of control (Luc) and UNRIP-depleted wild-type and knockout cells are shown. Probing with anti-tubulin, anti-histone 3 and anti-SmB (small nuclear ribonucleoprotein polypeptides B and B1) revealed the purity of the fractions. The right panel represents quantification of SMN1 in each fraction, normalized to the amount of proliferating cell nuclear antigen (PCNA) as the loading control. Uncropped images of blots are shown in Supplementary Fig. S7b. * Figure 4: NOLC1 differentially associates with UNRIP in wild-type and knockout cells and NOLC1 is a p53 target gene. () Comparative quantitative mass spectrometry of UNRIP protein complexes purified from wild-type and knockout cells. A scatter plot of the correlation of two independent pulldown experiments showing the heavy/light ratio of each identified protein purified from each cell type is illustrated. Immunoprecipitated proteins significantly over-presented (COIL) or under-presented (NOLC1) in the knockout versus the wild-type are highlighted (black dots). () Expression analysis of COIL and NOLC1 mRNA. qRT–PCR results for COIL and NOLC1 mRNA levels in HCT116 and RKO wild-type and knockout cells are shown. The values are normalized to the expression levels in the wild-type cells (mean±s.d., n=3). () Western blot analysis of COIL, NOLC1 and UNRIP–LAP in HCT116 and RKO wild-type and knockout cells. () Immunofluorescence microscopy analysis of NOLC1 (red) and COIL (green) in wild-type and knockout cells. The Cajal bodies are indicated with white arrows. Scale bars, 5 μm. () p53 Ch! IP. Wild-type and knockout cells were treated with DMSO or 5 μM Nutlin- 3α for 4 h, p53 was immunoprecipitated (DO-1 antibody) and the enrichment of selected promoter sequences was evaluated using quantitative PCR. The fold enrichment is calculated as a ratio of the percentage inputs from wild-type versus knockout samples. GAPDH was used to normalize this ratio. Shown is a representative experiment (n=3). () NOLC1 mRNA transcript levels are p53-dependent. HCT116 cells were treated with 5 μM Nutlin- 3α for 4 h and the expression of p21/WAF1 and NOLC1 mRNA was analysed by qRT–PCR. The values are normalized to DMSO-treated wild-type cells (mean±s.d., n=3). The response to DNA-damaging agents is presented in Supplementary Fig. S4c. Uncropped images of blots are shown in Supplementary Fig. S7c. * Figure 5: rRNA pseudouridylation is dependent on p53 and UNRIP. () snoRNP assembly assay. UNRIP-depleted HCT116 wild-type and knockout cells were transfected with a plasmid expressing GFP–NAF1. After 24 h, NAF1 was immunoprecipitated and the amount of associated pre-snoRNAs was investigated by qRT–PCR. The black bars represent mean enrichment values; significance was tested with Student's two-tailed t -test, * P<0.05, three independent experiments. () CMCT assay for pseudouridine (ψ) detection. After CMCT modification of total cellular RNA, primers were hybridized to targeted regions of the rRNA and reverse-transcriptase reactions determined the modified pseudouridine sites (black ball). Equal amounts of RNA were loaded in all reactions. The greyscale intensities are colour-coded for indicated treatments. (,) Quantification of the amount of pseudouridylation at 24 sites in 18S and 28S rRNA on a per site () or per treatment () basis. The colour-coding represents the knockout/wild-type ratio (blue, decreased; red, increased) of th! e intensities of the pseudouridine sites for indicated treatments, with the ratio of the Luc control for each site set to 1. The sites are clustered according to the primer with which they were detected (18S_1, green; 18S_2, purple; 28S_1, red; 28S_2, blue). The box plot shows the averaged values of knockout/wild-type pseudouridine ratios of the 24 tested sites for the indicated treatments. The whiskers denote the 10th and 90th percentiles of the values, respectively. The significance was tested with Student's paired two-tailed t -test, *** P<0.001,n=24, three independent experiments. () Model summarizing the synthetic interaction between TP53 and UNRIP. UNRIP is required to bring the SMN complex into the Cajal bodies. TP53 regulates the levels of NOLC1, which brings the immature snoRNPs into the Cajal bodies, where on interaction with COIL and SMN the mature snoRNPs are assembled. Thus, absence of TP53 synergizes with decreased UNRIP levels to perturb snoRNP assembly and! in turn this decreases the level of pseudouridylation of rRNA. Author information * Abstract * Author information * Supplementary information Affiliations * University of Technology Dresden, University Hospital and Medical Faculty Carl Gustav Carus, Department of Medical Systems Biology, Fetscherstraße 74, D-01307 Dresden, Germany * Dragomir B. Krastev, * Maciej Paszkowski-Rogacz & * Frank Buchholz * Max Planck Institute of Molecular Cell Biology and Genetics, Pfotenhauerstrasse 108, D-01307 Dresden, Germany * Dragomir B. Krastev, * Mikolaj Slabicki, * Maciej Paszkowski-Rogacz, * Magno Junqueira, * Andrej Shevchenko, * Karla M. Neugebauer & * Frank Buchholz * Max Planck Institute of Biochemistry, Am Klopferspitz 18, 82152 Martinsried, Germany * Nina C. Hubner & * Matthias Mann * Present address: Universitair Medisch Centrum Utrecht, 3508 AB Utrecht, The Netherlands * Nina C. Hubner * Present address: Brazilian Center for Protein Research, Department of Cell Biology, University of Brasilia, 70910-900 Brasilia, DF, Brazil * Magno Junqueira Contributions D.B.K., M.S., N.C.H., M.J. and K.M.N. carried out experiments, M.P-R. analysed data, A.S., M.M. and F.B. planned the project and D.B.K. and F.B. wrote the manuscript. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Frank Buchholz Author Details * Dragomir B. Krastev Search for this author in: * NPG journals * PubMed * Google Scholar * Mikolaj Slabicki Search for this author in: * NPG journals * PubMed * Google Scholar * Maciej Paszkowski-Rogacz Search for this author in: * NPG journals * PubMed * Google Scholar * Nina C. Hubner Search for this author in: * NPG journals * PubMed * Google Scholar * Magno Junqueira Search for this author in: * NPG journals * PubMed * Google Scholar * Andrej Shevchenko Search for this author in: * NPG journals * PubMed * Google Scholar * Matthias Mann Search for this author in: * NPG journals * PubMed * Google Scholar * Karla M. Neugebauer Search for this author in: * NPG journals * PubMed * Google Scholar * Frank Buchholz Contact Frank Buchholz Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Abstract * Author information * Supplementary information Excel files * Supplementary Information (3700K) Supplementary Table 1 * Supplementary Information (43K) Supplementary Table 2 * Supplementary Information (25K) Supplementary Table 3 * Supplementary Information (25K) Supplementary Table 4 PDF files * Supplementary Information (1600K) Supplementary Information Additional data - RasGRF suppresses Cdc42-mediated tumour cell movement, cytoskeletal dynamics and transformation
- Nat Cell Biol 13(7):819-826 (2011)
Nature Cell Biology | Article RasGRF suppresses Cdc42-mediated tumour cell movement, cytoskeletal dynamics and transformation * Fernando Calvo1, 2 * Victoria Sanz-Moreno3, 4 * Lorena Agudo-Ibáñez1 * Fredrik Wallberg3 * Erik Sahai2 * Christopher J. Marshall3 * Piero Crespo1 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:819–826Year published:(2011)DOI:doi:10.1038/ncb2271Received17 January 2011Accepted28 April 2011Published online19 June 2011 Abstract * Abstract * Author information * Supplementary information Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg Individual tumour cells move in three-dimensional environments with either a rounded or an elongated 'mesenchymal' morphology. These two modes of movement are tightly regulated by Rho family GTPases: elongated movement requires activation of Rac1, whereas rounded/amoeboid movement engages specific Cdc42 and Rho signalling pathways. In siRNA screens targeting the genes encoding guanine nucleotide exchange factors (GEFs), we found that the Ras GEF RasGRF2 regulates conversion between elongated- and rounded-type movement. RasGRF2 suppresses rounded movement by inhibiting the activation of Cdc42 independently of its capacity to activate Ras. RasGRF2 and RasGRF1 directly bind to Cdc42, outcompeting Cdc42 GEFs, thereby preventing Cdc42 activation. By this mechanism, RasGRFs regulate other Cdc42-mediated cellular processes such as the formation of actin spikes, transformation and invasion in vitro and in vivo. These results demonstrate a role for RasGRF GEFs as negative regulat! ors of Cdc42 activation. View full text Figures at a glance * Figure 1: Depletion of RasGRF2 affects cell morphology and movement in a Rac/Rho-independent manner. (,) Bottom left, histograms showing the percentage of elongated A375M2 () and A375P () cells, after transfection of three siRNAs against RasGRF2 (no. 7–10). (800 cells per experiment; n=4 in A375M2 and n=5 in A375P; error bars indicate s.e.m; *P<0.05, **P<0.01 and ***P<0.001 by Student's t-test.) Top left, immunoblots, representative of 4 or 5 independent experiments, showing the efficiency of RasGRF2 downregulation. Positive control cells (C+) were transfected with a cDNA for RasGRF2. Right, histograms (top) showing the increase in phospho-MLC2 after transfection of two siRNAs against RasGRF2 (no. 7, 10) relative to control (C) cells as quantified from western blots (representative blot shown at the bottom). (Error bars indicate s.e.m.; *P<0.05 and **P<0.01 by Student's t-test.) Middle, images of mock- or RasGRF2-siRNA-transfected A375M2 cells plated on a thick layer of collagen. Scale bars, 100 μm in low-magnification images and 20 μm in high-magnification imag! es. () Overexpression of RasGRF2 potentiates elongated morphology. A375M2 cells were stably transfected with vector or RasGRF2 wild type, ΔCdc25 and ΔDH mutants. RI, vector-transfected cells treated with Y27632. (800 cells per experiment; n=4; error bars indicate s.e.m.; *P<0.05, **P<0.01, ***P<0.001 and ns P>0.05 by Student's t-test.) Western blot confirms expression of indicated proteins. () Rac1 and Rho activation levels in A375M2 and HeLa cells after depletion (siRNA, si) or overexpression of RasGRF2, in cells exponentially growing or serum-starved for 24 h. Uncropped images of blots are shown in Supplementary Fig. S8. * Figure 2: RasGRF GEFs negatively regulate Cdc42 activation. () Cdc42 activation levels in A375M2, Jurkat and HeLa cells after depletion (siRNA, si) or overexpression of RasGRF2, in exponentially growing cells or in cells starved for 24 h. () Pak1 phosphorylation levels in A365M2 cells under the same conditions as in . (Error bars indicate s.e.m; n=5; **P<0.01 by Student's t-test, relative to the values in the respective controls (C).) Representative western blot used for quantification is shown. () Depletion of RasGRF2 increases Cdc42 activation under basal conditions and in response to some stimuli. HeLa cells, control or RasGRF2 depleted (siRNA, si), were serum-starved (St.) and stimulated with the indicated agonists. Iono., ionomycin. (,) RasGRF GEFs inhibit Cdc42 and Pak1 but not RhoA or Rac1 activation. COS-7 cells were transfected with HA-tagged GTPases in addition to Ras GEFs, plus vector (Vec.) or the indicated Rho family GEFs onco-Dbl, Ost, Dock10DHR2 domain and Vav1. Graph shows fold increase in phosphorylated PAK1 leve! ls as quantified from western blot performed as shown above the graph (error bars indicate s.e.m; n=3; *P<0.05, ***P<0.001 and ns P>0.05 by Student's t -test relative to the levels of Dbl-transfected cells). Uncropped images of blots are shown in Supplementary Fig. S8. * Figure 3: RasGRF GEFs associate with Cdc42 in vivo. () RasGRF and Cdc42 form a complex. Lysates from cell lines and mouse brain homogenates (Mm brain hom.) were immunoprecipitated for RasGRF (top panels) or Cdc42 (bottom panels). Immunoprecipitates (IP) and corresponding total lysates (TL) were probed by immunoblotting. Control immunoprecipitation using pre-immune serum (PI) was included. () RasGRF2 requires a full DH domain for binding to Cdc42. COS-7 cells were transfected with AU5–Cdc42N17 plus Flag-tagged RasGRF2 domain deletion mutants (see Supplementary Fig. S2a). Cdc42 was immunoprecipitated with anti-AU5 antibody and associated proteins were revealed by immunoblotting. () The RasGRF DH domain is sufficient to bind to Cdc42. COS-7 cells were transfected with HA–Cdc42N17 plus GFP, GFP-tagged RasGRF1 and 2, or their corresponding DH domains DH(1) or DH(2). Proteins bound to immunoprecipitated Cdc42 were revealed by immunoblotting. () The RasGRF DH but not the Cdc25 domain is required to inhibit Cdc42 activation. Cdc4! 