Disorder-Function Relationships for the Cell Cycle Regulatory Proteins P21 and P27

Review

Disorder-function relationships for the cell cycle regulatory proteins p21 and p27

Diana M. Mitrea, Mi-Kyung Yoon, Li Ou and structure. However, many proteins exhibit extensively dis- Richard W. Kriwacki * ordered regions and others are entirely unstructured under physiological conditions. These proteins are termed intrinsi- Department of Structural Biology , St. Jude Children ’ s cally disordered proteins (IDPs) or intrinsically unstructured Research Hospital, 262 Danny Thomas Place, Memphis, proteins (IUPs). IDPs are highly abundant in nature (Dunker TN 38103 , USA et al. , 2000 ); for example, up to a third of bacterial and * Corresponding author archeal proteins and 35 – 50 % of eukaryotic proteins are pre- e-mail: [email protected] dicted to contain disordered regions of 40 or more consecu- tive residues. IDPs lack stable tertiary or secondary structure in vitro under physiological conditions and exhibit extensive Abstract fl exibility and highly randomized conformations. Disordered proteins are involved in myriad cellular processes, including The classic structure-function paradigm has been challenged those involving movement and transport of molecules in cells by a recently identifi ed class of proteins: intrinsically disor- and signaling and regulation. However, although there is now dered proteins (IDPs). Despite their lack of stable secondary broad recognition of the biological importance of IDPs, our or tertiary structure, IDPs are prevalent in all forms of life understanding of the mechanisms that mediate their functions and perform myriad cellular functions, including signaling is incomplete. Here we review the basic properties of IDPs and regulation. Importantly, disruption of IDP homeostasis is and then discuss the structural features and ‘ disorder-func- associated with numerous human diseases, including cancer tion ’ relationships for two well characterized IDPs, the Cdk and neurodegeneration. Despite wide recognition of IDPs, regulators p21 and p27. These IDPs play critical roles in regu- the molecular mechanisms underlying their functions are not lating cell cycle progression and cell fate and are regarded fully understood. Here we review the structural features and as desirable drug targets against cancer (Funk and Galloway, disorder-function relationships for p21 and p27, two cyclin- 1998 ; Weiss , 2003 ; Borriello et al. , 2011a ). dependent kinase (Cdk) regulators involved in controlling cell division and fate. Studies of p21 bound to Cdk2/cyclin A revealed that a helix stretching mechanism mediates binding Functional features of IDPs promiscuity. Further, investigations of Tyr88-phosphorylated p27 identifi ed a signaling conduit that controls cell division Biological roles and is disrupted in certain cancers. These mechanisms rely upon a balance between nascent structure in the free state, Proteins predicted by bioinformatics analyses to be intrinsi- induced folding upon binding, and persistent fl exibility within cally disordered are involved in many biological roles, notably functional complexes. Although these disorder-function rela- in processes involving signaling and regulation (Iakoucheva et tionships are likely to be recapitulated in other IDPs, it is also al. , 2002 ). Several large-scale analyses of proteins expressed likely that the vocabulary of their mechanisms is much more in living cells also confi rm the broad biological roles of IDPs extensive than is currently understood. Further study of the in eukaryotes (Ward et al. , 2004 ; Galea et al. , 2006, 2009 ). physical properties of IDPs and elucidation of their links with For example, Galea et al. (2009) identifi ed more than 900 function are needed to fully understand the mechanistic lan- heat-stable, disordered proteins in mouse fi broblast cells that guage of IDPs. are involved in the regulation of cell proliferation, apoptosis, gene transcription, metabolic and biosynthetic processes, Keywords: cell cycle regulator; disorder-function; organelle and cytoskeletal structure, and cell migration. intrinsically disordered proteins; p21; p27; signaling conduit. Functional advantages of disorder

Introduction: intrinsically disordered proteins – a The ﬂ exible features of IDPs enable widely varied functions in distinct class of multifunctional proteins cells. However, detailed knowledge of the functional mechanisms associated with IDPs is limited. Nonetheless, studies of The classic structure-function paradigm posits that protein well characterized examples do provide insights into ‘ disor- function critically depends on a three-dimensional (3D) der-function relationships ’ from which general concepts have 260 D.M. Mitrea et al.

emerged. Domains within the cell cycle regulatory IDPs, p21Cip1 assembly of multicomponent, macromolecular complexes. (p21) and p27Kip1 (p27), bind several different cyclin-dependent Disordered scaffold proteins often contain multiple short kinase (Cdk)/cyclin complexes through structural adaptation to interaction motifs (Davey et al., 2011) that confer the ability accommodate similar but topologically distinct binding sites to interact with and promote the coassembly of many differ- (Wang et al., 2011). This system illustrates the ability of a dis- ent partners; examples include axin (Noutsou et al. , 2011 ), ordered domain to bind promiscuously to multiple targets, a CBP (Dyson and Wright, 2005), and BRCA1 (Mark et al., feature exhibited by many IDPs. p53 provides another example 2005 ). These examples illustrate the functional versatility of of promiscuous interactions, with a short domain within its dis- disordered proteins that is achieved through highly diverse ordered C-terminus binding, by assuming different conforma- physical mechanisms. A recent comprehensive review elabo- tions, to at least six different targets (Oldfi eld et al. , 2008 ). rates on this topic (Dyson , 2011 ). Interestingly, p27 can be switched from being an inhibitor to being a partial activator of the Cdks through phosphorylation. The extended, Cdk-bound conformation of p27 Structural features of IDPs enables several Tyr and Ser/Thr sites to be phosphorylated and to modulate its Cdk regulatory activity. This illustrates Primary structure another general feature of IDPs; their sequences are enriched in residues that offer sites for posttranslation modifi cation It is well established that the amino acid sequence of a protein (PTM), which allows their function(s) to be intimately con- determines its 3D structure. Recognition of the prevalence of trolled by modifying enzymes. Furthermore, the disordered IDPs has required reevaluation of the relationships between features of IDPs provide steric accessibility to these modify- primary structure and ‘ native ’ protein structure. IDPs exhibit ing enzymes (Iakoucheva et al. , 2004 ; Gnad et al. , 2011 ). For distinct amino acid compositions compared with those of example, the majority of known phosphorylation and acety- globular proteins. Based on an analysis of the IDPs and intrin- lation sites in retinoblastoma protein (pRb) occur within the sically disordered regions deposited in the DisProt database disordered C-terminal domain as well as in the linker regions. (Sickmeier et al. , 2007 ), the 20 natural amino acids can be Modifi cation of these sites alters the structure and function classifi ed as order promoting, disorder promoting, or struc- of pRb (Harbour et al. , 1999 ). The Lys- and Arg-rich N- and ture neutral (Romero et al., 2001; Radivojac et al., 2007). The C-terminal domains (termed ‘ tails ’ ) of histones can be modi- order-promoting residues are primarily (but not exclusively) fi ed through monomethylation, dimethylation, trimethyla- hydrophobic amino acids (C, W, Y, I, F, V, L, H, T, and N) and tion, acetylation, ubiquitination, sumoylation, as well as other the disorder-promoting residues are primarily polar and/or modifi cations, further illustrating the remarkable signaling charged amino acids (D, M, K, R, S, Q, P, and E). The struc- complexity – referred to as the ‘ histone code’ – that can be ture neutral residues (A and G) have no obvious preference achieved through the existence of multiple modifi able sites to be part of either ordered or disordered segments. In gen- within relatively short, disordered polypeptide segments. eral, IDPs are enriched in disorder-promoting residues, which Importantly, the clustering of PTM sites within disordered consequently limit their ability to fold into unique, globular polypeptide segments affords accessibility not only to the conformations. modifying enzymes but also to other proteins that interact specifi cally with the modifi ed sites to transmit biological Partially populated secondary structure of IDPs signals. Signaling via histone tail modifi cations is known, for example, to cause the assembly of a number of distinct, Although some IDPs completely lack secondary and ter- multiprotein complexes that mediate epigenetic phenomena tiary structure, others exhibit partially populated secondary (Suganuma and Workman , 2011 ). structure. NMR spectroscopy, molecular dynamics (MD) Other biological processes that are enabled by disorder are simulations and circular dichroism (CD) spectroscopy have molecular movement and transport. For example, transient shown that unbound p27 (Sivakolundu et al., 2005) and p21 disorder within the motor head is proposed to mediate the (Kriwacki et al. , 1996 ) exhibit partial populated secondary 8-nm-long steps of kinesin as it ‘ walks ’ along microtubules structure similar to the fully formed structure observed in the (Hyeon and Onuchic , 2007 ). Also, the motor protein dynein Cdk/cyclin bound state. This observation strongly suggests an depends on disorder-order transitions for interactions with association mechanism whereby bound state-like conforma- transported cargo and other aspects of its function (Morgan et tions, termed ‘ intrinsically folded structural units ’ (IFSUs) al. , 2011 ). Furthermore, the FG-Nup proteins, which occupy (Sivakolundu et al. , 2005 ), are populated and preferred for the pore of the nuclear pore complex (NPC), form a dynamic, interaction with a binding partner. This scenario is an exam- gel-like structure that facilitates passage of adapter-bound ple of conformational selection, which offers an energetic cargo through the NPC but otherwise serve as a barrier to advantage by reducing the entropic barrier to binding through macromolecular diffusion (Kriwacki and Yoon, 2011). The partial prefolding. ‘ fl atness ’ of the energy landscape associated with disordered proteins may be critical for enabling the conformational Coupled folding and binding fl uctuations associated with macromolecular movement and transport mechanisms. Another functional category associ- As the term suggests, coupled folding and binding refers to ated with disordered proteins is to serve as scaffolds for the the mechanism by which the folding of a disordered protein Disorder-function relationship for p21 and p27 261

is induced by association with a binding partner. This mech- (Figure 1 A). Expression of the D-type cyclins and their com- anism is distinct from the conformational selection mecha- plexation with Cdk4 and Cdk6 initiates entry into G1 phase. nism (discussed above), which involves the target binding Next, activation of Cdk2 bound to cyclin E and later cyclin to a partially folded structural element – an IFSU – within A promotes progression from G 1 to S phase. Finally, after an IDP. Examples of both mechanisms have been and con- the DNA replication is complete, activation of Cdk1/cyclin tinue to be documented, and it is most likely that during many B and Cdk1/cyclin A promotes progression from G2 to M binding events involving disordered proteins, there is synergy phase. between the conformational selection and coupled folding binding mechanisms (reviewed by Boehr et al. , 2009 ; Mittag Regulation of cyclin-dependent kinases et al. , 2010 ). The various nuclear Cdk/cyclin complexes are negatively regulated by the INK4 and Cip/Kip protein families. The Cell cycle inhibitors: p21 and p27 INK4 family members, including p16 (Serrano et al. , 1993 ), p15 (Hannon and Beach, 1994), p18 (Guan et al., 1994), and Biological functions p19 (Hirai et al., 1995), bind Cdk4 and Cdk6 and prevent their complexation with D-type cyclins, causing cell cycle

Cell cycle progression is regulated by the sequential acti- arrest in early G1 phase. In contrast, the Cip/Kip family mem- vation of various Cdk/cyclin complexes (Morgan , 1995 ) bers, including p21, p27, and p57Kip2 (p57), associate with all

Cdk1/ p21, p27, p57 A cyclin A

p21, Cdk1/ Cdk4 p27 cyclin B M (and Cdk6)/ p21, p27 cyclin D G2 G1 S

Cdk2 Cdk2 p21, and Cdk1/ (and Cdk1)/ p21, p27, p57 p27 cyclin A cyclin E B 1102030405060708090 -MSNVRVSNGSPSLERMDARQAEHPKPSACRNLFGPVDHEELTRDLEKHCRDMEEASQRKWNFDFQNHKPLEG--KYEWQEVEKGSLPEFYYRPPRPPKG p27 p21 ------MSEPAGDVRQNPCG-SKACRRLFGPVDSEQLSRDCDALMAGCIQEARERWNFDFVTETPLEG--DFAWERVRGLGLPKLYLPTG-PRRG p57 MSDASLRSTSTMERLVARGTFPVLVRTSACRSLFGPVDHEELSRELQARLAELNAEDQNRWDYDFQQDMPLRGPGRLQWTEVDSDSVPAFYRETVQVGRC RxLFG

D1 LH D2 310

Kinase inhibitory domain (KID)

17 81 PCNA 144 160 1 C p21 KID 164 142 NLS 157

28 Y74 90 152 169 1 T187 198 p27 KID NLS Y88 143 QT SCFSkp2 181 190

29 93 1 316 p57 KID PAPA QT 142 219 278 NLS 263PCNA 285

Figure 1 Cip/Kip proteins regulate cell cycle progression.

