p130-Angiomotin associates to actin and controls endothelial cell shape Mira Ernkvist1, Karin Aase1, Chinwe Ukomadu2,3, James Wohlschlegel2, Ryan Blackman2,3, Niina Veitonma¨ ki1, Anders Bratt1, Anindya Dutta2,3,4 and Lars Holmgren1
1 Department of Oncology-Pathology, Cancer Centre Karolinska Institute, Stockholm, Sweden 2 Department of Pathology, Brigham and Women’s Hospital, Boston, MA, USA 3 Department of Medicine, Brigham and Women’s Hospital, Boston, MA, USA 4 Department of Biochemistry and Molecular Genetics, University of Virginia, Charlottesville, VA, USA
Keywords actin fiber; angiogenesis; angiostatin; migration; tight junction
Correspondence L. Holmgren, Department of Oncology and Pathology, CCK R8 : 03, Karolinska University Hospital, SE-171 76 Stockholm, Sweden Fax: +46 8 339 031 Tel: +46 8 517 79317 E-mail: lars.holmgren@cck.ki.se
Database The nucleotide sequence reported here has been submitted to the GenBank database under bankit number 694853 and accession number AY987378.
(Received 26 January 2006, revised 6 March 2006, accepted 7 March 2006)
doi:10.1111/j.1742-4658.2006.05216.x
Angiomotin, an 80 kDa protein expressed in endothelial cells, promotes cell migration and invasion, and stabilizes tube formation in vitro. Angiomotin belongs to a new protein family with two additional members, Amotl-1 and Amotl-2, which are characterized by conserved coiled-coil domains and C-terminal PDZ binding motifs. Here, we report the identification of a 130 kDa splice isoform of angiomotin that is expressed in different cell types including vascular endothelial cells, as well as cytotrophoblasts of the placenta. p130-Angiomotin consists of a cytoplasmic N-terminal extension that mediates its association with F-actin. Transfection of p130-angiomotin into endothelial cells induces actin fiber formation and changes cell shape. The p130-angiomotin protein remained associated with actin after destabil- ization of actin fibers with cytochalasin B. In contrast to p80-angiomotin, p130-angiomotin does not promote cell migration and did not respond to angiostatin. We propose that p80- and p130-angiomotin play coordinating roles in tube formation by affecting cell migration and cell shape, respect- ively.
inhibitors of angiogenesis [1–3].
Abbreviations Amotl, angiomotin-like protein; BAE, bovine aortic endothelial; BCE, bovine capillary endothelial; DMEM, Dulbecco’s modified Eagle’s medium; EC, endothelial cell; FGF-2, basic fibroblast growth factor; MAE, mouse aortic endothelial; PEDF, pigment epithelium-derived factor; PmT, polyoma middle T.
FEBS Journal 273 (2006) 2000–2011 ª 2006 The Authors Journal compilation ª 2006 FEBS
2000
the process whereby new vessels are Angiogenesis, formed from pre-existing ones, plays a pivotal role in blood vessel formation during embryogenesis, organ growth and other physiological settings. Furthermore, dysregulated blood vessel formation may also contrib- ute to the pathogenesis of diseases like diabetic retino- pathy, atherosclerosis and endometriosis In addition, earlier findings that solid tumors must recruit endothelial cells (EC) to form their own microcircula- tion in order to be able to grow beyond a critical size [4,5] have led to the discovery of growth factors that trigger an angiogenic response [6,7]. Blood vessel formation is tightly regulated and rarely occurs in adults with the exception of the female reproductive system and wound healing. The strict control of blood vessel formation may be explained, at least partially, by endogenous such as thrombospondin [8], pigment epithelium-derived factor (PEDF) [9], endostatin [10], tumstatin [11], arresten [12], canstatin [13], and endorepellin [14] that negat- ively control vessel growth [15]. Some angiogenesis cleavage inhibitors are in that proteolytic latent
M. Ernkvist et al.
p130-Angiomotin controls cell shape
expression of p130-angiomotin controls cell shape, impairs migration and does not mediate responsiveness to angiostatin. endothelial migration and
inhibit angiogenesis. exposes cryptic fragments that One example of such a cryptic inhibitor is angiostatin, a 38 kDa proteolytic fragment of plasminogen that inhibits proliferation in vitro, and suppresses angiogenesis and the growth of primary tumors and metastasis in vivo [16,17]. We conclude that angiomotin is expressed as two isoforms, p80-angiomotin, which is involved in cell migration, and p130-angiomotin, which controls cell shape by the interaction with actin fibers.
Results
angiomotin-like protein (Amotl)-1 Identification of a novel splice isoform of angiomotin
In an effort to identify angiostatin-binding proteins, we previously screened a placenta library using angio- statin as bait. Using this approach, we identified an 80 kDa novel protein called p80-angiomotin [18] that belongs to a new protein family with two additional members, and Amotl-2. These proteins are characterized by conserved coiled-coil domains and C-terminal PDZ binding motifs [19,20]. It has previously been shown that alter- native splicing occurs within the angiomotin protein family and that some of these alternative transcripts are tissue specific [21]. Angiomotin differs from its related proteins as it contains an extracellular angio- statin-binding domain.
During an immunoprecipitation experiment designed for other purposes, we identified a novel isoform of angiomotin in 293T cells. The 293T cells were subjec- ted to immunoprecipitation analysis and the immuno- precipitates were separated by SDS ⁄ PAGE and detected by silver staining. Two bands corresponding to 80 and 130 kDa were cut out and identified by in-gel proteolysis and MS. Sequencing of the tryptic digests from the two proteins identified the 80 kDa band as p80-angiomotin, and the 130 kDa protein band as an unknown isoform of p80-angiomotin. Western blot analysis using an angiomotin-specific antibody confirmed that two isoforms of angiomotin are expressed in 293T cells (Fig. 1A).
