Regulation of the Cell Cycle

Basic Principles of Cell Cycle Progression

The essential function of cell cycle control is the regulated duplication of the cells’ genetic blueprint and the division of this genetic material such that one copy is provided to each daughter cell following division. The cell cycle can be divided conceptually into four individual phases. The “business” phases include S phase or synthesis phase, which is the period during which DNA is replicated, and mitosis ( M phase ), where DNA is packaged, the cells divide, and DNA is distributed to daughter cells. S phase and M phase are separated by Gap phases (G phase) to provide the cell with a proofreading period to ensure that DNA replication is completed and packaged appropriately prior to division. Separating M phase from S phase is the first Gap phase (G ₁ phase) and separating S phase from M phase is the second Gap phase (G ₂ phase). G ₀ or quiescence occurs when cells exit the cell cycle because of the absence of growth-promoting signals or the presence of prodifferentiation signals. Ordered progression through each phase is intricately regulated through both positive and negative regulatory signaling molecules and is the basis of normal organismal development. The consequences of deregulated growth control include failed or altered development and/or neoplastic/cancerous growth. Over the past two decades, a detailed picture of the major regulators of cell cycle control in both model organisms and higher eukaryotes has evolved. In this chapter, we describe the major regulators of cell division control and introduce current concepts regarding their participation in cell growth.

The Cyclin-Dependent Kinases

Cell cycle progression is positively regulated by a family of protein kinases referred to as the cyclin-dependent kinases (CDKs). In yeast, the organism in which early groundbreaking work defined many major cell cycle regulators, a single CDK regulates cell division: CDC2 ( Schizosaccharomyces pombe , fission yeast) and CDC28 ( Saccharomyces cerevisiae , budding yeast). In contrast, multicellular organisms use a distinct CDK whose activity promotes transition through each cell cycle phase ( Figure 11-1 ). CDKs are binary enzymes. The catalytic subunit, the CDK, coordinates adenosine triphosphate (ATP) and transfers phosphate to appropriate substrates. As a monomer, the CDK has no enzymatic activity; activation requires association with a specific allosteric activator called a cyclin. CDK subunits associate with specific cyclins ( Table 11-1 ) during distinct phases of the cell cycle and, as active protein kinases, trigger transition through cell cycle phases. Although some CDKs can form complexes with multiple cyclins, in most cases active complexes rely on specific partnerships.

Homology among CDKs, at the level of primary amino acid sequence, is in the range of 30% to 40%. The most highly conserved sequence, which contributes directly to cyclin binding, is the PSTAIRE sequence (CDK1, CDK2) or PV/ISTVRE (CDK4, CDK6) where letters refer to individual amino acids comprising this sequence (e.g., P = proline).

Cyclins associate with the CDK subunit through a conserved domain within the cyclin called the cyclin box . The crystal structure of cyclins has revealed that the cyclin box comprises two sets of five α helices that share little primary homology, but share significant homology with respect to structural and folding topology. Sequences N- and C-terminal to the cyclin box share little if any homology and contribute to substrate-specific interactions and to posttranslational regulation of cyclin protein accumulation (e.g., protein degradation).

Posttranslational Regulation of CDKs

Regulation of CDKs by Phosphorylation

Cyclin binding to the CDK contributes to kinase activation by inducing a conformational change wherein a C-terminal domain of the CDK, referred to as the T loop, is directed out of the substrate binding cleft. In the absence of cyclin binding, the T loop occludes substrate interactions. The cyclin-induced alteration, however, is not sufficient for complete CDK activation. T-loop displacement is ensured by direct phosphorylation of a conserved threonine residue within the T loop (Thr161, CDK1; Thr160, CDK2; Thr172, CDK4) by the CDK-activating kinase, CAK ( Figure 11-2 ). In mammalian cells, CAK itself is a cyclin-dependent kinase composed of CDK7 and cyclin H. CAK is constitutively active and contributes to CDK activation following cyclin binding via phosphorylation of the T-loop threonine.

Shortly after the identification of CDK7/cyclin H as CAK, CDK7/cyclin H was shown to be the previously identified activity referred to as TFIIH, demonstrating that CAK (CDK7/cyclin H) not only contributes to CDK activation but is also implicated in transcriptional regulation. TFIIH phosphorylates multiple serine/threonine residues located in the carboxyl-terminal domain (CTD) of the largest subunit of RNA polymerase II (RNAPII), thereby contributing to increased transcriptional initiation. CDK7 is also conserved in budding yeast. However, in yeast, CDK7 does not contribute to CDK activation. Rather, it is solely a regulator of RNA polymerase activity. Bona fide CAK activity in yeast is contributed by a distinct protein, CAK1.

