When animals develop from embryos to adults, or try to heal wounds later in life, their cells have to move. Moving means that the cells must invade into their surroundings, a dense network of proteins called the extracellular matrix. The cell first attaches to the extracellular matrix; degrades it; and then moves into the newly opened space. Cells have developed specialized structures called invadosomes to enable all these steps. Invadosomes are never static, they first assemble where cells interact with extracellular matrix, they then release proteins that loosen the matrix, and finally disassemble again to allow cells to move. Invadosomes in cancer cells often become overactive, and can allow the tumor cells to spread throughout the body.

A lot of different proteins are involved in controlling how and when cells move. p27 is a well-known protein usually found in a cell’s nucleus along with the cell’s DNA. Inside the nucleus, p27 suppresses tumor growth by stopping cells from dividing. However, often in cancer cells p27 moves outside of the cell’s nucleus where it contributes to cell movement via an unknown mechanism.

To answer how p27 controls cell invasion, Jeannot et al. used a biochemical technique to uncover which proteins p27 binds to when it is outside of the nucleus. One of its interaction partners was called Cortactin. This protein is known to be an important building block of invadosomes, and is involved in both the assembly and disassembly of these structures. In further experiments, Jeannot studied mouse cells with or without p27 and human cancer cells that can be grown in the laboratory. The experiments revealed that p27 promotes an enzyme called PAK1 to also bind to Cortactin. PAK1 then modified Cortactin, causing whole invadosomes to disassemble, which in turn allowed cells to de-attach from the matrix and move forward. In contrast, cells lacking p27 had more stable invadosomes, attached more strongly to the matrix and were better at degrading it, but could not invade as well as cells with p27.

Overall these experiments showed a new way that p27 promotes cell invasion. The next steps will include finding out exactly how the modification of Cortactin causes the invadosomes to disassemble. Furthermore, it will be important to study whether forcing p27 back into the nucleus can reduce the spread of cancer cells in the body.

p27^Kip1 is a cell cycle inhibitor that binds to a broad range of cyclin-CDK complexes (Besson et al., 2008). p27-mediated cyclin-CDK inhibition involves the N-terminal extremity (aa 28–89) of the protein that allows interaction with both cyclin and CDK subunits (Ou et al., 2012; Besson et al., 2008). The critical role of p27 as a negative regulator of cell proliferation is underscored by the phenotype of p27 knockout mice, which exhibit a hyperproliferative phenotype in multiple tissues and are more susceptible to tumor development than wild-type animals (Fero et al., 1996; Kiyokawa et al., 1996; Nakayama et al., 1996; Besson et al., 2006; Fero et al., 1998). While the tumor suppressive role of p27 is confirmed by the frequent loss of p27 expression in many types of cancers, p27 is also mislocalized in a fraction of tumors due to activation of various signaling pathways that cause its nuclear export and/or cytoplasmic retention (Chu et al., 2008). Cytoplasmic p27 has been associated with decreased patient survival, high tumor grade and metastasis in several tumor types including breast carcinomas, acute myeloid leukemia, glioblastoma, melanoma and non-small cell lung carcinomas (Lin et al., 2016; Chu et al., 2008; Liang et al., 2002; Yang et al., 2011; Min et al., 2004; Cheng et al., 2015). The possibility that cytoplasmic p27 may confer a pro-tumorigenic advantage was confirmed in a mouse model expressing a mutant form of p27 that cannot bind cyclin-CDK complexes (p27CK-) (Besson et al., 2007; Serres et al., 2011). These mice are more susceptible than p27-null and wild-type mice to both spontaneous and induced tumor development, thus uncovering an oncogenic role for p27 in vivo (Serres et al., 2011; Besson et al., 2007).

It now appears that p27 is a multifunctional protein involved in the regulation of multiple cellular processes, some of which are regulated by p27 in a CDK-independent manner (Besson et al., 2008; Sharma and Pledger, 2016). Indeed, p27 has been implicated in the control of cell migration, transcriptional repression, autophagy, stem cell specification and differentiation, cytokinesis, and apoptosis (Besson et al., 2008; Sharma and Pledger, 2016; Baldassarre et al., 2005; Besson et al., 2004b, 2007; Pippa et al., 2012; Serres et al., 2012; Nickeleit et al., 2008; Liang et al., 2007; Li et al., 2012; Nguyen et al., 2006; Jeannot et al., 2015). This functional versatility may stem from the intrinsically disordered structure of p27, which folds upon binding to other proteins, allowing p27 to interact with a wide variety of partners (Lacy et al., 2004; Galea et al., 2008). While these various functions participate in homeostasis in normal cells and tissues, they may be co-opted by tumor cells to promote oncogenesis.

Metastasis, the process whereby cells from a primary tumor disseminate throughout the body and form secondary tumors at distant sites, is the leading cause of cancer related death (Sethi and Kang, 2011). It requires tumor cells to migrate and invade through neighboring tissues, enter the blood flow and then extravasate to disseminate and form new tumors in host tissues (Sethi and Kang, 2011). Regulation of cell motility was the first CDK-independent role ascribed to p27 (Nagahara et al., 1998; Denicourt et al., 2007; Besson et al., 2004b). In the cytosol, p27 can interact with RhoA, preventing RhoA activation by guanine-nucleotide exchange factors (GEFs) and thereby modulating actin cytoskeleton dynamics and migration (Besson et al., 2004a, 2004b). Consequently, mouse embryo fibroblasts (MEFs) lacking p27 have more RhoA activity, increased numbers of stress fibers and focal adhesions and exhibit a defect in migration (Besson et al., 2004b). This pathway is important for proper migration of bone marrow macrophages and developing cortical neurons and for cancer cell migration and invasion in vivo (Godin et al., 2012; Nguyen et al., 2006; Papakonstanti et al., 2007; Gui et al., 2014; Wu et al., 2006; See et al., 2010; Larrea et al., 2009; Jin et al., 2013). p27 can also regulate cell migration by controlling microtubule stability through Stathmin or directly by binding to microtubules and promoting microtubule polymerization (Baldassarre et al., 2005; Godin et al., 2012). In a 3D environment, cells adopt different strategies to migrate and invade through surrounding matrix (Petrie and Yamada, 2016; Sethi and Kang, 2011). The mode of migration is influenced, at least in part, by the activities of Rho GTPases: RhoA activity favors an amoeboid migration, whereas Rac1 activity promotes mesenchymal migration (Sahai and Marshall, 2003; Vial et al., 2003). Accordingly, cytoplasmic p27 promotes mesenchymal migration via the inhibition of RhoA activity, while cells lacking p27 tend to adopt an amoeboid mode of migration (Gui et al., 2014; Belletti et al., 2010). More recently, p27 was reported to promote epithelial to mesenchymal transition by binding to JAK2, promoting STAT3 activation and the upregulation of Twist1 (Zhao et al., 2015).

Here, we describe a novel mechanism by which p27 directly regulates invadopodia formation and cell invasion through extracellular matrix (ECM) via Cortactin. Invadosomes (designating both invadopodia and podosomes) are thought to allow cells to coordinate ECM degradation with migration within the tissue microenvironment (Murphy and Courtneidge, 2011; Linder et al., 2011; Di Martino et al., 2016). Cortactin plays a key role in the formation of actin protrusive structures such as lamellipodia and invadosomes (Murphy and Courtneidge, 2011; Kirkbride et al., 2011; MacGrath and Koleske, 2012). Cortactin has been involved in all steps of the invadosome lifecycle, from assembly, maturation, proteolytic activity and disassembly (MacGrath and Koleske, 2012; Murphy and Courtneidge, 2011; Moshfegh et al., 2014). All these steps appear to be regulated by phosphorylation events on Cortactin (Oser et al., 2009; Moshfegh et al., 2014; Murphy and Courtneidge, 2011). Overall, Cortactin is a scaffold protein composed of an N-terminal acidic domain, followed by several actin-binding repeats allowing its interaction with F-actin, a Pro-rich region and a SH3 domain (MacGrath and Koleske, 2012; Kirkbride et al., 2011). Phosphorylation of Cortactin by Src and Abl family tyrosine kinases has two effects: first, this releases the actin severing protein Cofilin from an inhibitory interaction with Cortactin and generates actin barbed ends; second, Arp2/3, N-WASP and Nck1 are recruited onto Cortactin, promoting the polymerization of branched actin (Oser et al., 2009; Weaver et al., 2002; Oser et al., 2010). Dephosphorylation of Cortactin then allows the inhibition of Cofilin activity and stabilizes the invadopodia (Oser et al., 2009). Cortactin also plays a key role in matrix metalloproteinase secretion in mature invadopodia (Clark et al., 2007). Finally, invadopodia disassembly is induced by sequential activation of the Rac-GEF Trio, Rac1 and p21-Activated Kinase-1 (PAK1) pathway and presumably by PAK1-mediated phosphorylation of Cortactin on Ser113, since a Ser113 to Ala mutant blocked invadopodia disassembly (Moshfegh et al., 2014). How PAK-phosphorylated Cortactin mediates invadopodia disassembly is still unclear but could be due, at least in part, to a decreased affinity of Cortactin phosphorylated within its F-actin binding repeats (on S113, S150 and/or S282) for F-actin, potentially destabilizing the structure (Webb et al., 2006, 2005).

We found that p27 binds to Cortactin and localizes to invadopodia following serum or growth factor stimulation and promotes cell invasion. Paradoxically, p27-null cells have more invadopodia and degrade gelatin more efficiently, but exhibit impaired invasive capacity. In fact, we found that p27 promotes the recruitment of PAK1 to Cortactin and invadopodia turnover and this is dependent on phosphorylation of S113/S150/S282 of Cortactin, the sites targeted by PAK kinases. Thus, in absence of p27, the dynamics of invadopodia is altered and prevents efficient invasion through ECM. Altogether, we have identified a novel mechanism by which p27 directly controls cell invasion that could contribute to the increased invasive and metastatic capacity of tumors where p27 is mislocalized in the cytoplasm.

Our results indicate that p27 regulates invadopodia formation and activity by binding to Cortactin, probably via the regulation of invadopodia turnover mediated by a signaling pathway recently characterized whereby signaling through Trio/Rac1/PAK1/Cortactin induces the disassembly of invadopodia (Moshfegh et al., 2014). The main finding presented here is that p27 directly controls cellular invasion by impinging on this pathway through the promotion of the Cortactin/PAK1 interaction and the phosphorylation of Cortactin on Ser113, S150 and/or S282 by PAK. Indeed, inhibiting the pathway upstream of Cortactin, either with Rac1 and PAK1 inhibitors or siRNAs converted the phenotype of p27+/+ cells to that of p27−/− cells. Conversely, artificially activating this pathway at the level of Cortactin using phosphomimetic mutants for the PAK-targeted sites efficiently rescued the phenotype of p27-null cells. All three phosphorylation sites on Cortactin appear to act synergistically as the phenotype caused by the triple mutants (Cortactin TA and TD) was more profound than that of mutating S113 alone. While mimicking the phosphorylation of Cortactin on PAK1-targeted sites clearly causes the disassembly of invadopodia (Moshfegh et al., 2014; Webb et al., 2005), the mechanism involved is still unclear but may be due to decreased affinity of Cortactin phosphorylated within its Actin-binding repeats for F-Actin, which may destabilize invadopodia (Webb et al., 2006).