Angiomotin is primarily expressed in EC and medi- ates the inhibitory effect of angiostatin on EC migra- tion and tube formation in vitro [18]. Expression of p80-angiomotin in mouse aortic endothelial (MAE) cells increases the migratory response to chemotactic factors [18]. A role in migration is further emphasized by the findings from angiomotin knockout experiments in mice. Approximately 70% of the mice died during embryonic day 7–8 due to a migratory defect in the anterior visceral endoderm [22].
specifically recognize
In p130-angiomotin, amino acids 1–409 form the N-terminal part that is specific for p130-angiomotin and amino acids 410–1084 correspond to p80-angio- motin and form the C-terminal part of p130-angio- motin (Fig. 1B). p130-Angiomotin is generated by alternative splicing between exons 2 and 3, which gives rise to this isoform with an N-terminal exten- sion (Fig. 1C). We generated polyclonal antibodies specific for the N-terminal domain of p130-angio- motin. These antibodies the p130 and not the p80 isoform as analyzed by west- ern blot (Fig. 1D). We have recently shown that angiostatin binds angio- motin on the cell surface and we have proposed a model for the topology of the two angiomotin isoforms in which the angiostatin-binding domain is extracellular and the N- and C-terminal domains are intracellular [23]. In addition, we showed that the novel angiomotin isoform, p130-angiomotin, has a topology similar to that of p80-angiomotin and also localizes to cell–cell contacts where it regulates paracellular permeability [23].
FEBS Journal 273 (2006) 2000–2011 ª 2006 The Authors Journal compilation ª 2006 FEBS
2001
p130-Angiomotin contains an extended N-terminal domain but is otherwise identical to its shorter iso- form of 80 kDa and therefore contains an angiosta- tin-binding domain. The first 243 amino acids of the N-terminal domain of p130-angiomotin show 58% homology to the N-terminals of Amotl-1 and Amotl- 2. Within these 243 amino acids there are five islands of conserved regions rich in glutamine con- sisting of (cid:1) 8–10 amino acids, which are 100% con- served between the three proteins. The N-terminal domain of p130-angiomotin does not contain any signal or transmembrane sequences. The N-terminal domain is therefore predicted to localize to the cyto- plasm. In this study, we describe the isolation and characteri- zation of isoform of the p130-angiomotin splice angiomotin. Immunoprecipitation of endogenous angio- motin from 293T cells revealed two bands of molecular mass of 80 and 130 kDa. The bands were analyzed by MS and revealed that p130-angiomotin contains an extended N-terminal domain. This domain associates with actin in the cytoplasm, resulting in the stabilization of actin fibers, changing cell shape and increasing the cell area. We further show that p80-angiomotin and p130-angiomotin have different cellular functions. The p80-angiomotin isoform promotes cell migration and mediates responsiveness to angiostatin. However, over-
M. Ernkvist et al.
p130-Angiomotin controls cell shape
A
B
C
D
Fig. 1. Identification of p130-angiomotin, a novel splice isoform of angiomotin. (A) Rabbit polyclonal antibodies directed against the last 24 amino acids of angiomotin recognize two proteins of 80 and 130 kDa in 293T cells. (B) 293T cells were subjected to immunoprecipitation, separated on SDS ⁄ PAGE and proteins were visualized by silver staining. The 80 and 130 kDa bands were extracted and analyzed by MS. Red represents sequences from proteolytic fragments that are unique to the p130 isoform. Sequences in blue (410–1084) correspond to peptides that were also found in the p80 protein. Bold-italics were used where concurrent sequences were identified to distinguish individual peptide fragments. (C) A new open reading frame is generated by alternative splicing between exons 2 and 3 which results in the N-terminal extension of p130-angiomotin. Blue corresponds to the glutamine-rich domain in p130-angiomotin, red to the coiled-coil, yellow to the angi- ostatin-binding domain and turquoise to the PDZ binding domain. (D) Rabbit polyclonal antibodies generated against 266 amino acids in the N-terminus of p130-angiomotin specifically detect the 130 kDa isoform of angiomotin as analyzed by western blot. NIS, nonimmune serum; PI, preimmune sera; I, immune sera.
Angiomotin expression during development and in adult tissues E11 to the end of gestation, p130-angiomotin expres- sion was only detected in the placenta from E13 to E16 (Fig. 2B).
Analysis of angiomotin protein levels in adult mouse tissues showed that both p130- and p80-angiomotin are found in the lung, liver, brain and heart. In brain and heart, a stronger signal for p130-angiomotin than for p80-angiomotin was detected (Fig. 2C).
To determine the expression of p130- and p80-angi- omotin at a cellular level, we tested cell lysates from 293T cells, cytotrophoblasts, hTERT+-BCE cells and BAE cells. As shown in Fig. 2D, all tested cell types produce both isoforms of angiomotin.
The N-terminal domain of p130-angiomotin localizes to actin and stabilizes actin fibers
FEBS Journal 273 (2006) 2000–2011 ª 2006 The Authors Journal compilation ª 2006 FEBS
2002
In order to analyze the subcellular localization of angi- omotin, MAE cells that do not express detectable levels of endogenous angiomotin [18] were retrofected Angiomotin plays an important role in embryogenesis, as angiomotin knockout mice show abnormal visceral endoderm cell migration during gastrulation resulting in embryonic lethality in 70% of the mice [22]. We therefore analyzed the temporal expression pattern of the angiomotin isoforms during mouse embryogenesis using a western blot approach with affinity-purified antibodies. A blot of mouse embryos from embryonic day (E) 5–19 showed expression of p80-angiomotin from E6 with an increase of protein level during fetal development. No expression of p130-angiomotin was detected (Fig. 2A). However, overexposure of embry- onic lysates from E12 showed weak expression of p130-angiomotin (Fig. S1). We then analyzed angio- motin expression in extraembryonic tissues using pla- centa samples from different stages of gestation. In addition to p80-angiomotin that was expressed from
M. Ernkvist et al.
p130-Angiomotin controls cell shape
A
B
C
D
Fig. 2. Expression of p130-angiomotin in embryonic and adult mouse tissues. (A) Western blot analysis of angiomotin protein levels in mouse embryos from E5–19. p80-Angiomotin is expressed from E5–18 (arrowhead), whereas no p130-angiomotin protein could be detected. (B) Western blot analysis of mouse placental lysates show expression of p80-angiomotin from E11–19 (arrowhead). p130-Angiomotin is expressed from E13–16 (arrow). (C) Analysis of mouse adult tissues shows strong expression of p80- and p130-angiomotin in brain and weaker expression in lung, liver, and heart. (D) Western blot analysis of MAE cells retrofected with vector, p80- and p130-angiomotin and endo- genous expression of both isoforms in 293T cells, cytotrophoblasts, hTERT+-BCE cells and BAE cells. The arrows indicate the location of p130 and p80 bands. Lu, lung; Ki, kidney; Spl, spleen; Li, liver; Br, brain; Hrt, heart. Western blots were loaded with equally amount of pro- tein, measured using the method of Bradford.