(A) p21 and p27 inhibit the activity of Cdk1(2)/cyclin E(A) that are required for progression from G1 to S phase and Cdk1/cyclin B(A) required for mitosis. During G1 phase, p21 and p27 can mediate assembly and activation of Cdk4(6)/cyclin D, as well as inhibit these complexes. (B) Sequence alignment of the KID domains of the Cip/Kip proteins. The conserved subdomains are schematically represented: D1, which contains the cyclin-binding RxL motif (blue), LH (red), and D2, containing the 310 helix (green). (C) Domain organization in p21, p27, and p57. The kinase binding domains are depicted in gold, the PCNA binding regions within p27 and p57 in green, and the NLSs in blue. The phosphorylation sites involved in signal transduction along p27 (Tyr74, Tyr88, and Thr187) are also highlighted (adapted from Galea et al. , 2008b ). 262 D.M. Mitrea et al.

Cdk/cyclin complexes and inhibit their kinase activities at the et al. , 2005a ). Notably, this pathway of protein turnover at

G1 /S and G2 /M checkpoints (Figure 1A). the end of G1 phase is conserved between p21 and p27 and The binding promiscuity of the Cip/Kip proteins, in con- may be mediated by a similar NRTK-dependent mechanism. trast to the speciﬁ city of the INK4 proteins for Cdk4 and However, although Tyr88 and Thr187 of p27 are conserved as Cdk6, is a functional advantage afforded by their disordered Tyr77 and Ser130 in p21, it is not known whether the same features. Possibly due to their abilities to potently inhibit cell two-step phosphorylation cascade observed in the regulation cycle progression, the expression and activity of p21 and p27 of p27 ubiquitination occurs with p21. During the S phase are exquisitely regulated, as has been observed in general for and after UV irradiation, the ubiquitination and degradation IDPs (Gsponer et al. , 2008 ). Additionally, regulation of both of p21 are carried out by the CRL4Cdt2 ubiquitin ligase in a of these vital cell cycle proteins is altered in human cancer. proliferating cell nuclear antigen (PCNA)-dependent man- ner (Abbas et al., 2008; Kim et al., 2008; Nishitani et al.,

Regulation of p27 2008). The degradation of p21 at the G 2/M transition, when bound to Cdk1/cyclin A or Cdk1/cyclin B, is mediated by p27 mRNA expression is constant during cell cycle (Toyoshima APC/CCdc20 (Amador et al., 2007). Alternatively, free p21 is and Hunter , 1994 ), and protein level is controlled through eliminated from the cell by an ubiquitin-independent pathway regulation of translation and ubiquitination-dependent protein that involves N-terminal acetylation of p21 (Chen et al., 2004) degradation (Hengst and Reed , 1996 ; Lu and Hunter , 2010 ). or direct interaction between its C-terminal region and the C8 Furthermore, p27 function and subcellular localization are subunit of the 20S proteasome (Touitou et al. , 2001 ). regulated via posttranslational modifi cations. For example, ubiquitination-dependent p27 degradation is regulated by a p21 and p27 in the assembly of Cdk/cyclin complexes phosphorylation cascade, involving an intramolecular signal transduction mechanism (Galea et al., 2008a). Phosphorylation In addition to their well-documented roles in inhibiting of p27 on Tyr88 by nonreceptor tyrosine kinases (NRTKs) Cdk/cyclin complexes in the nucleus, p21 and p27 are (Grimmler et al. , 2007 ) relieves p27-dependent inhibition of reported to mediate the assembly and nuclear import of Cdk2/cyclin complexes, as previously mentioned, and pro- Cdk4(6)/cyclin D complexes (LaBaer et al. , 1997 ; Cheng motes Cdk2-dependent phosphorylation at Thr187. This latter et al. , 1999 ). p27 tyrosine phosphorylation regulates the acti- modifi cation is a specifi c signal for SCFSkp2 -dependent ubiq- vity of Cdk4/cyclin D and Cdk6/cyclin D complexes in early uitination of p27, which leads to degradation of p27 bound G1 phase. Phosphorylation of Tyr88 relieves Cdk4(6) inhibi- to complexes by the 26S proteasome. This regulatory cascade tion but also allows p27 to remain bound to and stabilize these causes depletion of p27 protein in late G1 phase and promotes complexes, providing an explanation for how p27 can serve the transition to S phase. A Thr187 phosphorylation-inde- as an assembly factor for Cdk4/cyclin D and Cdk6/cyclin pendent ubiquitination pathway, involving Kip1 ubiquitina- D complexes (Blain, 2008). p21 and p27 are representative tion-promoting complex (KPC), has been implicated in p27 examples of IDPs that function as scaffolds for the assem- degradation at the G 0/G 1 transition (Kamura et al. , 2004 ). In bly of multiprotein complexes, in which interaction motifs contrast to SCFSkp2 -mediated ubiquitination, which targets p27 within their extended structure mediate the specifi c assembly within Cdk/cyclin complexes, KPC only targets free cytoplas- process. Polypeptide fl exibility is advantageous for recruit- mic p27 (Kotoshiba et al. , 2005 ). In addition, the Cul4A-DDB1 ment of the various subunits that experience assembly and for E3 ligase, when complexed with DDB1 and Artemis, mediates avoiding steric clashes during the assembly process. ubiquitination and proteasomal degradation of nuclear p27 in cultured cells at the G1 to S transition, independent of phos- Cip/Kip family of proteins and cancer phorylation of Thr157 or Thr187 (Yan et al., 2011). Maintenance of normal cell proliferation requires appropriate Regulation of p21 regulation of the proteins that control cell cycle and genetic alteration and/or dysregulation of these proteins is often asso- In contrast to the constant expression of p27, expression of ciated with tumorigenesis. For example, many cancer-asso- p21 is transcriptionally regulated by p53 in response to DNA ciated mutations disable the transactivation function of p53, damage, oxidative stress, and other cellular insults. p21 preventing stress-induced up-regulation of p21 and associ- expression is induced by p53 binding to a p53-responsive ated cell cycle arrest. Alternatively, D-type cyclins, the criti- DNA element within the p21 gene promoter and causes cell cal positive regulators of entry into the cell division cycle, are cycle arrest (El -Deiry et al., 1994 ). commonly overexpressed in human cancers. Apart from the In addition to transcriptional regulation, p21 levels are also regulatory link with p53 noted above, other types of altera- controlled through posttranslational modifi cations. Pools of tions to p21 and p27 regulation are commonly observed in Cdk/cyclin-complexed p21 are targeted for degradation via human cancer. Interestingly, the p21 and p27 genes do not the ubiquitin-proteasome pathway by three, cell-cycle stage- experience cancer-associated missense mutations, possibly specifi c E3 ubiquitin ligase complexes. At the G 1 /S transition, due to the general high tolerance of IDP amino acid sequences p21 bound to Cdk2/cyclin E or Cdk2/cyclin A is ubiquitinated to such mutations (Brown et al. , 2010 ). Two predominant by the SCFSkp2 complex after phosphorylation on Ser130 by mechanisms of dysregulation of p21 and p27 in cancer are Cdk2 itself (Yu et al., 1998; Bornstein et al., 2003; Wang altered subcellular localization and degradation. p57 is the Disorder-function relationship for p21 and p27 263

least studied member of the Cip/Kip family, but its down-reg- by others to be due to direct interaction with the Rho GTPase, ulation both at the protein and transcriptional levels is asso- RhoA, and inhibition of its interactions with guanine nucleo- ciated with human malignancies, including liver, pancreatic, tide exchange factors (GEFs), which promote RhoA activa- and colorectal cancers (reviewed by Borriello et al. , 2011a ). tion (Besson et al. , 2004 ). This regulatory mechanism has Phosphorylation of p27 on Thr157 within the nuclear local- been shown to contribute to tumor invasion and progression ization signal (NLS) by Akt was observed in approximately (Larrea et al., 2009b) and is also thought to mediate tumor 40 % of breast carcinomas and was associated with cytoplas- metastasis (Wander et al. , 2011 ). Recently, it was shown that mic localization of p27 (Liang et al., 2002; Shin et al., 2002; phosphorylation of p27 on Thr198, in this case by the kinase Viglietto et al., 2002). Cytoplasmic localization of p27 leads RSK1, was also associated with cytoplasmic localization to unchecked Cdk activity in the nucleus, which drives pro- and RhoA-dependent promotion of cell migration (Larrea liferation of these cancer cells. It was proposed that phos- et al. , 2009a ). It was suggested that activation of the PI3K phorylation of Thr157 blocks interactions with importin-α , and MAPK pathways in cancer cells, both of which activate which otherwise mediates transport of cargo (like p27) from RSK1, provides a mechanism for p27-dependent up-regula- the cytoplasm to the nucleus through the nuclear pore com- tion of cell motility and metastatic potential. plex (Blain and Massague, 2002). Interestingly, essentially The nuclear function of p27 is also regulated by phosphory- the same Akt-dependent, localization-altering mechanism lation, and these pathways are corrupted in cancer. Grimmler was previously discovered to affect p21 in breast cancer et al. (2007) discovered that Tyr88 of p27 was phosphorylated cells through phosphorylation of Thr145 (Zhou et al., 2001). by the BCR-ABL fusion oncoprotein in chronic myelogenous However, the infl uence of Thr145 phosphorylation on p21 leukemia (CML) cells and that this modifi cation was associ- localization was shown by others to be cell type-depen- ated with p27 ubiquitination and 26S proteasome-dependent dent. For example, Rossig et al. (2001) , demonstrated that degradation. This oncoprotein-driven mechanism activates the Akt-dependent phosphorylation of Thr145 inhibited inter- Cdk enzymes that drive progression into the S phase of the cell actions between nuclear p21 and PCNA as well as with division cycle and is associated with cell hyperproliferation. Cdk/cyclin complexes, counteracting p21-dependent inhi- This mechanism is also operative in some breast cancers, in bition of both DNA replication and cell cycle progression. which up-regulation of the NRTK, Src, catalyzes phosphory- Phosphorylation of p21 on Ser146 was also demonstrated to lation of both Tyr74 and Tyr88 of p27 (Chu et al., 2007). The inhibit binding to PCNA (Scott et al. , 2000 ). IDPs often con- structural details of these mechanisms of p27 elimination in tain short motifs within their sequences that mediate interac- cancer cells will be discussed below. Many other types of can- tions with other macromolecules and thus control signaling and cer exhibit reduced p27 levels, which is a general marker of regulation (Davey et al., 2011); these motifs are often subject poor clinical prognosis (Frescas and Pagano, 2008). In these to a posttranslational modifi cation that can alter interactions other cases, p27 elimination is often driven by up-regulation and also their roles in signaling and regulation. These observa- of the E3 ubiquitin ligase, SCF Skp2, which specifi cally targets tions for p27 and p21 illustrate a general vulnerability of IDPs p27 for ubiquitination (reviewed by Frescas and Pagano , to dysregulation through aberrant posttranslational modifi ca- 2008). Finally, results from genetically modifi ed mice have tion, exemplifi ed here in association with tumorigenesis. demonstrated that loss of a single p27 allele, resulting in p27 Cytoplasmic mislocalization of p27 and p21 is also asso- protein levels that are ∼ 50 % of wild-type, was associated ciated with gain of function. Independent of modifi cation with heightened DNA damage stress-induced tumorigenesis of Thr157, phosphorylation of p27 on Ser10 also enforces in multiple organs (Philipp -Staheli et al., 2001 ). Together, cytoplasmic localization in cultured cells overexpressing p27 the results above demonstrate that maintaining normal levels (Ishida et al. , 2002 ; Connor et al. , 2003 ). However, in another and subcellular localization of p27 is critical for tumor sup- study (Kotake et al. , 2005 ), whereas Ser10 phosphorylation pression and that heightened degradation or mislocalization was shown to selectively stabilize p27 in the G0 phase in pri- through several different mechanisms is associated with the mary mouse cells, the Ser10 to Ala (S10A) phosphorylation- genesis of many cancer types in humans. defi cient mutant did not disrupt cytoplasmic translocation of nuclear p27. Moreover, the extent of p27 cytoplasmic localization appeared to be tissue- and cell-type specifi c, suggest- Disorder-function relationships for p27 ing a variable role for this posttranslational modifi cation in regulating p27 function. Although shown to be the major site Primary structure of p27 of phosphorylation on p27 (Ishida et al. , 2000 ), Kotake et al. (2005) showed that Ser10 was dispensable for p27-dependent p27 is the most thoroughly studied member of the Cip/Kip cell cycle regulation based on results from a p27 S10A mutant protein family and whose other members include p21 and knock-in mouse model. p27 S10A/S10A animals exhibited normal p57. Cdk/cyclin regulation is mediated by an N-terminal body and organ size, in striking contrast with p27 -/- mice that kinase inhibitory domain (KID) that is conserved in p21, p27, developed severe hyperplasia of multiple organs (Nakayama and p57 (Figure 1B,C). The structure of p27-KID bound to et al. , 1996 ). Importantly, Ser10 was also noted to promote Cdk2/cyclin A (Russo et al. , 1996 ) revealed three func- migration of hepatocellular carcinoma cells through altera- tional subdomains within the KID. Subdomain D1, spanning tion of actin dynamics (McAllister et al., 2003). Cytoplasmic 10 residues at the N-terminus of the KID, interacts with cyclin p27-dependent promotion of cell migration was later shown A through the conserved RxL cyclin binding motif. The ∼ 30 264 D.M. Mitrea et al.