Results from Moshfegh et al. (2014) and the present study show that inhibition of the Trio/Rac1/PAK1/Phospho-Cortactin pathway stabilizes invadopodia and is associated with a dramatic increase in proteolytic degradation of ECM. Surprisingly, this phenotype is accompanied by a sharp decrease in the ability of cells to invade through matrix in 3D invasion assays. Thus, it appears that the turnover of invadopodia rather than elevated ECM degradation is required for efficient invasion. Affecting the dynamics of invadopodia formation, maturation and disassembly is therefore detrimental to invasion, as observed here when this last step is inhibited by interfering with Rac1/PAK1 signaling or when p27 is absent. Interestingly, p27 also regulates cell migration in a similar manner by interfering with RhoA activation by its GEFs (Besson et al., 2004a, 2004b). In absence of p27, cells accumulate actins stress fibers and focal adhesions and exhibit a migration defect due to an imbalance in the dynamic cycles of RhoA activation and inactivation (and concomitant Rac1 inactivation and activation) required for F-actin turnover, focal adhesion disassembly and efficient migration (Besson et al., 2004a, 2004b; Ren et al., 2000; Arthur and Burridge, 2001; Cox et al., 2001; Sahai et al., 2001; Vial et al., 2003). Migration and invasion are two intimately linked processes (Friedl and Wolf, 2003; Petrie and Yamada, 2016). It is quite remarkable that p27 appears to regulate both of these processes, although by distinct mechanisms. Indeed, p27 binds to RhoA and Cortactin via the same domain located within the last 8 aa of the protein (Figure 1E) (Larrea et al., 2009; Godin et al., 2012), but while RhoA preferentially binds p27 phosphorylated on T198 (Larrea et al., 2009), the interaction of p27 with Cortactin did not require the presence of T198 (Figure 1E), suggesting that different pools of the protein are involved in their regulation. In addition, the regulation of invadopodia by p27 did not appear to involve RhoA, as RhoA silencing affected both p27−/− and p27+/+ cells in a similar manner (Figure 7—figure supplement 1). p27 also regulates invasion by promoting EMT, notably through JAK2/STAT3 signaling and induction of the transcription factor Twist (Zhao et al., 2015).

We found that p27 binds to Cortactin on its actin-binding repeats and this interaction promoted the association of Cortactin with PAK1. Since we were unable to detect an interaction between p27 and PAK1 using either endogenous or overexpressed proteins, it remains unknown how p27 promotes the PAK1/Cortactin interaction. Evidence suggest that Cortactin is an intrinsically disordered protein (IDP), especially its actin-binding repeats (Shvetsov et al., 2009) and p27 is a well-known IDP (Galea et al., 2008; Lacy et al., 2004; Ou et al., 2012). Since these proteins typically fold upon binding to their partners, an attractive possibility is that p27 binding causes a conformation change in Cortactin that would favor its interaction with PAK1.

p27 is a cyclin-CDK inhibitor and although p27 regulates migration and invasion in a CDK-independent manner (Besson et al., 2004b; Larrea et al., 2009; Zhao et al., 2015; Gui et al., 2014), there is evidence to suggest that some cell cycle independent roles of CDKs and cyclins may be mediated via CDK inhibitors such as p27 (Hydbring et al., 2016). For instance, the regulation of cell migration by Cyclin D1 requires p27 (Li et al., 2006) and the role of CDK5 in promoting neuronal progenitor migration in vivo is mediated via the phosphorylation of p27 on Ser10 which regulates the RhoA/ROCK/LIMK/Cofilin pathway (Nguyen et al., 2006; Kawauchi et al., 2006). Interestingly, CDK5 was recently found to play an important role in invadopodia formation and cancer cell invasion, and CDK5 inhibition blocked both these processes (Quintavalle et al., 2011; Bisht et al., 2015). CDK5 phosphorylated Caldesmon, causing its degradation by the proteasome and promoting invadopodia formation and invasion (Quintavalle et al., 2011). Since p27 is a poor inhibitor of CDK5 complexes (Lacy et al., 2005; Lee et al., 1996), it would be interesting to test whether this role of CDK5 may involve p27. Indeed, an attractive hypothesis is that p27 could simultaneously interact with CDK5 complexes, via its N-terminal cyclin-CDK binding domain, and with Cortactin via its C-terminus, thereby recruiting CDK5 in proximity to Caldesmon bound to Cortactin and allowing Caldesmon phosphorylation.

Our study further defines the complex role played by p27 in the regulation of cytoskeletal dynamics, migration, EMT and invasion and provides another element to explain why the Cdkn1b gene is rarely mutated in cancer (Chu et al., 2008; Besson et al., 2008; Kandoth et al., 2013). Indeed, p27 is either downregulated, mostly via increased proteasomal degradation, or excluded from the nuclei of cancer cells. Given that upon cytoplasmic relocalization, p27 promotes both migration and invasion and may serve to coordinately regulate these processes, it appears likely that this feature may be selected for during tumor progression and could participate in the acquisition of a metastatic behavior by cancer cells.

1. 1. Arthur WT
   2. Burridge K

   (2001) RhoA inactivation by p190RhoGAP regulates cell spreading and migration by promoting membrane protrusion and polarity

   Molecular Biology of the Cell 12:2711–2720.

   https://doi.org/10.1091/mbc.12.9.2711

   • PubMed
   • Google Scholar
2. 1. Baldassarre G
   2. Belletti B
   3. Nicoloso MS
   4. Schiappacassi M
   5. Vecchione A
   6. Spessotto P
   7. Morrione A
   8. Canzonieri V
   9. Colombatti A

   (2005) p27(Kip1)-stathmin interaction influences sarcoma cell migration and invasion

   Cancer Cell 7:51–63.

   https://doi.org/10.1016/j.ccr.2004.11.025

   • PubMed
   • Google Scholar
3. 1. Belletti B
   2. Pellizzari I
   3. Berton S
   4. Fabris L
   5. Wolf K
   6. Lovat F
   7. Schiappacassi M
   8. D'Andrea S
   9. Nicoloso MS
   10. Lovisa S
   11. Sonego M
   12. Defilippi P
   13. Vecchione A
   14. Colombatti A
   15. Friedl P
   16. Baldassarre G

   (2010) p27kip1 controls cell morphology and motility by regulating microtubule-dependent lipid raft recycling

   Molecular and Cellular Biology 30:2229–2240.

   https://doi.org/10.1128/MCB.00723-09

   • PubMed
   • Google Scholar
4. 1. Besson A
   2. Assoian RK
   3. Roberts JM

   (2004a) Regulation of the cytoskeleton: an oncogenic function for CDK inhibitors?

   Nature Reviews Cancer 4:948–955.

   https://doi.org/10.1038/nrc1501

   • PubMed
   • Google Scholar
5. 1. Besson A
   2. Gurian-West M
   3. Schmidt A
   4. Hall A
   5. Roberts JM

   (2004b) p27Kip1 modulates cell migration through the regulation of RhoA activation

   Genes & Development 18:862–876.

   https://doi.org/10.1101/gad.1185504

   • PubMed
   • Google Scholar
6. 1. Besson A
   2. Gurian-West M
   3. Chen X
   4. Kelly-Spratt KS
   5. Kemp CJ
   6. Roberts JM

   (2006) A pathway in quiescent cells that controls p27Kip1 stability, subcellular localization, and tumor suppression

   Genes & Development 20:47–64.

   https://doi.org/10.1101/gad.1384406

   • PubMed
   • Google Scholar
7. 1. Besson A
   2. Hwang HC
   3. Cicero S
   4. Donovan SL
   5. Gurian-West M
   6. Johnson D
   7. Clurman BE
   8. Dyer MA
   9. Roberts JM

   (2007) Discovery of an oncogenic activity in p27Kip1 that causes stem cell expansion and a multiple tumor phenotype

   Genes & Development 21:1731–1746.

   https://doi.org/10.1101/gad.1556607

   • PubMed
   • Google Scholar
8. 1. Besson A
   2. Dowdy SF
   3. Roberts JM

   (2008) CDK inhibitors: cell cycle regulators and beyond

   Developmental Cell 14:159–169.

   https://doi.org/10.1016/j.devcel.2008.01.013

   • PubMed
   • Google Scholar
9. 1. Bian Y
   2. Song C
   3. Cheng K
   4. Dong M
   5. Wang F
   6. Huang J
   7. Sun D
   8. Wang L
   9. Ye M
   10. Zou H

   (2014) An enzyme assisted RP-RPLC approach for in-depth analysis of human liver phosphoproteome

   Journal of Proteomics 96:253–262.

   https://doi.org/10.1016/j.jprot.2013.11.014

   • PubMed
   • Google Scholar
10. 1. Bisht S
    2. Nolting J
    3. Schütte U
    4. Haarmann J
    5. Jain P
    6. Shah D
    7. Brossart P
    8. Flaherty P
    9. Feldmann G

    (2015) Cyclin-Dependent kinase 5 (CDK5) Controls melanoma cell motility, invasiveness, and metastatic Spread-Identification of a promising novel therapeutic target

    Translational Oncology 8:295–307.

    https://doi.org/10.1016/j.tranon.2015.06.002

    • PubMed
    • Google Scholar
11. 1. Boehm M
    2. Yoshimoto T
    3. Crook MF
    4. Nallamshetty S
    5. True A
    6. Nabel GJ
    7. Nabel EG

    (2002) A growth factor-dependent nuclear kinase phosphorylates p27^(Kip1) and regulates cell cycle progression

    The EMBO Journal 21:3390–3401.

    https://doi.org/10.1093/emboj/cdf343

    • PubMed
    • Google Scholar
12. 1. Castro-Castro A
    2. Janke C
    3. Montagnac G
    4. Paul-Gilloteaux P
    5. Chavrier P

    (2012) ATAT1/MEC-17 acetyltransferase and HDAC6 deacetylase control a balance of acetylation of alpha-tubulin and cortactin and regulate MT1-MMP trafficking and breast tumor cell invasion

    European Journal of Cell Biology 91:950–960.

    https://doi.org/10.1016/j.ejcb.2012.07.001

    • PubMed
    • Google Scholar
13. 1. Chen L
    2. Giorgianni F
    3. Beranova-Giorgianni S

    (2010) Characterization of the phosphoproteome in LNCaP prostate Cancer cells by in-gel isoelectric focusing and tandem mass spectrometry

    Journal of Proteome Research 9:174–178.

    https://doi.org/10.1021/pr900338q

    • PubMed
    • Google Scholar
14. 1. Cheng Y
    2. Lu J
    3. Chen G
    4. Ardekani GS
    5. Rotte A
    6. Martinka M
    7. Xu X
    8. McElwee KJ
    9. Zhang G
    10. Zhou Y

    (2015) Stage-specific prognostic biomarkers in melanoma

    Oncotarget 6:4180–4189.

    https://doi.org/10.18632/oncotarget.2907

    • PubMed
    • Google Scholar
15. 1. Chong C
    2. Tan L
    3. Lim L
    4. Manser E

    (2001) The mechanism of PAK activation. autophosphorylation events in both regulatory and kinase domains control activity

    Journal of Biological Chemistry 276:17347–17353.

    https://doi.org/10.1074/jbc.M009316200

    • PubMed
    • Google Scholar
16. 1. Chu IM
    2. Hengst L
    3. Slingerland JM