CDK phosphorylation is not solely an activating event. Phosphorylation of N-terminal threonine and tyrosine residues near the ATP binding pocket is inhibitory. Phosphorylation of threonine 14/tyrosine 15 is catalyzed by two enzymes, Wee1 and Myt1 (see Figure 11-2 ). Although Wee1 is a cytosolic enzyme, Myt1 is localized to endoplasmic reticulum structures. The significance of the differential localization of Wee1 versus Myt1 remains to be established. Threonine 14/tyrosine 15 is located adjacent to the ATP binding pocket of CDKs, providing a structural basis for how phosphorylation of these residues prevents ATP binding. Both threonine and tyrosine residues are conserved in CDK1-3, but only the tyrosine residue is present in CDK4-6. Although phosphorylation of CDK1-2 contributes to the timing of their activation during a normal cell cycle, the CDK4/6 enzymes appear to be subject to this inhibitory phosphorylation only when cells incur DNA damage.

In mammalian cells, the removal of N-terminal inhibitory phosphates is catalyzed by any one of three highly related dual-specificity protein phosphatases: CDC25A, CDC25B, or CDC25C. In contrast, yeast cells harbor a single CDC25 isoform that carries out all relevant functions. CDC25 isoforms are expressed in a cell cycle–dependent manner, and the A-B-C designation corresponds to their order of expression during the cell cycle. CDC25 A is expressed first, with its expression peaking at the G ₁ /S boundary. CDC25B expression follows that of CDC25A, with the highest levels detected during S phase. Finally, CDC25C is expressed during late G ₂ and mitosis. From this expression pattern, substrate specificity was inferred, with CDC25A targeting the G ₁ CDK _S (CDK4/6 or CDK2-cyclin E), CDC25B regulating the S-phase CDKs (CDK2-cyclin A), and CDC25C regulating mitotic CDKs (CDK1-cyclin B). Consistent with this hypothesis, inhibition of CDC25A resulted in increased CDK2-cyclin E tyrosine phosphorylation. Also consistent with substrate specificity being determined by the timing of expression, CDC25 enzymes do not exhibit any specificity toward distinct CDK substrates in vitro. However, timing of expression is not the sole determinant. Deletion of CDC25B or CDC25C, or even the combined deletion does not impair mouse development or cell proliferation in vitro. It appears from this analysis that CDC25A expression is sufficient to drive cell cycle expression.

CDK Regulation by Small-Polypeptide Inhibitors

In addition to CDK regulation via phosphorylation, CDKs are subject to direct regulation by small-polypeptide inhibitory proteins referred to as CDK inhibitors, or CKIs ( Figure 11-3 ). These regulators bind directly to and inactivate CDK-cyclin complexes. There are two families of CKIs that have distinct biochemical activities. The Ink4 (inhibitors of CDK4) family proteins bind exclusively to G ₁ CDKs CDK4 and CDK6. Binding can directly inhibit an active CDK4/6-cyclin complex, or Ink4 protein can bind to monomeric CDK4/6 and prevent cyclin association. The second family, Cip/Kip family proteins, bind to a broad range of CDK-cyclin complexes but functionally appear to be negative regulators of CDK2 complexes.

Ink4 Family

The Ink4 family of proteins consists of four members: p16 ^Ink4a , p15 ^Ink4b , p18 ^Ink4c , and p19 ^Ink4d . All four proteins bind exclusively to and inhibit D-type cyclin-dependent kinases CDK4 and CDK6. The founding member of the Ink4a family was discovered as a protein that interacted with CDK4 in co-immunoprecipitation experiments, subsequently identified as MTS1.

Ink4 proteins are homologous in primary structure, sharing the presence of four or five ankyrin repeats, which are responsible for protein-protein interactions with CDK4/6. Each repeat consists of an extended strand connected by a helix-loop-helix (HLH) motif to the next extended strand. The crystal structure of the p19 ^Ink4d -CDK6 complex has been solved and provided valuable details about the mechanism of CDK inhibition by Ink proteins ( Figure 11-4 ). α-Helices and β-turns of p19 ^Ink4d ankyrin repeats form a “cap” over the N-terminal domain of CDK6 and induce its spatial movement away from the C terminus. This event inhibits productive ATP binding but does not interfere with the formation of CDK-cyclin complex. As expected from their structure, all four Ink proteins exhibit similar biochemical activities toward CDK4 and CDK6. Interestingly, a short peptide that was derived from one of the ankyrin motifs had the ability to bind and inhibit CDK4, implying the importance of these domains in Ink4 functionality.