2 activation was analysed in cells transfected with HA–Cdc42 and onco-Dbl, plus RasGRF2, and its ΔCdc25 and ΔDH mutants. () Cdc42 regulation by RasGRF2 is not related to its Ras GEF or Rac GEF activities. Cdc42–, Ras–, Rac1– and RhoA–GTP levels were determined in HeLa cells transiently transfected with siRNAs: control (C), RasGRF2 (GRF2 siRNA) or Flag-tagged RasGRF2 and its ΔDH or ΔCdc25 mutants. () RasGRFs outcompete Cdc42 GEFs for binding with Cdc42. COS-7 cells were transfected with AU5–Cdc42N17 plus Dbl in addition to Ras GEFs. Cdc42 was anti-AU5 immunoprecipitated and associated proteins were revealed by immunoblotting. () As in , RasGRFs outcompete Dock10 for binding to Cdc42. () The RasGRF DH but not the Cdc25 domain is required to inhibit Dbl/Cdc42 association. Cells were transfected with AU5–Cdc42N17 and onco-Dbl, plus RasGRF2 or its Cdc25 and DH deletion mutants. Proteins bound to immunoprecipitated Cdc42 were revealed by immunoblotting. Vec., ve! ctor-transfected cells. Uncropped images of blots are shown in! Supplementary Fig. S8. * Figure 4: RasGRF GEFs regulate Cdc42-mediated transformation and filopodia formation in NIH3T3 cells. () RasGRF expression blocks Cdc42-mediated transformation. NIH3T3 cells were transfected with onco-Dbl plus the indicated cDNAs. The bars represent the number of transformed foci, mean ± s.d. of five independent experiments, expressed as a percentage relative to those induced by onco-Dbl alone. (***P<0.001, **P<0.01, *P<0.05 and ns P>0.05 by Student's t -test.) () RasGRFs impair Dock10-induced filopodia formation. Representative confocal micrographs of NIH3T3 cells transfected with Dock10DHR2–GFP plus the indicated constructs. Cells were immunostained to monitor the expression of RasGRFs (blue) and stained with phalloidin (red) to mark filamentous actin. The symbols in the bottom left illustrate the confocal plane shown. Scale bar, 10 μm. * Figure 5: Endogenous RasGRF2 regulates Cdc42-mediated filopodia formation in HeLa cells () Downregulation of endogenous RasGRF2 levels affects filopodia formation in HeLa cells. Top, representative confocal micrographs of HeLa cells co-transfected with pEGFP plus RasGRF2 siRNA, Cdc42 shRNA or Cdc42N17 and transfected with GFP-tagged versions of RasGRF1 and RasGRF2. Cells were starved and stained with phalloidin (red). Scale bar, 20 μm. Bottom, magnifications (×5) of the indicated areas. Filopodia numbers were compared between transfected versus untransfected cells in the same preparation. () Immunoblots showing the specificity of Cdc42 knockdown. Uncropped images of blots are shown in Supplementary Fig. S8. * Figure 6: RasGRF2 regulates invasion and lung colonization by melanoma cells. () Knockdown of RasGRF2 promotes collagen I matrix invasion. Invasion index (number of invading cells at 50 μm/number of cells at 3 μm) of the indicated melanoma cell lines transfected with scrambled control (mock) or with two siRNAs against RasGRF2 (no. 7, 10). (n=4; error bars indicate s.e.m.; **P<0.01 and *P<0.05 by Student's t-test.) () Percentage of rounded GFP–WM266.4 cells invading into collagen I matrix after RasGRF2 depletion. (n=3; more than 1,000 cells were imaged per condition; **P<0.01, ***P<0.001 by Student's t-test.) () Images of GFP–WM266.4 cells transfected with scrambled siRNAs (mock) or siRNA no. 7 against RasGRF2 after 24 h of invasion into collagen I. Scale bar, 100 μm. () Knockdown of RasGRF2 promotes lung colonization by melanoma cells. Left, representative images of mouse lungs at 2 and 20 h after tail injection with WM266.4 cells transfected with mRFP–control siRNA (control) together with GFP–control siRNA (mock) or GFP–GRF! 2 siRNA. Scale bars, 50 mm. Top right, graph showing the relative proportions of mock- and GRF2-siRNA-transfected cells within the lungs. Each point represents the ratio of experimental (mock or GRF2-depleted cells) versus the control cells of a pair of lungs. Student's t-test was used to generate the P value. Bottom right, RasGRF2 knockdown efficiency was evaluated by western blotting. () As in , but using WM1361 cells. Uncropped images of blots are shown in Supplementary Fig. S8. Author information * Abstract * Author information * Supplementary information Affiliations * Instituto de Biomedicina y Biotecnologı´a de Cantabria (IBBTEC), Consejo Superior de Investigaciones Cientı´ficas (CSIC) — IDICAN — Universidad de Cantabria. Departamento de Biologı´a Molecular, Facultad de Medicina, Santander, 39011, Cantabria, Spain * Fernando Calvo, * Lorena Agudo-Ibáñez & * Piero Crespo * Tumour Cell Biology Laboratory, Cancer Research UK, London Research Institute, 44 Lincoln's Inn Fields, London WC2A 3PX, UK * Fernando Calvo & * Erik Sahai * Section of Cell and Molecular Biology, CR-UK Tumour Cell Signalling Unit, Institute of Cancer Research, 237 Fulham Road, London SW3 6JB, UK * Victoria Sanz-Moreno, * Fredrik Wallberg & * Christopher J. Marshall * Present address: Randall Division of Cell and Molecular Biophysics, New Hunt's House, King's College London, Guy's Campus, London SE1 1UL, UK * Victoria Sanz-Moreno Contributions F.C. carried out all of the experiments with the exception of those shown in Figs 1 and 6, which were carried out by F.C. and V.S-M. V.S-M. and F.W. carried out the invasion assays and imaging of the invading cells. L.A-I. carried out the in vitro GTPase binding experiments and the soft-agar colonies. F.C. and V.S-M. also prepared the figures and carried out the statistical analyses. E.S. contributed in the lung colonization assays. F.C., V.S-M., C.J.M. and P.C. conceived the study and C.J.M. and P.C. wrote the manuscript. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Piero Crespo Author Details * Fernando Calvo Search for this author in: * NPG journals * PubMed * Google Scholar * Victoria Sanz-Moreno Search for this author in: * NPG journals * PubMed * Google Scholar * Lorena Agudo-Ibáñez Search for this author in: * NPG journals * PubMed * Google Scholar * Fredrik Wallberg Search for this author in: * NPG journals * PubMed * Google Scholar * Erik Sahai Search for this author in: * NPG journals * PubMed * Google Scholar * Christopher J. Marshall Search for this author in: * NPG journals * PubMed * Google Scholar * Piero Crespo Contact Piero Crespo Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Abstract * Author information * Supplementary information Movies * Supplementary Information (2M) Supplementary Movie 1 * Supplementary Information (2M) Supplementary Movie 2 * Supplementary Information (2M) Supplementary Movie 3 * Supplementary Information (2M) Supplementary Movie 4 * Supplementary Information (2M) Supplementary Movie 5 * Supplementary Information (2M) Supplementary Movie 6 PDF files * Supplementary Information (1700K) Supplementary Information Additional data - A direct role for Met endocytosis in tumorigenesis
- Nat Cell Biol 13(7):827-837 (2011)
Nature Cell Biology | Article A direct role for Met endocytosis in tumorigenesis * Carine Joffre1, 4 * Rachel Barrow1 * Ludovic Ménard1 * Véronique Calleja2 * Ian R. Hart3 * Stéphanie Kermorgant1 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:827–837Year published:(2011)DOI:doi:10.1038/ncb2257Received01 February 2011Accepted07 April 2011Published online05 June 2011 Abstract * Abstract * Author information * Supplementary information Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg Compartmentalization of signals generated by receptor tyrosine kinase (RTK) endocytosis has emerged as a major determinant of various cell functions. Here, using tumour-associated Met-activating mutations, we demonstrate a direct link between endocytosis and tumorigenicity. Met mutants exhibit increased endocytosis/recycling activity and decreased levels of degradation, leading to accumulation on endosomes, activation of the GTPase Rac1, loss of actin stress fibres and increased levels of cell migration. Blocking endocytosis inhibited mutants' anchorage-independent growth, in vivo tumorigenesis and metastasis while maintaining their activation. One mutant resistant to inhibition by a Met-specific tyrosine kinase inhibitor was sensitive to endocytosis inhibition. Thus, oncogenicity of Met mutants results not only from activation but also from their altered endocytic trafficking, indicating that endosomal signalling may be a crucial mechanism regulating RTK-dependent tumorig! enesis. View full text Figures at a glance * Figure 1: Active Met mutants accumulate in intracellular compartments. () Western blots for Met, phosphorylated Met (Tyr 1234/1235 and 1349) and the constitutive heat shock protein 70 (HSC70). Numbers represent mean ± s.e.m. (arbitrary units, n=6) of Met p-145/HSC70 ratios. () P-Met-Y1349/Met ratios identified by densitometric analysis of western blots (not shown) in the indicated cells stimulated, or not, with HGF for 15 min. Data are mean (arbitrary units) ± s.e.m. (n=3). #, *P<0.05; * compared with wild type, no HGF. () The indicated cells were treated with the Met inhibitor PHA-665752 (PHA, 100 nM) or DMSO. The amount of phosphorylated Met (phosphorylated at Y1234/1235) was calculated as a percentage of the total Met, as assessed from densitometric analysis of western blots (shown in Supplementary Fig. S1c). Data are mean (arbitrary units) ± s.e.m. (n=3). *P<0.05 . () Confocal sections of cells, stimulated, or not, with HGF and stained for Met (green) and for phospho-tyrosine clone 4G10 (red), and with DAPI (blue). Arrows show pl! asma membrane Met. Scale bars, 10 μm. (,) Proportion of intracellular pools versus total cellular Met as determined from a biotin surface removal assay. () Cells were surface-biotinylated and the surface protein fraction was removed by streptavidin pulldown. The total sample before pulldown (total), the supernatant corresponding to the intracellular fraction (unbound; outlined in red) and the surface fractions (bound) were analysed by western blotting for Met (left panels) and phosphorylated Met (Tyr 1234/1235; right panels). () The percentage of intracellular Met was calculated as a ratio to the total. Data are mean ± s.e.m. (n=3). *P<0.05; **P<0.01. Uncropped images of blots are shown in Supplementary Fig. S9. * Figure 2: Met mutants shuttle between the plasma membrane and endosomes and are protected from degradation. (,) Cells were pretreated with cycloheximide for 4 h. () Confocal sections of cells stimulated, or not, with HGF and stained for Met (green) and for EEA1 (red), and with DAPI (blue). Scale bars, 10 μm. () Cells were incubated with or without Cy3–transferrin for 30 min before fixation and stained for Met and EEA1 or Rab11 and stained with DAPI (Supplementary Fig. S2a,c). Percentage of co-localization between Met and EEA1 or transferrin (mean ± s.e.m., n=3) or Rab11 (mean, n=2) (left graph) and percentage of cells that were positive for perinuclear Met staining (mean ± s.e.m., n=3, right graph). () Biotinylation internalization/recycling assay. In the indicated cells, the levels of surface biotinylated Met internalized (15 min) and