    (2008) The cdk inhibitor p27 in human Cancer: prognostic potential and relevance to anticancer therapy

    Nature Reviews Cancer 8:253–267.

    https://doi.org/10.1038/nrc2347

    • PubMed
    • Google Scholar
17. 1. Clark ES
    2. Whigham AS
    3. Yarbrough WG
    4. Weaver AM

    (2007) Cortactin is an essential regulator of matrix metalloproteinase secretion and extracellular matrix degradation in invadopodia

    Cancer Research 67:4227–4235.

    https://doi.org/10.1158/0008-5472.CAN-06-3928

    • PubMed
    • Google Scholar
18. 1. Connor MK
    2. Kotchetkov R
    3. Cariou S
    4. Resch A
    5. Lupetti R
    6. Beniston RG
    7. Melchior F
    8. Hengst L
    9. Slingerland JM

    (2003) CRM1/Ran-mediated nuclear export of p27(Kip1) involves a nuclear export signal and links p27 export and proteolysis

    Molecular Biology of the Cell 14:201–213.

    https://doi.org/10.1091/mbc.E02-06-0319

    • PubMed
    • Google Scholar
19. 1. Cox EA
    2. Sastry SK
    3. Huttenlocher A

    (2001) Integrin-mediated adhesion regulates cell polarity and membrane protrusion through the rho family of GTPases

    Molecular Biology of the Cell 12:265–277.

    https://doi.org/10.1091/mbc.12.2.265

    • PubMed
    • Google Scholar
20. 1. Denicourt C
    2. Saenz CC
    3. Datnow B
    4. Cui XS
    5. Dowdy SF

    (2007) Relocalized p27Kip1 tumor suppressor functions as a cytoplasmic metastatic oncogene in melanoma

    Cancer Research 67:9238–9243.

    https://doi.org/10.1158/0008-5472.CAN-07-1375

    • PubMed
    • Google Scholar
21. 1. Di Martino J
    2. Henriet E
    3. Ezzoukhry Z
    4. Goetz JG
    5. Moreau V
    6. Saltel F

    (2016) The microenvironment controls invadosome plasticity

    Journal of Cell Science 129:1759–1768.

    https://doi.org/10.1242/jcs.182329

    • PubMed
    • Google Scholar
22. 1. Fero ML
    2. Rivkin M
    3. Tasch M
    4. Porter P
    5. Carow CE
    6. Firpo E
    7. Polyak K
    8. Tsai LH
    9. Broudy V
    10. Perlmutter RM
    11. Kaushansky K
    12. Roberts JM

    (1996) A syndrome of Multiorgan Hyperplasia with features of gigantism, Tumorigenesis, and female sterility in p27(Kip1)-deficient mice

    Cell 85:733–744.

    https://doi.org/10.1016/S0092-8674(00)81239-8

    • PubMed
    • Google Scholar
23. 1. Fero ML
    2. Randel E
    3. Gurley KE
    4. Roberts JM
    5. Kemp CJ

    (1998) The murine gene p27Kip1 is haplo-insufficient for tumour suppression

    Nature 396:177–180.

    https://doi.org/10.1038/24179

    • PubMed
    • Google Scholar
24. 1. Friedl P
    2. Wolf K

    (2003) Tumour-cell invasion and migration: diversity and escape mechanisms

    Nature Reviews Cancer 3:362–374.

    https://doi.org/10.1038/nrc1075

    • PubMed
    • Google Scholar
25. 1. Galea CA
    2. Nourse A
    3. Wang Y
    4. Sivakolundu SG
    5. Heller WT
    6. Kriwacki RW

    (2008) Role of intrinsic flexibility in signal transduction mediated by the cell cycle regulator, p27 Kip1

    Journal of Molecular Biology 376:827–838.

    https://doi.org/10.1016/j.jmb.2007.12.016

    • PubMed
    • Google Scholar
26. 1. Gao Y
    2. Dickerson JB
    3. Guo F
    4. Zheng J
    5. Zheng Y

    (2004) Rational design and characterization of a rac GTPase-specific small molecule inhibitor

    PNAS 101:7618–7623.

    https://doi.org/10.1073/pnas.0307512101

    • PubMed
    • Google Scholar
27. 1. Godin JD
    2. Thomas N
    3. Laguesse S
    4. Malinouskaya L
    5. Close P
    6. Malaise O
    7. Purnelle A
    8. Raineteau O
    9. Campbell K
    10. Fero M
    11. Moonen G
    12. Malgrange B
    13. Chariot A
    14. Metin C
    15. Besson A
    16. Nguyen L

    (2012) p27(Kip1) is a microtubule-associated protein that promotes microtubule polymerization during neuron migration

    Developmental Cell 23:729–744.

    https://doi.org/10.1016/j.devcel.2012.08.006

    • PubMed
    • Google Scholar
28. 1. Gui P
    2. Labrousse A
    3. Van Goethem E
    4. Besson A
    5. Maridonneau-Parini I
    6. Le Cabec V
    7. Goethem, E VAN
    8. Cabec LE V

    (2014) Rho/ROCK pathway inhibition by the CDK inhibitor p27(kip1) participates in the onset of macrophage 3D-mesenchymal migration

    Journal of Cell Science 127:4009–4023.

    https://doi.org/10.1242/jcs.150987

    • PubMed
    • Google Scholar
29. 1. Hydbring P
    2. Malumbres M
    3. Sicinski P

    (2016) Non-canonical functions of cell cycle cyclins and cyclin-dependent kinases

    Nature Reviews Molecular Cell Biology 17:280–292.

    https://doi.org/10.1038/nrm.2016.27

    • PubMed
    • Google Scholar
30. 1. Ishida N
    2. Hara T
    3. Kamura T
    4. Yoshida M
    5. Nakayama K
    6. Nakayama KI

    (2002) Phosphorylation of p27Kip1 on serine 10 is required for its binding to CRM1 and nuclear export

    Journal of Biological Chemistry 277:14355–14358.

    https://doi.org/10.1074/jbc.C100762200

    • PubMed
    • Google Scholar
31. 1. Jeannot P
    2. Callot C
    3. Baer R
    4. Duquesnes N
    5. Guerra C
    6. Guillermet-Guibert J
    7. Bachs O
    8. Besson A

    (2015) Loss of p27Kip¹ promotes metaplasia in the pancreas via the regulation of Sox9 expression

    Oncotarget 6:35880–35892.

    https://doi.org/10.18632/oncotarget.5770

    • PubMed
    • Google Scholar
32. 1. Jin X
    2. Jin X
    3. Sohn YW
    4. Yin J
    5. Kim SH
    6. Joshi K
    7. Nam DH
    8. Nakano I
    9. Kim H

    (2013) Blockade of EGFR signaling promotes glioma stem-like cell invasiveness by abolishing ID3-mediated inhibition of p27(KIP1) and MMP3 expression

    Cancer Letters 328:235–242.

    https://doi.org/10.1016/j.canlet.2012.09.005

    • PubMed
    • Google Scholar
33. 1. Kandoth C
    2. McLellan MD
    3. Vandin F
    4. Ye K
    5. Niu B
    6. Lu C
    7. Xie M
    8. Zhang Q
    9. McMichael JF
    10. Wyczalkowski MA
    11. Leiserson MD
    12. Miller CA
    13. Welch JS
    14. Walter MJ
    15. Wendl MC
    16. Ley TJ
    17. Wilson RK
    18. Raphael BJ
    19. Ding L

    (2013) Mutational landscape and significance across 12 major Cancer types

    Nature 502:333–339.

    https://doi.org/10.1038/nature12634

    • PubMed
    • Google Scholar
34. 1. Katsube T
    2. Togashi S
    3. Hashimoto N
    4. Ogiu T
    5. Tsuji H

    (2004) Filamentous actin binding ability of cortactin isoforms is responsible for their cell-cell junctional localization in epithelial cells

    Archives of Biochemistry and Biophysics 427:79–90.

    https://doi.org/10.1016/j.abb.2004.04.015

    • PubMed
    • Google Scholar
35. 1. Kawauchi T
    2. Chihama K
    3. Nabeshima Y
    4. Hoshino M

    (2006) Cdk5 phosphorylates and stabilizes p27kip1 contributing to actin organization and cortical neuronal migration

    Nature Cell Biology 8:17–26.

    https://doi.org/10.1038/ncb1338

    • PubMed
    • Google Scholar
36. 1. Kirkbride KC
    2. Sung BH
    3. Sinha S
    4. Weaver AM

    (2011) Cortactin: a multifunctional regulator of cellular invasiveness

    Cell Adhesion & Migration 5:187–198.

    https://doi.org/10.4161/cam.5.2.14773

    • PubMed
    • Google Scholar
37. 1. Kiyokawa H
    2. Kineman RD
    3. Manova-Todorova KO
    4. Soares VC
    5. Hoffman ES
    6. Ono M
    7. Khanam D
    8. Hayday AC
    9. Frohman LA
    10. Koff A

    (1996) Enhanced growth of mice lacking the cyclin-dependent kinase inhibitor function of p27(Kip1)

    Cell 85:721–732.

    https://doi.org/10.1016/S0092-8674(00)81238-6

    • PubMed
    • Google Scholar
38. 1. Klammer M
    2. Kaminski M
    3. Zedler A
    4. Oppermann F
    5. Blencke S
    6. Marx S
    7. Müller S
    8. Tebbe A
    9. Godl K
    10. Schaab C

    (2012) Phosphosignature predicts dasatinib response in non-small cell lung Cancer

    Molecular & Cellular Proteomics 11:651–668.

    https://doi.org/10.1074/mcp.M111.016410

    • PubMed
    • Google Scholar
39. 1. Lacy ER
    2. Filippov I
    3. Lewis WS
    4. Otieno S
    5. Xiao L
    6. Weiss S
    7. Hengst L
    8. Kriwacki RW

    (2004) p27 binds cyclin-CDK complexes through a sequential mechanism involving binding-induced protein folding

    Nature Structural & Molecular Biology 11:358–364.

    https://doi.org/10.1038/nsmb746

    • PubMed
    • Google Scholar
40. 1. Lacy ER
    2. Wang Y
    3. Post J
    4. Nourse A
    5. Webb W
    6. Mapelli M
    7. Musacchio A
    8. Siuzdak G
    9. Kriwacki RW

    (2005) Molecular basis for the specificity of p27 toward cyclin-dependent kinases that regulate cell division

    Journal of Molecular Biology 349:764–773.

    https://doi.org/10.1016/j.jmb.2005.04.019

    • PubMed
    • Google Scholar
41. 1. Larrea MD
    2. Hong F
    3. Wander SA
    4. da Silva TG
    5. Helfman D
    6. Lannigan D
    7. Smith JA
    8. Slingerland JM

    (2009) RSK1 drives p27Kip1 phosphorylation at T198 to promote RhoA inhibition and increase cell motility

    PNAS 106:9268–9273.

    https://doi.org/10.1073/pnas.0805057106

    • PubMed
    • Google Scholar
42. 1. Lee MH
    2. Nikolic M
    3. Baptista CA
    4. Lai E
    5. Tsai LH
    6. Massagué J

    (1996) The brain-specific activator p35 allows Cdk5 to escape inhibition by p27Kip1 in neurons

    PNAS 93:3259–3263.