middle T (PmT)–EC cells and 293T cells. The cells were either stained with the antibodies directed against the C-terminus of angiomotin or with the polyclonal antibodies towards the specific N-terminal domain of p130-angiomotin. The staining pattern in both cell types confirms the colocalization of p130-angiomotin with F-actin (Fig. S2B–E). as we reported [18]. However,
of the full-length staining patterns were
Next, we investigated whether the N-terminal por- tion that is unique to p130-angiomotin is mediating the association with actin. For this purpose, MAE cells were transiently transfected with a construct encoding the N-terminal domain. Immunofluorescent staining of the N-terminal fragment showed a perfect colocaliza- tion with actin fibers but lacked the punctuated pattern p130-angiomotin characteristic (Fig. 3G–I). Similar also observed in M21 cells and HeLa cells transfected with the N-terminal domain (Fig. S2F,G). These results show that the N-terminus of p130-angiomotin is locali- zed in the cytoplasm and that it associates with actin fibers. Similar to the p130-angiomotin retrofected cells, the N-terminal transfected cells contained more actin fibers than vector and p80-angiomotin cells, indicating that the N-terminus of p130-angiomotin induces actin fiber formation.
with constructs encoding either the p80- or the p130- angiomotin isoforms. Polyclonal positive cells were selected with puromycin and expression was detected by western blot analysis (Fig. 2D). The MAE cells ret- rofected with the angiomotin isoforms were stained with C-terminal-specific antibodies that bind to both isoforms. The p80-angiomotin isoform localized to the lamellipodia, previously have (Fig. 3A–C) the positive staining of p130-angiomotin MAE cells showed a very regular pattern of small linear dots (Fig. 3D). The filamentous expression pattern of p130-angiomotin suggested that p130-angiomotin may colocalize with F-actin. To assess potential actin association, MAE cells expressing p130-angiomotin were stained for F-actin with Texas red-conjugated phalloidin as well as for p130-angiomo- tin. The punctuated p130-angiomotin staining showed an overlap with F-actin in most locations (Fig. 3D–F). This staining pattern was not restricted to MAE cells as M21 melanoma cells showed similar structures when transfected with p130-angiomotin (Fig. S2A). Further- more, the phalloidin staining of polymerized actin revealed a marked increase in actin fibers in p130- angiomotin cells compared with p80-angiomotin and vector-transfected cells (Fig. 3). To further analyze the colocalization of p130-angiomotin with F-actin, we used computerized deconvolution together with a 3D-imaging software to determine the overlap of p130-angiomotin and F-actin. Figure 3M shows the 3D imaging of p130-angiomotin and F-actin in one single actin fiber.
FEBS Journal 273 (2006) 2000–2011 ª 2006 The Authors Journal compilation ª 2006 FEBS
2003
Endogenous localization of p130-angiomotin was analyzed by immunofluorescent staining in polyoma To investigate whether p130-angiomotin is associ- ated with F-actin, we treated the cells with the actin depolymerizing agent cytochalasin B. As shown in Fig. 4D, the punctuated staining pattern of p130- angiomotin was disrupted in cells treated with cyto- chalasin B. The disrupted p130-angiomotin staining overlapped with the phalloidin staining of aggregated structures of actin (Fig. 4D–F). Next, we transfected
M. Ernkvist et al.
p130-Angiomotin controls cell shape
B
C
A
E
F
D
H
I
G
K
L
J
M
Fig. 3. The N-terminal domain of p130-angiomotin induces actin fiber formation. (A–C) MAE cells retrofected with p80-angiomotin double- stained with angiomotin rabbit polyclonal antibodies (green) and Texas red-conjugated phalloidin (red) show that the p80 isoform localized to the lamellipodia of migrating cells. (D–F) The same staining of MAE p130-angiomotin retrofected cells yields a different subcellular localiz- ation. p130-Angiomotin localizes to actin fibers in a punctuated pattern. (G–I) MAE cells transfected with the flag-tagged construct encoding the N-terminal domain of p130-angiomotin were stained with a flag antibody and phalloidin. The flag antibody staining shows complete over- lap with F-actin. (J–L) MAE vector cells stained with angiomotin polyclonal antibodies and phalloidin. (M) Three-dimensional imaging of p130- angiomotin and phalloidin staining show colocalization of p130-angiomotin and F-actin aggregates in the actin fibers. Scale bar: A–L, 20 lm; M, 1 lm.
FEBS Journal 273 (2006) 2000–2011 ª 2006 The Authors Journal compilation ª 2006 FEBS
2004
M. Ernkvist et al.
p130-Angiomotin controls cell shape
A
B
C
D
E
F
G
H
I
J
K
L
Fig. 4. p130-Angiomotin remained assoc- iated with F-actin after disruption of actin fibers. (A–C) MAE–p130-angiomotin cells stained for angiomotin (green) and F-actin (phalloidin, red). (D–F) Same staining of MAE–p130-angiomotin cells after 1 h of 50 lgÆmL)1 cytochalasin B treatment. Aggregated structures of p130-angiomotin entirely overlap with the disrupted actin fibers. (G–I) MAE cells transfected with the flag tagged N-terminus of p130-angiomotin stained with the flag antibody and phalloidin. (J–L) Same staining of the N-terminus trans- fected MAE cells 1 h after the addition of 50 lgÆmL)1 cytochalasin B. Aggregated structures of the N-terminus entirely overlap with the disrupted actin fibers. Scale bar: 20 lm.
staining of either
the N-terminal domain of p130-angiomotin into MAE cells (Fig. 4G–I) and treated these cells with cytochala- sin B. Again, the N-terminal staining overlapped with the aggregated structures of actin (Fig. 4J–L). To ensure that the staining was not due to a cell collapse, we also stained the treated cells for tubulin and cytochrome C. Cytochalasin B treatment did not effect tubulin or cytochrome C (Fig. S3A,B). We also treated p80-angiomotin-trans- fected MAE cells with cytochalasin B. Here, the la- mellipodia staining of p80-angiomotin was disrupted, whereas the general staining detected in all p80-angi- omotin-positive cells was not affected (Fig. S3C). of the cells. Cells positive for angiomotin or N-ter- minal staining were photographed and the cellular area and the number of actin fibers were measured as des- cribed in Experimental procedures. The median area of the MAE cells expressing p130-angiomotin or the N-terminal domain was more than two times larger than the p80-angiomotin and the vector cells as shown in the box plot diagram in Fig. 5A. The numbers of actin fibers in MAE cells expressing p130-angiomotin or the N-terminal domain were quadrupled compared with p80-angiomotin or vector cells (Fig. S4). These results show that p130-angiomotin regulates cytoskele- ton organization and cell shape through the N-ter- minal part of the protein.