recycled (15 min) were measured. Experiments were carried out in duplicate. Left, Met western blots after streptavidin pulldown. Right, quantification of Met internalization and recycling through densitometric analysis of bands in! western blots. Data are mean ± s.e.m. (n=4, left graph) or mean (n=1, right graph). () Expression of Met at the plasma membrane evaluated by flow cytometry in the indicated cells transfected with wild-type GFP–Rab11 or the dominant-negative mutant GFP–Rab11S25N. Percentage of Met downregulation at the plasma membrane in transfected cells, compared with non-transfected cells (GFP-negative cells; mean ± s.e.m., n=3). () Met/HSC70 ratios obtained by densitometric analysis of western blots and plotted as a percentage of the initial content. The indicated cells were pretreated with cycloheximide and stimulated with HGF for the indicated times (mean ± s.e.m., n=3). Statistical results are shown for 8 h versus 0 h. () Biotinylation degradation assay. The indicated cells were surface biotinylated and incubated at 37 °C. At the indicated times, cells were lysed and the remaining biotinylated Met, after streptavidin pulldown, was analysed by western blotting for ! Met. () Biotinylation internalization assay. Cells expressing ! MetM1268T were pretreated (+), or not (−), with PHA for 30 min, surface biotinylated and incubated at 37 °C for 15 min with or without PHA. Top, Met or phospho-Met western blots after streptavidin pulldown. Bottom, the graph represents the percentage of internalization (mean ± s.e.m. n=3). TS, total surface; Int., internalized. *P<0.05; **P<0.01; NS, not significant. Uncropped images of blots are shown in Supplementary Fig. S9. * Figure 3: Endocytosis of Met mutant is dependent on clathrin, dynamin, c-Cbl and Grb2. () Biotinylation internalization assay with cells pretreated with DMSO (−) or 80 μM dynasore (+) for 30 min. () Confocal sections of cells stained for Met (green) and DAPI (blue). Arrows show plasma membrane Met. Cells expressing MetM1268T were transfected with control or CHC RNAi. Scale bars, 10 μm. (,) M1268T Met mutant expressing cells were transfected with control, c-Cbl or Grb2 RNAi. () Left, biotinylation internalization assay. Each experiment was done in duplicate. Right, data are mean ± s.e.m.for control and Grb2 RNAi (n=3) and mean for Cbl RNAi (n=1). () c-Cbl, Grb2 and tubulin western blots. () Grb2 immunoprecipitation from cells expressing wild-type Met and MetM1268T. Met and Grb2 western blots from immunoprecipitates and from total lysates (TL). Controls include incubation with beads alone or coupled to an isotype-matched IgG. () Left, confocal sections of cells expressing human Met constructs, hWt, hM1268T and hM1268T/N1358H, pretreated with cyclohe! ximide for 4 h and stained for Met (green), EEA1 (red) and DAPI (blue). Scale bars, 10 μm. Right, percentage of co-localization (mean ± s.e.m., n=3). #, *P<0.05. Uncropped images of blots are shown in Supplementary Fig. S9. * Figure 4: Met mutants require endocytosis to control actin cytoskeleton remodelling. (, and -) Indicated cells were stained with Cy3–phalloidin (red) and DAPI (blue). (,) The indicated cells were treated with DMSO or PHA for 90 min or transfected with control or Met RNAi. () Confocal sections. Scale bars, 10 μm. () Percentage of cells lacking stress fibres. Data are mean ± s.e.m. (n=3). Control is the average of DMSO (n=3) and RNAi control (n=3). () Western blots showing Met and tubulin expression after control or Met RNAi transfections in the indicated cells. () Confocal sections of D1246N and M1268T cells treated with dynasore for 90 min. Scale bars, 10 μm. () Percentage of MetM1268T-expressing cells lacking stress fibres when transfected with dynamin-2K44A–GFP (n=3), or transduced with CHC shRNA (n=3) or transfected with c-Cbl RNAi (n=4), Grb2 RNAi (n=3), Grb2-89A–myc (n=3) or Grb2-49L/203R–myc (n=3). Data are normalized to appropriate controls (GFP, shRNA control, RNAi control, myc–Grb2-Wt) and are mean ± s.e.m. () Percentage of! cells expressing human Met forms, hWt, hM1268T and hM1268T/N1358H, that lack stress fibres. hM1268T cells were treated, or not, with PHA for 90 min. Data are mean ± s.e.m. (n=3). #, *P<0.05; ##, **P<0.01; ###, ***P<0.001; NS: not significant. Uncropped images of blots are shown in Supplementary Fig. S9. * Figure 5: Met mutants induce Rac1 activation, which is dependent on endocytosis. () Western blots showing the levels of Rac1–GTP, measured by GST–CRIB pulldown, and total Rac1 (see Methods). () Percentage of wild-type and M1268T cells lacking stress fibres when transfected with control or Rac1 RNAi. Data are mean ± s.e.m. (n=3). () Confocal sections of the indicated cells treated with DMSO or PHA for 60 min and stained for Rac1 (red) and EEA1 (green) and with DAPI (blue). Arrows indicate Rac1 at membrane protrusions. Scale bars, 10 μm. () Percentage of the indicated cells with marked Rac1 staining present at membrane protrusions when transfected with control, Met, c-Cbl or Grb2 RNAi. Data are mean ± s.e.m. (n=3). () Western blots showing the levels of CHC, Rac1–GTP (measured by GST–CRIB pulldown), total Rac1 and HSC70 in MetM1268T-expressing cells transfected with control or CHC RNAi. () The localization of active Rac1 in cells was determined by incubating fixed cells with purified GST–PAK-PBD (binds to Rac–GTP). Cells were further! stained for Met (blue), Rac1 (green) and GST (red). A representative confocal picture of MetM1268T-expressing cells is shown. Arrows indicate activated Rac1 at the plasma membrane. Right panels, Magnifications 1 and 2 correspond to cellular location number 1 and 2 in the merged picture of cells treated with GST–PAK-PBD (upper right picture). Plots of the intensity profiles, at different emission wavelengths corresponding to the signals of Met, Rac1 and GST, of a set of pixels distributed on a line drawn across vesicles (in the magnifications) are shown. Scale bars, 10 μm. () Percentage of cells, expressing human Met forms, hWt, hM1268T and hM1268T/N1358H, that exhibit Rac1 at membrane protrusions. hM1268T were treated, or not, with PHA for 90 min. Data are mean ± s.e.m. (n=3). #, * P<0.05; ##, **P<0.01; NS: not significant. Uncropped images of blots are shown in Supplementary Fig. S9. * Figure 6: Met mutants require endocytosis to stimulate cell migration. () Average number of the indicated cells that have migrated through Transwells over 2 h of incubation in the presence of DMSO or dynasore. Each experiment was done in triplicate. Data are mean ± s.e.m. (n=3). () Percentage of MetM1268T-expressing cells that migrated through Transwells when treated with PHA (n=2), transfected with Met RNAi (n=3), dynamin-2K44A–GFP (n=3) or CHC RNAi (n=3), transduced with shRNA CHC (n=2), transfected with c-Cbl (n=6) or Grb2 RNAi (n=3). The last column represents the percentage of cells expressing hM1268T/N1358H Met (n=3) that migrated. Data are normalized to appropriate controls (DMSO, GFP, shRNA control, RNAi control and cells expressing hM1268T Met, respectively). Each experiment was done in triplicate. Data are mean ± s.e.m. () Percentage of M1268T cells that have migrated through Transwells when treated with the Rac inhibitor NSC23766 or transfected with Rac1 RNAi over appropriate controls (DMSO and RNAi control, respectively). ! Each experiment was done in triplicate. Data are mean ± s.e.m. (n=3). () MetM1268T-expressing cells (500,000) transduced with control or CHC shRNA were injected intravenously in nude mice and, after 10 days, their lungs were analysed and weighed. Data are mean ± s.e.m. of n=5 mice. *P<0.05; ##, **P<0.01; ***P<0.001; NS: not significant. * Figure 7: Blocking endocytosis reduces in vitro tumour transformation stimulated by Met mutants. Cells were cultured in soft agar. Experiments were done in triplicate. () From day 5, DMSO or dynasore was added daily to the medium. Pictures at day 9. Numbers are the average total number of colonies per plate ± s.e.m. (n=3; see colony area Supplementary Fig. S7a). () The indicated cells were transfected with control, Met, c-Cbl or Grb2 RNAi. Data are the average colony areas ± s.e.m. (a.u., arbitrary units). Pictures (not shown) were taken at day 6 (n=3). () Data are the mean of the percentage of reduction ± s.e.m.in colony number formed by cells expressing MetM1268T when treated with dynasore or transfected with Met, c-Cbl or Grb2 RNAi versus their appropriate controls (DMSO, RNAi control; n=3). The last column represents cells expressing hM1268T/N1358H Met versus cells expressing hM1268T Met (mean ± s.e.m., n=3). *P<0.05; **P<0.01; ***P<0.001. * Figure 8: Blocking endocytosis reduces in vivo tumour transformation stimulated by Met mutants. The indicated cells (500,000) were injected subcutaneously into nude mice (minimum 5 mice per group in each experiment). () Tumour growth curves of the different cell lines over time. Data are mean ± s.e.m. of n=10 mice from two independent experiments. () Control DMSO diluents, PHA (100 nM) or dynasore (80 μM) was applied topically to the skin over tumours once they had achieved 50 mm3 size. Percentage of tumour volume (data are mean ± s.e.m. of n=10 mice from two independent experiments) derived from MetD1246N- and MetM1268T-expressing cells, after four and five days of treatment with PHA or dynasore relative to DMSO treatment (see Methods). () Percentage of tumour volume derived from MetM1268T-expressing cells transduced with control (mean ± s.e.m. of n=6 mice) or CHC shRNA (mean ± s.e.m. of n=7 mice), after 9 days post-graft. () PHA (100 nM), dynasore (80 μM) or control DMSO diluents were applied topically to the skin over tumours once they had ac! hieved 50 mm3 size. The bars represent the percentage of tumour volume derived from wild-type Met-expressing cells, after 4 and 5 days of treatment with PHA or dynasore relative to DMSO treatment. Data are mean ± s.e.m. of n=10 mice and are pooled from two independent experiments. () Confocal sections of tumour sections from wild-type and M1268T-expressing cells (from and ) stained for Met (green) and EEA1 (red), and with DAPI (blue). Arrow heads show plasma membrane Met. Scale bars, 10 μm. (,) Western blots for phospho-Met (Tyr 1234/1235), total Met and HSC70 carried out on tumour samples from three different mice for each condition (from or ; Dyn, dynasore). The numbers are the fold increases (mean ± s.e.m.) in the level of phosphorylated Met (P-Met/Met) for dynasore treatment versus DMSO () or for CHC shRNA cells versus control shRNA cells (). *P<0.05; **P<0.01; NS: not significant. Uncropped images of blots are shown in Supplementary Fig. S9. Author information * Abstract * Author information * Supplementary information Affiliations * Spatial Signalling Team, Centre for Tumour Biology, Barts Cancer Institute, Queen Mary University of London, John Vane Science Centre, Charterhouse Square, London EC1M 6BQ, UK * Carine Joffre, * Rachel Barrow, * Ludovic Ménard & * Stéphanie Kermorgant * Cell Biophysics Laboratory, Cancer Research UK, London Research Institute, 44 Lincoln's Inn Fields, London WC2A 3PX, UK * Véronique Calleja * Centre for Tumour Biology, Barts Cancer Institute, Queen Mary University of London, John Vane Science Centre, Charterhouse Square, London EC1M 6BQ, UK * Ian R. Hart * Present address: Institut Curie, INSERM U-830, 26 rue d'Ulm, 75248 Paris Cedex 05, France * Carine Joffre Contributions C.J. carried out and analysed most experiments using the NIH3T3 cells both in vitro and in vivo, including characterization of the cell lines, western blots, immunofluorescence analyses, Transwell migration, soft-agar assays, flow cytometry analysis, immunoprecipitation and most separation procedures investigating internalization, degradation and recycling. R.B. did most RNAi knockdowns and transfections followed by the subsequent migration assays, soft-agar assays, western blots and immunofluorescence analysis. R.B. carried out transfection and characterization of hM1268T and hM1268T/N1358H mutants in NIH3T3 cells. L.M. carried out Rac1 activation assays following Grb2 RNAi and PHA treatment, established the detection of active Rac and helped with immunofluorescence analyses and analysis of hM1268T and hM1268T/N1358H cells. V.C. designed and developed the hM1268T and hM1268T/N1358H mutants. S.K. conceived the project, designed experiments, interpreted the data and carried o! ut some of the immunofluorescence and confocal microscopy analyses. I.R.H. advised on the design of the in vivo experiments, trained C.J. and carried out inoculations of tumour cells. S.K. and C.J. wrote the manuscript, with additional input from I.R.H. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Stéphanie Kermorgant Author Details * Carine Joffre Search for this author in: * NPG journals * PubMed * Google Scholar * Rachel Barrow Search for this author in: * NPG journals * PubMed * Google Scholar * Ludovic Ménard Search for this author in: * NPG journals * PubMed * Google Scholar * Véronique Calleja Search for this author in: * NPG journals * PubMed * Google Scholar * Ian R. Hart Search for this author in: * NPG journals * PubMed * Google Scholar * Stéphanie Kermorgant Contact Stéphanie Kermorgant Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Abstract * Author information * Supplementary information Excel files * Supplementary Information (30K) Supplementary Table 1 PDF files * Supplementary Information (1200K) Supplementary Information Additional data - Inhibition of glycogen synthase kinase-3 alleviates Tcf3 repression of the pluripotency network and increases embryonic stem cell resistance to differentiation
- Nat Cell Biol 13(7):838-845 (2011)
Nature Cell Biology | Letter Inhibition of glycogen synthase kinase-3 alleviates Tcf3 repression of the pluripotency network and increases embryonic stem cell resistance to differentiation * Jason Wray1, 4 * Tüzer Kalkan1 * Sandra Gomez-Lopez1, 4 * Dominik Eckardt2, 4 * Andrew Cook3 * Rolf Kemler2 * Austin Smith1 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:838–845Year published:(2011)DOI:doi:10.1038/ncb2267Received10 January 2011Accepted20 April 2011Published online19 June 2011 Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg Self-renewal of rodent embryonic stem cells is enhanced by partial inhibition of glycogen synthase kinase-3 (Gsk3; refs 1, 2). This effect has variously been attributed to stimulation of Wnt signalling by β-catenin1, stabilization of Myc protein3 and global de-inhibition of anabolic processes4. Here we demonstrate that β-catenin is not necessary for embryonic stem cell identity or expansion, but its absence eliminates the self-renewal response to Gsk3 inhibition. Responsiveness is fully restored by truncated β-catenin lacking the carboxy-terminal transactivation domain5. However, requirement for Gsk3 inhibition is dictated by expression of T-cell factor 3 (Tcf3) and mediated by direct interaction with β-catenin. Tcf3 localizes to many pluripotency genes6 in embryonic stem cells. Our findings confirm that Tcf3 acts as a transcriptional repressor and reveal that β-catenin directly abrogates Tcf3 function. We conclude that Gsk3 inhibition stabilizes the embryonic stem cell! state primarily by reducing repressive influence on the core pluripotency network. View full text Figures at a glance * Figure 1: Suppression of Gsk3 mediates enhanced embryonic stem cell self-renewal but β-catenin is dispensable for embryonic stem cell maintenance. () Histogram showing the number of undifferentiated (alkaline phosphatase positive, AP+) colonies formed from 600 E14IVC embryonic stem cells plated in N2B27 with Mek inhibitor PD (1 μM) plus CH; or alternative Gsk3 inhibitors A, B, C, D, E, F and G (see Methods and Supplementary Fig. S1 for details). () Histogram showing the relative number of undifferentiated (alkaline phosphatase positive, AP+) colonies formed from 600 Gsk3α+/−,Gsk3β−/− (3/4KO) embryonic stem cells plated in N2B27 plus PD. Cells were untreated (−) or transfected with control siRNA or siRNAs against Gsk3α (α), Gsk3β (β) or both (αβ). Mean of two biological replicates. () Phase-contrast microscopy images showing the typical morphology of primary colonies isolated from GFP-negative (left) and -positive (right) fractions of βcatfl/− embryonic stem cells transiently transfected with Cre–IRES–GFP. Note the lack of cell–cell contacts in colonies from the GFP-positive fraction. Scale b! ar, 200 μm. () Phase-contrast and fluorescence micrographs showing immunostaining of βcatfl/− and βcatΔ/− embryonic stem cells for Oct4 and Nanog. DAPI, 4,6-diamidino-2-phenylindole. Scale bar, 100 μm. () Histogram showing gene expression in βcatfl/− and βcatΔ/− embryonic stem cells cultured in 2i+LIF relative to EpiSC-like cells derived by culture in activin+FGF2 (A+F) for seven passages. Klf4 and Klf2 are specific for naive embryonic stem cells, whereas Fgf5 is upregulated and Nanog downregulated in EpiSCs (ref. 28). Mean±s.d. of three biological replicates. * Figure 2: βcatΔ/− embryonic stem cells do not resist differentiation on Gsk3 inhibition. () Histogram showing the number of undifferentiated (alkaline phosphatase positive, AP+), colonies formed by parental βcatΔ/− or βcatΔ/− embryonic stem cells expressing a Ctnnb1 transgene (Res) or the corresponding vector control (Vec). Cells were plated in N2B27 plus 2i, PD+LIF (PL) or CH+LIF (CL). Data are expressed relative to the number of colonies in 2i+LIF. Mean±s.d. of three biological replicates. () Flow cytometry analysis of Rex1GFP expression in βcatfl/− (CRd2) or βcatΔ/− (CreC and CreD) Rex1GFP reporter embryonic stem cells cultured in 2i+LIF, PD+LIF (PL), CH+LIF (CL) or 2i for 96 h. () Histogram showing the number of undifferentiated (alkaline phosphatase positive, AP+), mixed and differentiated (AP−) colonies formed from 600 βcatfl/− or βcatΔ/− embryonic stem cells plated in 2i+LIF following 48 h culture in N2B27 alone (N) or plus 2i+LIF (2i+L), 2i, PD+LIF (PL) or CH+LIF (CL). () Flow cytometry analysis of Rex1GFP expression in βcat! fl/− (CRd2) or βcatΔ/− (CreC and CreD) Rex1GFP reporter embryonic stem cells cultured in N2B27 alone or plus 2i+LIF (2i+L) or CH for 48 h. () Histograms showing expression of pluripotency marker genes Nanog and Klf4 and early differentiation marker genes Fgf5 and Sox3 in βcatfl/− (CRd2) or βcatΔ/− (CreC and CreD) Rex1GFP reporter embryonic stem cells cultured in N2B27 alone (N) or plus CH for 48 h relative to 2i+LIF. Mean±s.d. of three biological replicates. * Figure 3: β-catenin inhibits differentiation independently of its transcriptional activation domain. () Western blot showing β-catenin and α-tubulin (loading control) expression in wild-type (E14), βcatfl/− or 'Rescue' βcatΔ/− embryonic stem cells expressing randomly integrated wild-type or C-terminal-deleted Ctnnb1 (ΔC) transgenes or the corresponding vector control (Vec). ΔC1 and ΔC2 are independent clones. Uncropped images of blots are shown in Supplementary Fig. S9. () Histogram showing TOPFlash and FOPFlash reporter activity in βcatΔ/− cells expressing randomly integrated wild-type or C-terminal-deleted (ΔC) Ctnnb1 transgenes. Cells cultured in PD+LIF with or without (NON) CH. ΔC1 and ΔC2 are independent clones. Mean±s.d. of three biological replicates. () Histogram showing the number of undifferentiated colonies formed from 600 βcatΔ/− cells expressing randomly integrated wild-type or C-terminal-deleted (ΔC) Ctnnb1 transgenes. Cells were plated in 2i+LIF after 48 h in N2B27 alone or plus CH. () Histograms showing expression of Nanog, Klf! 4 and Fgf5 in βcatΔ/− cells expressing randomly integrated wild-type or C-terminal-deleted Ctnnb1 (ΔC) transgenes. Cells were cultured in N2B27 alone or plus CH for 24 h. Expression is shown relative to levels in 2i+LIF or N2B27 alone (N). ΔC1 and ΔC2 are independent clones. () Histogram showing relative expression of Axin2 and Cdx1 in βcatΔ/− cells expressing randomly integrated wild-type or C-terminal-deleted (ΔC) Ctnnb1 transgenes. Cells were cultured in PD+LIF (PL) or 2i+LIF (2i+L) for 48 h. Data in – are mean of two biological replicates. * Figure 4: β-catenin functions by abrogating Tcf3 repression. () Flow cytometry analysis showing the profile of Rex1GFP expression in βcatfl/− (CRd2) or βcatΔ/− (CreC and CreD) Rex1GFP reporter embryonic stem cells mock transfected (Mock) or transfected with control siRNAs or siRNA against Tcf3 and cultured for 48 h in N2B27 alone. GFP-ve; wild-type embryonic stem cells that do not express GFP. () Histograms showing relative expression of Axin2, Cdx1, Nanog and Klf4 in βcatfl/− (CRd2) or βcatΔ/− (CreC and CreD) Rex1GFP reporter embryonic stem cells. Cells were cultured for 24 h in N2B27 alone and were transfected with control siRNAs or siRNAs against Tcf3. Expression is shown relative to untreated. Mean±s.d. of three biological replicates. () Phase-contrast and fluorescent micrographs showing Nanog and βIII-tubulin (TuJ1) expression in Tcf3-null embryonic stem cells stably expressing a randomly integrated Tcf3 transgene or the corresponding vector control. Cells were cultured in N2B27 plus PD or plus 2i for the indi! cated number of passages (p). Scale bar, 100 μm. () Histogram showing the number of undifferentiated (AP+), mixed and differentiated (AP−) colonies formed from 600 Tcf3-null embryonic stem cells in N2B27 alone or plus PD, LIF or PD+LIF (PL) in the presence or absence (NON) of CH. * Figure 5: Gsk3 inhibition relieves the core pluripotency network from repression by Tcf3 and complements Mek inhibition and/or Stat3 activation to stabilize embryonic stem cell self-renewal. () Histogram showing TOPFlash and FOPFlash activation in Tcf3-null embryonic stem cells expressing wild-type or amino-terminal-deleted (ΔN) Tcf3 or the vector control (Vec) in PD+LIF (PL) or 2i+LIF (2i+L). Mean±s.d. of three biological replicates. () Histogram showing relative expression of Cdx1 in Tcf3-null embryonic stem cells expressing wild-type (Tcf3WT) or N-terminal-deleted (Tcf3ΔN) Tcf3 or the vector control in 2i+LIF (2i+L) or N2B27 alone (N). Mean±s.d. of three biological replicates. () Histogram showing the number of undifferentiated colonies formed from 600 Tcf3-null embryonic stem cells expressing wild-type (Tcf3WT) or N-terminal-deleted (Tcf3ΔN) Tcf3 or the vector control in N2B27 plus PD, 2i, PD+LIF (PL) or 2i+LIF (2i+L). Mean of two biological replicates. () Histograms showing relative enrichment of Klf2 and Nodal promoter regions following chromatin immunoprecipitation for Tcf3 in Tcf3-null embryonic stem cells expressing Tcf3 or the vector control. Mean! ±s.d. of three replicates. () Histograms showing the response to CH of Klf2 and Nodal in βcatΔ/− cells expressing wild-type or C-terminal-deleted (ΔC) Ctnnb1 cultured in N2B27 alone for 24 h followed by 8 h in N2B27 plus PD in the presence or absence of CH. Mean ratio (+CH/−CH) of two biological replicates. () In the presence of Gsk3 inhibitor (Gski) and a Mapk (Erk) kinase inhibitor (Meki), repressive effects on the pluripotent gene regulatory network are abolished. The pluripotent circuitry is also positively regulated by Stat3 acting primarily through Klf4. Any two of these three effects are sufficient to stabilize the network and sustain embryonic stem cell self-renewal. Gski generates intracellular β-catenin, which interacts with Tcf3 and abolishes its repressor effect on multiple genes in the pluripotent network. Gski also supports embryonic stem cell propagation through stimulatory effects on metabolic and biosynthetic processes (dashed arrow). () In the! absence of inhibitors, Tcf3 repression and activated Erk driv! e embryonic stem cells into differentiation. When embryonic stem cells are maintained in serum using LIF without inhibitors, cultures are heterogeneous and metastable owing to the coexistence of the states in and . Author information * Author information * Supplementary information Affiliations * Wellcome Trust Centre for Stem Cell Research & Department of Biochemistry, University of Cambridge, Tennis Court Road, Cambridge CB2 1QR, UK * Jason Wray, * Tüzer Kalkan, * Sandra Gomez-Lopez & * Austin Smith * Max-Planck Institute of Immunobiology, Stubeweg 51, D-79108 Freiburg, Germany * Dominik Eckardt & * Rolf Kemler * World Wide Medicinal Chemistry, Pfizer Ltd, Sandwich CT13 9NJ, UK * Andrew Cook * Present addresses: Cancer Institute, University College London, Paul O'Gorman Building, 72 Huntley Street, London WC1E 6BT, UK (J.W.); Instituto de Fisiologı´a Celular, División de Neurociencias, UNAM, Circuito Exterior s/n, Ciudad Universitaria, México, DF 04510, Mexico (S.G-L.); Miltenyi Biotec GmbH, Friedrich-Ebert-Straße 68, 51429 Bergisch Gladbach, Germany (D.E.) * Jason Wray, * Sandra Gomez-Lopez & * Dominik Eckardt Contributions J.W. carried out, analysed and interpreted experiments, T.K. created and validated the Rex1GFPd2 reporter, S.G-L. generated Cre–Ires–fluorescent protein plasmids, D.E. and R.K. generated floxed β-catenin embryonic stem cells, A.C. selected and provided Gsk3 inhibitors, and A.S. supervised the study and wrote the paper together with J.W. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Austin Smith Author Details * Jason Wray Search for this author in: * NPG journals * PubMed * Google Scholar * Tüzer Kalkan Search for this author in: * NPG journals * PubMed * Google Scholar * Sandra Gomez-Lopez Search for this author in: * NPG journals * PubMed * Google Scholar * Dominik Eckardt Search for this author in: * NPG journals * PubMed * Google Scholar * Andrew Cook Search for this author in: * NPG journals * PubMed * Google Scholar * Rolf Kemler Search for this author in: * NPG journals * PubMed * Google Scholar * Austin Smith Contact Austin Smith Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Author information * Supplementary information PDF files * Supplementary Information (1800K) Supplementary Information Additional data - The bidirectional depolymerizer MCAK generates force by disassembling both microtubule ends
- Nat Cell Biol 13(7):846-852 (2011)
Nature Cell Biology | Letter The bidirectional depolymerizer MCAK generates force by disassembling both microtubule ends * Yusuke Oguchi1 * Seiichi Uchimura2 * Takashi Ohki1 * Sergey V. Mikhailenko1 * Shin'ichi Ishiwata1, 3, 4 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:846–852Year published:(2011)DOI:doi:10.1038/ncb2256Received13 September 2010Accepted06 April 2011Published online22 May 2011 Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg During cell division the replicated chromosomes are segregated precisely towards the spindle poles1, 2. Although many cellular processes involving motility require ATP-fuelled force generation by motor proteins, most models of the chromosome movement invoke the release of energy stored at strained (owing to GTP hydrolysis) plus ends of microtubules3, 4. This energy is converted into chromosome movement through passive couplers5, 6, 7, whereas the role of molecular motors is limited to the regulation of microtubule dynamics. Here we report, that the microtubule-depolymerizing activity of MCAK (mitotic centromere-associated kinesin), the founding member of the kinesin-13 family, is accompanied by the generation of significant tension—remarkably, at both microtubule ends. An MCAK-decorated bead strongly attaches to the microtubule side, but readily slides along it in either direction under weak external loads and tightly captures and disassembles both microtubule ends. We sho! w that the depolymerization force increases with the number of interacting MCAK molecules and is ~1 pN per motor. These results provide a simple model for the generation of driving force and the regulation of chromosome segregation by the activity of MCAK at both kinetochores and spindle poles through a 'side-sliding, end-catching' mechanism. View full text Figures at a glance * Figure 1: Analysis of MCAK-dependent bead movement. () Experimental set-up for the observation of microtubule depolymerization induced by MCAK-decorated beads. Two MCAK-decorated beads, held by optical tweezers, are brought close to the ends of a microtubule (one per each end), which they readily capture. One of the trapped beads is then released, and fast shortening of the microtubule is observed. () Time-lapse fluorescence images of the microtubule shortening by MCAK-decorated beads at three different MCAK-to-bead ratios (hereafter, these values indicate molar mixing ratios of MCAK and beads). For the corresponding video, see Supplementary Movie S2. The white and the red arrowheads indicate the trapped and the released bead, respectively. Scale bar, 3 μm. () Two-dimensional plot of the positions of two MCAK-decorated beads obtained by phase-contrast microscopy imaging (MCAK:bead = 124,000:1). The y axis corresponds to the orientation of the microtubule's long axis before the bead release. () Time courses of the bead-to! -bead distance at various MCAK-to-bead ratios. The traces are noisier at low ratios, indicating larger fluctuations of the positions of the attached beads owing to weaker bead–microtubule attachment by a smaller number of MCAK molecules. Beads coated with an antibody against β -tubulin are used as a control. The red lines represent linear fits to the data. () Microtubule depolymerization rate, which is a sum of the depolymerization rates at the plus and the minus ends, as a function of the MCAK-to-bead ratio. n=1, 8, 18, 19, 13, 2, and 14 for the MCAK-to-bead ratio being 3,100, 4,130, 6,200, 12,400, 37,200, 74,400, and 124,000, respectively. Error bars (absent when n=1 or 2) represent the s.e.m. Data are fitted with the Hill equation. * Figure 2: Tension generation depends on MCAK density. () Experimental set-up for the measurement of tension generation (left) and the typical time courses of the displacement of two trapped beads at a trap stiffness of 0.078 pN nm−1 (right). The depolymerization force (marked by horizontal arrows) is defined as the value of the external load at which one of MCAK-decorated beads detaches from the microtubule and both beads return to the trap centres. For the three MCAK-to-bead ratios shown in this panel, the average velocities determined from linear fits to the raw data traces were (from left to right) 1.1±0.3, 3.6±0.7 and 2.8±0.3 nm s−1 (mean ± s.e.m.; n=14, 14 and 44, respectively). () Histograms of the depolymerization force at various MCAK-to-bead ratios. The ratio and the force values (mean ± s.d.) are shown in each plot. () Average depolymerization force as a function of the MCAK-to-bead ratio in the absence of glycerol (solid line) and in the presence of 30% (v/v) glycerol (dashed line). Data are fitted w! ith the Hill equation. The maximum depolymerization force was 2.6±0.2 pN (Km=4,930, nH=1.0) and 4.2±0.4 pN (Km=4,190, nH=1.7) in the absence or presence of glycerol, respectively. The error bars represent the s.e.m. () Histogram of the depolymerization forces in the absence of glycerol combined for all MCAK-to-bead ratios fitted with two Gaussian distributions. * Figure 3: Measurement of the depolymerization force generated on two simultaneously captured microtubules. () Schematic representations (top) and time-lapse fluorescence images (bottom) showing the 'chewing' of two simultaneously captured microtubules (MCAK:bead = 124,000:1; scale bar, 3 μm). () The corresponding time courses for the displacement of one of the beads. () Depolymerization forces generated on two microtubules (2 MT; n=2) and on one microtubule (1 MT; n=4) after the other microtubule had detached from the bead. The depolymerization force on one microtubule is larger than the peak value of the distribution for the same ratio (MCAK:bead = 124,000:1) in Fig. 2b, which may be due to a more complex experimental geometry and loading conditions when two microtubules were captured simultaneously. () Schematic model of the regulation of tension balance on the chromosome in the spindle by the number of attached microtubules. * Figure 4: MCAK–microtubule interaction markedly depends on the region of the microtubule. () Experimental set-up for measuring the interaction force between an MCAK-decorated bead and a microtubule. A microtubule was immobilized on a large bead (diameter 2 μm) that was coated with an antibody against β -tubulin and attached to the glass surface, and an MCAK-decorated bead (MCAK:bead = 124,000:1) was brought to its side by optical tweezers. When weak external loads were exerted on the MCAK-decorated bead by slowly displacing the stage along the microtubule's long axis, the trapped bead readily slid along the microtubule in either direction. The arrows indicate the on-axis direction in Supplementary Movies S5–S7. The upper and the lower parts in and correspond to the plus- and minus-end-directed loads, respectively. Scale bar, 3 μm. () Time courses of the displacement of the bead (red trace) and the stage (black trace). Areas coloured in grey correspond to the sliding motion of the bead on the microtubule lattice, and the value of the frictional force is! calculated as Ff=kΔx, where Δx is the average bead displacement from the trap centre within the grey areas and k is the trap stiffness. () The frictional force under the plus- or the minus-end-directed loads and the rupture force at both microtubule ends plotted against the stage velocity. The error bars represent the s.e.m. () The rupture force at both microtubule ends plotted against the loading rate. The error bars represent the s.e.m. (For the values of n in and , see Supplementary Tables S1 and S2.) * Figure 5: Impact of free tubulin and the model of MCAK action. () Dependence of the depolymerization force on free-tubulin concentration (the error bars represent the s.e.m.); () the corresponding histograms. MCAK:bead = 12,400:1. The fitted curve in is defined as F=Fmax[M′]/(Km+[M′]), where Fmax and Km were determined from the data shown in Fig. 2c (no glycerol); [M′ ] is the concentration of MCAK without bound tubulin determined by [M′]=[M]Kd/(Kd+[Tub]), where [M] and [Tub] are the concentrations of MCAK and free tubulin, respectively; Kd is the dissociation constant of MCAK from the microtubules. () Model of force production by the depolymerization activity of MCAK. The MCAK molecule binds to the microtubule end (1) and undergoes conformational change, which induces deformation of a protofilament (2). Before it detaches from the microtubule, another MCAK molecule binds a different protofilament at larger separation from the bead (3). The MCAK molecule detaches from the microtubule (4) and is recycled after releasing the tubul! in dimer. When the first MCAK molecule dissociates from the microtubule before the second binds it, the bead–microtubule attachment ruptures (5). The black arrowhead is a visual marker. Author information * Author information * Supplementary information Affiliations * Department of Physics, Faculty of Science and Engineering, Waseda University, 3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, Japan * Yusuke Oguchi, * Takashi Ohki, * Sergey V. Mikhailenko & * Shin'ichi Ishiwata * Advanced Technology Development Core, Brain Science Institute, RIKEN, Saitama 351-0198, Japan * Seiichi Uchimura * Advanced Research Institute for Science and Engineering, Waseda University, Tokyo 169-8555, Japan * Shin'ichi Ishiwata * Waseda Bioscience Research Institute in Singapore (WABIOS), Singapore 138667, Republic of Singapore * Shin'ichi Ishiwata Contributions Y.O. carried out the experiments and analysed the results. Y.O., S.V.M. and S.I. designed the experiments and wrote the manuscript. Y.O., S.U. and T.O. prepared DNA constructs and purified proteins. All authors discussed the results and commented on the manuscript. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Shin'ichi Ishiwata Author Details * Yusuke Oguchi Search for this author in: * NPG journals * PubMed * Google Scholar * Seiichi Uchimura Search for this author in: * NPG journals * PubMed * Google Scholar * Takashi Ohki Search for this author in: * NPG journals * PubMed * Google Scholar * Sergey V. Mikhailenko Search for this author in: * NPG journals * PubMed * Google Scholar * Shin'ichi Ishiwata Contact Shin'ichi Ishiwata Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Author information * Supplementary information Excel files * Supplementary Information (20K) Supplementary Information Movies * Supplementary Information (2M) Supplementary Movie 1 * Supplementary Information (700K) Supplementary Movie 2 * Supplementary Information (500K) Supplementary Movie 3 * Supplementary Information (2M) Supplementary Movie 4 * Supplementary Information (5M) Supplementary Movie 5 * Supplementary Information (4M) Supplementary Movie 6 * Supplementary Information (1M) Supplementary Movie 7 PDF files * Supplementary Information (2M) Supplementary Information Additional data - A membrane trafficking pathway regulated by the plant-specific RAB GTPase ARA6