    https://doi.org/10.1073/pnas.93.8.3259

    • PubMed
    • Google Scholar
43. 1. Li Z
    2. Jiao X
    3. Wang C
    4. Ju X
    5. Lu Y
    6. Yuan L
    7. Lisanti MP
    8. Katiyar S
    9. Pestell RG

    (2006) Cyclin D1 induction of cellular migration requires p27(KIP1)

    Cancer Research 66:9986–9994.

    https://doi.org/10.1158/0008-5472.CAN-06-1596

    • PubMed
    • Google Scholar
44. 1. Li H
    2. Collado M
    3. Villasante A
    4. Matheu A
    5. Lynch CJ
    6. Cañamero M
    7. Rizzoti K
    8. Carneiro C
    9. Martínez G
    10. Vidal A
    11. Lovell-Badge R
    12. Serrano M

    (2012) p27(Kip1) directly represses Sox2 during embryonic stem cell differentiation

    Cell Stem Cell 11:845–852.

    https://doi.org/10.1016/j.stem.2012.09.014

    • PubMed
    • Google Scholar
45. 1. Liang J
    2. Zubovitz J
    3. Petrocelli T
    4. Kotchetkov R
    5. Connor MK
    6. Han K
    7. Lee JH
    8. Ciarallo S
    9. Catzavelos C
    10. Beniston R
    11. Franssen E
    12. Slingerland JM

    (2002) PKB/Akt phosphorylates p27, impairs nuclear import of p27 and opposes p27-mediated G1 arrest

    Nature Medicine 8:1153–1160.

    https://doi.org/10.1038/nm761

    • PubMed
    • Google Scholar
46. 1. Liang J
    2. Shao SH
    3. Xu ZX
    4. Hennessy B
    5. Ding Z
    6. Larrea M
    7. Kondo S
    8. Dumont DJ
    9. Gutterman JU
    10. Walker CL
    11. Slingerland JM
    12. Mills GB

    (2007) The energy sensing LKB1-AMPK pathway regulates p27(kip1) phosphorylation mediating the decision to enter autophagy or apoptosis

    Nature Cell Biology 9:218–224.

    https://doi.org/10.1038/ncb1537

    • PubMed
    • Google Scholar
47. 1. Licciulli S
    2. Maksimoska J
    3. Zhou C
    4. Troutman S
    5. Kota S
    6. Liu Q
    7. Duron S
    8. Campbell D
    9. Chernoff J
    10. Field J
    11. Marmorstein R
    12. Kissil JL

    (2013) FRAX597, a small molecule inhibitor of the p21-activated kinases, inhibits tumorigenesis of neurofibromatosis type 2 (NF2)-associated schwannomas

    Journal of Biological Chemistry 288:29105–29114.

    https://doi.org/10.1074/jbc.M113.510933

    • PubMed
    • Google Scholar
48. 1. Lin TC
    2. Tsai LH
    3. Chou MC
    4. Chen CY
    5. Lee H

    (2016) Association of cytoplasmic p27 expression with an unfavorable response to cisplatin-based chemotherapy and poor outcomes in non-small cell lung Cancer

    Tumor Biology 37:4017–4023.

    https://doi.org/10.1007/s13277-015-4272-7

    • PubMed
    • Google Scholar
49. 1. Linder S
    2. Wiesner C
    3. Himmel M

    (2011) Degrading devices: invadosomes in proteolytic cell invasion

    Annual Review of Cell and Developmental Biology 27:185–211.

    https://doi.org/10.1146/annurev-cellbio-092910-154216

    • PubMed
    • Google Scholar
50. 1. Lu W
    2. Katz S
    3. Gupta R
    4. Mayer BJ

    (1997) Activation of pak by membrane localization mediated by an SH3 domain from the adaptor protein nck

    Current Biology 7:85–94.

    https://doi.org/10.1016/S0960-9822(06)00052-2

    • PubMed
    • Google Scholar
51. 1. Lu W
    2. Mayer BJ

    (1999) Mechanism of activation of Pak1 kinase by membrane localization

    Oncogene 18:797–806.

    https://doi.org/10.1038/sj.onc.1202361

    • PubMed
    • Google Scholar
52. 1. MacGrath SM
    2. Koleske AJ

    (2012) Cortactin in cell migration and Cancer at a glance

    Journal of Cell Science 125:1621–1626.

    https://doi.org/10.1242/jcs.093781

    • PubMed
    • Google Scholar
53. 1. Malliri A
    2. Ten KJP
    3. Olivio C
    4. Collard JG

    (2000) Determination of the activity of Rho-like GTPases in cells

    Methods in Molecular Biology 189:99–109.

    https://doi.org/10.1385/1-59259-281-3:099

    • Google Scholar
54. 1. Mertins P
    2. Yang F
    3. Liu T
    4. Mani DR
    5. Petyuk VA
    6. Gillette MA
    7. Clauser KR
    8. Qiao JW
    9. Gritsenko MA
    10. Moore RJ
    11. Levine DA
    12. Townsend R
    13. Erdmann-Gilmore P
    14. Snider JE
    15. Davies SR
    16. Ruggles KV
    17. Fenyo D
    18. Kitchens RT
    19. Li S
    20. Olvera N
    21. Dao F
    22. Rodriguez H
    23. Chan DW
    24. Liebler D
    25. White F
    26. Rodland KD
    27. Mills GB
    28. Smith RD
    29. Paulovich AG
    30. Ellis M
    31. Carr SA

    (2014) Ischemia in tumors induces early and sustained phosphorylation changes in stress kinase pathways but does not affect global protein levels

    Molecular & Cellular Proteomics 13:1690–1704.

    https://doi.org/10.1074/mcp.M113.036392

    • PubMed
    • Google Scholar
55. 1. Mertins P
    2. Mani DR
    3. Ruggles KV
    4. Gillette MA
    5. Clauser KR
    6. Wang P
    7. Wang X
    8. Qiao JW
    9. Cao S
    10. Petralia F
    11. Kawaler E
    12. Mundt F
    13. Krug K
    14. Tu Z
    15. Lei JT
    16. Gatza ML
    17. Wilkerson M
    18. Perou CM
    19. Yellapantula V
    20. Huang KL
    21. Lin C
    22. McLellan MD
    23. Yan P
    24. Davies SR
    25. Townsend RR
    26. Skates SJ
    27. Wang J
    28. Zhang B
    29. Kinsinger CR
    30. Mesri M
    31. Rodriguez H
    32. Ding L
    33. Paulovich AG
    34. Fenyö D
    35. Ellis MJ
    36. Carr SA
    37. NCI CPTAC

    (2016) Proteogenomics connects somatic mutations to signalling in breast Cancer

    Nature 534:55–62.

    https://doi.org/10.1038/nature18003

    • PubMed
    • Google Scholar
56. 1. Min YH
    2. Cheong JW
    3. Kim JY
    4. Eom JI
    5. Lee ST
    6. Hahn JS
    7. Ko YW
    8. Lee MH

    (2004) Cytoplasmic mislocalization of p27Kip1 protein is associated with constitutive phosphorylation of akt or protein kinase B and poor prognosis in acute myelogenous leukemia

    Cancer Research 64:5225–5231.

    https://doi.org/10.1158/0008-5472.CAN-04-0174

    • PubMed
    • Google Scholar
57. 1. Moshfegh Y
    2. Bravo-Cordero JJ
    3. Miskolci V
    4. Condeelis J
    5. Hodgson L

    (2014) A Trio-Rac1-Pak1 signalling Axis drives invadopodia disassembly

    Nature Cell Biology 16:574–586.

    https://doi.org/10.1038/ncb2972

    • PubMed
    • Google Scholar
58. 1. Murphy DA
    2. Courtneidge SA

    (2011) The 'ins' and 'outs' of podosomes and invadopodia: characteristics, formation and function

    Nature Reviews Molecular Cell Biology 12:413–426.

    https://doi.org/10.1038/nrm3141

    • PubMed
    • Google Scholar
59. 1. Nagahara H
    2. Vocero-Akbani AM
    3. Snyder EL
    4. Ho A
    5. Latham DG
    6. Lissy NA
    7. Becker-Hapak M
    8. Ezhevsky SA
    9. Dowdy SF

    (1998) Transduction of full-length TAT fusion proteins into mammalian cells: tat-p27kip1 induces cell migration

    Nature Medicine 4:1449–1452.

    https://doi.org/10.1038/4042

    • PubMed
    • Google Scholar
60. 1. Nakayama K
    2. Ishida N
    3. Shirane M
    4. Inomata A
    5. Inoue T
    6. Shishido N
    7. Horii I
    8. Loh DY
    9. Nakayama K

    (1996) Mice lacking p27(Kip1) display increased body size, multiple organ Hyperplasia, retinal dysplasia, and pituitary tumors

    Cell 85:707–720.

    https://doi.org/10.1016/S0092-8674(00)81237-4

    • PubMed
    • Google Scholar
61. 1. Ndubaku CO
    2. Crawford JJ
    3. Drobnick J
    4. Aliagas I
    5. Campbell D
    6. Dong P
    7. Dornan LM
    8. Duron S
    9. Epler J
    10. Gazzard L
    11. Heise CE
    12. Hoeflich KP
    13. Jakubiak D
    14. La H
    15. Lee W
    16. Lin B
    17. Lyssikatos JP
    18. Maksimoska J
    19. Marmorstein R
    20. Murray LJ
    21. O'Brien T
    22. Oh A
    23. Ramaswamy S
    24. Wang W
    25. Zhao X
    26. Zhong Y
    27. Blackwood E
    28. Rudolph J

    (2015) Design of selective PAK1 inhibitor G-5555: improving properties by employing an unorthodox Low-pK a polar moiety

    ACS Medicinal Chemistry Letters 6:1241–1246.

    https://doi.org/10.1021/acsmedchemlett.5b00398

    • PubMed
    • Google Scholar
62. 1. Nguyen L
    2. Besson A
    3. Heng JI
    4. Schuurmans C
    5. Teboul L
    6. Parras C
    7. Philpott A
    8. Roberts JM
    9. Guillemot F

    (2006) p27kip1 independently promotes neuronal differentiation and migration in the cerebral cortex

    Genes & Development 20:1511–1524.

    https://doi.org/10.1101/gad.377106

    • PubMed
    • Google Scholar
63. 1. Nickeleit I
    2. Zender S
    3. Sasse F
    4. Geffers R
    5. Brandes G
    6. Sörensen I
    7. Steinmetz H
    8. Kubicka S
    9. Carlomagno T
    10. Menche D
    11. Gütgemann I
    12. Buer J
    13. Gossler A
    14. Manns MP
    15. Kalesse M
    16. Frank R
    17. Malek NP

    (2008) Argyrin a reveals a critical role for the tumor suppressor protein p27(kip1) in mediating antitumor activities in response to proteasome inhibition

    Cancer Cell 14:23–35.

    https://doi.org/10.1016/j.ccr.2008.05.016

    • PubMed
    • Google Scholar
64. 1. Olsen JV
    2. Vermeulen M
    3. Santamaria A
    4. Kumar C
    5. Miller ML
    6. Jensen LJ
    7. Gnad F
    8. Cox J
    9. Jensen TS
    10. Nigg EA
    11. Brunak S
    12. Mann M

    (2010) Quantitative phosphoproteomics reveals widespread full phosphorylation site occupancy during mitosis