residues at the C-terminus of the KID, termed the D2 sub- (Bienkiewicz et al. , 2002 ), which was subsequently attributed 13 domain, bind and inhibit Cdk2. The D1 and D2 domains are to the LH subdomain through NMR analysis of C α chemical connected by subdomain LH, which adopts a dynamic, helical shift values (Lacy et al., 2004). Furthermore, 1 H-15 N hetero- conformation. nuclear NOE (hetNOE) values showed that residues within In contrast to their N-terminal KIDs, the C-terminal the LH and D2 subdomains exhibited partially restricted domains (CTDs) of Cip/Kip proteins exhibit more variability, conformations. Interestingly, only subdomain D1 exhibited exhibiting only partially conserved features. For instance, all negative hetNOE values consistent with a high degree of ﬂ ex- three members of the family contain a conserved NLS within ibility. Subsequently, analysis using 1 H-1 H NOESY NMR and their CTDs (Figure 1C). Additionally, p21 and p57 exhibit MD computations (Sivakolundu et al. , 2005 ) revealed that motifs that confer the ability to bind and inhibit the DNA several segments within unbound p27 exhibited structural polymerase δ processivity factor, PCNA (Waga et al., 1994; features similar to those observed in the bound state – the Watanabe et al. , 1998 ). so-called IFSUs discussed above. For example, subdomain 13 Furthermore, p27 and p57 exhibit a conserved ‘ QT ’ LH adopted helical structure also indicated by C α chemi- domain within their CTDs. These domains contain phospho- cal shift and hetNOE analyses, and subdomain D2 formed a rylation sites, Thr187 in p27 and Thr310 in p57 (Figure 1C), β -hairpin and a single turn of helix. These results indicated that that mediate interactions with the SCF Skp2 E3 ubiquitin ligase, p27-KID does not completely lack secondary structure before ubiquitination, and 26S proteasomal degradation (Montagnoli binding Cdk/cyclin complexes; rather, several sub domains et al., 1999; Nguyen et al., 1999; Kamura et al., 2003). This are transiently folded in bound state-like conformations. This observation suggests that a conserved mechanism mediates observation suggests that p27-KID associates with its binding degradation of the two proteins. In summary, the sequences partners through a conformational selection mecha nism. This of these three proteins are rich in signaling and interaction may be advantageous by reducing kinetic barriers to folding motifs that encode diverse functionality within relatively upon binding and by reducing the entropic penalty associated short amino acid sequences. Different motifs are utilized dif- with this process. However, as will be discussed below, other ferently in the three proteins, giving rise to their distinct bio- regions of p27 experience binding-induced folding, giving logical roles: p21 as a mediator of p53-dependent cell cycle rise to an association mechanism with features of both con- arrest, p27 as a constitutive regulator of cell cycle entry, and formational selection and induced folding. p57 as cell-type-speciﬁ c regulator of terminal cell differen- tiation (Borriello et al., 2011b). These features illustrate how Crystal structure of p27-KID/Cdk2/cyclin A functional complexity can be achieved through evolution of closely related IDP sequences. The structure of the p27-KID bound to Cdk2/cyclin A, solved more than 15 years ago (Russo et al. , 1996 ), provided insight Partial secondary structure within p27 in solution into the mechanisms by which p27 inhibits the kinase activity of Cdk2 and also promotes its assembly with cyclin A. Although highly disordered, as indicated by the limited res- p27 binds in a highly extended conformation to both subunits onance dispersion observed in the 2D 1 H-15 N HSQC NMR of the Cdk2/cyclin A complex (Figure 2 A), burying more spectrum (Lacy et al., 2004), p27 retains partial secondary than 2000 Å2 of solvent-exposed surface (Russo et al. , 1996 ). structure in solution. Analysis by CD indicated the exis- These extensive interactions are associated with a highly α Δ tence of a small percentage of -helical secondary structure favorable Gibbs free energy of binding ( G) giving a K d

ABLH D2 A Slow K2 D1 310 helix + p27 A A K2 K2 Cdk2 1 2 Cyclin A Binding of D1 Binding and folding of LH and D2 (Fast) A K2 (Slow)

Figure 2 p27-KID binds to Cdk2/cyclin A complex in a two-step mechanism. (A) Crystal structure of p27-KID in complex with Cdk2 (cyan) and cyclin A (magenta); PDB accession number 1JSU. Subdomain D1 (blue) binds and inhibits substrate binding to cyclin A, whereas subdomain D2 binds Cdk2, inserting a 3 10 helix, marked by the arrow, into the ATP-binding pocket of the kinase. Subdomain LH (red) connects the other two subdomains. (B) Sequential mechanism of p27-KID association with Cdk2/cyclin A; ﬁ rst, the highly disordered and dynamic RxL motif within subdomain D1 rapidly associates with cyclin A; next, subdomain LH fully folds, positioning subdomain D2 near Cdk2; ﬁ nally, subdomain extensively folds upon binding and remodeling Cdk2 (adapted from Lacy et al. , 2004 ). Cdk2 and cyclin A are colored as in panel A. Disorder-function relationship for p21 and p27 265

value of approximately 5 nm (Lacy et al., 2004). Subdomain evolved to bind the full repertoire of Cdk/cyclin complexes D1, containing the RxL motif, binds in an extended confor- that regulate progression of the cell division cycle. Although mation on the surface of cyclin A. Subdomain LH forms a the sequences of the different Cdk/cyclin complexes are very 22-residue-long α -helix, spanning the ∼ 40- Å gap between similar, a recent ITC analysis of p27 binding to Cdk4/cyclin Cdk2 and cyclin A. Subdomain D2 forms a β -hairpin and an D1 revealed thermodynamic differences in comparison with intermolecular β -sheet on the surface of the N-terminal lobe binding to Cdk2/cyclin A. Although the Δ G for p27 binding of Cdk2 (Figure 2A). Through these interactions, p27 inhibits to Cdk4/cyclin D1 was similar to that observed for binding to Cdk2/cyclin A in several different ways. First, the RxL motif Cdk2/cyclin A, the contributions from enthalpy and entropy within subdomain D1, a mimic of the substrate recognition (enthalpy/entropy compensation) were markedly different sequence (Schulman et al. , 1998 ), blocks the substrate bind- (Ou et al. , 2011 ). As discussed above, a ﬂ at conformational ing site on cyclin A. Next, subdomain D2 displaces the ﬁ rst energy landscape in the free state is a characteristic of IDPs. β -strand of Cdk2, remodeling the catalytic cleft. Finally, the The different contributions of enthalpy and entropy in the

310 helix observed in solution occupies the ATP-binding pocket formation of the two p27/Cdk/cyclin complexes suggested of Cdk2. The structure of this ternary complex illustrates how differences in the extent of the interaction between the dif- several different short segments (termed subdomains here) ferent subdomains of p27 and the various surfaces of the two within an IDP can achieve complex functionality: inhibition Cdk/cyclin complexes. In particular, the ITC data suggested of substrate recruitment, inhibition of kinase activity through that subdomain D2 was more dynamically associated with structural remodeling and inhibition of ATP binding. Cdk4 than with Cdk2. These results illustrate that, even with closely related binding partners, conserved motifs within Sequential binding mechanism IDPs can utilize different blends of enthalpy and entropy to drive binding. Further, these different thermodynamic blends Although the p27/Cdk2/cyclin A structure provided detailed result in Cdk/cyclin complexes that exhibit different dynamic insights into the molecular basis of inhibition of Cdk/cyclin properties (Lacy et al. , 2005 ). complexes by p27 (Russo et al., 1996), questions remained regarding the kinetics and thermodynamics of complex for- Regulation of p27 function by phosphorylation mation. Isothermal titration calorimetry (ITC) (Lacy et al., 2004) and surface plasmon resonance (SPR) (Lacy et al., The localization, turnover, and activity of p27 are regulated 2005 ) were used to measure thermodynamic and kinetic through posttranslational modifi cations, the most notewor- parameters associated with p27 binding to Cdk2/cyclin A, thy being phosphorylation. For example, phosphorylation revealing a sequential binding mechanism. First, the highly of Thr187 by the Cdk2/cyclin E (and A) complexes (Sheaff disordered and dynamic D1 subdomain rapidly binds to a et al., 1997; Vlach et al., 1997) creates a phospho-degron conserved binding pocket on the surface of cyclin A. Next, (Carrano et al. , 1999 ), which signals the recruitment and the LH subdomain fully folds into an α -helix, which spans the binding of the SCFSkp2 E3 ubiquitin ligase complex for ubiq- gap from cyclin A to Cdk2. Three IFSUs, a β -hairpin, an inter- uitination and subsequent proteasomal degradation of p27. β molecular -strand, and a 310 helix, sequentially fully fold as Surprisingly, phosphorylation of Thr187 occurs although the p27 binds to Cdk2. Cdk2 is dramatically remolded during p27 KID (of p27) remains bound to Cdk2/cyclin E (and A), an binding, including displacement of a β -strand by a segment interaction known to inhibit the kinase activity of Cdk2. This of p27. We envision that the highly fl exible D1 subdomain of apparent paradox was resolved by recent studies that uncov- p27 rapidly scans protein surfaces, becoming engaged on the ered a mechanism by which p27 functions as a conduit for surface of cyclin A after encountering its specifi c binding site. transmitting intramolecular signals through posttranslational The rapid association kinetics of this interaction (Lacy et al., modifi cations. Specifi cally, Grimmler et al. (2007) showed 2004) may be due to the large capture radius of disordered that phosphorylation of Cdk2/cyclin A-bound p27 by NRTKs, p27 à la the ‘ fl y casting’ mechanism (Shoemaker et al., 2000). such as Lyn and Abl, on Tyr88 relieved inhibition of Cdk2 This fast initial step is followed by the slow component of the and promoted Cdk2-mediated phosphorylation of Thr187 via binding mechanism, when the longer, less fl exible D2 subdo- a pseudo unimolecular reaction mechanism. main folds onto the surface of Cdk2 and becomes positioned As discussed above, crystallographic analysis of the within the active site (Figure 2B). This sequential mechanism, p27/Cdk2/cyclin A complex (Figure 2A) showed that Tyr88 which combines the induced fi t and conformational selec- binds within the Cdk2-active site and blocks ATP binding. This tion mechanisms, is conserved when p27 interacts with other residue, however, is able to fl uctuate between ATP-binding Cdk/cyclin complexes. For example, kinetic analysis revealed pocket-bound and solvent-exposed conformations, thus that a truncated form of p27 lacking subdomain D1 bound to allowing phosphorylation by NRTKs when solvent exposed Cdk4/cyclin D at a much lower rate than did a p27 construct (Ou et al. , 2011 ). NMR studies showed that phosphorylation containing this subdomain. This suggests that subdomain of Tyr88 of p27 by Abl leads to ejection of the inhibitory 310 D1 accelerates binding to this Cdk/cyclin complex as well helix of p27 from the Cdk2 ATP-binding pocket, leaving other (L. Ou, B. Waddell, and R. Kriwacki, unpublished results). subdomains unperturbed (Grimmler et al. , 2007 ). Sequence analysis of proteins in the Cip/Kip family as well The CTD of p27 is a critical component of this intramo- as the Cdks and cyclins that regulate cell division suggests lecular signaling system, mediating signal transmission that the sequential binding mechanism is conserved. p27 has between the fi rst (phosphorylation of Tyr88) and second 266 D.M. Mitrea et al.