- Nat Cell Biol 13(7):853-859 (2011)
Nature Cell Biology | Letter A membrane trafficking pathway regulated by the plant-specific RAB GTPase ARA6 * Kazuo Ebine1 * Masaru Fujimoto1 * Yusuke Okatani1 * Tomoaki Nishiyama2 * Tatsuaki Goh1, 7 * Emi Ito1 * Tomoko Dainobu1 * Aiko Nishitani3 * Tomohiro Uemura1 * Masa H. Sato3 * Hans Thordal-Christensen4 * Nobuhiro Tsutsumi5 * Akihiko Nakano1, 6 * Takashi Ueda1 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:853–859Year published:(2011)DOI:doi:10.1038/ncb2270Received19 November 2010Accepted28 April 2011Published online12 June 2011 Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg Endosomal trafficking plays an integral role in various eukaryotic cell activities and serves as a basis for higher-order functions in multicellular organisms. An understanding of the importance of endosomal trafficking in plants is rapidly developing1, 2, but its molecular mechanism is mostly unknown. Several key regulators of endosomal trafficking, including RAB5, which regulates diverse endocytic events in animal cells3, 4, are highly conserved. However, the identification of lineage-specific regulators in eukaryotes indicates that endosomal trafficking is diversified according to distinct body plans and lifestyles. In addition to orthologues of metazoan RAB5, land plants possess a unique RAB5 molecule, which is one of the most prominent features of plant RAB GTPase organization5, 6. Plants have also evolved a unique repertoire of SNAREs, the most distinctive of which are diverse VAMP7-related longins, including plant-unique VAMP72 derivatives7. Here, we demonstrate that ! a plant-unique RAB5 protein, ARA6, acts in an endosomal trafficking pathway in Arabidopsis thaliana. ARA6 modulates the assembly of a distinct SNARE complex from conventional RAB5, and has a functional role in the salinity stress response. Our results indicate that plants possess a unique endosomal trafficking network and provide the first indication of a functional link between a specific RAB and a specific SNARE complex in plants. View full text Figures at a glance * Figure 1: The land-plant-unique RAB5 (ARA6) and conventional RAB5 proteins localize to different endosomes. () Schematic representation of the primary structures of Homo sapiens RAB5c (HsRAB5c) and Arabidopsis thaliana RAB5 proteins. () Maximum-likelihood tree for selected RAB GTPases. Bootstrap values of 75% or more are shown. () Subcellular localization of ARA6–Venus (ARA6–V, red) and GFP–ARA7 (G–ARA7, green) in root epidermal cells treated with dimethylsulphoxide (upper left), 33 μM wortmannin (upper right) or 50 μM BFA (lower left) as observed by confocal microscopy. The ARA6 compartments are endosomal organelles, as indicated by BOR1–GFP (BOR1–G, green) co-localization en route from the plasma membrane to the vacuole (lower right). () ARA6–Venus did not localize to the mRFP–SYP43 (R-SYP43, red)-positive TGN in dimethylsulphoxide-treated (left) and wortmannin-treated (right) cells. * Figure 2: Genetic interactions between SYP22 and RAB5 genes. () The ara7 and rha1 mutations aggravated the abnormal morphological phenotypes of syp22-1. Plants were grown for 30 days (30 d). Leaves from wild-type and each mutant are shown in the inset. () ara6 mutations suppressed the abnormal phenotypes of the syp22-1 mutant and the plants with mutations in SYP22 and the conventional RAB5 genes. () Expression of constitutively active ARA7 (ARA7Q69L) and the ara6-1 mutation suppressed the growth defect of vps9a-2, which was aggravated by the heterozygous mutation of ARA7 (ara7−/+). () Immunoblotting of seed proteins using the anti-12S globulin antibody. Arrowheads indicate unprocessed precursors. () GFP–CT24 (green) was secreted into the intercellular spaces in rha1 syp22-1 and ara6-1 rha1 syp22-1 mutants, but it accumulated in protein storage vacuoles with autofluorescence emission (red) in wild-type and ara6-1 syp22-1 cells. The vps9a-2 mutant also secreted GFP–CT24, a phenotype that was not suppressed by ara6-1. () ! The delayed endocytosis of BOR1–GFP to vacuoles in syp22-1 plants was suppressed by ara6-1. Root endodermal cells were observed at the indicated times after transfer from low-boron (1 μM) to high-boron (100 μM) medium. Uncropped images of blots are shown in Supplementary Fig. S6. * Figure 3: ARA6 and VAMP727 act at the plasma membrane. () Constitutively active ARA6Q93L, but not ARA7Q69L, accumulates on the plasma membrane after wortmannin treatment. The plasma membranes that exhibit fluorescence emission are indicated by arrowheads. () ARA6 co-localizes with the vacuolar SNARE SYP22 near the vacuole (arrowheads in upper row), but not at the plasma membrane (lower row). () Root epidermal cells expressing pairs of fluorescently tagged proteins were observed by TIRFM. ARA6–GFP (green) co-localized with TagRFP–VAMP727 (red) at punctate foci near the plasma membrane (arrowheads in left panel). ARA6–mRFP (red) also co-localized with GFP–SYP121 (green) near the plasma membrane (arrowheads in right panel). () TIRFM revealed TagRFP–SYP121 and GFP–VAMP727 co-localization at some locations on the plasma membrane (arrowheads). () GFP–VAMP727 co-immunoprecipitated with SYP121 and vacuolar Q-SNAREs (SYP22, VTI11 and SYP5). () GFP–VAMP713 co-immunoprecipitated with the vacuolar Q-SNAREs but not a plasma m! embrane SNARE (SYP121). GFP alone did not co-precipitate with any of the SNAREs. () FRET occurred between plasma membrane SNAREs but not between SNARE and non-SNARE proteins (PIP2a and Lti6a). Results are presented as mean FRET efficiency ± standard deviation as evaluated by bleaching TagRFP to 70% of the original fluorescence intensity (n=11). Uncropped images of blots are shown in Supplementary Fig. S6. * Figure 4: ARA6 promotes VAMP727–SYP121 complex formation at the plasma membrane. () Re-localization of GFP–VAMP727 to the plasma membrane in the vamp727 mutant required ARA6 activity. Plasma membranes that exhibit fluorescence emission are indicated by arrowheads. () TIRFM of TagRFP–VAMP727 and CLC–GFP in close proximity to the plasma membrane. Some VAMP727 foci overlapped with CLC (arrowhead). () Immunoprecipitation of SNARE complexes at the vacuole (SYP22, VTI11 and SYP5) and plasma membrane (SYP121) from the vamp727 and ara6-1 vamp727 mutants expressing GFP–VAMP727 using anti-GFP antibody. () Relative signal intensities in independent immunoblots (including results shown in ) were quantified. The results are presented as mean ± standard error (n=7 immunoblots for SYP22, n=8 for VTI11 and n=9 for SYP5 and SYP121). Wild type is the vamp727 mutant complemented with GFP–VAMP727. The amount of SYP121 bound to GFP–VAMP727 was significantly decreased in ara6-1 mutants when compared with the wild type. **P<0.01, Student's t -test. () Lysate f! rom vamp727 mutants complemented with GFP-VAMP727, and plants expressing ARA6Q93L and GFP–VAMP727 were immunoprecipitated and immunoblotted as in . The amount of SYP121 complexed with GFP–VAMP727 significantly increased with co-expression of ARA6Q93L. *P<0.05, Student's t-test. Results are presented as mean ± standard error (n=5 immunoblots). Uncropped images of blots are shown in Supplementary Fig. S6. * Figure 5: ARA6 is required for salinity stress tolerance. () ARA6Q93L–GFP (green) formed discrete speckles on the plasma membrane following NaCl treatment. Plants were grown on MS plates for 4 days, then cultivated on MS (−NaCl) or MS plus 100 mM NaCl (+NaCl) plates for 12 days. Root epidermal cells were observed by TIRFM. () ara6-1, vamp727 and ara6-1 vamp727 mutant plants were more sensitive to 50 mM NaCl stress than wild-type plants after 17 days of growth. () Fresh weight of plants grown under the same conditions as in . Results are presented as mean relative fresh weight (ratio of fresh weight to fresh weight of control plants) ± standard deviation (fresh weight of 32 plants measured for each genotype, which was repeated four times). The fresh weight of ara6 and ara6 vamp727 was significantly decreased when compared with wild type. *P<0.05, Student's t -test. () Wild-type and transgenic plants expressing ARA6Q93L–GFP were grown on MS plates for 4 days, then transferred to MS or MS+100 mM NaCl plates an! d cultivated for 12 additional days. () Fresh weight of plants grown under the same conditions as in . Results are presented as mean relative fresh weight ± standard deviation (n=20 in each genotype/condition combination). Author information * Author information * Supplementary information Affiliations * Department of Biological Sciences, Graduate School of Science, The University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan * Kazuo Ebine, * Masaru Fujimoto, * Yusuke Okatani, * Tatsuaki Goh, * Emi Ito, * Tomoko Dainobu, * Tomohiro Uemura, * Akihiko Nakano & * Takashi Ueda * Advanced Science Research Center, Kanazawa University, 13-1 Takara-machi, Kanazawa, Ishikawa 920-0934, Japan * Tomoaki Nishiyama * Graduate School of Life and Environmental Sciences, Kyoto Prefectural University, 1-5, Shimogamo-nakaragi-cho, Sakyo-ku, Kyoto, Kyoto 606-8522, Japan * Aiko Nishitani & * Masa H. Sato * Department of Agriculture and Ecology, Faculty of Life Sciences, University of Copenhagen Thorvaldsensvej 40, 1871 Frederiksberg C, Denmark * Hans Thordal-Christensen * Laboratory of Plant Molecular Genetics, Graduate School of Agricultural and Life Sciences, The University of Tokyo, Bunkyo-ku, Tokyo 113-8657, Japan * Nobuhiro Tsutsumi * Molecular Membrane Biology Laboratory, RIKEN Advanced Science Institute, Wako, Saitama 351-0198, Japan * Akihiko Nakano * Present address: Department of Biology, Graduate School of Science, Kobe University, Kobe 657-8501, Japan * Tatsuaki Goh Contributions T. Ueda designed the study; K.E. carried out the main parts of the genetic, biochemical and confocal microscopy experiments; M.F. and N.T. conducted TIRFM; T.G. carried out the experiments presented in Fig. 2c; T.N. carried out the phylogenetic analysis; Y.O., T.D., E.I., A. Nishitani, M.H.S. and T. Uemura constructed the transgenic plants used in this study; H.T-C. prepared the anti-SYP121 antibody; and A. Nakano and T. Ueda supervised the study. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Takashi Ueda Author Details * Kazuo Ebine Search for this author in: * NPG journals * PubMed * Google Scholar * Masaru Fujimoto Search for this author in: * NPG journals * PubMed * Google Scholar * Yusuke Okatani Search for this author in: * NPG journals * PubMed * Google Scholar * Tomoaki Nishiyama Search for this author in: * NPG journals * PubMed * Google Scholar * Tatsuaki Goh Search for this author in: * NPG journals * PubMed * Google Scholar * Emi Ito Search for this author in: * NPG journals * PubMed * Google Scholar * Tomoko Dainobu Search for this author in: * NPG journals * PubMed * Google Scholar * Aiko Nishitani Search for this author in: * NPG journals * PubMed * Google Scholar * Tomohiro Uemura Search for this author in: * NPG journals * PubMed * Google Scholar * Masa H. Sato Search for this author in: * NPG journals * PubMed * Google Scholar * Hans Thordal-Christensen Search for this author in: * NPG journals * PubMed * Google Scholar * Nobuhiro Tsutsumi Search for this author in: * NPG journals * PubMed * Google Scholar * Akihiko Nakano Search for this author in: * NPG journals * PubMed * Google Scholar * Takashi Ueda Contact Takashi Ueda Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Author information * Supplementary information Movies * Supplementary Information (3M) Supplementary Movie 1 PDF files * Supplementary Information (1600K) Supplementary Information Additional data - Willin and Par3 cooperatively regulate epithelial apical constriction through aPKC-mediated ROCK phosphorylation