Despite similar biochemical activities and comparable tertiary structures of Ink proteins, their regulation is distinct. p16 ^Ink4a is not expressed in most tissues. Rather, it is induced in response to expression of oncogenic or transforming proteins and during cellular senescence. Several oncogenes as well as tumor suppressors regulate p16 ^Ink4a expression. For example, overexpression of Ras increases p16 ^Ink4a levels in primary rodent cells. Inactivation of the retinoblastoma susceptibility protein, Rb, or tumor suppressor p53 can also promote p16 ^Ink4a expression. In contrast, p15 ^Ink4b expression is regulated by growth-inhibitory factors (anti-mitogens) such as TGF-β. Only p18 ^Ink4c and p19 ^Ink4d expression seems to be regulated during various phases of the cell cycle, with expression peaking during S phase. The expression patterns of Ink4 proteins are also differentially regulated during development.

The structure of the genomic Ink4a locus is unique. Transcription through this locus gives rise to two biochemically distinct proteins, p16 ^Ink4a and p19 ^ARF , as a result of alternative exon utilization. Although p16 ^Ink4a regulates CDK4/CDK6 activity, thereby indirectly regulating the Rb tumor suppressor, p19 ^ARF regulates the p53 tumor suppressor. p19 ^ARF acts by attenuating Mdm2-mediated degradation of p53 and is known as an activator of the p53 pathway. Therefore, loss of p19 ^ARF leads to impairment of p53 signaling. Elimination of the Ink4a/ARF genetic locus in mice makes the animals highly prone to tumor development.

Cip/Kip Family

The Cip/Kip family of CKIs includes three members: p21 ^Cip1 , p27 ^Kip1 , and p57 ^Kip2 . Unlike the Ink4 family of CKIs, Cip/Kip inhibitors bind to and efficiently inhibit various CDKs. Members of the Cip/Kip family are highly homologous and share approximately 50% of their sequences. The amino terminus of both p21 ^Cip1 and p27 ^Kip1 contains an RXL (where X is typically basic) sequence that is responsible for binding to cyclins and is called the cyclin-binding motif . Cip/Kip inhibitors also contain a domain that is responsible for the binding to CDKs (N-terminal in p21 ^Cip1 and p27 ^Kip1 ). The crystal structure of the p27 ^Kip1 /cyclin A/cdk2 complex ( Figure 11-5 ) revealed that p27 ^Kip1 binds CDK2 at its N terminus and inserts into the catalytic cleft, thus mimicking ATP. On cyclin A/CDK2, p27 ^Kip1 binds to the groove of the cyclin box. Because both Ink and Cip/Kip proteins occupy almost the same binding sites on CDKs, binding is mutually exclusive. For example, in vitro, p15 ^Ink4b inhibits binding of p27 ^Kip1 . However, in cells, which protein gets to the CDK first is often determined by the coordinated cellular localization of the inhibitors and cyclin-CDK complexes.

p27 ^Kip1 is responsible for induction and maintenance of the quiescent state. p27 ^Kip1 expression is induced in response to growth factor withdrawal and on contact inhibition in cell cultures. p27 ^Kip1 levels are decreased on addition of the mitogens by various mechanisms described in subsequent paragraphs. Overexpression of p27 ^Kip1 in cells leads to cell cycle arrest in G ₁ phase. Unlike p27 ^Kip1 , p21 ^Cip1 is present at high levels in cycling cells, keeping CDK activities under tight control. p21 ^Cip1 levels are induced in response to DNA damage and genotoxic stress as a result of activation of p53. Similar to p21 ^Cip1 and p27 ^Kip1 , induction of p57 ^Kip2 can mediate cell cycle arrest in G ₁ phase. In addition, p57 ^Kip2 also participates in the M-to-G ₁ transition through activation by p73. Abrogation of p73 or its downstream effector p57 ^KIP2 perturbs mitotic progression and transition to G ₁ phase.

Transcriptional Regulation by the E2F Transcription Factors

E2F was originally identified as a cellular DNA binding activity that regulated expression of the viral E2 promoter. Since this seminal work, molecular analysis has revealed that the E2F activity is encoded by a family of DNA binding proteins, which includes transcriptional activators and repressors. Mammalian cells encode eight known E2F proteins (E2F1-8; Figure 11-6 ). Further complication ensues from the fact that E2F associates with DNA as a heterodimer; the two known heterodimeric partners for E2F are DP1 and DP2. Indeed, the founding member, E2F1, can drive S-phase entry in the absence of growth factor stimulation. The ability of E2F1 to drive S phase derives from its role in the regulation of genes whose protein products play essential roles in S-phase progression. Established E2F targets include components of DNA replication complexes (MCMs) and S-phase cyclins (E and A). E2F family members were initially considered requisite regulators of S-phase entry. E2F1, E2F2, and E2F3 accumulate during G ₁ phase and play critical roles in promoting expression of S-phase targets. Strikingly, E2F4 through E2F7 function as transcriptional repressors; E2F3b, an alternatively spliced isoform of E2F3, is also a transcriptional repressor. The E2F repressors function to maintain cells in a quiescent or resting state. In addition to DP1, E2F complexes are further modulated by members of the retinoblastoma protein (pRb) family (pRb, p107, p130; Figure 11-7 ). The Rb family member functions to recruit chromatin-remodeling enzymes, such as histone deacetylases, to E2F complexes. As a consequence, increased activity of E2F1 through E2F3 requires dissociation of “pRb” from the E2F/DP1 heterodimer. As illustrated in the following sections, the G ₁ CDK/cyclin kinase triggers this through direct phosphorylation of the associated pRb family member.