(phosphorylation of Thr187) steps. NMR analysis showed cell migration. As discussed above, under certain circum- that this domain is highly disordered both in isolation and stances phosphorylation on Ser10 (Ishida et al. , 2000 ) or when the KID is bound to Cdk2/cyclin A (Galea et al. , Thr157 (Liang et al., 2002; Shin et al., 2002; Viglietto et al., 2008a). MD simulations showed that the p27 C-terminal 2002) causes cytoplasmic localization, wherein p27 interacts segment immediately following the KID protruded from the with RhoA and modulates activation of ROCK and other surface of Cdk2/cyclin A and exhibited somewhat restricted downstream signaling molecules that ultimately control the fl exibility. Electrostatic repulsion between this segment con- dynamics of actin stress fi bers (Besson et al., 2004). p27 necting the KID and other regions of the CTD (residues directly binds RhoA and inhibits GDP/GTP exchange by sev- 110– 140) and Cdk2/cyclin A may dislodge portions of p27 eral GEFs. The CTD of p27 was reported to mediate these from the complex after Tyr88 phosphorylation, enabling the inhibitory interactions with RhoA, but the structural details C-terminal segment containing Thr187 to fold back into the of these interactions have not yet been elucidated. Although catalytic site (Figure 3 A). Biochemical analysis showed that incompletely understood, these observations demonstrate intracomplex signal transduction was 8-fold more effi cient how an IDP can perform multiple functions in distinct subcel- than the bimolecular, intercomplex mechanism (Grimmler et lular compartments and how different domains can evolve to al., 2007). The conformational gymnastics of the p27-CTD, mediate these various functions. within the complex with Cdk2/cyclin A and when Tyr88 is phosphorylated, mediate intramolecular signal transmission. The fi nal steps of the signaling cascade involve recognition of Disorder-function relationships: lessons from p21 phosphorylated Thr187 by the SCFSkp2 E3 ligase followed by p27 ubiquitination and degradation. The structure of a p27- p21: Early insights into the functional role of derived peptide containing phosphorylated Thr187 bound to polypeptide disorder Skp1/Skp2/Cks1 revealed how this phospho-degron is recog- nized by the E3 ubiquitin ligase complex (Figure 3B) (Hao p21 is the smallest member of the Cip/Kip family of IDPs et al. , 2005 ). The p27/Cdk2/cyclin A signaling system illus- (164 amino acids) and, as discussed above, binds and inhibits trates how functional complexity can be encoded within a the full repertoire of Cdk/cyclin complexes that regulate cell relatively small disordered protein: promiscuous binding division. p21 is expressed in response to genotoxic stress- to the full repertoire of Cdk/cyclin complexes, inhibition induced activation of p53 and causes cell cycle arrest at the of several aspects of Cdk/cyclin function, and sensitivity to G1 /S and G2 /M checkpoints (Xiong et al., 1993; Harper et al., extrinsic signals enabling inhibitory function to be relieved 1995). p21 was among the fi rst examples of a protein that was and participation in signal transduction. Unfortunately, this fully disordered under physiological conditions and biologi- complexity exposes the system to functional alteration, and cally functional. In 1996, Kriwacki et al. reported, based on can contribute to cancer development and progression, as NMR data, that p21 experiences a disorder to order transi- discussed above. tion upon association with Cdk2. Because p21 is a universal regulator of Cdk/cyclin complexes (Xiong et al. , 1993 ), Other functions of p27 these authors proposed that the ability to fold upon binding conferred ‘ binding promiscuity,’ enabling p21 to regulate the The nuclear Cdk regulatory functions of p27 are comple- full Cdk/cyclin repertoire. It was imagined that p21 folded mented by cytoplasmic functions associated with control of into related but different conformations to regulate different

ABY74 Y88 p27 Skp2 Cyclin A Skp1 Step 2

Step 1 p27-CTD

C T187 Cdk2 Cks1

Figure 3 p27 functions as a signaling conduit to prime its C-terminus for recruitment of the degradation machinery. (A) Snapshot from an MD trajectory illustrating the dynamic character of the p27 CTD (yellow tube) and the mechanism of intracomplex signal transduction (taken from Galea et al. , 2008b ). Phosphorylation of Tyr88 (green) or Tyr74 (red) and Tyr88 relieves inhibition of Cdk2 by ejecting the 310 helix (containing Tyr88) from the active site (step 1) and allows Thr187 (orange) within the CTD to engage the active site and become phosphorylated (step 2), which leads to subsequent recruitment of Skp2. The p27 KID is represented as a gray tube (adapted from Galea et al., 2008b ). (B) Crystal structure of a portion of the p27 CTD bound to the E3 ubiquitin ligase, Skp1/Skp2/Cks1, involved in polyubiquitination and proteasomal degradation of p27. Residues 181 – 190 of p27-CDT (orange) interact directly with both Skp2 (olive) and Cks1 (pink) and indirectly with Skp1 (wheat) (PDB accession number 2AST). Disorder-function relationship for p21 and p27 267

Cdk/cyclin complexes. This speculation was borne out in a within crystals revealed heightened disorder, in contrast to recent study, as discussed in the following. a high degree of order observed for subdomains D1 and D2 (Russo et al., 1996). Furthermore, in this crystal structure, the Structural adaptation by p21 as a mechanism of LH subdomain of p27 exhibited a conformation correspond- binding promiscuity ing to a stretched α -helix, elongated over its 22 residue length by about 4 Å . Wang et al. (2011) deduced that helix stretching The sequences of the D1 and D2 subdomains of the was responsible for the dynamic features of p21 when bound Cip/Kip family members are highly conserved, whereas those to Cdk2/cyclin A and that this ‘ structural adaptation ’ by both of the LH subdomains exhibit variation (Figure 1A) (Lacy p21 and p27 was associated with their ability to bind promis- et al. , 2004 ). Structural studies using NMR demonstrated cuously to the entire Cdk/cyclin repertoire. that the conformation of p21-KID bound to Cdk2/cyclin Experiments with p21 variants with different length LH A in solution (Wang et al., 2005b) was similar to that of subdomains validated this hypothesis. Also, the crystal struc- p27-KID bound to the same Cdk/cyclin pair in crystals tures of Cdk4/cyclin D1 (Day et al. , 2009 ) and Cdk4/cyclin D3 (Russo et al. , 1996 ). However, while some residues of the (Takaki et al., 2009) revealed that the LH subdomain would p21 LH subdomain exhibited helical resonance features, have to contract and pivot in order for the D1 and D2 subdo- others exhibited resonance broadening due to conformational mains to bind conserved surfaces on the D-type cyclins and dynamics. This was a critical clue regarding the origins of Cdk4, respectively (Figure 4 ). The lack of tertiary contacts binding promiscuity. Analysis of 1 H-15 N hetNOE values for between the different subdomains of the p21 and p27 KIDs residues in p21-KID bound to Cdk2/cyclin A revealed that, creates a ﬂ at energy landscape that enables structural adap- although amide dynamics for residues within subdomains D1 tation while folding upon binding to Cdk/cyclin complexes. and D2 were restricted, residues within the LH subdomain This study also illustrates how different regions of an IDP for which resonances were observed experienced height- are needed for different aspects of function; in this case, sub- ened dynamics. Inspection of B-factors for this region of p27 domains D1 and D2 have evolved for speciﬁ c recognition of

A Standard helix

Sub-domain LH is flexible and able to stretch Stretched helix

p21-KID

Structural adaptation, binding diversity B

36.2 Å 34.0 Å Leu 101 ? Å Leu 255 Other Cdk/cyclin Val 30 Val 32 complexes that regulate cell division Cyclin D3 Cyclin A Cdk2 Cdk4

Figure 4 Structural adaptation by the LH subdomain helix in p21-mediated promiscuous binding to the Cdk/cyclin repertoire that regulates cell division. (A) Comparison of a standard α -helix, 22 residues in length, with an NMR-validated model of the LH subdomain of p21; structural adaptation through helix stretching accompanies binding to Cdk2/cyclin A. (B) The binding of p21 to different members of the Cdk/cyclin repertoire requires structural adaptation by the LH subdomain through helix stretching/contraction and pivoting; the distance between conserved features of the p21 binding sites within the Cdk and cyclin subunits are illustrated for Cdk2/cyclin A (magenta/cyan) and Cdk4/cyclin D3 (red/green). The locations of and distance between Val30 and Leu255 and between Val32 and Leu101, respectively, within these two complexes are illustrated. 268 D.M. Mitrea et al.