- Nat Cell Biol 13(7):860-866 (2011)
Nature Cell Biology | Letter Willin and Par3 cooperatively regulate epithelial apical constriction through aPKC-mediated ROCK phosphorylation * Takashi Ishiuchi1, 2 * Masatoshi Takeichi1 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:860–866Year published:(2011)DOI:doi:10.1038/ncb2274Received06 December 2010Accepted06 May 2011Published online19 June 2011 Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg Apical-domain constriction is important for regulating epithelial morphogenesis. Epithelial cells are connected by apical junctional complexes (AJCs) that are lined with circumferential actomyosin cables. The contractility of these cables is regulated by Rho-associated kinases (ROCKs). Here, we report that Willin (a FERM-domain protein) and Par3 (a polarity-regulating protein) cooperatively regulate ROCK-dependent apical constriction. We found that Willin recruits aPKC and Par6 to the AJCs, independently of Par3. Simultaneous depletion of Willin and Par3 completely removed aPKC and Par6 from the AJCs and induced apical constriction. Induced constriction was through upregulation of the level of AJC-associated ROCKs, which was due to loss of aPKC. Our results indicate that aPKC phosphorylates ROCK and suppresses its junctional localization, thereby allowing cells to retain normally shaped apical domains. Thus, we have uncovered a Willin/Par3–aPKC–ROCK pathway that controls! epithelial apical morphology. View full text Figures at a glance * Figure 1: Willin recruits aPKC to AJCs in MDCK cells. () Top, parental cells (left) and stable transfectants expressing Willin–GFP (Willin–GFP/MDCK; right) immunostained for GFP and aPKC or Par3. Scale bar, 50 μm. Bottom left, cross-sections of the transfectants immunostained for GFP and aPKC. Bottom right, signal intensity of aPKC or Par3 at AJCs (mean ± s.e.m., ***P<0.001, Student's t-test, n>20 cells). Bottom middle, western blots showing total aPKC expression levels in the cells. () Left, Willin–GFP was immunoprecipitated from Willin–GFP/MDCK cells, and the immunoprecipitates (IPs) were analysed by immunoblotting for Par3, aPKC and GFP. Mouse IgG was used as a control. The arrows indicate three isoforms of Par3. Only the longest (relative molecular mass 180,000) isoform is co-precipitated with Willin. Right, endogenous Willin or aPKC was immunoprecipitated from EpH4 cells, and the immunoprecipitates (IPs) were analysed by immunoblotting for Willin, aPKC and Par3. Rabbit IgG was used as a control. () Deleti! on mutants of Willin and their ability to recruit aPKC to AJCs. aa, amino acids. () Willin–GFP or WillinΔJFR–GFP was immunoprecipitated by GFP antibody from its stable transfectants, and the immunoprecipitates (IPs) were analysed by immunoblotting for aPKC, Par3 and GFP. Mouse IgG was used as a control. () A GST-pulldown assay was carried out to examine the direct interaction between Willin and aPKC by using recombinant His-tagged PKCι and GST–Willin(330–440). The resulting samples were separated by SDS–PAGE, and analysed by western blotting (IB, immunoblot) to detect aPKC or by CBB staining. () Par3 was depleted in parental and Willin–GFP/MDCK cells. Cells were immunostained for Par3, GFP and aPKC. Yellow dashed lines indicate the borders between Par3-positive and -negative cells. Scale bar, 50 μm. Uncropped images of blots and gels are shown in Supplementary Fig. S9. * Figure 2: Willin and Par3 cooperatively regulate aPKC localization and apical constriction in EpH4 cells. () Top, Willin, Par3 or both were depleted by siRNA in EpH4 cells. Cells were immunostained for Willin, Par3 or aPKC. Localization of aPKC at AJCs indicated by arrows is magnified in the right panels. Scale bar, 50 μm. Bottom, signal intensity of aPKC at AJCs (mean ± s.e.m., ***P<0.001, Student's t-test, n>20 cells). () Knockdown efficiency for Willin, PKCλ and Par3. Knockdown of Willin and Par3 does not affect aPKC protein level. In and , DKD represents Par3 and Willin double knockdown. () Top, Willin, Par3 or PKCλ was depleted. Cells were immunostained for Willin, Par3, aPKC or ZO-1 (a AJC marker). Dashed lines indicate the borders between normal and knockdown cells. Scale bar, 50 μm. Bottom, the apical area of each cell was measured using ZO-1-stained samples throughout the experiments. The asterisks indicate that the mean is significantly different from the siRNA control (mean ± s.e.m., **P<0.01, ***P<0.001, Student's t-test, n>30 cells). Uncropped ! images of blots are shown in Supplementary Fig. S9. * Figure 3: Interaction between aPKC and Par3/Willin is necessary to regulate apical actomyosin activity. () Left, full-length Par3 and Par3ΔBD used in this study. aa, amino acids. Right, expression of GFP, Par3ΔBD–GFP and Par3–GFP in Par3-shRNA-27 MDCK cells stably expressing each molecule, as examined by western blotting for GFP. () Par3-shRNA-27 MDCK cells expressing GFP, Par3ΔBD–GFP or Par3–GFP were immunostained for GFP, aPKC, ZO-1 or 1P-MLC (pS19-MLC). The regions outlined by the dashed square are magnified below. Scale bars, 50 μm. () Apical area in Par3-shRNA-27 MDCK cells expressing the indicated proteins. The asterisks indicate that the mean is significantly different (mean ± s.e.m., ***P<0.001, Student's t-test, n>50 cells). Signal intensity of P-MLC at AJCs is also indicated in Supplementary Fig. S4c. () Left, expression of WillinΔJFR–GFP and Willin–GFP in Par3-shRNA-27 MDCK cells stably expressing each molecule, as examined by western blotting for GFP. Right, these transfectants were immunostained for GFP, aPKC, ZO-1 or 1P-MLC. The regions ! outlined by a dashed square are magnified below. Scale bars, 50 μm. Apical area is indicated in . Uncropped images of blots are shown in Supplementary Fig. S9. * Figure 4: aPKC phosphorylates ROCK1. () Left, Willin and Par3 or PKCλ were depleted by siRNA in EpH4 cells. Cells were immunostained for aPKC and ROCK1. The arrows point to cells that are magnified in the right panels. Scale bar, 50 μm. Top right, signal intensity of ROCK1 at AJCs in these cells. The asterisks indicate that the mean is significantly different from the siRNA control (mean ± s.e.m., ***P<0.001, Student's t-test, n>20 cells). Bottom right, total ROCK1 expression level examined by western blotting. DKD, Par3–Willin double knockdown. () Cell homogenates from EpH4 cells were left untreated, or were treated with λ phosphatase (λPPase) or λPPase and Na3VO4, a phosphatase inhibitor. After treatment, ROCK1 was analysed by western blotting. () Left, fragments of ROCK1 used in this study. Phosphorylation sites identified commonly by in vitro and in vivo assays are indicated by arrows. Right, CBB staining of GST-tagged ROCK1A, ROCK1B, ROCK1C and ROCK1D. () In vitro kinase assay against ROCK1! fragments was carried out by incubating them with recombinant (Rec.) PKCι. Western blotting was done for GST. Phos-tag was used in SDS–PAGE. () Proteins indicated were expressed in HEK293 cells, and the ROCK1-fragment band shift was examined by western blotting. Phos-tag was used in SDS–PAGE. The arrows and the arrowhead indicate different shifted bands. exp., exposure. Uncropped images of blots and gels are shown in Supplementary Fig. S9. * Figure 5: ROCK1 phosphorylation regulates its distribution. () shRNA expression vector against ROCK1 or both ROCK1 and ROCK2 (ROCKs) was co-transfected with GFP expression vector into Par3-shRNA-27 MDCK cells. Left, cells were immunostained for ZO-1 and GFP. Arrowheads indicate the disruption of cell–cell junctions occasionally observed in ROCKs-shRNA-transfected cells. Scale bar, 50 μm. Right, apical area in Par3-shRNA-27 MDCK cells transfected with the indicated vectors. The asterisks indicate that the mean is significantly different from control shRNA-1 (mean ± s.e.m., ***P<0.001, Student's t-test, n>50 cells). () GFP–ROCK1 wild type, -4A, -4E, -5A or -5E was co-transfected with ROCKs shRNA vector in Par3-shRNA-27 MDCK cells. Cells were immunostained for GFP and ZO-1. Scale bar, 30 μm. Right, apical area in Par3-shRNA-27 MDCK cells transfected with the indicated vectors. The asterisks indicate that the mean is significantly different from GFP–ROCK1 wild type (mean ± s.e.m., **P<0.01, ***P<0.001, Student's! t-test, n>50 cells). () GFP, GFP–ROCK1 wild-type, -4A, -4E, -5A or -5E was expressed in HEK293 cells and fractionation was carried out. Top, equal amounts of samples from supernatant (S) and pellet (P) fractions were loaded and western blotting was carried out. GAPDH (glyceraldehyde 3-phosphate dehydrogenase) and N-cadherin were used as positive controls for fractionation. Bottom, measured S/P ratio of GFP–ROCK1. GAPDH was used to normalize the results. The histogram shows an average result of five independent experiments. The asterisks indicate that the mean is significantly different from wild type (mean ± s.e.m, **P<0.01, Mann–Whitney's U-test, n=5 experiments). Uncropped images of blots are shown in Supplementary Fig. S9. Author information * Author information * Supplementary information Affiliations * RIKEN Center for Developmental Biology, Chuo-ku, Kobe 650-0047, Japan * Takashi Ishiuchi & * Masatoshi Takeichi * Graduate School of Biostudies, Kyoto University, Sakyo-ku, Kyoto 606-8502, Japan * Takashi Ishiuchi Contributions T.I. designed and conducted experiments. T.I. and M.T. wrote the paper. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Masatoshi Takeichi Author Details * Takashi Ishiuchi Search for this author in: * NPG journals * PubMed * Google Scholar * Masatoshi Takeichi Contact Masatoshi Takeichi Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Author information * Supplementary information PDF files * Supplementary Information (3M) Supplementary Information Additional data - The PIAS homologue Siz2 regulates perinuclear telomere position and telomerase activity in budding yeast
- Nat Cell Biol 13(7):867-874 (2011)