p130-Angiomotin and p80-angiomotin display different effects on cell shape and motility
FEBS Journal 273 (2006) 2000–2011 ª 2006 The Authors Journal compilation ª 2006 FEBS
2005
The p130-angiomotin retrofected MAE cells and the transfected MAE cells contained more N-terminal actin fibers and displayed a flattened morphology com- pared with vector and p80-angiomotin-expressing cells. This observation prompted us to measure the cell area and quantify actin fibers. The cells were stained for angiomotin or N-terminal expression together with phalloidin to visualize the outline and the actin fibers Our previous studies have shown that p80-angio- motin promotes migration of transfected cells [18]. We therefore used an in vitro wound-healing assay to investigate the effect of p130-angiomotin expression on the migration rate of MAE cells. In this assay, MAE cells transfected with p80-angiomotin, p130-angiomo- tin or vector were grown to confluence and wounds were subsequently generated by scraping with a pip- ette. The migration rate was estimated by analyzing the distance of the leading edge of the migrating cells from the edge of the wound. In these experiments,
M. Ernkvist et al.
p130-Angiomotin controls cell shape
B
A
C
Fig. 5. p130-Angiomotin and p80-angiomotin display different effects on cell shape, cell motility and the response to angiostatin. (A) MAE–p130-angiomotin, MAE cells transfect- ed with the N-terminal domain, MAE–p80- angiomotin cells and MAE–vector were stained and cell area was estimates as des- cribed in Experimental procedures. Cells expressing p130-angiomotin or its N-term- inal fragment are more than twice as large as p80-angiomotin and vector cells. (B) MAE cells expressing p130-, p80-angiomotin and vector were grown to confluency and wounds were generated by scraping with a pipette. The migration rate was measured after 3 and 6 h. p80-Angiomotin cells migrat- ed almost twice as fast as p130-angiomotin cells and vector cells. (C) MAE cells expres- sing p130-angiomotin, p80-angiomotin and vector were analyzed in the Boyden cham- ber assay. Migration of p80-angiomotin was stimulated by FGF-2 and inhibited by angio- statin, whereas p130-angiomotin and vector cells were stimulated by FGF-2 but not inhibited by angiostatin. Plasminogen, the parent molecule of angiostatin, did not affect migration in any cell line. **P < 0.01, ***P < 0.001.
expression of p80-angiomotin stimulated migration almost twofold, whereas p130-angiomotin did not enhance the migration rate compared with vector cells (Fig. 5B).
Cell shape and migration of p130-angiomotin expressing cells is not affected by angiostatin
[18]. We therefore compared the migration of p130- angiomotin, p80-angiomotin, and vector cells in the presence or absence of angiostatin. Migration of vec- tor-transfected cells was stimulated in the presence of basic fibroblast growth factor (FGF-2), but was not inhibited in the presence of angiostatin or plasminogen (Fig. 5C). The FGF-2-stimulated migration of p80- angiomotin-expressing cells was inhibited by angio- statin, whereas plasminogen had no effect (Fig. 5C). In contrast to p80-angiomotin cells, the basal migration of p130-angiomotin-expressing cells was reduced, and there was no significant difference in the migration rate of p130-angiomotin cells treated with angiostatin.
Discussion
We have recently shown that both isoforms of angio- motin are membrane proteins and that their angiosta- tin-binding domains are localized on the cell surface [23]. Therefore, we analyzed whether MAE cells expressing p130-angiomotin respond to angiostatin. First, we examined if angiostatin could affect the cell shape of the p130-angiomotin positive cells. The cells were incubated with 5 lgÆmL)1 angiostatin overnight and fixed and analyzed the following day. The results showed that angiostatin treatment did not significantly affect the cell area of p130-angiomotin cells (Fig. S5).
FEBS Journal 273 (2006) 2000–2011 ª 2006 The Authors Journal compilation ª 2006 FEBS
2006
We identified a novel splice isoform of angiomotin that associates with actin and regulates cell shape. p130- Angiomotin contains an extended N-terminus that is localized to the cytoplasm, otherwise, the protein is identical to p80-angiomotin. p130-Angiomotin is an alternative splice variant of p80-angiomotin. The p130- angiomotin sequence contains no potential proteolytic cleavage sites that could mediate the conversion from It has previously been shown that angiostatin does not inhibit EC migration in the wound-healing assay [24], therefore, we assessed the response to angiostatin in the Boyden chamber assay. In this assay, MAE cells expressing p80-angiomotin are inhibited by angiostatin
M. Ernkvist et al.
p130-Angiomotin controls cell shape
the that demonstrates
p130-angiomotin to p80-angiomotin. Therefore, we conclude that p80-angiomotin is not mediated by pro- teolytic cleavage of p130-angiomotin. the C-terminus
C-terminal angiomotin region, corresponding to p80-angiomotin, organizes p130-angiomotin in compartments along the actin fibers. Furthermore, concentrates the cell, p130-angiomotin to the central part of whereas expression of the N-terminus alone is localized over the whole cell. The different
intracellular, whereas the
The first 243 amino acids in the N-terminal domain show 58% homology to Amotl-1 and Amotl-2. Little is known regarding the cellular localization and func- tion of Amotl-2, but Amotl-1 (JEAP) has been shown to be expressed in exocrine cells and to localize to ZO-1 in tight junctions in vitro and in vivo [19]. We have recently shown that both isoforms of angiomotin colocalize with ZO-1 in cell contacts [23]. We also showed that angiostatin binds to angiomotin on the cell surface and we proposed a model of the topology in which both the N- and C-terminus of angiomotin are angiostatin-binding domain is extracellular [23]. In this article, we have characterized the different functions of p130-angiomo- tin compared with p80-angiomotin.
therefore, rule out
intracellular expression pattern of retrofected p80-angiomotin and p130-angiomotin in MAE cells also argues that the isoforms have distinct cellular functions. We have previously shown that MAE cells overexpressing p80-angiomotin increase the migration rate twofold compared with vector cells [18]. Furthermore, p80-angiomotin expression promotes tumor growth and invasion into surrounding muscle tissue in vivo [29]. A role for angiomotin in cell migra- tion is further emphasized by the finding that angiom- otin-deficient mice die in utero at E7–8 due to a migration defect in the anterior visceral endoderm [22]. In a previous report [23], we showed that Chinese hamseter ovary cells transfected with p130-angiomotin respond to angiostatin. However, the transfected cells also expressed a low level of p80-angiomotin. We could not, that p80-angiomotin could be responsible for the observed antimigratory effect of angiostatin. In this report, we isolated clones that exclusively express p130-angiomotin and we were able to show that these cells migrated slower and did not respond to angiostatin.