conserved features of cyclin and Cdk surfaces, respectively, p21 was unknown, the activity of p27- and p21-containing whereas subdomain LH has evolved to adaptively span the Cdk/cyclin complexes is likely due at least in part to tyrosine ∼ 35- Å gap between these subunits. phosphorylation and not variable binding stoichiometry as proposed by Zhang et al. (1994) . p21 as a signaling conduit Subdomain D1 of p21, containing the RxL motif (Schulman et al. , 1998 ) is essentially identical to that of p27 and therefore As described above, p27 acts as a fl exible signaling conduit mediates binding to a conserved binding pocket on the sur- by conducting phosphorylation and ubiquitination signals face of cyclin A (Wang et al. , 2005b ). In addition, p21 exhibits along its length. The fi rst signal, Tyr88 phosphorylation, a second RxL motif within its C-terminal domain (residues causes transmission of the second phosphorylation of Thr187 R155 RL) (Adams et al. , 1996 ), which has been shown to medi- through reactivation of Cdk2 and the fl exibility of the p27 ate binding to cyclins A and E (Adams et al., 1996; Chen et CTD (Figure 3A). Finally, the Thr187 phospho-degron is rec- al. , 1996 ), although with reduced affi nity relative to that of ognized by SCFSkp2 for ubiquitination and p27 is targeted for the N-terminal KID (Chen et al. , 1996 ). The existence of a degradation. The motifs in p27 that are critical for this signal- secondary cyclin-binding site creates the opportunity for the ing conduit are conserved in p21. For example, Tyr77 within binding of multiple Cdk/cyclin complexes to p21. Because subdomain D2 of p21 mimics Tyr88 of p27, and the region the affi nity of the N-terminal KID of p21 for Cdk/cyclin com- near Ser130 of p21 is a Cdk2 recognition sequence, similar to plexes (e.g., Cdk2/cyclin A) is extremely high (IC50 , 300 p m ), that near Thr187 in p27. Further, disorder within the CTD of Cdk/cyclin complexes will fi rst bind to and be inhibited by p21 (Esteve et al., 2003; Yoon et al., 2009) may mediate signal the N-terminal KID. However, if Cdk/cyclin complexes are transduction from step 1 to step 2 as does that of p27 (Figure in excess of p21 molecules, they will additionally bind to the 3A). Interestingly, it has been reported that Ser130 phospho- secondary C-terminal RxL motif. Zhu et al. (2005) showed rylation by Cdk2/cyclin E is required for ubiquitination of through cellular overexpression studies that this secondary p21 by SCF Skp2 and its subsequent proteasomal degradation interaction serves to recruit active Cdk2/cyclin E complexes (Bornstein et al. , 2003 ; Zhu et al. , 2005 ). This experimental for phosphorylation of p21 on Ser130. As discussed above, evidence supports the hypothesis that the signaling conduit phosphorylation of Ser130 is a signal for p21 ubiquitination observed for p27 is conserved in p21 and that NRTKs control and 26S proteosomal degradation. What is the physiologi- p21 activity and stability. However, this hypothesis is only cal relevance of this observation? The secondary RxL motif beginning to be experimentally tested (see below). will be populated with active Cdk/cyclin complexes only The possibility that the Cdk inhibitory activity of p21 is after the primary N-terminal KID is populated with inhibited regulated by tyrosine phosphorylation provides a plausible complexes; this will occur only when Cdk/cyclin complexes explanation for past results that demonstrated that Cdk/cyclin are in stoichiometric excess over p21. This condition exists complexes containing p21 were catalytically active (Zhang when p21 levels initially begin to rise after p53-dependent et al., 1994; LaBaer et al., 1997; Cheng et al., 1999). Zhang activation due to DNA damage. We propose that the second- et al. (1994) originally proposed that multiple p21 molecules ary cyclin-binding site exists to prime p21 for degradation could bind to Cdk/cyclin complexes, with binding of the fi rst through Ser130 phosphorylation to provide a mechanism for associated with kinase activity and that of others with kinase rapid p21 turnover (Zhu et al. , 2005 ) and transient p21-depen- inhibition. However, the requirement for the binding of mul- dent cell cycle arrest. This hypothesis will require verifi cation tiple p21 molecules to achieve Cdk inhibition was challenged through studies of endogenous proteins in cells responding to by two later reports (Hengst et al., 1998; Adkins and Lumb, DNA damage. 2000). The former of these clearly demonstrated that the catalytic activity of Cdk2/cyclin A was completely inhibited Interactions of the C-terminal domain of p21 with through binding of a single p21 molecule. Based on recent PCNA and their regulation by phosphorylation studies of tyrosine phosphorylation of p27, it is possible that the p21 protein under study in these earlier reports (Zhang et In addition to its previously discussed role of causing cell al., 1994; LaBaer et al., 1997; Cheng et al., 1999) was phos- cycle arrest through inhibition of Cdk2/cyclin A or E com- phorylated on Tyr77. By analogy with p27, complexes of a plexes at the G 1 to S phase transition, p21 also inhibits DNA single molecule of p21 with Cdk2/cyclin A (Grimmler et al. , replication during S phase by inhibiting the DNA polymerase 2007 ) or Cdk4/cyclin D (James et al. , 2008 ) could, in prin- δ processivity factor, PCNA (Waga et al. , 1994 ). A portion of ciple, be ‘ activated ’ through phosphorylation of Tyr77 and the CTD of p21 (Gly139 to Ser160) binds to PCNA with high displacement of this portion of p21 from the ATP-binding affi nity. p21 inhibits the DNA replication function of PCNA pocket of Cdk2. This model of activation of p21-containing by blocking its interaction with other DNA replication factors Cdk/cyclin complexes is supported by structural and in vitro but not its role in DNA repair (Li et al., 1994; Chen et al., biochemical data showing that, upon phosphorylation, Tyr77 1995; Luo et al., 1995). Interactions with PCNA are inhib- is ejected from the Cdk2-active site and that Cdk2 regains ited through phosphorylation of p21 on Thr145, Ser146, or signifi cant catalytic activity (Yoon, Park, and Kriwacki, Ser160, as discussed below. manuscript in preparation). Therefore, although the multiple Similar to p21-KID, the disordered p21 CTD experiences p21 molecule-binding model was compelling at a time when folding upon binding to its partners, including PCNA. In the regulatory role of tyrosine phosphorylation of p27 and 1996, Gulbis et al. (1996) reported the crystal structure of Disorder-function relationship for p21 and p27 269

three copies of a 22 amino acids C-terminal peptide of p21 performs its oncogenic functions by interacting with cytosolic in complex with trimeric PCNA (Figure 5 A). The PCNA- proteins. For example, activated Akt resulted in the increase bound p21 peptide assumes a generally extended conforma- of the p21 stability (Li et al. , 2002 ), perhaps mediated by the tion and can be divided into three segments. The fi rst fi ve formation of Cdk4/cyclin D1/p21 ternary complexes in the N-terminal residues, Gly139 to Arg143, are mostly basic and cytosol (Coleman et al. , 2003 ). mediate dynamic electrostatic interactions with complemen- Protein kinase C, a downstream effector of Akt, was found tary charges on the PCNA surface (Figure 5A). The central to phosphorylate p21 on Ser146 (Scott et al., 2000; Li et al., six residues, comprising Ser146 to Tyr151, form one turn of 2002 ), Ser153 (Rodriguez -Vilarrupla et al., 2005 ), and Ser160 a 310 helix and occupy a hydrophobic groove on the PCNA (Scott et al. , 2000 ). Modifi cations at Ser146 and Ser160 are surface. The remaining amino acids, His152 to Ser160, form associated with inhibition of PCNA binding (Scott et al., an intermolecular antiparallel β -strand with an interdomain 2000; Li et al., 2002). Furthermore, phosphorylation of Ser146 loop of PCNA, stabilized by intermolecular electrostatic and mediates the degradation of p21 by serving as a binding site hydrogen bonds (Figure 5A). As was observed for p27-KID, for the C8 subunit of the proteasome (Touitou et al. , 2001 ). some regions within the unbound, isolated p21 CTD exhibit Ser153 phosphorylation regulates subcellular localization by conformational bias toward the bound conformation in solu- targeting p21 to the cytoplasm (Rodriguez -Vilarrupla et al., tion (Yoon et al. , 2009 ). 2005 ). In summary, signals from multiple kinase pathways are Interestingly, the region of the p21 CTD shown to fold into integrated within the dynamic structure of p21, altering Cdk a β -strand when associated with PCNA (Thr145 to Pro164) inhibitory function, stability, and subcellular localization. This also associates with calmodulin (CaM). Analyses by NMR regulatory complexity is achieved because the p21 sequence (Yoon et al., 2009) and molecular modeling (Esteve et al., contains many residues that can be posttranslationally modi- 2003 ) support the conclusion that this segment of p21 is fi ed and because these residues remain dynamic (to varied α -helical when bound to CaM (Figure 5B), providing an extents) even when p21 is bound to Cdk/cyclin complexes. example of structural polymorphism within a disordered protein domain mediating binding to diverse targets. As previously discussed, phosphorylation events within the Concluding remarks CTD of p21 regulate several aspects of p21 function, including PCNA-binding activity and subcellular localization, and The fi eld of IDPs, as a branch of the broader fi eld of struc- dysregulation of these regulatory signals has been linked to tural biology, has come of age in the past 5 years, as the disease. Akt, a regulator of cell proliferation and survival, prevalence and biological importance of this distinct class of can phosphorylate Thr145 (Rossig et al. , 2001 ), modifi cation proteins have become broadly appreciated. From a systems shown to inhibit the interaction of p21 with PCNA (Scott et biology perspective, it is now known that IDPs, as a class, al. , 2000 ; Rossig et al. , 2001 ). The structure of p21 C-terminal are more tightly controlled compared with the folded proteins peptide bound to PCNA (Gulbis et al., 1996) suggests that (Gsponer et al., 2008). Although IDPs perform diverse func- phosphorylation of p21 on Thr145 might disrupt the intramo- tions in cells, their roles in signaling and regulation are partic- lecular and intermolecular hydrogen bonding, thus decreasing ularly critical for controlling cellular behavior. Consequently, the interaction between p21 and PCNA. Also, Thr145 phos- disruption of IDP homeostasis is more likely to be detrimen- phorylation has been found to result in the subcellular relocal- tal to cells than is disruption of their folded counterparts ization of p21 from the nucleus to the cytoplasm, in which p21 (Vavouri et al., 2009). The increasing number of disordered proteins implicated in disease onset and progression supports this hypothesis (Babu et al. , 2011 ). The system-scale studies noted above were enabled by algorithms that accurately AB p21-CTD identify proteins, or regions of proteins, that are likely to be CaM disordered. However, although these and related studies have broadcast the biological importance of IDPs, they do not pro- PCNA vide insights into the rich landscape of physical mechanisms p21-CTD through which IDPs perform their functions. In this review, we have summarized the biophysical, structural, and functional properties of two IDPs, p21 and p27. They regulate a variety of targets that play key roles in essential biological processes. For example, they control the function of Cdks in cell cycle progression and modulate activity of proteins that function within DNA replication complexes, regulate actin Figure 5 The p21 CTD exhibits structural polymorphism. dynamics, and regulate cell migration. Although structural (A) A portion of the p21 CTD (residues 144– 164) adopts an extended conformation upon binding to three equivalent sites on the surface and dynamic characterization of the free states of IDPs can of the PCNA trimer (PDB accession number 1AXC). (B) NMR and defi ne their energy landscape, this information alone does computational studies suggest that the same region of p21 CTD (res- not provide insights into the mechanism(s) of their func- idues 145– 164) adopts an α -helical conformation upon binding to tion. These fundamental physical properties, the time evolu- calmodulin (reproduced from Esteve et al. , 2003 , with permission). tion of structure, must be understood in the context of their 270 D.M. Mitrea et al.