    Science Signaling 3:ra3.

    https://doi.org/10.1126/scisignal.2000475

    • PubMed
    • Google Scholar
65. 1. Ong CC
    2. Gierke S
    3. Pitt C
    4. Sagolla M
    5. Cheng CK
    6. Zhou W
    7. Jubb AM
    8. Strickland L
    9. Schmidt M
    10. Duron SG
    11. Campbell DA
    12. Zheng W
    13. Dehdashti S
    14. Shen M
    15. Yang N
    16. Behnke ML
    17. Huang W
    18. McKew JC
    19. Chernoff J
    20. Forrest WF
    21. Haverty PM
    22. Chin SF
    23. Rakha EA
    24. Green AR
    25. Ellis IO
    26. Caldas C
    27. O'Brien T
    28. Friedman LS
    29. Koeppen H
    30. Rudolph J
    31. Hoeflich KP

    (2015) Small molecule inhibition of group I p21-activated kinases in breast Cancer induces apoptosis and potentiates the activity of microtubule stabilizing agents

    Breast Cancer Research 17:59.

    https://doi.org/10.1186/s13058-015-0564-5

    • PubMed
    • Google Scholar
66. 1. Oser M
    2. Yamaguchi H
    3. Mader CC
    4. Bravo-Cordero JJ
    5. Arias M
    6. Chen X
    7. Desmarais V
    8. van Rheenen J
    9. Koleske AJ
    10. Condeelis J

    (2009) Cortactin regulates cofilin and N-WASp activities to control the stages of invadopodium assembly and maturation

    The Journal of Cell Biology 186:571–587.

    https://doi.org/10.1083/jcb.200812176

    • PubMed
    • Google Scholar
67. 1. Oser M
    2. Mader CC
    3. Gil-Henn H
    4. Magalhaes M
    5. Bravo-Cordero JJ
    6. Koleske AJ
    7. Condeelis J

    (2010) Specific tyrosine phosphorylation sites on cortactin regulate Nck1-dependent actin polymerization in invadopodia

    Journal of Cell Science 123:3662–3673.

    https://doi.org/10.1242/jcs.068163

    • PubMed
    • Google Scholar
68. 1. Ou L
    2. Waddell MB
    3. Kriwacki RW

    (2012) Mechanism of cell cycle entry mediated by the intrinsically disordered protein p27(Kip1)

    ACS Chemical Biology 7:678–682.

    https://doi.org/10.1021/cb200487h

    • PubMed
    • Google Scholar
69. 1. Papakonstanti EA
    2. Ridley AJ
    3. Vanhaesebroeck B

    (2007) The p110delta isoform of PI 3-kinase negatively controls RhoA and PTEN

    The EMBO Journal 26:3050–3061.

    https://doi.org/10.1038/sj.emboj.7601763

    • PubMed
    • Google Scholar
70. 1. Perkins DN
    2. Pappin DJ
    3. Creasy DM
    4. Cottrell JS

    (1999) Probability-based protein identification by searching sequence databases using mass spectrometry data

    Electrophoresis 20:3551–3567.

    https://doi.org/10.1002/(SICI)1522-2683(19991201)20:18<3551::AID-ELPS3551>3.0.CO;2-2

    • PubMed
    • Google Scholar
71. 1. Petrie RJ
    2. Yamada KM

    (2016) Multiple mechanisms of 3D migration: the origins of plasticity

    Current Opinion in Cell Biology 42:7–12.

    https://doi.org/10.1016/j.ceb.2016.03.025

    • PubMed
    • Google Scholar
72. 1. Pippa R
    2. Espinosa L
    3. Gundem G
    4. García-Escudero R
    5. Dominguez A
    6. Orlando S
    7. Gallastegui E
    8. Saiz C
    9. Besson A
    10. Pujol MJ
    11. López-Bigas N
    12. Paramio JM
    13. Bigas A
    14. Bachs O

    (2012) p27Kip1 represses transcription by direct interaction with p130/E2F4 at the promoters of target genes

    Oncogene 31:4207–4220.

    https://doi.org/10.1038/onc.2011.582

    • PubMed
    • Google Scholar
73. 1. Quintavalle M
    2. Elia L
    3. Price JH
    4. Heynen-Genel S
    5. Courtneidge SA

    (2011) A cell-based high-content screening assay reveals activators and inhibitors of Cancer cell invasion

    Science Signaling 4:ra49.

    https://doi.org/10.1126/scisignal.2002032

    • PubMed
    • Google Scholar
74. 1. Ren XD
    2. Kiosses WB
    3. Sieg DJ
    4. Otey CA
    5. Schlaepfer DD
    6. Schwartz MA

    (2000)

    Focal adhesion kinase suppresses rho activity to promote focal adhesion turnover

    Journal of Cell Science 113 ( Pt 20:3637–3678.

    • PubMed
    • Google Scholar
75. 1. Rinschen MM
    2. Yu MJ
    3. Wang G
    4. Boja ES
    5. Hoffert JD
    6. Pisitkun T
    7. Knepper MA

    (2010) Quantitative phosphoproteomic analysis reveals vasopressin V2-receptor-dependent signaling pathways in renal collecting duct cells

    PNAS 107:3882–3887.

    https://doi.org/10.1073/pnas.0910646107

    • PubMed
    • Google Scholar
76. 1. Rodier G
    2. Montagnoli A
    3. Di Marcotullio L
    4. Coulombe P
    5. Draetta GF
    6. Pagano M
    7. Meloche S

    (2001) p27 cytoplasmic localization is regulated by phosphorylation on Ser10 and is not a prerequisite for its proteolysis

    The EMBO Journal 20:6672–6682.

    https://doi.org/10.1093/emboj/20.23.6672

    • PubMed
    • Google Scholar
77. 1. Spector DL

    (1997)

    Cells: A Laboratory Manual

    Isolation of fibroblasts, Cells: A Laboratory Manual, New York, Cold Spring Harbor Laboratory Press.

    • Google Scholar
78. 1. Sahai E
    2. Olson MF
    3. Marshall CJ

    (2001) Cross-talk between ras and rho signalling pathways in transformation favours proliferation and increased motility

    The EMBO Journal 20:755–766.

    https://doi.org/10.1093/emboj/20.4.755

    • PubMed
    • Google Scholar
79. 1. Sahai E
    2. Marshall CJ

    (2003) Differing modes of tumour cell invasion have distinct requirements for rho/ROCK signalling and extracellular proteolysis

    Nature Cell Biology 5:711–719.

    https://doi.org/10.1038/ncb1019

    • PubMed
    • Google Scholar
80. 1. Sander EE
    2. van Delft S
    3. ten Klooster JP
    4. Reid T
    5. van der Kammen RA
    6. Michiels F
    7. Collard JG

    (1998) Matrix-dependent Tiam1/Rac signaling in epithelial cells promotes either cell-cell adhesion or cell migration and is regulated by phosphatidylinositol 3-kinase

    The Journal of Cell Biology 143:1385–1398.

    https://doi.org/10.1083/jcb.143.5.1385

    • PubMed
    • Google Scholar
81. 1. Seals DF
    2. Azucena EF
    3. Pass I
    4. Tesfay L
    5. Gordon R
    6. Woodrow M
    7. Resau JH
    8. Courtneidge SA
    9. Pass JR
    10. Resau M
    11. J. H

    (2005) The adaptor protein Tks5/Fish is required for podosome formation and function, and for the protease-driven invasion of Cancer cells

    Cancer Cell 7:155–165.

    https://doi.org/10.1016/j.ccr.2005.01.006

    • PubMed
    • Google Scholar
82. 1. See WL
    2. Heinberg AR
    3. Holland EC
    4. Resh MD

    (2010) p27 deficiency is associated with migration defects in PDGF-expressing gliomas in vivo

    Cell Cycle 9:1562–1567.

    https://doi.org/10.4161/cc.9.8.11259

    • PubMed
    • Google Scholar
83. 1. Serres MP
    2. Zlotek-Zlotkiewicz E
    3. Concha C
    4. Gurian-West M
    5. Daburon V
    6. Roberts JM
    7. Besson A

    (2011) Cytoplasmic p27 is oncogenic and cooperates with ras both in vivo and in vitro

    Oncogene 30:2846–2858.

    https://doi.org/10.1038/onc.2011.9

    • PubMed
    • Google Scholar
84. 1. Serres MP
    2. Kossatz U
    3. Chi Y
    4. Roberts JM
    5. Malek NP
    6. Besson A

    (2012) p27(Kip1) controls cytokinesis via the regulation of Citron kinase activation

    Journal of Clinical Investigation 122:844–858.

    https://doi.org/10.1172/JCI60376

    • PubMed
    • Google Scholar
85. 1. Sethi N
    2. Kang Y

    (2011) Unravelling the complexity of metastasis - molecular understanding and targeted therapies

    Nature Reviews Cancer 11:735–748.

    https://doi.org/10.1038/nrc3125

    • PubMed
    • Google Scholar
86. 1. Sharma SS
    2. Pledger WJ

    (2016) The non-canonical functions of p27(Kip1) in normal and tumor biology

    Cell Cycle 15:1–13.

    https://doi.org/10.1080/15384101.2016.1157238

    • PubMed
    • Google Scholar
87. 1. Shvetsov A
    2. Berkane E
    3. Chereau D
    4. Dominguez R
    5. Reisler E

    (2009) The actin-binding domain of cortactin is dynamic and unstructured and affects lateral and longitudinal contacts in F-actin

    Cell Motility and the Cytoskeleton 66:90–98.

    https://doi.org/10.1002/cm.20328

    • PubMed
    • Google Scholar
88. 1. Sun Y
    2. Sun J
    3. Lungchukiet P
    4. Quarni W
    5. Yang S
    6. Zhang X
    7. Bai W

    (2015) Fe65 suppresses breast Cancer cell migration and invasion through Tip60 mediated cortactin acetylation

    Scientific Reports 5:11529.

    https://doi.org/10.1038/srep11529

    • PubMed
    • Google Scholar
89. 1. Vial E
    2. Sahai E
    3. Marshall CJ

    (2003) ERK-MAPK signaling coordinately regulates activity of Rac1 and RhoA for tumor cell motility

    Cancer Cell 4:67–79.

    https://doi.org/10.1016/S1535-6108(03)00162-4

    • PubMed
    • Google Scholar
90. 1. Vidal C
    2. Geny B
    3. Melle J
    4. Jandrot-Perrus M
    5. Fontenay-Roupie M

    (2002) Cdc42/Rac1-dependent activation of the p21-activated kinase (PAK) regulates human platelet lamellipodia spreading: implication of the cortical-actin binding protein cortactin

    Blood 100:4462–4469.

    https://doi.org/10.1182/blood.V100.13.4462

    • PubMed
    • Google Scholar
91. 1. Weaver AM
    2. Heuser JE
    3. Karginov AV
    4. Lee WL
    5. Parsons JT
    6. Cooper JA

    (2002) Interaction of cortactin and N-WASp with Arp2/3 complex

    Current Biology 12:1270–1278.

    https://doi.org/10.1016/S0960-9822(02)01035-7

    • PubMed
    • Google Scholar
92. 1. Webb BA
    2. Eves R
    3. Crawley SW
    4. Zhou S
    5. Côté GP
    6. Mak AS