The subcellular localization of p130-angiomotin was distinct from that of p80-angiomotin in MAE cells. In contrast to the lamellipodia staining observed with p80-angiomotin in subconfluent cells, p130-angiomotin was expressed in a regular punctuated pattern overlap- ping with F-actin. This staining pattern did not over- lap with that of tensin found in fibrillar adhesions [25] (Fig. S6A) or a-actinin, an actin-binding protein [26,27] (Fig. S6B). We also tested if p130-angiomotin was colocalizing with the focal adhesion protein paxil- lin. In the end of the p130-angiomotin fibers, some overlap could be found, but in general, they did not colocalize (Fig. S6C).
than cells
The different effects of angiostatin on the two differ- ent isforms indicate that the signaling pathways from the extracellular angiostatin-binding domain differ between the two isoforms. In line with the migration data, angiostatin did not affect the actin fiber formation or the cell shape of p130-angiomotin-expressing cells.
p130-Angiomotin is expressed during embryogenesis, albeit at a lower level than p80-angiomotin. This differ- is not detected in the adult ence in expression level mouse, as both isoforms are expressed in the brain, heart, lung and liver and p130-angiomotin has a higher expression in some of the tissues. In the placenta, angio- motin is expressed by EC and cytotrophoblasts.
fragment
FEBS Journal 273 (2006) 2000–2011 ª 2006 The Authors Journal compilation ª 2006 FEBS
2007
Cells expressing p130-angiomotin contained more actin fibers expressing p80-angiomotin, resulting in altered cell shape and an approximately twofold increase in cell area. Staining of MAE cells transfected with the cytoplasmic N-terminal domain of p130-angiomotin showed a uniform colocalization with F-actin, as opposed to the punctuated pattern observed with cells retrofected with p130-angiomotin plasmid. Expression of the N-terminal fragment resul- ted in a similar accumulation of actin fibers, which resulted in increased cell shape as seen with p130- angiomotin. The staining of both p130-angiomotin and the N-terminal remained associated with phalloidin staining after treatment with cytochalasin B, an actin-depolymerizing agent. This does not contra- dict the result that the angiostatin-binding domain of p130-angiomotin is exposed at the cell surface because it has been shown that the localization of other trans- membrane proteins such as E-cadherin, Claudin-1 and JAM-1 is disrupted after treatment with cyto- chalasin D, another inhibitor of actin polymerization [28]. The punctuated membrane localization of p130- A vascular tube is formed by signaling from sur- rounding tissue that induces the directed migration of cells [30]. Tube formation also involves reshaping of the cytoskeleton of cells to form a closed monolayer, which surrounds a hollow fluid filled lumen [31]. Here we show that ECs express two isoforms of angiomotin, p80 and p130. We speculate that these isoforms play distinct roles during angiogenesis in which p80-angio- motin promotes cell migration and p130-angiomotin is involved in changing the morphology of ECs during tubulogenesis.
M. Ernkvist et al.
p130-Angiomotin controls cell shape
Experimental procedures
phate coprecipitation. The culture supernatant was added to MAE cells and incubated for 24 h. Subpopulations of the p130-angiomotin-expressing MAE cells also express low levels of p80-angiomotin. This is due to the usage of the internal ATG within the p130 cDNA. To get pure p130- angiomotin-expressing cells we subcloned the cells and veri- fied their expression with western blot.
Cell lines and reagents
Spontaneously immortalized MAE cells [32] and ecotropic retrovirus producing Phoenix eco cells provided by G. Nolan, (Stanford University, Palo Alto, CA) were grown in Dulbecco’s modified Eagle’s medium (DMEM) (Sigma, St Louis, MO) with 10% fetal bovine serum, 1% glutamine and 1% penicillin ⁄ streptomycin.
Stably transfected MAE cells were grown in the presence of 5 lgÆmL)1 puromycin (vector, p80-angiomotin, and p130-angiomotin) or 800 lgÆmL)1 G418 (p130-specific N-terminal).
Rabbit polyclonal antibodies against different domains of angiomotin were generated. A fragment of the 266 amino acids specific for p130-angiomotin was used to produce the N-terminal antibodies. The angiostatin-binding domain of angiomotin was used to generate antibodies against the extracellular part of angiomotin [18]. Peptides correspond- ing to the 24 most C-terminal amino acids of angiomotin were used to generate C-terminal-specific antibodies. Anti- bodies towards the angiostatin-binding domain and the C-terminus were affinity purified.
Angiomotin antibodies
293T cells and HeLa cells were grown in DMEM supple- mented with 10% donor bovine serum and antibiotics. M21 melanoma cells (a kind gift from Dr Staffan Stro¨ mblad, Karolinska Institute, Sweden) were grown in RPMI-1640 containing 5–10% fetal calf serum and 2 mm glutamine. Bovine aortic endothelial (BAE) cells were grown in DMEM with 10% fetal bovine serum, 1% glutamine and 1% penicil- lin ⁄ streptomycin. hTERT+ bovine capillary endothelial (BCE) cells were grown in DMEM with 10% fetal bovine serum, 2 ngÆmL)1 FGF-2, 1% glutamine and 1% penicil- lin ⁄ streptomycin. PmT-EC cells (endothelial cells derived from embryonic stem cells and immortalized with polyoma middle T virus) [33] were grown in DMEM with 10% fetal bovine serum (HyClone, Logan, UT), 1% glutamine, 1% penicillin ⁄ streptomycin, 50 lgÆmL)1 endothelial cell growth supplement (Sigma) and 100 lgÆmL)1 heparin (Sigma).
Angiostatin (K1–4) was generated from human-plasma derived plasminogen. Plasminogen was digested by elastase and then purified using lysine sepharose. Residual kringle 4 fragment was removed by gel filtration chromatography [16]. All preparations used in this report were tested for endotoxin levels using the limulus assay (BioWhittaker, Walkersville, MD). Preparations containing < 1 ngÆmL)1 were used for the experiments.