Step 1 Step 2 Step 3 Step 4 Y88 ATP Cdk2 A pY88 pY88 pY88 NRTKs Cdk2 SCFSkp2 26S Ub Proteasome Ub Ub Ub T187 Ub Ub Ub Ub Imatinib Ub T187 pT187 pT187

Cdk2/cyclin A Partial Intra-complex Ubiquitination Degradation of p27 inhibited, reactivation of phosphorylation of p27 by 26S cell cycle Cdk2/cyclin A of T187 by SCFSkp2 proteasome, arrest Cdk2/cyclin A is fully active, cell cycle progression

Figure 6 p27 serves as a signaling conduit that controls cell cycle progression. Cell cycle is arrested by p27 inhibition of Cdk/cyclin complexes. The inhibitory effect is released through phosphorylation of p27 at Tyr88 by NRTKs, which are up-regulated in cancer and can be suppressed with kinase inhibitors drugs, such as imatinib (step 1). The now liber- ated kinase active site is able to phosphorylate p27 at Thr187 (step 2), which in turn recruits SCFSkp2 ubiquitin ligase complex (step 3). Polyubiquitinated p27 is fi nally degraded by the 26S proteasome (step 4). Note that the ubiquitination pattern shown in the Figure is solely for illustrative purposes (adapted from Galea et al. , 2008b ). functional complexes. For example, studies of p21 bound recognition motif present in substrates and p21-like cyclin-de- to Cdk2/cyclin A revealed that a simple helix-stretching pendent kinase inhibitors. Mol. Cell. Biol. 16 , 6623 – 6633. mechanism mediated Cdk-binding promiscuity. Similarly, Adkins, J.N. and Lumb, K.J. (2000). Stoichiometry of cylin A-cyclin Cip1/Waf1 studies of Tyr88-phosphorylated p27 bound to this same dependent kinase 2 inhibition by p21 . Biochemistry 39 , 13925 – 13930. Cdk/cyclin complex lead to the discovery of the signaling con- Amador, V., Ge, S., Santamaria, P.G., Guardavaccaro, D., and Pagano, duit mechanism (Figure 6 ) that controls cell cycle entry and M. (2007). APC/C(Cdc20) controls the ubiquitin-mediated deg- is disrupted in CML and breast cancer. The physical mecha- radation of p21 in prometaphase. Mol. Cell 27 , 462 – 473. nisms underlying function are understood for only a small Babu, M., van der Lee, R., de Groot, N., and Gspooner, J. (2011). fraction of proteins that constitute the disordered proteome. Intrinsically disordered proteins: regulation and disease. Curr. Just as interdisciplinary studies allowing correlations of struc- Opin. Struct. Biol. 21 , 432 – 440. ture and function have empowered understanding of the func- Besson, A., Gurian-West, M., Schmidt, A., Hall, A., and Roberts, tional mechanisms of folded proteins, broad approaches suited J.M. (2004). p27 Kip1 modulates cell migration through the regula- to the dynamic features of IDPs must be applied to understand tion of RhoA activation. Genes Dev. 18 , 862 – 876. relationships between disorder and function. Expanding our Bienkiewicz, E.A., Adkins, J.N., and Lumb, K.J. (2002). Functional consequences of preorganized helical structure in the intrinsi- understanding of disorder-function relationships for IDPs has cally disordered cell-cycle inhibitor p27Kip1 . Biochemistry 41 , great potential to impact human health. Currently, IDPs are 752 – 759. generally considered to be undruggable targets for therapeutic Blain, S.W. (2008). Switching cyclin D-Cdk4 kinase activity on and disease intervention. Several recent studies (Metallo , 2010 ), off. Cell Cycle 7 , 892 – 898. however, demonstrate that small organic molecules can bind Blain, S.W. and Massague, J. (2002). Breast cancer banishes p27 tightly and specifi cally to regions within IDPs, providing from nucleus. Nat. Med. 8 , 1076 – 1078. incentive for more intensive effort toward the development Boehr, D.D., Nussinov, R., and Wright, P.E. (2009). The role of of drugs that modulate IDP function. Detailed investigation dynamic conformational ensembles in biomolecular recognition. of the functional mechanisms of IDPs together with drug dis- Nat. Chem. Biol. 5 , 789 – 796. covery efforts to target those IDP mechanisms that go awry Bornstein, G., Bloom, J., Sitry-Shevah, D., Nakayama, K., Pagano, M., and Hershko, A. (2003). Role of the SCFSkp2 ubiquitin ligase in diseases such as cancer and neurodegeneration will expand in the degradation of p21 Cip1 in S phase. J. Biol. Chem. 278 , our knowledge of the molecular basis of living systems and 25752 – 25757. possibly enhance human health in the future. Borriello, A., Bencivenga, D., Criscuolo, M., Caldarelli, I., Cucciolla, V., Tramontano, A., Borgia, A., Spina, A., Oliva, A., Naviglio, S., et al. (2011a). Targeting p27Kip1 protein: its relevance in the ther- References apy of human cancer. Exp. Opin. Ther. Targets. 15 , 677 – 693. Borriello, A., Caldarelli, I., Bencivenga, D., Criscuolo, M., Cucciolla, Abbas, T., Sivaprasad, U., Terai, K., Amador, V., Pagano, M., and V., Tramontano, A., Oliva, A., Perrotta, S., and Della Rangione, Dutta, A. (2008). PCNA-dependent regulation of p21 ubiquityla- F. (2011b). p57Kip2 and cancer: time for a critical appraisal. Mol. tion and degradation via the CRL4Cdt2 ubiquitin ligase complex. Cancer Res. 9 , 1269 – 1284. Genes Dev. 22 , 2496 – 2506. Brown, C.J., Johnson, A.K., and Daughdrill, G.W. (2010). Comparing Adams, P.D., Sellers, W.R., Sharma, S.K., Wu, A.D., Nalin, C.M., models of evolution for ordered and disordered proteins. Mol. and Kaelin, W.G. Jr. (1996). Identifi cation of a cyclin-cdk2 Biol. Evol. 27 , 609 – 621. Disorder-function relationship for p21 and p27 271

Carrano, A.C., Eytan, E., Hershko, A., and Pagano, M. (1999). SKP2 Galea, C.A., Wang, Y., Sivakolundu, S.G., and Kriwacki, R.W. is required for ubiquitin-mediated degradation of the CDK inhib- (2008b). Regulation of cell division by intrinsically unstructured itor p27. Nat. Cell Biol. 1 , 193 – 199. proteins: intrinsic fl exibility, modularity, and signaling conduits. Chen, J., Jackson, P.K., Kirschner, M.W., and Dutta, A. (1995). Biochemistry 47 , 7598 – 7609. Separate domains of p21 involved in the inhibition of Cdk kinase Galea, C.A., High, A.A., Obenauer, J.C., Mishra, A., Park, C.G., and PCNA. Nature 374 , 386 – 388. Punta, M., Schlessinger, A., Ma, J., Rost, B., Slaughter, C.A., Chen, J., Saha, P., Kornbluth, S., Dynlacht, B.D., and Dutta, A. et al. (2009). Large-scale analysis of thermostable, mammalian (1996). Cyclin-binding motifs are essential for the function of proteins provides insights into the intrinsically disordered pro- p21 CIP1 . Mol. Cell. Biol. 16 , 4673 – 4682. teome. J. Proteome Res. 8 , 211 – 226. Chen, X., Chi, Y., Bloecher, A., Aebersold, R., Clurman, B.E., and Gnad, F., Gunawardena, J., and Mann, M. (2011). PHOSIDA 2011: Roberts, J.M. (2004). N-acetylation and ubiquitin-independent the posttranslational modifi cation database. Nucleic Acids Res. proteasomal degradation of p21Cip1 . Mol. Cell 16 , 839 – 847. 39 , D253 – D260. Cheng, M., Olivier, P., Diehl, J.A., Fero, M., Roussel, M.F., Roberts, Grimmler, M., Wang, Y., Mund, T., Cilensek, Z., Keidel, E.M., Waddell, J.M., and Sherr, C.J. (1999). The p21Cip1 and p27Kip1 CDK ‘ inhib- M.B., Jakel, H., Kullmann, M., Kriwacki, R.W., and Hengst, L. Kip1 itors’ are essential activators of cyclin D-dependent kinases in (2007). Cdk-inhibitory activity and stability of p27 are directly murine fi broblasts. EMBO J. 18 , 1571 – 1583. regulated by oncogenic tyrosine kinases. Cell 128 , 269 – 280. Chu, I., Sun, J., Arnaout, A., Kahn, H., Hanna, W., Narod, S., Gsponer, J., Futschik, M.E., Teichmann, S.A., and Babu, M.M. Sun, P., Tan, C.K., Hengst, L., and Slingerland, J. (2007). p27 (2008). Tight regulation of unstructured proteins: from transcript Phosphorylation by Src regulates inhibition of cyclin E-Cdk2. synthesis to protein degradation. Science 322 , 1365 – 1368. Cell 128 , 281 – 294. Guan, K.L., Jenkins, C.W., Li, Y., Nichols, M.A., Wu, X., O’Keefe, Coleman, M.L., Marshall, C.J., and Olson, M.F. (2003). Ras pro- C.L., Matera, A.G., and Xiong, Y. (1994). Growth suppression motes p21Waf1/Cip1 protein stability via a cyclin D1-imposed block by p18, a p16INK4/MTS1 and p14INK4B/MTS2-related CDK6 in proteasome-mediated degradation. EMBO J. 22 , 2036 – 2046. inhibitor, correlates with wild-type pRb function. Genes Dev. 8 , Connor, M.K., Kotchetkov, R., Cariou, S., Resch, A., Lupetti, R., 2939 – 2952. Beniston, R.G., Melchior, F., Hengst, L., and Slingerland, J.M. Gulbis, J.M., Kelman, Z., Hurwitz, J., O ’ Donnell, M., and Kuriyan, J. (1996). Structure of the C-terminal region of p21WAF1/CIP1 com- (2003). CRM1/Ran-mediated nuclear export of p27 Kip1 involves plexed with human PCNA. Cell 87 , 297 – 306. a nuclear export signal and links p27 export and proteolysis. Mol. Hannon, G.J. and Beach, D. (1994). p15ink4b is a potential effector Biol. Cell 14 , 201 – 213. of TGF-β -induced cell cycle arrest. Nature 371 , 257 – 261. Davey, N.E., Trave, G., and Gibson, T.J. (2011). How viruses hijack Hao, B., Zheng, N., Schulman, B.A., Wu, G., Miller, J.J., Pagano, cell regulation. Trends Biochem. Sci. 36 , 159 – 169. M., and Pavletich, N.P. (2005). Structural basis of the Day, P.J., Cleasby, A., Tickle, I.J., O ’ Reilly, M., Coyle, J.E., Holding, Cks1-dependent recognition of p27 Kip1 by the SCF Skp2 ubiquitin F.P., McMenamin, R.L., Yon, J., Chopra, R., Lengauer, C., et al. ligase. Mol. Cell 20 , 9 – 19. (2009). Crystal structure of human CDK4 in complex with a Harbour, J.W., Luo, R.X., Dei Santi, A., Postigo, A.A., and Dean, D-type cyclin. Proc. Natl. Acad. Sci. USA 106 , 4166 – 4170. D.C. (1999). Cdk phosphorylation triggers sequential intramo- Dunker, A.K., Obradovic, Z., Romero, P., Garner, E.C., and Brown, lecular interactions that progressively block Rb functions as cells C.J. (2000). Intrinsic. protein disorder in complete genomes. move through G1. Cell 98 , 859 – 869. Genome Inform. Ser. Workshop Genome Inform. 11 , 161 – 171. Harper, J.W., Elledge, S.J., Keyomarsi, K., Dynlacht, B., Tsai, Dyson, H.J. (2011). Expanding the proteome: disordered and alterna- L.H., Zhang, P., Dobrowolski, S., Bai, C., Connell-Crowley, L., tively folded proteins. Q. Rev. Biophys. 44 , 467 – 518. Swindell, E., et al. (1995). Inhibition of cyclin-dependent kinases Dyson, H.J. and Wright, P.E. (2005). Intrinsically unstructured pro- by p21. Mol. Biol. Cell 6 , 387 – 400. teins and their functions. Nat. Rev. Mol. Cell. Biol. 6 , 197 – 208. Hengst, L. and Reed, S.I. (1996). Translational control of p27Kip1 El-Deiry, W.S., Harper, J.W., Connor, P.M., Velculescu, V.E., Canman, accumulation during the cell cycle. Science 271 , 1861 – 1864. C.E., Jackman, J., Pietenpol, J.A., Burrell, M., Hill, D.E., Wang, Hengst, L., Gopfert, U., Lashuel, H.A., and Reed, S.I. (1998). Y., et al. (1994). WAF1/CIP1 in induced in p53-mediated G1 Complete inhibition of Cdk/cyclin by one molecule of p21Cip1 . arrest and apoptosis. Cancer Res. 54, 1169 – 1174. Genes Dev. 12 , 3882 – 3888. Esteve, V., Canela, N., Rodrigues-Vilarrupla, A., Aligue, R., Agell, Hirai, H., Roussel, M.F., Kato, J.Y., Ashmun, R.A., and Sherr, C.J. N., Mingarro, I., Bachs, O., and Perez-Paya, E. (2003). The struc- (1995). Novel INK4 proteins, p19 and p18, are specifi c inhibitors tural plasticity of the C terminus of p21Cip is a determinant for of the cyclin D-dependent kinases CDK4 and CDK6. Mol. Cell. target protein recognition. ChemBioChem 4 , 863 – 869. Biol. 15 , 2672 – 2681. Frescas, D. and Pagano, M. (2008). Deregulated proteolysis by the Hyeon, C. and Onuchic, J.N. (2007). Mechanical control of the direc- F-box proteins SKP2 and β -TrCP: tipping the scales of cancer. tional stepping dynamics of the kinesin motor. Proc. Natl. Acad. Nat. Rev. Cancer 8 , 438 – 449. Sci. USA 104 , 17382 – 17387. Funk, J.O. and Galloway, D.A. (1998). Inhibiting CDK inhibitors: Iakoucheva, L.M., Brown, C.J., Lawson, J.D., Obradovic, Z., and new lessons from DNA tumor viruses. Trends Biochem. Sci. 23 , Dunker, A.K. (2002). Intrinsic disorder in cell-signaling and can- 337 – 341. cer-associated proteins. J. Mol. Biol. 323 , 573 – 584. Galea, C.A., Pagala, V.R., Obenauer, J.C., Park, C.G., Slaughter, Iakoucheva, L.M., Radivojac, P., Brown, C.J., O’ Connor, T.R., Sikes, C.A., and Kriwacki, R.W. (2006). Proteomic studies of the intrin- J.G., Obradovic, Z., and Dunker, A.K. (2004). The importance sically unstructured mammalian proteome. J. Proteome Res. 5 , of intrinsic disorder for protein phosphorylation. Nucleic Acids 2839 – 2848. Res. 32 , 1037 – 1049. Galea, C.A., Nourse, A., Wang, Y., Sivakolundu, S.G., Heller, W.T., Ishida, N., Kitagawa, M., Hatakeyama, S., and Nakayama, K. and Kriwacki, R.W. (2008a). Role of intrinsic fl exibility in signal (2000). Phosphorylation at serine 10, a major phosphorylation transduction mediated by the cell cycle regulator, p27Kip1 . J. Mol. site of p27 Kip1 , increases its protein stability. J. Biol. Chem. 275 , Biol. 376 , 827 – 838. 25146 – 25154. 272 D.M. Mitrea et al.