Nature Cell Biology | Letter The PIAS homologue Siz2 regulates perinuclear telomere position and telomerase activity in budding yeast * Helder C. Ferreira1 * Brian Luke2, 3 * Heiko Schober1, 4 * Véronique Kalck1 * Joachim Lingner2 * Susan M. Gasser1 * Affiliations * Contributions * Corresponding authorJournal name:Nature Cell BiologyVolume: 13,Pages:867–874Year published:(2011)DOI:doi:10.1038/ncb2263Received02 September 2010Accepted19 April 2011Published online12 June 2011 Article tools * Full text * Print * Email * Download PDF * Download citation * Order reprints * Rights and permissions * Share/bookmark * Connotea * CiteULike * Facebook * Twitter * Delicious * Digg Budding yeast telomeres are reversibly bound at the nuclear envelope through two partially redundant pathways that involve the Sir2/3/4 silencing complex and the Yku70/80 heterodimer1, 2. To better understand how this is regulated, we studied the role of SUMOylation in telomere anchoring. We find that the PIAS-like SUMO E3 ligase Siz2 sumoylates both Yku70/80 and Sir4 in vivo and promotes telomere anchoring to the nuclear envelope. Remarkably, loss of Siz2 also provokes telomere extension in a telomerase-dependent manner that is epistatic with loss of the helicase Pif1. Consistent with our previously documented role for telomerase in anchorage3, normal telomere anchoring in siz2Δ is restored by PIF1 deletion. By live-cell imaging of a critically short telomere, we show that telomeres shift away from the nuclear envelope when elongating. We propose that SUMO-dependent association with the nuclear periphery restrains bound telomerase, whereas active elongation correlates with! telomere release. View full text Figures at a glance * Figure 1: Yeast telomeres are delocalized by deletion of SIZ2. () Visualizing individual telomere position within living yeast by a lacO/LacI system. lacO arrays were integrated 12 kb and 18 kb from the ends of Tel6R and Tel8L, respectively. Telomere position was quantified by scoring the nuclear zone occupied. A three-dimensional (3D) stack of images was collected and the focus position determined relative to three zones of equal area (zone 1 being ≤0.2 μm from Nup49–GFP fluorescence emission). A random distribution (33%) is indicated with a dotted red line in the graphs in –. Confocal planes of representative single cells containing Tel6R-tagged loci in either a wild-type or siz2Δ background are shown. Scale bars, 2 μm. () SIZ2 deletion caused delocalization of Tel6R. The distribution of Tel6R foci in both G1- and S-phase cells in wild-type (GA-1459) and siz2Δ (GA-5162) strains was analysed as described above. The number of cells analysed in each cell cycle stage for wild-type and siz2Δ respectively is indicated wit! hin the graph. Statistical significance of zone 1 enrichment is determined relative to a random distribution and is indicated above the bars, *=P>0.05. () Tel8L is also delocalized in siz2Δ cells. Wild-type Tel8L-tagged (GA-1986) and siz2Δ (GA-5378) strains were analysed as in . () Tel6R delocalization is specific to siz2Δ strains. The fraction of Tel6R foci in zone 1 (periphery) is shown for wild-type (GA-1459), cst9Δ (GA-4881), siz1Δ (GA-4882) and mms21-ch (GA-5738) strains. The number of cells analysed in each cell cycle stage for wild-type and mutant strains respectively is indicated underneath the graph. The percentage of wild-type foci in zone 1 is compared with siz1Δ, cst9Δ and mms21-ch, *=P>0.05. * Figure 2: Siz2-dependent SUMOylation of Sir4 and Yku70/80 promotes chromatin anchoring. () Minimal anchor assay. The position of ARS607 within living yeast was visualized by a lacO/LacI system and quantified by scoring the nuclear zone they occupied, with respect to the periphery as in Fig. 1a. The presence of adjacent lexA operator sequences allows for the recruitment of LexA-tagged proteins. LexA-fusion constructs of the anchoring proteins Yku70, Yku80 and the Sir4-PAD are illustrated with the relative position of putative and experimentally verified SUMO conjugation sites based on SUMOplot and ref. 14. () LexA-fusion constructs were transformed into wild-type (GA-1461) or siz2Δ (GA-4447) cells and the fraction of cells with ARS607 anchored at the periphery is shown. Top left, LexA–Sir4-PAD (plasmid 1454) anchoring is partially inhibited by SIZ2 deletion in both G1 and S phase. The number of cells analysed in each cell cycle stage per strain is indicated underneath the graph. In all graphs, the statistical significance of the difference in zone 1 distribut! ion between the indicated columns is shown above the bars, *=P>0.05. Top right, anchoring of LexA–yku80-4 (plasmid 1362) is impaired in S phase in siz2Δ cells. This can be restored using LexA–yku80-4–SUMO (plasmid 2274). Bottom left, a similar analysis for LexA–Yku70 (plasmid 2325). LexA–Yku70–SUMO (plasmid 2334) restores G1 anchoring siz2Δ cells. Bottom right, LexA–Yif1 (plasmid 1453)-mediated chromatin anchoring is unaffected by SIZ2 deletion. () Siz2 sumoylates Yku70/80 and Sir4 in vivo. SUMOylated proteins were purified from yeast strains expressing His-tagged Smt3 by Ni affinity chromatography. Slower-migrating species of Yku70/80 (using anti-Myc) and Sir4 (with anti-Sir4 antibody) are detected only when Smt3 is His tagged. Deletion of SIZ2 results in a marked loss of SUMOylation for both Yku80 (80%) and Sir4 (87%) and a partial loss for Yku70 (40%). Uncropped images of blots are shown in Supplementary Fig. S7. * Figure 3: SUMO modification of Yku70 and Yku80 enhances telomere anchoring. () The indicated Yku70/80 constructs were transformed into wild-type (GA-1986), yku70Δ (GA-1916) or yku80Δ (GA-6561) cells and the fraction of cells with Tel8L anchored at the periphery is indicated. Left, SUMO modification of Yku70 promotes Tel8L anchoring in G1 phase. LexA (plasmid 1452)-, LexA–Yku70 (plasmid 2325)- and LexA–Yku70–SUMO (plasmid 2334)-containing strains were analysed as in Fig. 1d. Right, SUMO modification of Yku80 promotes Tel8L anchoring in G1 and S phase. Control vector (plasmid 723), Yku80 (plasmid 2838) and Yku80–SUMO (plasmid 2839) were transformed into the indicated genotypes and analysed as above. In all graphs, the statistical significance of the difference in zone 1 distribution between the indicated columns is shown above the bars. () C-terminal Yku70/80–SUMO fusion constructs are functional and suppress the temperature sensitivity of yku70Δ and yku80Δ strains. The same yeast strains and plasmids as in were plated at the indicated t! emperature as a tenfold dilution series onto selective media. () Restoration of telomere length using Yku70/80–SUMO fusion constructs. Genomic DNA from the strains as in was prepared and analysed for bulk telomere length. The terminal TG tract is indicated with a bar and the asterisks indicate internal Y′ fragments. Uncropped images of blots are shown in Supplementary Fig. S7. * Figure 4: Pif1 links Siz2-dependent telomere length and localization regulation. () Loss of Siz2 results in a longer terminal TG repeat. The bulk telomere length of isogenic wild-type, siz1Δ, siz2Δ and siz1Δ siz2Δ strains (yBL7, yBL224, yBL226 and yBL254) was measured by Southern blotting on XhoI-digested genomic DNA using a Y'+TG probe. The terminal TG tract is indicated with a bar and the asterisks indicate internal Y′ fragments. The change in telomere length when compared with the wild type is indicated below each lane. () Siz2-dependent telomere homeostasis is independent of homologous recombination. The same as in , except isogenic strains carrying rad52Δ or siz2Δ rad52Δ were used (yBL229, yBL252). () PIF1 is epistatic with SIZ2 for telomere length homeostasis. SIZ2 was deleted in rif1Δ, rif2Δ and pif1Δ backgrounds (yBL359, yBL361, yBL363) and the telomere length of double mutants was monitored as in Fig. 3c. () PIF1 deletion restores anchoring of Tel6R in siz2Δ strains. Left, Tel6R was monitored in wild-type (GA-1459), pif1Δ (GA-494! 0), siz2Δ (GA-5162) and siz2Δ pif1Δ (GA-5411) strains and was analysed as in Fig. 1d. Right, comparison of the localization of Tel6R in siz2Δ pif1Δ (GA-5411) and siz2Δ pif1Δ est2Δ (sporulated from GA-5650) strains showing that S-phase telomere anchoring in GA-5411 requires telomerase. In all graphs, the statistical significance of the difference in zone 1 distribution between the indicated columns is shown above the bars. Uncropped images of blots are shown in Supplementary Fig. S7. * Figure 5: Critically short telomeres detach from the periphery on elongation during the first S phase. () Visualizing elongating telomeres. Telomerase-positive or -negative donors (GA-3362 and GA-3363) were mated to an est1Δ telomerase-deficient recipient (GA-3361) that has short telomeres. GA-3361 carries both LacI–GFP and Nup49–GFP, and Tel5R is lacO tagged. Cell cycle stages are determined by the timing of nuclear fusion, bud emergence and the onset of mitosis. A late S-phase cell (large bud) is shown. Scale bar, 2 μM. () In the zygotes described in , we monitored telomere localization by taking imaging stacks every 3 min. Shown here is a time-lapse series of overexposed fluorescence imaging on top of differential interference images to reveal nuclear shape and timing of key cell cycle events. Note that zygotic nuclei appear elongated in the first cell cycle, whereas they are round in the following cell cycles. () The absolute distance of tagged Tel5R relative to the nuclear periphery was scored and all measurements were synchronized relative to the cell's entr! y into mitosis. The distance of Tel5R from the nuclear periphery was scored for more than 50 cells at each time point (EST1,n=55;est1Δ,n=65). The mean and standard deviation values are plotted for the time period spanning late S phase. () Summary of data from this and other papers into a model for how telomere position may affect elongation efficiency by telomerase. When telomeres are long, Siz2-mediated SUMOylation (red circles) of Yku70/Yku80 and Sir4 favours telomere anchorage, which in turn may restrict elongation efficiency. Anchorage may also protect the telomere from recognition as a double-strand break. () When telomeres become critically short, loss of Siz2-mediated anchoring, perhaps through de-SUMOylation by Ulp1, releases telomeres from the periphery for efficient elongation. Pif1 does not affect tethering directly but alters the efficiency of Est2 association with the telomere and thus antagonizes elongation. Control of telomerase through other pathways is not! included for simplicity. Author information * Author information * Supplementary information Affiliations * Friedrich Miescher Institute for Biomedical Research, Maulbeerstrasse 66, 4058 Basel, Switzerland * Helder C. Ferreira, * Heiko Schober, * Véronique Kalck & * Susan M. Gasser * Swiss Institute for Experimental Cancer Research (ISREC), School of Life Sciences, Ecole Polytechnique Fédérale de Lausanne (EPFL), 1015 Lausanne, Switzerland * Brian Luke & * Joachim Lingner * Present address: ZMBH-Universität Heidelberg, Im Neuenheimer Feld 282, Heidelberg D-69120, Germany * Brian Luke * Present address: Institute of Cell Biology, Swiss Federal Institute of Technology (ETH) Zurich, 8093 Zurich, Switzerland * Heiko Schober Contributions S.M.G. and J.L. directed the study. H.C.F. did the biochemical experiments and with the help of V.K. carried out all telomere localization and minimal anchor experiments. B.L. did most of the telomere length assays and H.S. did the short-telomere mating assay. The manuscript was prepared by H.C.F. and S.M.G. with contributions from J.L., B.L. and H.S. Competing financial interests The authors declare no competing financial interests. Corresponding author Correspondence to: * Susan M. Gasser Author Details * Helder C. Ferreira Search for this author in: * NPG journals * PubMed * Google Scholar * Brian Luke Search for this author in: * NPG journals * PubMed * Google Scholar * Heiko Schober Search for this author in: * NPG journals * PubMed * Google Scholar * Véronique Kalck Search for this author in: * NPG journals * PubMed * Google Scholar * Joachim Lingner Search for this author in: * NPG journals * PubMed * Google Scholar * Susan M. Gasser Contact Susan M. Gasser Search for this author in: * NPG journals * PubMed * Google Scholar Supplementary information * Author information * Supplementary information Excel files * Supplementary Information (14K) Supplementary Table 1 PDF files * Supplementary Information (900K) Supplementary Information Additional data
No comments:
Post a Comment