    (2005) PAK1 induces podosome formation in A7r5 vascular smooth muscle cells in a PAK-interacting exchange factor-dependent manner

    AJP: Cell Physiology 289:C898–C907.

    https://doi.org/10.1152/ajpcell.00095.2005

    • PubMed
    • Google Scholar
93. 1. Webb BA
    2. Zhou S
    3. Eves R
    4. Shen L
    5. Jia L
    6. Mak AS

    (2006) Phosphorylation of cortactin by p21-activated kinase

    Archives of Biochemistry and Biophysics 456:183–193.

    https://doi.org/10.1016/j.abb.2006.06.011

    • PubMed
    • Google Scholar
94. 1. Wiśniewski JR
    2. Nagaraj N
    3. Zougman A
    4. Gnad F
    5. Mann M

    (2010) Brain phosphoproteome obtained by a FASP-based method reveals plasma membrane protein topology

    Journal of Proteome Research 9:3280–3289.

    https://doi.org/10.1021/pr1002214

    • PubMed
    • Google Scholar
95. 1. Wu FY
    2. Wang SE
    3. Sanders ME
    4. Shin I
    5. Rojo F
    6. Baselga J
    7. Arteaga CL

    (2006) Reduction of cytosolic p27(Kip1) inhibits Cancer cell motility, survival, and tumorigenicity

    Cancer Research 66:2162–2172.

    https://doi.org/10.1158/0008-5472.CAN-05-3304

    • PubMed
    • Google Scholar
96. 1. Yang J
    2. Liao D
    3. Wang Z
    4. Liu F
    5. Wu G

    (2011) Mammalian target of rapamycin signaling pathway contributes to glioma progression and patients' Prognosis

    Journal of Surgical Research 168:97–102.

    https://doi.org/10.1016/j.jss.2009.06.025

    • Google Scholar
97. 1. Zhang X
    2. Yuan Z
    3. Zhang Y
    4. Yong S
    5. Salas-Burgos A
    6. Koomen J
    7. Olashaw N
    8. Parsons JT
    9. Yang XJ
    10. Dent SR
    11. Yao TP
    12. Lane WS
    13. Seto E

    (2007) HDAC6 modulates cell motility by altering the acetylation level of cortactin

    Molecular Cell 27:197–213.

    https://doi.org/10.1016/j.molcel.2007.05.033

    • PubMed
    • Google Scholar
98. 1. Zhang Y
    2. Zhang M
    3. Dong H
    4. Yong S
    5. Li X
    6. Olashaw N
    7. Kruk PA
    8. Cheng JQ
    9. Bai W
    10. Chen J
    11. Nicosia SV
    12. Zhang X

    (2009) Deacetylation of cortactin by SIRT1 promotes cell migration

    Oncogene 28:445–460.

    https://doi.org/10.1038/onc.2008.388

    • PubMed
    • Google Scholar
99. 1. Zhao D
    2. Besser AH
    3. Wander SA
    4. Sun J
    5. Zhou W
    6. Wang B
    7. Ince T
    8. Durante MA
    9. Guo W
    10. Mills G
    11. Theodorescu D
    12. Slingerland J

    (2015) Cytoplasmic p27 promotes epithelial-mesenchymal transition and tumor metastasis via STAT3-mediated Twist1 upregulation

    Oncogene 34:5447–5459.

    https://doi.org/10.1038/onc.2014.473

    • PubMed
    • Google Scholar

[Editors’ note: a previous version of this study was rejected after peer review, but the authors submitted for reconsideration. The first decision letter after peer review is shown below.]

Thank you for submitting your work entitled "p27^Kip1 promotes invadopodia turnover and invasion through the regulation of the PAK1/Cortactin pathway" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by Fiona Watt as the Senior Editor.

Our decision has been reached after consultation between the reviewers. Based on these discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife.

This study demonstrates that p27 binding to Cortactin promotes PAK1 interactions with Cortactin, and invadopodia disassembly by phosphorylation of Cortactin on Ser-113. Biochemical analysis is presented to support these conclusions, including the demonstration that a phosphomimetic mutant of Cortactin can complement the p27 deficiency phenotype on invadopodia disassembly. Together, these aspects of the manuscript represent significant strengths. However, the detailed reviews (appended) identify a number of weaknesses in the data used to support the presented conclusions. Given the extensive new experimentation that will be required to address these weaknesses, a decision to reject your manuscript was made. Please be assured that eLife is interested in this area of research. Indeed, we would welcome the submission of a new manuscript on this topic when findings from new experimentation are obtained.

Reviewer #1:

This study demonstrates that p27 binding to Cortactin promotes PAK1 interactions with Cortactin, and invadopodia disassembly by phosphorylation of Cortactin on Ser-113. In general, the biochemical analysis is strong and convincing. Indeed, the demonstration that a phosphomimetic mutant of Cortactin can complement the p27 deficiency phenotype on invadopodia disassembly is very interesting. However, there are a number of areas that require strengthening.

1) The siRNA studies (Rac and RhoA) were performed with a single oligonucleotide. To exclude off0target effects, phenotypes should be confirmed with two independent siRNAs.

2) Some care should be taken concerning the PAK1/Cortactin connection. First, the studies presented are focussed on mutational analysis of the Cortactin phosphorylation site Ser-113 rather than the kinase. Second, in vitro studies demonstrate phosphorylation of Cortactin by PAK3 at more than one site (Webb et al.). Third, what is the evidence that PAK1 phosphorylates Cortactin in vivo? Fourth, the in vivo studies are reliant on a drug.

3) The conclusion that p27 regulates Cortactin/PAK mediated invadopodia disassembly are based on static images of invadopodia number. Conclusions about the dynamic nature of invadopodia would be best supported by images documenting the dynamics.

Reviewer #2:

This work, building on prior publications from Webb et al. and Moshfegh et al., 2014 attempts to provide a mechanistic basis for the effects of cytosolic p27 on invadopodia dynamics and cell invasion. For the most part, the data appear sound, but there are a few key issues that are not adequately addressed.

1) The only experiment that identifies Pak1 as a link between p27 and invadopodia is treatment of MEFs with FRAX597 (Figure 6). There are two issues here: this compound affects all three group A Paks (thus one can't say the effects are through Pak1) and FRAX597 has poor selectivity towards STE20 kinases such as Group A Paks as compared to tyrosine kinases. There are much more selective compounds (e.g., FRAX1036 and G5555) as well as other approaches (siRNA, CRISPR) that could and should be used to establish the link claimed by the authors. As things stand, the data do not adequately support the author's contentions.

2) There really isn't any proof that Cortactin gets phosphorylated by Pak at S113, only that substitutions at this residue affect invadopodia dynamics (Figure 7). What would have made the case much stronger are phospho-S113 blots and/or IFs showing that phosphorylation of this site is modulated by Rac/Pak. As far as I can tell, the initial paper from Webb et al., 2006, on which the present manuscript relies, only provided in vitro evidence that recombinant Pak3 could phosphorylate Cortactin at S113. As there are no available phospho S113 antibodies, the authors might instead IP Cortactin from p27-/- MEFs and blot with p-Ser antibodies or, better, ID sites by mass spec. I know this sort of experiments isn't fun or easy, but the use of phospho-mimics alone is not adequate to support their conclusions.

Reviewer #3:

This manuscript identifies p27^kip1 as a regulator of invadopodia formation and cell invasion, increasing the surprising number of cellular processes known to be regulated by this protein. While cytoplasmic p27^kip1 was previously established to limit RhoA activity and stress fiber formation to promote cell migration, Jeannot et al., 2015 describe a new mechanism, where p27 promotes Cortactin phosphorylation by Pak1 to reduce invadopodia and increase cell invasion through collagen. While the authors include a strong set of experiments that convincingly demonstrates a p27/Cortactin/PAK1 pathway exists, the manuscript overall feels somewhat preliminary because it is not clear how inhibition of this pathway can increase invadopodia formation and prevent cell invasion. Presenting a mechanism to explain this puzzling observation and more clearly demonstrating p27 is localized to matrix degrading structures would strengthen the novel, and important conclusions made by the authors. Additional concerns are listed below.

Major Concerns:

The pre-fixation permeabilization/extraction of the cytoplasm prior to staining for p27 does not give a lot of confidence that p27 is localized to invadopodia (as in Figure 2A, for example). Additionally, the Cortactin phosphomimetic construct decreases the number of invadopodia without significantly reducing the area of gelatin degradation, raising the strong possibility that the structures being identified by Cortactin (eg. Figure 3A) are not actual invadopodia, despite their punctate appearance. To make confirm these structures are invadopodia, the authors should show p27 is enriched in Cortactin or Tsk5 positive puncta which coincide with regions of matrix degradation.

The authors' measure number of cells with invadopodia (eg. Figure 3B), yet often conclude that cells have more invadopodia. If the authors are going to conclude that there are more invadopodia per cell, this is something that must be explicitly measured.

Importantly, the strikingly counterintuitive conclusion that increasing invadopodia inhibits invasion was not sufficiently justified. It remains a possibility that another pathway mediated by p27 is responsible for the invasion phenotype. One way to address this possibility would be to show expression of p27 1-190 is unable to rescue cell invasion.

It is not clear if the migration defect resulting from loss of p27Kip (Besson et al., 2006), is mechanistically distinct from the invasion phenotype reported here. In their previous publication, it was shown that inhibition of Rho kinase rescued migration. Does Rho kinase inhibition also restore the ability of the p27 -/- cells to invade through thin collagen layers? If not, it would strengthen the conclusion that the authors are investigating a novel migration mechanism downstream of p27^kip1.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for resubmitting your work entitled "p27^Kip1 promotes invadopodia turnover and invasion through the regulation of the PAK1/Cortactin pathway" for further consideration at eLife. Your revised article has been favorably evaluated by Fiona Watt (Senior editor) and three reviewers, one of whom is a member of our Board of Reviewing Editors.

The manuscript has been improved but there are some remaining issues that need to be addressed before acceptance, as outlined below:

1) While the colocalization of p27 and Tks5 in regions of degraded gelatin is now included in the manuscript, increasing confidence that these structures are bona fide invadopodia, it is not clear how specific the colocalization is compared to regions of the cell that are not associated with regions of gelatin degradation. I suggest the authors include line scans of regions of the cell not associated with gelatin degradation to demonstrate the specificity and import of the colocalization they now measure near regions of degraded gelatin.

2) It is not clear if statistical tests were performed on the data presented in Figure 2C, Figure 2—figure supplement 2B, Figure 5 C and E, and Figure 8—figure supplement 1B.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "p27^Kip1 promotes invadopodia turnover and invasion through the regulation of the PAK1/Cortactin pathway" for further consideration at eLife. Your revised article has been favorably evaluated by Fiona Watt (Senior editor) and a Reviewing editor.

The manuscript has been improved but there is one minor issue that needs to be addressed before acceptance. The new statistical analysis of Figure 2C and Figure 2—figure supplement 2B stated in the figure legends indicates that differences are not significant. However, in the text referring to these figures, it is stated that increases were observed. This is contradictory. Please clarify.

[Editors’ note: the author responses to the first round of peer review follow.]