Ishida, N., Hara, T., Kamura, T., Yoshida, M., Nakayama, K., and (2002). PKB/Akt phosphorylates p27, impairs nuclear import of p27 Nakayama, K.I. (2002). Phosphorylation of p27Kip1 on serine and opposes p27-mediated G1 arrest. Nat. Med. 8 , 1153 – 1160. 10 is required for its binding to CRM1 and nuclear export. J. Lu, Z. and Hunter, T. (2010). Ubiquitylation and proteasomal deg- Biol. Chem. 277 , 14355 – 14358. radation of the p21Cip1 , p27 Kip1 and p57Kip2 CDK inhibitors. Cell James, M.K., Ray, A., Leznova, D., and Blain, S.W. (2008). Cycle 9 , 12. Differential modifi cation of p27(Kip1) controls its cyclin D-cdk4 Luo, Y., Hurwitz, J., and Massague, J. (1995). Cell-cycle inhibition inhibitory activity. Mol. Cell. Biol. 28 , 498 – 510. by independent CDK and PCNA binding domains in p21Cip1. Kamura, T., Hara, T., Kotoshiba, S., Yada, M., Ishida, N., Imaki, H., Nature 375 , 159 – 161. Hatakeyama, S., Nakayama, K., and Nakayama, K.I. (2003). Mark, W.Y., Liao, J.C., Lu, Y., Ayed, A., Laister, R., Szymczyna, B., Degradation of p57Kip2 mediated by SCFSkp2 -dependent ubiquity- Chakrabartty, A., and Arrowsmith, C.H. (2005). Characterization lation. Proc. Natl. Acad. Sci. USA 100 , 10231 – 10236. of segments from the central region of BRCA1: an intrinsically Kamura, T., Hara, T., Matsumoto, M., Ishida, N., Okumura, F., disordered scaffold for multiple protein-protein and protein-DNA Hatakeyama, S., Yoshida, M., Nakayama, K., and Nakayama, interactions ? J. Mol. Biol. 345 , 275 – 287. K.I. (2004). Cytoplasmic ubiquitin ligase KPC regulates prote- McAllister, S.S., Becker-Hapak, M., Pintucci, G., Pagano, M., and olysis of p27Kip1 at G1 phase. Nat. Cell Biol. 6 , 1229 – 1235. Dowdy, S.F. (2003). Novel p27 kip1 C-terminal scatter domain Kim, Y., Starostina, N.G., and Kipreos, E.T. (2008). The CRL4Cdt2 mediates Rac-dependent cell migration independent of cell cycle ubiquitin ligase targets the degradation of p21Cip1 to control arrest functions. Mol. Cell. Biol. 23 , 216 – 228. replication licensing. Genes Dev. 22 , 2507 – 2519. Metallo, S.J. (2010). Intrinsically disordered proteins are potential Kotake, Y., Nakayama, K., Ishida, N., and Nakayama, K.I. (2005). drug targets. Curr. Opin. Chem. Biol. 14 , 481 – 488. Role of serine 10 phosphorylation in p27 stabilization revealed Mittag, T., Kay, L.E., and Forman-Kay, J.D. (2010). Protein dynam- by analysis of p27 knock-in mice harboring a serine 10 mutation. ics and conformational disorder in molecular recognition. J. Mol. J. Biol. Chem. 280 , 1095 – 1102. Recognit. 23 , 105 – 116. Kotoshiba, S., Kamura, T., Hara, T., Ishida, N., and Nakayama, K.I. Montagnoli, A., Fiore, F., Eytan, E., Carrano, A.C., Draetta, G.F., (2005). Molecular dissection of the interaction between p27 and Hershko, A., and Pagano, M. (1999). Ubiquitination of p27 is Kip1 ubiquitylation-promoting complex, the ubiquitin ligase that regulated by Cdk-dependent phosphorylation and trimeric complex formation. Genes Dev. 13 , 1181 – 1189. regulates proteolysis of p27 in G1 phase. J. Biol. Chem. 280 , Morgan, D.O. (1995). Principles of CDK regulation. Nature 374 , 17694 – 17700. 131 – 134. Kriwacki, R.W. and Yoon, M.K. (2011). Cell biology. Fishing in the Morgan, J.L., Song, Y., and Barbar, E. (2011). Structural dynamics nuclear pore. Science 333 , 44 – 45. and multi-region interactions in dynein-dynactin recognition. J. Kriwacki, R.W., Hengst, L., Tennant, L., Reed, S.I., and Wright, P.E. Biol. Chem. 286 , 39349 – 39359. (1996). Structural studies of p21(waf1/cip1/sdi1) in the free and Nakayama, K., Ishida, N., Shirane, M., Inomata, A., Inoue, T., Cdk2-bound state: conformational disorder mediates binding Shishido, N., Horii, I., Loh, D.Y., and Nakayama, K. (1996). Mice diversity. Proc. Natl. Acad. Sci. USA 93 , 11504 – 11509. lacking p27 Kip1 display increased body size, multiple organ hyper- LaBaer, J., Garrett, M.D., Stevenson, L.F., Slingerland, J.M., Sandhu, plasia, retinal dysplasia, and pituitary tumors. Cell 85 , 707 – 720. C., Chou, H.S., Fattaey, A., and Harlow, E. (1997). New func- Nguyen, H., Gitig, D.M., and Koff, A. (1999). Cell-free degrada- tional activities for the p21 family of Cdk inhibitors. Genes Dev. tion of p27 kip1, a G1 cyclin-dependent kinase inhibitor, is depen- 11 , 847 – 862. dent on CDK2 activity and the proteasome. Mol. Cell. Biol. 19 , Lacy, E.R., Filippov, I., Lewis, W.S., Otieno, S., Xiao, L., Weiss, S., 1190 – 1201. Hengst, L., and Kriwacki, R.W. (2004). p27 binds cyclin-CDK Nishitani, H., Shiomi, Y., Iida, H., Michishita, M., Takami, T., and complexes through a sequential mechanism involving binding- Tsurimoto, T. (2008). CDK inhibitor p21 is degraded by a pro- induced protein folding. Nat. Struct. Mol. Biol. 11 , 358 – 364. liferating cell nuclear antigen-coupled Cul4-DDB1Cdt2 pathway Lacy, E.R., Wang, Y., Post, J., Nourse, A., Webb, W., Mapelli, during S phase and after UV irradiation. J. Biol. Chem. 283 , M., Musacchio, A., Siuzdak, G., and Kriwacki, R.W. (2005). 29045 – 29052. Molecular basis for the specifi city of p27 toward cyclin-de- Noutsou, M., Duarte, A.M., Anvarian, Z., Didenko, T., Minde, D.P., pendent kinases that regulate cell division. J. Mol. Biol. 349 , Kuper, I., de Ridder, I., Oikonomou, C., Friedler, A., Boelens, R., 764 – 773. et al. (2011). Critical scaffolding regions of the tumor suppressor Larrea, M.D., Hong, F., Wander, S.A., da Silva, T.G., Helfman, D., Axin1 are natively unfolded. J. Mol. Biol. 405 , 773 – 786. Lannigan, D., Smith, J.A., and Slingerland, J.M. (2009a). RSK1 Oldfi eld, C.J., Meng, J., Yang, J.Y., Yang, M.Q., Uversky, V.N., and drives p27 Kip1 phosphorylation at T198 to promote RhoA inhibi- Dunker, A.K. (2008). Flexible nets: disorder and induced fi t in tion and increase cell motility. Proc. Natl. Acad. Sci. USA 106 , the associations of p53 and 14-3-3 with their partners. BMC 9268 – 9273. Genomics 9 (Suppl), S1. Larrea, M.D., Wander, S.A., and Slingerland, J.M. (2009b). p27 as Ou, L., Ferreira, A.M., Otieno, S., Xiao, L., Bashford, D., and Jekyll and Hyde: regulation of cell cycle and cell motility. Cell Kriwacki, R.W. (2011). Incomplete folding upon binding medi- Cycle 8 , 3455 – 3461. ates Cdk4/cyclin D complex activation by tyrosine phosphoryla- Li, R., Waga, S., Hannon, G.J., Beach, D., and Stillman, B. (1994). tion of inhibitor p27 protein. J. Biol. Chem. 286 , 30142 – 30151. Differential effects by the p21 CDK inhibitor on PCNA-dependent Philipp-Staheli, J., Payne, S.R., and Kemp, C.J. (2001). p27Kip1 : regu- DNA replication and repair. Nature 371 , 534 – 537. lation and function of a haploinsuffi cient tumor suppressor and Li, Y., Dowbenko, D., and Lasky, L.A. (2002). AKT/PKB phos- its misregulation in cancer. Exp. Cell. Res. 264 , 148 – 168. phorylation of p21Cip/WAF1 enhances protein stability of Radivojac, P., Iakoucheva, L.M., Oldfi eld, C.J., Obradovic, Z., p21Cip /WAF1 and promotes cell survival. J. Biol. Chem. 277 , Uversky, V.N., and Dunker, A.K. (2007). Intrinsic disorder and 11352 – 11361. functional proteomics. Biophys. J. 92 , 1439 – 1456. Liang, J., Zubovitz, J., Petrocelli, T., Kotchetkov, R., Connor, M.K., Rodriguez-Vilarrupla, A., Jaumot, M., Abella, N., Canela, N., Brun, S., Han, K., Lee, J.H., Ciarallo, S., Catzavelos, C., Beniston, R., et al. Diaz, C., Estanyol, J.M., Bachs, O., and Agell, N. (2005). Binding Disorder-function relationship for p21 and p27 273