In addition to the regulation of S-phase entry and progression, E2F transcriptional activators can trigger apoptosis or cell suicide. The mechanisms whereby E2F induces cell death remain unclear. However, pro-apoptotic genes have been identified as E2F target genes. Examples include the p19 ^ARF protein, which is a known regulator of the p53 tumor suppressor. In addition, E2F can increase expression of pro-apoptotic proteins Puma, Noxa, and Bim and repress the anti-apoptotic Bcl2 family protein, Mcl1.

Two new E2F family members recently identified, E2F7 and E2F8, provide an important constraint against excessive E2F1 activation. Unlike the other mammalian E2Fs, E2F7 and E2F8 have two DNA-binding domains and do not require a DP partner to bind to DNA and as such are classified as atypical E2F family members. These atypical E2Fs bind to the consensus E2F recognition sequence and can repress expression from a subgroup of cell cycle–regulated E2F targets. An E2F7 and E2F8 double knockout is an embryonic lethal resulting from massive apoptosis; this phenotype can be bypassed by removing E2F1 or p53. Our current level of understanding underscores E2F7 and E2F8 as a distinct arm of the E2F network involved in repression of transcription during S-G ₂ and control of the E2F1-p53 apoptotic axis.

G ₁ Regulation/Restriction Point Control

During the first Gap phase or G ₁ , cells prepare for DNA replication. They must synthesize proteins necessary to replicate their genome, and once these are made, assemble the various components of the DNA replication machinery on chromatin at so-called origins of replication. This is coordinated with nutrient and growth factor availability to ensure that the cell is in an environment that supports cell division. The G ₁ phase of the cell cycle is unique in that it represents the only time wherein cells are sensitive to signals from their extracellular environment. These signals are in the form of adhesion to substratum and growth factors. Cells require growth factor–dependent signals up to a point in late G ₁ referred to as the restriction point (“start” in yeast).

Progression through G ₁ phase is driven by the collective activities of two distinct CDKs. The first is CDK4 or CDK6 in combination with a D-type cyclin. Mammalian cells encode three distinct D cyclins (D1, D2, D3), which are expressed in a tissue-specific manner. Whereas CDK4 and CDK6 are constitutively expressed, D cyclins are expressed in response to growth factor signaling. Following accumulation of active cyclin D/CDK4 or CDK6, the CDK2 kinase in combination with cyclin E accumulates to facilitate the transition from G ₁ to S phase.

A key protein that regulates G ₁ -phase progression in the mammalian cell cycle is retinoblastoma protein, Rb. The Rb family consists of three members, Rb, p107, and p130. In quiescent cells, Rb proteins associate with E2F transcription factors to repress E2F-dependent transcription. E2F targets include genes responsible for regulation of cell cycle and DNA replication, such as cyclins E and A (see Figure 11-7 ). Rb activity is regulated at the level of posttranslational modification, specifically phosphorylation. Hypophosphorylated Rb is active and binds to E2F, thereby silencing E2F-dependent activity. Hypophosphorylated Rb family proteins therefore play a central role in maintaining cells in a resting or quiescent state. Quiescent cells reenter the cell cycle in response to mitogenic growth factors. Growth factor signaling induces the expression of D-type cyclins at transcriptional and posttranslational levels, leading to activation of cyclin D–dependent kinases CDK4 and -6 and subsequent Rb phosphorylation. Cyclin D/CDK4 or -6 complexes also have a kinase-independent function. They sequester p21 ^Cip1 and p27 ^Kip1 CDK1s from CDK2 kinases and allow activation of basal CDK2/cyclin E kinases, which further phosphorylate Rb family proteins. Phosphorylation of Rb promotes its dissociation from E2F, allowing transcriptional activation of E2F targets such as cyclin E. The E2F-dependent spike in cyclin E, and thus CDK2/cyclin E activity, represents the transition from mitogen-dependent to mitogen-independent cell cycle progression (or passage through the restriction point). In addition to maintaining Rb proteins in a hyperphosphorylated (inactive) state, the activation of cyclin E/CDK2 promotes proteasome-dependent degradation of its own inhibitor p27 ^Kip1 (described in a subsequent section). These changes, which include cyclin D/CDK4/6 and cyclin E/CDK2 activation, Rb phosphorylation, and destruction of p27 ^Kip1 , render cells with decreased mitogen dependency and are irreversibly committed to enter S phase of the cell cycle.