Immunoprecipitation
Cells were extracted with lysis buffer (50 mm, Tris ⁄ HCl pH 7.4, 0.2% Nonidet P40, 150 mm NaCl, 1 mm EDTA and protease inhibitors). Lysates were immunoprecipitated with either preimmune or immune sera for 1 h at 4 (cid:1)C. Pel- lets were washed three times with lysis buffer and resus- pended in 1· sample buffer and analyzed by SDS ⁄ PAGE. For in gel digestion and MS, 5 g of 293T cells were solubi- lized in 50 mL of lysis buffer. The supernatant was clarified by centrifugation at 17000 r.p.m. for 20 min. Five hundred microliters of a 50% slurry of preimmune and immune antibodies coupled to Sepharose CL-4B was used for immunoprecipitation at 4 (cid:1)C for 90 min. The pellet was washed three times with RIPA buffer (50 mm Tris pH 7.4, 150 mm NaCl, 1 mm EDTA, 1% Nonidet P40, 1% deoxy- cholic acid, 0.1% SDS and protease inhibitors). Samples were eluted with 100 mm ethanolamine for 2 min at room temperature, denatured with Lamelli sample buffer and analyzed on SDS ⁄ PAGE. Following silver staining, bands were excised and subjected to MS.
Plasmid construction and transfection
The p80- and p130-angiomotin were subcloned into the pENTR-2B vector (Gateway, Invitrogen, Carlsbad, CA) and then recombined into the converted Gateway destination vec- tor pBABE (provided by Dr H. Land, LCRF, London, UK) using the LR recombination reaction (Gateway, Invitrogen). The N-terminal construct was a kind gift from Dr A. Shi- mono, Center for Developmental Biology, Kobe, Japan.
Cells were transfected using Lipofectamine 2000 accord- ing to the protocol of the manufacturer (Life Technologies, Grand Island, NY).
MS of proteolytic fragments from p130 sequence showed peptides that came from p80 and from what was then con- sidered the 5¢ untranslated region of p80 angiomotin (bold- face in Fig. 1B) and from two ESTs in the data base (BI 058583 and CK001189.1). The extension matched a sequence in the genomic DNA, 7 kb upstream from the 5¢-end of from an the p80 cDNA suggesting it was upstream exon. Because we were unable to acquire these ESTs, we used a PAC clone (AC004827.1) containing the
For retrofection, Phoenix Ecotropic packaging cell line was transfected with the pBABE, pBABE-p80-angiomotin, or pBABE-p130-angiomotin using standard calcium phos-
FEBS Journal 273 (2006) 2000–2011 ª 2006 The Authors Journal compilation ª 2006 FEBS
2008
Cloning of p130-angiomotin
M. Ernkvist et al.
p130-Angiomotin controls cell shape
genomic sequence; PCR amplified the upstream exon and ligated it to the 5¢-end of the angiomotin p80 cDNA. The resulting cDNA contained the published sequence for ESTs BI058583 and CK001189.1 and the entire open reading frame of p130 angiomotin. The database used was http:// www.ncbi.nlm.nih.gov/entrez/querry.
a Zeiss Axioplan 2 microscope, and images collected using an AxioCam HRm Camera and axiovision 4.2 software. Cells treated with cytochalasin B (Sigma) were treated with 50 lgÆmL)1 cytochalasin B 1 h prior to fixation. Cells treated with Y-27632 (Calbiochem) were treated with 10 lgÆmL)1 Y-27632 3 h prior to fixation. Then the cells were stained as described above.
Cells were stained with the C-terminal-specific angiomotin polyclonal antibodies for detection of positive cells and then stained with phalloidin to visualize the whole cell and the actin fibers. Photos were taken with the AxioCam HRm camera attached to the Zeiss Axioplan 2 micro- scope. Eighty-six N-terminus expressing cells and (cid:1) 180 p130-angiomotin, p80-angiomotin and vector cells were measured. The cell areas were measured in Fujifilm image gauge version 4.0 for the cell sizes. The actin fibers were manually counted. Box plots and statistical calculations were done in statview 5.0.1.
Western blot Quantification of cell area and actin fibers
Cell lysates were analyzed by SDS ⁄ PAGE and proteins were transferred to nitrocellulose membrane. Non-specific bind- ing was blocked for 1 h in 10% dried milk in NaCl ⁄ Pi con- taining 0.1% Tween (NaCl ⁄ Pi-T). The filter was incubated over night at 4 (cid:1)C in 5% dried milk in NaCl ⁄ Pi-T with ang- iomotin C-terminal-specific polyclonal antibodies (1 : 1000) or polyclonal antibodies towards the extracellular domain of angiomotin (1 : 2000). Secondary anti-(rabbit-HRP) was diluted 1 : 10 000 (Amersham Pharmacia Biotech, Piscata- way, NJ) and incubated for 1 h at room temperature. After- wards, the nitrocellulose membrane was visualized using a detection system from Santa Cruz Biotechnology Inc (Santa Cruz, CA) in an intelligent dark box (Fujifilm) with the program las1000 (Fujifilm). The membrane containing placenta and embryo proteins purchased from RNWAY laboratories (Korea) was incubated as above.
In vitro wound healing
Cells were plated in chamber slides and grown over night to confluency. A wound was made with a 10 lL pipette. Photos of the wounds were taken after 30 min (time point 0), 3.5 h (time point 3 h) and 6.5 h (time point 6 h). The distance between the cells in the wound edges was meas- ured. Time points 3 and 6 h were compared with time point 0 to show rate of wound closure. Statistical calculations were done in statveiw 5.0.1.
Preparation of tissues for western blot
Different tissues were collected from a C57 ⁄ B6 mouse that was killed according to directive N403 ⁄ 04, following proce- dures approved by the ethical committee of Stockholm North, Sweden. The tissues were put in a lysis buffer con- taining 1% Triton X-100, 1% sodiumdeoxychelate, 50 mm Tris pH 8.0, 100 mm NaCl, 5 mm EDTA and protease inhibitors. A two times loading buffer containing 100 mm Tris pH 6.8, 4% SDS and 20% glycerol was used. The samples were sonicated, boiled for 10 min and then centri- fuged for 10 min. The western blot was loaded with equal amount of protein for every organ, measured by Bradford.