Reviewer #1:

This study demonstrates that p27 binding to Cortactin promotes PAK1 interactions with Cortactin, and invadopodia disassembly by phosphorylation of Cortactin on Ser-113. In general, the biochemical analysis is strong and convincing. Indeed, the demonstration that a phosphomimetic mutant of Cortactin can complement the p27 deficiency phenotype on invadopodia disassembly is very interesting. However, there are a number of areas that require strengthening.

1) The siRNA studies (Rac and RhoA) were performed with a single oligonucleotide. To exclude off0target effects, phenotypes should be confirmed with two independent siRNAs.

To strengthen the data obtained with single siRNAs for Rac1 and RhoA, we have repeated these experiments (# of cells forming invadopodia and area degraded/cell) with a pool of 4 siRNA against Rac1 (ON-TARGETplus Mouse Rac1 siRNA – SMARTpool (L041170000005)) and with a new pool of 3 siRNAs against RhoA (sc-36414). These results have been incorporated into Figure 7B-D for Rac1 and Figure 7—figure supplement 1A-C for RhoA. In addition, siRac1 results are also replicated using a pharmacological inhibitor of Rac1 (Figure 7E-F), arguing against an off-target effect of the siRNA.

2) Some care should be taken concerning the PAK1/Cortactin connection. First, the studies presented are focussed on mutational analysis of the Cortactin phosphorylation site Ser-113 rather than the kinase. Second, in vitro studies demonstrate phosphorylation of Cortactin by PAK3 at more than one site (Webb et al.). Third, what is the evidence that PAK1 phosphorylates Cortactin in vivo? Fourth, the in vivo studies are reliant on a drug.

We have modified the text and use more careful wording when describing the PAK1/Cortactin connection. We also have taken several steps to strengthen this point:

1) To investigate the importance of PAK1, we had data using the FRAX597 inhibitor, which targets the ATP-binding pocket in the kinase domain of PAK1-3 and inhibits the kinase activity of PAK, therefore suggesting that the kinase activity of PAK is involved.

As suggested by reviewer 2, we replicated these experiments (# of cells forming invadopodia and area degraded/cell) and with other, more specific inhibitors, of PAK1-3 (FRAX1036 and G5555). These inhibitors gave similar results as FRAX597, thus confirming the importance of PAK activity in p27-mediated control of invadopodia dynamics. This data has been included in the manuscript as Figure 6—figure supplement 1.

Since these inhibitors cannot discriminate between PAK1-3, we also performed similar experiments with siRNAs (a pool of 4 siRNAs) against PAK1 (ON-TARGETplus Mouse PAK1 (18479) siRNA – SMARTpool (L048101000005). Results with PAK1 siRNA are similar to those obtained with PAK pharmacological inhibitors, supporting the involvement of PAK1 in the pathway we describe. These data have been included in Figure 6A-C.

Altogether, we now have data from 3 different PAK1-3 pharmacological inhibitors and from PAK1-specific siRNA that give similar results, thus strongly suggesting that PAK1 is indeed involved in the pathway regulating invadopodia dynamics.

2) Webb and colleagues (2006) have reported the phosphorylation of Cortactin by PAK3 in vitro on three different residues (S113, S150 and S282).

As suggested by the reviewer, we tested whether the other sites known to be phosphorylated by PAK on Cortactin were important in the phenotype observed, We generated triple S113/S150/S282 unphosphorylatable and phosphomimetic Cortactin mutants and measured the effect on # of cells forming invadopodia and area degraded/cell. Triple mutants give similar results as S113 Cortactin mutants, with a stronger effect of the triple unphosphorylatable mutant in p27+/+ cells (13.5% cells forming invadopodia with S113A versus 23.3% with tripleA mutant; 6.17 fold increased area degraded with S113A versus 10.02 fold with TripleA) suggesting that these sites (or at least one of them) also contribute to Cortactin regulation of invadopodia dynamics. These results have been included in Figure 7F-J. These results also reinforce the idea that phosphorylation of Cortactin is important in the control of invadopodia dynamics.

3) We now provide 4 lines of evidence indicating that these sites are indeed phosphorylated in vivo:

a) By IP Cortactin and anti Phospho-Ser immunoblot we found that Ser phosphorylation of Cortactin was reduced in presence of PAK1-3 inhibitor FRAX597. This data is provided as Figure 8—figure supplement 1A and B.

b) By IP Cortactin and anti Phospho-Ser immunoblot we found that expression of Cortactin S113A and Cortactin S113A/S150A/S282A exhibit decreased phospho-Ser signal compared to wild-type Cortactin, arguing that phosphorylation of these sites contribute to the phospho-Ser signal detected. This data is provided as Figure 8—figure supplement 1C.

c) Our initial attempt to obtain mass spectrometry data showing that Cortactin was phosphorylated in vivo was with a private company. We since established a collaboration with a group specialized in MS work at the Institut Cochin. We were able to detect a phosphorylation of Cortactin at Ser150 (one of the 3 sites reported to be phosphorylated by PAK1-3) when a myristoylated form of the second SH3 domain of Nck1 was co-expressed (previously shown to activate PAK1 by Bruce Mayer's lab: Lu et al., 1997 and Lu and Mayer, 1999) but not when only Cortactin was expressed. Therefore PAK activation in these cells appear to induce Cortactin S150 phosphorylation. This mass spectrometry data is provided as Figure 8—figure supplement 2.

d) Database mining (PhosphositePlus) shows that Cortactin phosphorylation on S113, S150 and/or S282, the three sites targeted by PAK1-3, has previously been detected in vivo by mass spectrometry and this is described in 8 publications of global proteome analyses:

1) Chen et al., J Proteome Res 2010, 9:174-8

Shows Cortactin S113 phosphorylation in LNCaP prostate cancer cells

2)Olsen et al., Sci Signal 2010, 3(104):ra3

Shows Cortactin S282 phosphorylation in mitotic cells

3) Rinschen et al., PNAS 2010, 107: 3882-7

Shows Cortactin S150 phosphorylation in mpkCCD renal cells

4) Wisniewski et al., J Proteome Res 2010, 9:3280-9

Shows Cortactin S113 phosphorylation in mouse brain

5) Klammer et al., Mol Cell Proteomics 2012, 11: 651-88

Shows Cortactin S282 phosphorylation in breast tumors

6) Bian et al., J Proteomics 2014, 96: 253-62

Shows Cortactin S150 phosphorylation in human liver

7) Mertins et al., Mol Cell Proteomics 13: 1690-704

Show Cortactin S113, S150 and S282 phosphorylation in ovarian and breast xenograft tissues

8) Mertins et al., Nature 2016, 534: 55-62

Shows Cortactin S113 and S282 phosphorylation in breast cancer

These references are now discussed in our revised manuscript.

We believe that the fact several other groups have reported the phosphorylation of those sites in vivo in various models in independent studies provides strong evidence that Cortactin is actually phosphorylated on these three sites in vivo.

4) As detailed in point #1, our in vivo studies are now relying on three different PAK1-3 pharmacological inhibitor as well as similar experiments performed with PAK1 siRNA treated cells.

3) The conclusion that p27 regulates Cortactin/PAK mediated invadopodia disassembly are based on static images of invadopodia number. Conclusions about the dynamic nature of invadopodia would be best supported by images documenting the dynamics.

To strengthen our conclusions and provide direct evidence that p27 affects invadopodia stability, we determined invadopodia lifetime in p27+/+ and p27-/- MEFS by videomicroscopy using eGFP-Tks5. This data has been added as Figure 5A and shows that invadopodia in p27-/- cells have a longer lifetime than in p27+/+ cells.

Reviewer #2:

This work, building on prior publications from Webb et al. and Moshfegh et al., 2014 attempts to provide a mechanistic basis for the effects of cytosolic p27 on invadopodia dynamics and cell invasion. For the most part, the data appear sound, but there are a few key issues that are not adequately addressed.

1) The only experiment that identifies Pak1 as a link between p27 and invadopodia is treatment of MEFs with FRAX597 (Figure 6). There are two issues here: this compound affects all three group A Paks (thus one can't say the effects are through Pak1) and FRAX597 has poor selectivity towards STE20 kinases such as Group A Paks as compared to tyrosine kinases. There are much more selective compounds (e.g., FRAX1036 and G5555) as well as other approaches (siRNA, CRISPR) that could and should be used to establish the link claimed by the authors. As things stand, the data do not adequately support the author's contentions.

As suggested by the reviewer, we replicated these experiments (# of cells forming invadopodia and area degraded/cell) with siRNAs against PAK1 and with the more specific inhibitors of PAK1-3, FRAX1036 and G5555, to better support the link between PAK1 and Cortactin in p27-mediated regulation of invadopodia dynamics. These results are provided in Figure 6A-C and in Figure 6—figure supplement 1.

2) There really isn't any proof that Cortactin gets phosphorylated by Pak at S113, only that substitutions at this residue affect invadopodia dynamics (Figure 7). What would have made the case much stronger are phospho-S113 blots and/or IFs showing that phosphorylation of this site is modulated by Rac/Pak. As far as I can tell, the initial paper from Webb et al., 2006, on which the present manuscript relies, only provided in vitro evidence that recombinant Pak3 could phosphorylate Cortactin at S113. As there are no available phospho S113 antibodies, the authors might instead IP Cortactin from p27-/- MEFs and blot with p-Ser antibodies or, better, ID sites by mass spec. I know this sort of experiments isn't fun or easy, but the use of phospho-mimics alone is not adequate to support their conclusions.

As suggested by the reviewer, we performed additional experiments indicating that:

a) By IP Cortactin and anti Phospho-Ser immunoblot we found that Ser phosphorylation of Cortactin was reduced in presence of PAK1-3 inhibitor FRAX597. This suggests that PAK1-3 contribute to the phosphorylation of Cortactin. This data is provided as Figure 8—figure supplement 1A and B.

b) By IP Cortactin and anti Phospho-Ser immunoblot we found that expression of Cortactin S113A and Cortactin S113A/S150A/S282A (the sites targeted by PAK) exhibit decreased phospho-Ser signal compared to wild-type Cortactin, arguing that phosphorylation of these sites contribute to the phospho-Ser signal detected. This data is provided as Figure 8—figure supplement 1C.

c) Our initial attempt to obtain mass spectrometry data showing that Cortactin was phosphorylated in vivo was with a private company. We since established a collaboration with a group specialized in MS work at the Institut Cochin. We were able to detect a phosphorylation of Cortactin at Ser150 (one of the 3 sites reported to be phosphorylated by PAK1-3) when a myristoylated form of the second SH3 domain of Nck1 was co-expressed (previously shown to activate PAK1 by Bruce Mayer's lab: Lu et al., 1997 and Lu and Mayer, 1999) but not when only Cortactin was expressed. Therefore PAK activation in these cells appear to induce Cortactin S150 phosphorylation. This mass spectrometry data is provided as Figure 8—figure supplement 2.

d) Database mining (PhosphositePlus) shows that Cortactin phosphorylation on S113, S150 and/or S282, the three sites targeted by PAK1-3, has previously been detected in vivo by mass spectrometry and this is described in 8 publications of global proteome analyses:

1) Chen et al., J Proteome Res 2010, 9:174-8

Shows Cortactin S113 phosphorylation in LNCaP prostate cancer cells

2) Olsen et al., Sci Signal 2010, 3(104):ra3

Shows Cortactin S282 phosphorylation in mitotic cells

3) Rinschen et al., PNAS 2010, 107: 3882-7

Shows Cortactin S150 phosphorylation in mpkCCD renal cells

4) Wisniewski et al., J Proteome Res 2010, 9:3280-9

Shows Cortactin S113 phosphorylation in mouse brain

5) Klammer et al., Mol Cell Proteomics 2012, 11: 651-88

Shows Cortactin S282 phosphorylation in breast tumors

6) Bian et al., J Proteomics 2014, 96: 253-62

Shows Cortactin S150 phosphorylation in human liver

7) Mertins et al., Mol Cell Proteomics 13: 1690-704

Show Cortactin S113, S150 and S282 phosphorylation in ovarian and breast xenograft tissues

8) Mertins et al., Nature 2016, 534: 55-62

Shows Cortactin S113 and S282 phosphorylation in breast cancer

These references are now discussed in our revised manuscript.