of calmodulin to the carboxy-terminal region of p21 induces A., et al. (2002). Cytoplasmic relocalization and inhibition of the nuclear accumulation via inhibition of protein kinase C-mediated cyclin-dependent kinase inhibitor p27Kip1 by PKB/Akt-mediated phosphorylation of Ser153. Mol. Cell. Biol. 25 , 7364 – 7374. phosphorylation in breast cancer. Nat. Med. 8 , 1136 – 1144. Romero, P., Obradovic, Z., Li, X., Garner, E.C., Brown, C.J., and Vlach, J., Hennecke, S., and Amati, B. (1997). Phosphorylation- Dunker, A.K. (2001). Sequence complexity of disordered pro- dependent degradation of the cyclin-dependent kinase inhibitor tein. Proteins 42 , 38 – 48. p27. EMBO J. 16 , 5334 – 5344. Rossig, L., Jadidi, A.S., Urbich, C., Badorff, C., Zeiher, A.M., and Waga, S., Hannon, G.J., Beach, D., and Stillman, B. (1994). The p21 Dimmeler, S. (2001). Akt-dependent phosphorylation of p21Cip1 inhibitor of cyclin-dependent kinases controls DNA replication regulates PCNA binding and proliferation of endothelial cells. by interaction with PCNA. Nature 369 , 574 – 578. Mol. Cell. Biol. 21 , 5644 – 5657. Wander, S.A., Zhao, D., and Slingerland, J.M. (2011). p27: a barom- Russo, A.A., Jeffrey, P.D., Patten, A.K., Massague, J., and Pavletich, eter of signaling deregulation and potential predictor of response N.P. (1996). Crystal structure of the p27Kip1 cyclin-dependent- to targeted therapies. Clin. Cancer Res. 17 , 12 – 18. kinase inhibitor bound to the cyclin A-Cdk2 complex. Nature Wang, W., Nacusi, L., Sheaff, R.J., and Liu, X. (2005a). Ubiquitination 382 , 325 – 331. of p21Cip1/WAF1 by SCFSkp2: substrate requirement and ubiquit- Schulman, B.A., Lindstrom, D.L., and Harlow, E. (1998). Substrate ination site selection. Biochemistry 44 , 14553 – 14564. recruitment to cyclin-dependent kinase 2 by a multipurpose docking Wang, Y., Filippov, I., Richter, C., Luo, R., and Kriwacki, R.W. site on cyclin A. Proc. Natl. Acad. Sci. USA 95 , 10453 – 10458. (2005b). Solution NMR studies of an intrinsically unstructured Scott, M.T., Morrice, N., and Ball, K.L. (2000). Reversible phospho- protein within a dilute, 75 kDa eukaryotic protein assembly; rylation at the C-terminal regulatory domain of p21Waf1/Cip1 modu- probing the practical limits for effi ciently assigning polypeptide lates proliferating cell nuclear antigen binding. J. Biol. Chem. backbone resonances. ChemBioChem 6 , 2242 – 2246. 275 , 11529 – 11537. Wang, Y., Fisher, J.C., Mathew, R., Ou, L., Otieno, S., Sublet, J., Serrano, M., Hannon, G.J., and Beach, D. (1993). A new regulatory Xiao, L., Chen, J., Roussel, M.F., and Kriwacki, R.W. (2011). motif in cell-cycle control causing specifi c inhibition of cyclin Intrinsic disorder mediates the diverse regulatory functions of the D/CDK4. Nature 366 , 704 – 707. Cdk inhibitor p21. Nat. Chem. Biol. 7 , 214 – 221. Sheaff, R.J., Groudine, M., Gordon, M., Roberts, J.M., and Clurman, Ward, J.J., Sodhi, J.S., McGuffi n, L.J., Buxton, B.F., and Jones, D.T. B.E. (1997). Cyclin E-CDK2 is a regulator of p27Kip1. Genes (2004). Prediction and functional analysis of native disorder Dev. 11 , 1464 – 1478. in proteins from the three kingdoms of life. J. Mol. Biol. 337 , Shin, I., Yakes, F.M., Rojo, F., Shin, N.-Y., Bakin, A.V., Baselga, J., 635 – 645. and Arteaga, C.L. (2002). PKB/Akt mediates cell-cycle progress- Watanabe, H., Pan, Z.Q., Schreiber-Agus, N., DePinho, R.A., sion by phosphorylation of p27KIP1 at threonine 157 and modula- Hurwitz, J., and Xiong, Y. (1998). Suppression of cell trans- tion of its cellular localization. Nat. Med. 8 , 1145 – 1152. formation by the cyclin-dependent kinase inhibitor p57KIP2 Shoemaker, B.A., Portman, J.J., and Wolynes, P.G. (2000). Speeding requires binding to proliferating cell nuclear antigen. Proc. Natl. molecular recognition by using the folding funnel: the fl y-casting Acad. Sci. USA 95 , 1392 – 1397. mechanism. Proc. Natl. Acad. Sci. USA 97 , 8868 – 8873. Weiss, R.H. (2003). p21Waf1/Cip1 as a therapeutic target in breast Sickmeier, M., Hamilton, J.A., LeGall, T., Vacic, V., Cortese, M.S., and other cancers. Cancer Cell 4 , 425 – 429. Tantos, A., Szabo, B., Tompa, P., Chen, J., Uversky, V.N., et al. Xiong, Y., Hannon, G.J., Zhang, H., Casso, D., Kobayashi, R., and (2007). DisProt: the database of disordered proteins. Nucleic Beach, D. (1993). p21 is a universal inhibitor of cyclin kinases. Acids Res. 35 , D786 – D793. Nature 366 , 701 – 704. Sivakolundu, S.G., Bashford, D., and Kriwacki, R.W. (2005). Yan, Y., Zhang, X., and Legerski, R.J. (2011). Artemis interacts with Disordered p27(Kip1) Exhibits Intrinsic structure resembling the Cul4A-DDB1DDB2 ubiquitin E3 ligase and regulates degra- the Cdk2/cyclin A-bound conformation. J. Mol. Biol. 353 , dation of the CDK inhibitor p27. Cell Cycle 10 , 4098 – 4109. 1118 – 1128. Yoon, M.K., Venkatachalam, V., Huang, A., Choi, B.S., Stultz, C.M., Suganuma, T. and Workman, J.L. (2011). Signals and combinato- and Chou, J.J. (2009). Residual structure within the disordered rial functions of histone modifi cations. Annu. Rev. Biochem. 80 , C-terminal segment of p21Waf1/Cip1/Sdi1 and its implications for 473 – 499. molecular recognition. Protein Sci. 18 , 337 – 347. Takaki, T., Echalier, A., Brown, N.R., Hunt, T., Endicott, J.A., and Yu, Z.K., Gervais, J.L., and Zhang, H. (1998). Human CUL-1 associ- Noble, M.E. (2009). The structure of CDK4/cyclin D3 has impli- ates with the SKP1/SKP2 complex and regulates p21CIP1/WAF1 and cations for models of CDK activation. Proc. Natl. Acad. Sci. cyclin D proteins. Proc. Natl. Acad. Sci. USA 95 , 11324 – 11329. USA 106 , 4171 – 4176. Zhang, H., Hannon, G.J., and Beach, D. (1994). p21-copntaining Touitou, R., Richardson, J., Bose, S., Nakanishi, M., Rivett, J., cyclin kinases exist in both active and inactive states. Genes Dev. and Allday, M.J. (2001). A degradation signal located in the 8 , 1750 – 1758. C-terminus of p21WAF1/CIP1 is a binding site for the C8 alpha-sub- Zhou, B.P., Liao, Y., Xia, W., Spohn, B., Lee, M.-H., and Hung, unit of the 20S proteasome. EMBO J. 20 , 2367 – 2375. M.-C. (2001). Cytoplasmic localization of p21Cip1/WAF1 by Toyoshima, H. and Hunter, T. (1994). p27, a novel inhibitor of G1 Akt-induced phosphorylation in Her – 2/neu-overexpressing cells. cyclin-Cdk protein kinase activity, is related to p21. Cell 78 , Nat. Cell Biol. 3 , 245 – 252. 67 – 74. Zhu, H., Nie, L., and Maki, C.G. (2005). Cdk2-dependent Inhibition Vavouri, T., Semple, J.I., Garcia-Verdugo, R., and Lehner, B. (2009). of p21 stability via a C-terminal cyclin-binding motif. J. Biol. Intrinsic protein disorder and interaction promiscuity are widely Chem. 280 , 29282 – 29288. associated with dosage sensitivity. Cell 138 , 198 – 208. Viglietto, G., Motti, M.L., Bruni, P., Melillo, R.M., D ’ Alessio, A., Califano, D., Vinci, F., Chiappetta, G., Tsichlis, P., Bellacosa, Received November 8, 2011; accepted December 21, 2011 274 D.M. Mitrea et al.

Dr. Diana M. Mitrea, PhD, Dr. Li Ou, PhD, received obtained her BS degree in his BS degree in Traditional biochemical engineering from Chinese Medicine from Babes-Bolyai University, the Hunan University Cluj-Napoca, Romania, and of Traditional Chinese moved to the United States Medicine. He earned his MS to further her education in degree in the laboratory of protein folding and protein Dr. Lingyi Kong at China engineering in Dr. Stewart Pharmaceutical University. Loh’s laboratory at SUNY He received his PhD in Dr. Upstate Medical University, Houming Wu’s laboratory at Syracuse, NY. After receiv- Shanghai Institute of Organic ing her doctoral degree at the Chemistry under the Chinese end of 2009, Diana continued her training with a short post- Academy of Sciences. doctoral fellowship in Dr. Loh’s laboratory. In the fall of 2010 Currently, Dr. Li works in Dr. Kriwacki’s Laboratory at St. she joined Dr. Richard Kriwacki’s group at St. Jude Children’s Jude Children’s Research Hospital, and focuses on structural Research Hospital, Memphis, TN. Her research interests focus studies of interactions between p27Kip1 and Cdk4/cyclin D1, on understanding the correlation between structure, or lack and the general role of phosphorylation in the regulation of thereof, and the biological function of intrinsically disordered intrinsically disordered proteins. and polymorphic proteins. Dr. Richard W. Kriwacki, Dr. Mi-Kyung Yoon obtained PhD, was trained in biophys- her PhD degree from the ics at Yale University in New Department of Chemistry at Haven, CT, and in struc- the Korea Advanced Institute tural biology at The Scripps of Science and Technology Research Institute in La Jolla, (KAIST), Daejeon, South CA. While a postdoctoral fel- Korea, in 2008. After spend- low with Dr. Peter E. Wright ing two years at KAIST as a at Scripps, he discovered post-doctoral researcher, she that the cell cycle regula- moved to United States for tor, p21Waf1/Cip1/Sdi1 (p21), was further training in 2010. Since unstructured under physi- then, she has been working as ological conditions. This a post-doctoral fellow at St. observation led to the demonstration that p21 functioned by Jude Children’s Research Hospital with Dr. Kriwacki. Her folding upon binding to its cyclin-dependent kinase (Cdk)/ research interest is to understand structure/function relation- cyclin regulatory targets, and that this mechanism conferred ships for the intrinsically disordered proteins. the ability to promiscuously regulate the full Cdk/cyclin repertoire that controls eukaryotic cell division. Subsequently, Dr. Kriwacki has led an independent research laboratory at St. Jude Children’s Research Hospital in Memphis, TN. His con- tinuing studies center on understanding the physical mechanisms that mediate the functions of disordered proteins. A primary objective is to understand “disorder/function relationships” and the evolutionary advantages associated with disordered proteins.