Migration assay
Migration assays were performed in a modified Boyden chamber using a 48-well chemotaxis chamber (Neuroprobe Inc., Gaithersburg, MD) as described earlier [18]. Briefly, 8 lm Nucleopore polyvinylpirrolidine-free polycarbonate filters were coated with 100 lgÆmL)1 of collagen type 1 (Cohesion, Palo Alto, CA) overnight. hTERT+-BCE cells were starved in 0.2% FCS–DMEM for 16 h. In order to test the inhibitory activity of angiostatin, hTERT+-BCE, and vector and angiomotin MAE cells were preincubated with 5 lgÆmL)1 of angiostatin for 1 h. The cells were trypsi- nized, resuspended in DMEM containing 0.1% bovine serum albumin (BSA) and 30 000 cells were added with or without angiostatin to each well of the upper chamber. FGF-2 (Peprotech EC Ltd, Rocky Hill, NJ) at 20 ngÆmL)1 was used as a chemoattractant in the lower chambers. The chemotaxis chambers were incubated for 3–5 h at 37 (cid:1)C with 10% CO2 to allow cells to migrate through the colla- gen-coated polycarbonate filter. Nonmigrated cells on the upper surface of the filter were removed and the filter was
Cultured cells were plated in chamber slides and allowed to grow and adhere over night. The cells were fixed in 4% PFA for 10 min at room temperature. When needed, the cells were permeabilized in 0.1% Triton X-100 (Sigma) for one minute. Nonspecific reactivity was blocked by incuba- ting with 5% horse serum in NaCl ⁄ Pi for 1 h before addi- tion of primary antibody in blocking buffer for 1 h. Antibody binding was detected with fluorescent-labeled sec- ondary antibodies (Dako, Carpinteria, CA and Molecular Probes, Eugene, OR). F-Actin was visualized with Texas red phalloidin (Molecular Probes). The slides were mounted with mounting media from Vector Laboratories, viewed on
FEBS Journal 273 (2006) 2000–2011 ª 2006 The Authors Journal compilation ª 2006 FEBS
2009
Immunofluorescence
M. Ernkvist et al.
p130-Angiomotin controls cell shape
Herman S, Sarkar PK et al. (2000) Anti-angiogenic cues from vascular basement membrane collagen. Cancer Res 60, 2520–2526.
stained with Giemsa Stain (VWR International Ltd, West Chester, PA). The total number of migrated cells per field was counted at ·20 magnification; each sample was tested in quadruplicates in three independent experiments.
Acknowledgements
13 Kamphaus GD, Colorado PC, Panka DJ, Hopfer H, Ramchandran R, Torre A, Maeshima Y, Mier JW, Sukhatme VP & Kalluri R (2000) Canstatin, a novel matrix-derived inhibitor of angiogenesis and tumor growth. J Biol Chem 275, 1209–1215.
14 Mongiat M, Sweeney SM, San Antonio JD, Fu J & Iozzo RV (2003) Endorepellin, a novel inhibitor of angiogenesis derived from the C terminus of perlecan. J Biol Chem 278, 4238–4249.
15 Folkman J (2004) Endogenous angiogenesis inhibitors.
This work was supported by grants from the Swedish Cancer Society, Swedish Society of Medicine, Cancer- fo¨ reningen Stockholm, the Karolinska Institute, RO1 CA60499 and RO1 HL 64597.
References
Apmis 112, 496–507.
1 Carmeliet P & Jain RK (2000) Angiogenesis in cancer
and other diseases. Nature 407, 249–257.
16 O’Reilly MS, Holmgren L, Shing Y, Chen C, Rosenthal RA, Moses M, Lane WS, Cao Y, Sage EH & Folkman J (1994) Angiostatin: a novel angiogenesis inhibitor that mediates the suppression of metastases by a Lewis lung carcinoma. Cell 79, 315–328.
2 Folkman J (1995) Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1, 27–31. 3 Folkman J (1995) Seminars in Medicine of the Beth
17 O’Reilly MS, Holmgren L, Chen C & Folkman J (1996) Angiostatin induces and sustains dormancy of human primary tumors in mice. Nat Med 2, 689–692.
Israel Hospital, Boston. Clinical applications of research on angiogenesis. N Engl J Med 333, 1757–1763. 4 Folkman J (1971) Tumor angiogenesis: therapeutic
implications. N Engl J Med 285, 1182–1186. 5 Holmgren L, O’Reilly MS & Folkman J (1995)
18 Troyanovsky B, Levchenko T, Mansson G, Matvijenko O & Holmgren L (2001) Angiomotin: an angiostatin binding protein that regulates endothelial cell migration and tube formation. J Cell Biol 152, 1247–1254. 19 Nishimura M, Kakizaki M, Ono Y, Morimoto K,
Dormancy of micrometastases: balanced proliferation and apoptosis in the presence of angiogenesis suppres- sion. Nat Med 1, 149–153.
Takeuchi M, Inoue Y, Imai T & Takai Y (2002) JEAP, a novel component of tight junctions in exocrine cells. J Biol Chem 277, 5583–5587.
6 Hanahan D & Folkman J (1996) Patterns and emerging mechanisms of the angiogenic switch during tumorigen- esis. Cell 86, 353–364.
7 Bouck N, Stellmach V & Hsu SC (1996) How tumors
become angiogenic. Adv Cancer Res 69, 135–174. 8 Good DJ, Polverini PJ, Rastinejad F, Le Beau MM,
20 Bratt A, Wilson WJ, Troyanovsky B, Aase K, Kessler R, Van Meir EG & Holmgren L (2002) Angiomotin belongs to a novel protein family with conserved coiled-coil and PDZ binding domains. Gene 298, 69–77.
Lemons RS, Frazier WA & Bouck NP (1990) A tumor suppressor-dependent inhibitor of angiogenesis is immu- nologically and functionally indistinguishable from a fragment of thrombospondin. Proc Natl Acad Sci USA 87, 6624–6628.
21 Moreau J, Lord M, Boucher M, Belleau P & Fernandes MJ (2005) Protein diversity is generated within the motin family of proteins by alternative pre-mRNA splicing. Gene 350, 137–148.
9 Dawson DW, Volpert OV, Gillis P, Crawford SE,
22 Shimono A & Behringer RR (2003) Angiomotin regu- lates visceral endoderm movements during mouse embryogenesis. Curr Biol 13, 613–617.
Xu H, Benedict W & Bouck NP (1999) Pigment epithe- lium-derived factor: a potent inhibitor of angiogenesis. Science 285, 245–248.
23 Bratt A, Birot O, Sinha I, Veitonmaki N, Aase K,
10 O’Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G,
Ernkvist M & Holmgren L (2005) Angiomotin regulates endothelial cell-cell junctions and cell motility. J Biol Chem 280, 34859–34869.