Reviewer #3:

This manuscript identifies p27^kip1 as a regulator of invadopodia formation and cell invasion, increasing the surprising number of cellular processes known to be regulated by this protein. While cytoplasmic p27^kip1 was previously established to limit RhoA activity and stress fiber formation to promote cell migration, Jeannot et al, 2015. describe a new mechanism, where p27 promotes Cortactin phosphorylation by Pak1 to reduce invadopodia and increase cell invasion through collagen. While the authors include a strong set of experiments that convincingly demonstrates a p27/Cortactin/PAK1 pathway exists, the manuscript overall feels somewhat preliminary because it is not clear how inhibition of this pathway can increase invadopodia formation and prevent cell invasion. Presenting a mechanism to explain this puzzling observation and more clearly demonstrating p27 is localized to matrix degrading structures would strengthen the novel, and important conclusions made by the authors. Additional concerns are listed below.

Major Concerns:

The pre-fixation permeabilization/extraction of the cytoplasm prior to staining for p27 does not give a lot of confidence that p27 is localized to invadopodia (as in Figure 2A, for example). Additionally, the Cortactin phosphomimetic construct decreases the number of invadopodia without significantly reducing the area of gelatin degradation, raising the strong possibility that the structures being identified by Cortactin (eg. Figure 3A) are not actual invadopodia, despite their punctate appearance. To make confirm these structures are invadopodia, the authors should show p27 is enriched in Cortactin or Tsk5 positive puncta which coincide with regions of matrix degradation.

We performed co-immunostaining for Tks5, p27 and fluorescent gelatin providing evidence that p27 and Tks5 colocalize in structures capable of degrading ECM. These data in p27+/+ MEFs and in A549 lung adenocarcinoma cells are now presented in Figure 3—figure supplement 2.

We had some variability in the last set of experiments using S113 Cortactin mutants. More experiments have now been performed and these experiments have been incorporated in the existing figure panels, improving the quality of the data presented (Figure 8B-E). In addition, we now provide new data using Cortactin mutants in which the three sites reported to be phosphorylated by PAK (S113/S150/S282) have been mutated to either unphosphorylatable and phosphomimetic residues. The triple unphosphorylatable mutant (Cortactin TA) give more robust responses in p27+/+ cells (13.5% cells forming invadopodia with S113A versus 23.3% with tripleA mutant; 6.17 fold increased area degraded with S113A versus 10.02 fold with TripleA) suggesting that these sites have a synergistic effect in Cortactin regulation of invadopodia dynamics. This data is presented in Figure 8F-J.

The authors' measure number of cells with invadopodia (eg. Figure 3B), yet often conclude that cells have more invadopodia. If the authors are going to conclude that there are more invadopodia per cell, this is something that must be explicitly measured.

We have modified the text to ensure that our conclusions are referring only to the number of cells forming invadopodia.

Importantly, the strikingly counterintuitive conclusion that increasing invadopodia inhibits invasion was not sufficiently justified. It remains a possibility that another pathway mediated by p27 is responsible for the invasion phenotype. One way to address this possibility would be to show expression of p27 1-190 is unable to rescue cell invasion.

As suggested by the reviewer, we performed invasion assays with p27-/- MEFs infected with either empty vector, p27CK- or p27CK- 1-190 mutants. While p27CK- expression clearly increased the invasive capacity of p27-/- cells, the p27CK- 1-190 mutant had no effect, thus confirming that the ability of p27 to interact with Cortactin is required for the regulation of invasion. This data is now presented in Figure 4E.

It is not clear if the migration defect resulting from loss of p27Kip (Besson et al., 2006), is mechanistically distinct from the invasion phenotype reported here. In their previous publication, it was shown that inhibition of Rho kinase rescued migration. Does Rho kinase inhibition also restore the ability of the p27 -/- cells to invade through thin collagen layers? If not, it would strengthen the conclusion that the authors are investigating a novel migration mechanism downstream of p27^kip1.

As the reviewer suggested, we performed transwell invasion assays with cells treated with the ROCK inhibitor Y26632. ROCK inhibition had no effect on the invasive capacity of p27+/+ and p27-/- MEFs (Figure 7—figure supplement 1D). This is in agreement with our data indicating that siRNAs against RhoA did not significantly affect the number of cells forming invadopodia and only slightly increased the ability of both p27+/+ and p27-/- MEFs to degrade ECM (Figure 7—figure supplement 1A-C), suggesting that the regulation of RhoA by p27 is not involved in this phenotype.

[Editors' note: the author responses to the re-review follow.]

The manuscript has been improved but there are some remaining issues that need to be addressed before acceptance, as outlined below:

1) While the colocalization of p27 and Tks5 in regions of degraded gelatin is now included in the manuscript, increasing confidence that these structures are bona fide invadopodia, it is not clear how specific the colocalization is compared to regions of the cell that are not associated with regions of gelatin degradation. I suggest the authors include line scans of regions of the cell not associated with gelatin degradation to demonstrate the specificity and import of the colocalization they now measure near regions of degraded gelatin.

As requested, we added line scans on the same images not associated with regions of gelatin degradation and the results indicate that individual spots of Tks5 and p27 do not colocalize extensively outside of these regions of gelatin degradation. Two panels (C and D) have been added to Figure 3—figure supplement 2 to illustrate this point.

2) It is not clear if statistical tests were performed on the data presented in Figure 2C, Figure 2—figure supplement 2B, Figure 5 C and E, and Figure 8—figure supplement 1B.

Statistical analyses had not been performed on the graphs. We have now performed these analyses:

For Figure 2C, Figure 2— figure supplement 2B and Figure 5E, the differences of each time-point with the control condition (time 0) were not significant. We added a sentence in the figure legends of these panels to mention that fact. For the graphs in Figure 5B and Figure 8—figure supplement 1B, the differences were statistically significant (p<0.01) and we added ‘**’ on the graphs and legends. All statistical analyses for these panels are now included in the source data files accompanying the manuscript.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

The manuscript has been improved but there is one minor issue that needs to be addressed before acceptance. The new statistical analysis of Figure 2C and Figure 2—figure supplement 2B stated in the figure legends indicates that differences are not significant. However, in the text referring to these figures, it is stated that increases were observed. This is contradictory. Please clarify.

Original sentence:

“Consistent with the translocation of p27 in the cytoplasm and the formation of invadopodia after serum or growth factor stimulation, an increased amount of Cortactin co-immunoprecipitated with p27 at 1 h and 3 h post-stimulation in p27+/+ E6 MEFs (Figure 2C) and in Hela cells (Figure 2—figure supplement 2).”

New sentence:

“Consistent with the translocation of p27 in the cytoplasm and the formation of invadopodia after serum or growth factor stimulation, an increased association between p27 and Cortactin by co-immunoprecipitation at 1 h and 3 h post-stimulation in p27+/+ E6 MEFs (Figure 2C) and in Hela cells (Figure 2—figure supplement 2) was observed in five and three independent experiments, respectively. However, due to variability in signal intensities among independent experiments, these differences were not statistically significant.”

Author details

1. Pauline Jeannot

   1. INSERM UMR1037, Cancer Research Center of Toulouse, Toulouse, France
   2. Université Toulouse III Paul Sabatier, Toulouse, France
   3. CNRS ERL5294, Toulouse, France

   Contribution

   PJ, Conceptualization, Data curation, Formal analysis, Validation, Investigation, Methodology, Writing—original draft

   Competing interests

   The authors declare that no competing interests exist.

2. Ada Nowosad

   1. INSERM UMR1037, Cancer Research Center of Toulouse, Toulouse, France
   2. Université Toulouse III Paul Sabatier, Toulouse, France
   3. CNRS ERL5294, Toulouse, France

   Contribution

   AN, Conceptualization, Data curation, Investigation, Methodology

   Competing interests

   The authors declare that no competing interests exist.

3. Renaud T Perchey

   1. INSERM UMR1037, Cancer Research Center of Toulouse, Toulouse, France
   2. Université Toulouse III Paul Sabatier, Toulouse, France
   3. CNRS ERL5294, Toulouse, France

   Contribution

   RTP, Data curation, Investigation, Methodology

   Competing interests

   The authors declare that no competing interests exist.

4. Caroline Callot

   1. INSERM UMR1037, Cancer Research Center of Toulouse, Toulouse, France
   2. Université Toulouse III Paul Sabatier, Toulouse, France
   3. CNRS ERL5294, Toulouse, France

   Contribution

   CC, Investigation, Methodology

   Competing interests

   The authors declare that no competing interests exist.

5. Evangeline Bennana

   3P5 proteomics facility of the Université Paris Descartes, Inserm U1016 Institut Cochin, Sorbonne Paris Cité, Paris, France

   Contribution

   EB, Data curation, Formal analysis, Investigation, Methodology

   Competing interests

   The authors declare that no competing interests exist.

6. Takanori Katsube

   National Institute of Radiological Sciences, Chiba, Japan

   Contribution

   TK, Resources, Methodology

   Competing interests

   The authors declare that no competing interests exist.

7. Patrick Mayeux

   3P5 proteomics facility of the Université Paris Descartes, Inserm U1016 Institut Cochin, Sorbonne Paris Cité, Paris, France

   Contribution

   PM, Conceptualization, Data curation, Formal analysis, Investigation

   Competing interests

   The authors declare that no competing interests exist.

8. François Guillonneau

   3P5 proteomics facility of the Université Paris Descartes, Inserm U1016 Institut Cochin, Sorbonne Paris Cité, Paris, France

   Contribution

   FG, Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing—review and editing

   Competing interests

   The authors declare that no competing interests exist.

9. Stéphane Manenti

   1. INSERM UMR1037, Cancer Research Center of Toulouse, Toulouse, France
   2. Université Toulouse III Paul Sabatier, Toulouse, France
   3. CNRS ERL5294, Toulouse, France

   Contribution

   SM, Funding acquisition, Writing—original draft, Project administration

   Competing interests

   The authors declare that no competing interests exist.

10. Arnaud Besson

    1. INSERM UMR1037, Cancer Research Center of Toulouse, Toulouse, France
    2. Université Toulouse III Paul Sabatier, Toulouse, France
    3. CNRS ERL5294, Toulouse, France

    Contribution

    AB, Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Methodology, Writing—original draft, Project administration, Writing—review and editing

    For correspondence

    arnaud.besson@inserm.fr

    Competing interests

    The authors declare that no competing interests exist.

    "This ORCID iD identifies the author of this article:" 0000-0002-9599-3943

• 1,653

  Page views

• 479

  Downloads

• 32

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, Scopus, PubMed Central.