24 Shichiri M & Hirata Y (2001) Antiangiogenesis signals
Lane WS, Flynn E, Birkhead JR, Olsen BR & Folkman J (1997) Endostatin: an endogenous inhibitor of angio- genesis and tumor growth. Cell 88, 277–285. 11 Hamano Y, Zeisberg M, Sugimoto H, Lively JC,
by endostatin. Faseb J 15, 1044–1053.
25 Zamir E, Katz BZ, Aota S, Yamada KM, Geiger B & Kam Z (1999) Molecular diversity of cell–matrix adhe- sions. J Cell Sci 112 (11), 1655–1669.
26 Belkin AM & Koteliansky VE (1987) Interaction of
Maeshima Y, Yang C, Hynes RO, Werb Z, Sudhakar A & Kalluri R (2003) Physiological levels of tumstatin, a fragment of collagen IV alpha3 chain, are generated by MMP-9 proteolysis and suppress angiogenesis via alphaV beta3 integrin. Cancer Cell 3, 589–601.
iodinated vinculin, metavinculin and alpha-actinin with cytoskeletal proteins. FEBS Lett 220, 291–294.
12 Colorado PC, Torre A, Kamphaus G, Maeshima Y, Hopfer H, Takahashi K, Volk R, Zamborsky ED,
FEBS Journal 273 (2006) 2000–2011 ª 2006 The Authors Journal compilation ª 2006 FEBS
2010
M. Ernkvist et al.
p130-Angiomotin controls cell shape
27 Otey CA, Pavalko FM & Burridge K (1990) An interac- tion between alpha-actinin and the beta 1 integrin subu- nit in vitro. J Cell Biol 111, 721–729.
28 Yamada A, Irie K, Fukuhara A, Ooshio T & Takai Y (2004) Requirement of the actin cytoskeleton for the association of nectins with other cell adhesion molecules at adherens and tight junctions in MDCK cells. Genes Cells 9, 843–855.
29 Levchenko T, Bratt A, Arbiser JL & Holmgren L
(2004) Angiomotin expression promotes hemangio- endothelioma invasion. Oncogene 23, 1469–1473. 30 Carmeliet P (2000) Mechanisms of angiogenesis and
arteriogenesis. Nat Med 6, 389–395.
31 Bazzoni G & Dejana E (2004) Endothelial cell-to-cell junctions: molecular organization and role in vascular homeostasis. Physiol Rev 84, 869–901.
32 Bastaki M, Nelli EE, Dell’Era P, Rusnati M, Molinari- Tosatti MP, Parolini S, Auerbach R, Ruco LP, Possati L & Presta M (1997) Basic fibroblast growth factor- induced angiogenic phenotype in mouse endothelium. A study of aortic and microvascular endothelial cell lines. Arterioscler Thromb Vasc Biol 17, 454–464.
33 Balconi G, Spagnuolo R & Dejana E (2000) Development of endothelial cell lines from embryonic stem cells: a tool for studying genetically manipulated endothelial cells in vitro. Arterioscler Thromb Vasc Biol 20, 1443–1451.
Supplementary material
is available
F-actin was also found when the 293T cells were stained with antibodies towards the specific N-terminal of p130-angiomotin. (F) Transfection of the construct encoding the N-terminal domain of p130-angiomotin into M21 cells shows that the N-terminus forms a fiber-like pattern that colocalizes with actin fibers. (G) The same staining pattern is also found in HeLa cells transfected with the N-terminus. Scale bar: 10 lm. Fig. S3. Cytochalasin B treatment does not affect the localization of tubulin or cytochrome c, but disrupts the lamellipodia localization of p80-angiomotin. (A) Untreated cells show the same tubulin staining as cells treated with cytochalasin B, whereas the actin staining was disrupted when treated with cytochalasin B. (B) Untreated cells show the same cytochrome C staining as cells treated with cytochalasin B, whereas the actin staining was disrupted when treated with cytochala- sin B. (C) Untreated stably retrofected p80-angiomotin MAE cells show a lamellipodia staining. Treatment with cytochalsin B disrupted lamellipodia staining, whereas the general staining detected in all p80-angio- motin positive cells was not affected. Scale bar: 20 lm. Fig. S4. p130-Angiomotin induces stress fiber forma- tion. The same cells that were measured for cell size were also quantified for the amount of actin fibers. Cells expressing the N-terminal domain of p130-angi- omotin and the full-length p130-angiomotin contained about four times more actin fibers than p80-angiomo- tin-expressing cells and vector cells. ***P<0.001. Fig. S5. Angiostatin treatment does not affect the cell area of p130-angiomotin cells. The cells were incubated with 5 mgÆmL)1 angiostatin overnight, and fixed and analyzed the following day. Angiostatin treatment had no effect on the cell area of p130-angiomotin cells. Fig. S6. The p130-angiomotin staining does not over- lap with the staining of tensin, a-actinin or paxillin. (A) Double staining of p130-angiomotin and tensin in MAE cells show no overlap. (B) The staining pattern of a-actinin is very similar to that of p130-angiomotin. However, double staining of a-actinin and p130-angio- motin shows that they do not colocalize. (C) Double staining of p130-angiomotin and paxillin shows no direct overlap between the two proteins, but some- times, paxillin is located just in the end of the p130- angiomotin fibers. This material is available as part of the online article from http://www.blackwell-synergy.com
FEBS Journal 273 (2006) 2000–2011 ª 2006 The Authors Journal compilation ª 2006 FEBS
2011
The following supplementary material online: Fig. S1. Expression of angiomotin during E12. Overex- posure of embryonic lysate from E12 shows that p130- angiomotin (arrow head) is weakly expressed. Fig. S2. M21 and HeLa cells transfected with p130- angiomotin and the N-terminal domain of p130-angi- omotin, and endogenous expression of angiomotin in PmT-EC cells and 293T cells. (A) Transfection of p130-angiomotin into M21 cells shows a regular pat- tern of small, linear dots, the same staining pattern as in stably retrofected MAE cells. (B) PmT-EC cells stained with the C-terminal-specific antibodies (green) show a punctuated pattern that overlap with F-actin (red). (C) PmT-EC cells stained with the N-terminal- specific antibodies show the same punctuated pattern (green) that overlaps with F-actin (red). (D) 293T cells were stained with the polyclonal antibodies towards both isoforms and with phalloidin. The staining shows a punctuated pattern overlapping F-actin. (E) The localization that colocalizes with same subcellular
