CDK2 in Complex with Synthetic Inhibitors

CDK2 in Complex with Synthetic Inhibitors

Jennifer L. Howard '09 and Emily K. Staudenmaier '10

I. Introduction

Protein kinases belong to the enzyme family of transferases that catalyze the transfer of a phosphoryl group from a donor to an acceptor. Cyclin-dependent kinases (CDKs) are serine/threonine kinases that require association with a cyclin regulatory protein for activation [4] . They are responsible for regulating progression through the cell division cycle, helping to ensure that the eukaryotic genome is replicated only once per cell cycle. CDKs are required for the correct timing and order of events of cell division [1,2,4,7].

There are four phases in the cell division cycle: G₁, S, G₂, and M. Click here for a schematic drawing of the cell cycle. CDK2, a crucial component of the CDK complex, is responsible for the G₁/S phase transition [4]. Cells prepare for divison during the G₁ phase and the actual chromosome replication occurs during the synthesis or S phase. Each chromosome is replicated, resulting in two sister chromatids. CDK levels are low during the G₁ phase but elevated during S, G₂, and M phases, limiting opportunities for cyclin binding and replication [7].

The retinoblastoma protein is inactivatedthrough phosphorylation by CDK2, committing the cell to division during the G₁ phase. CDK2 is regulated by the A and E forms of cyclin. Cyclin E binds to CDK2 for progression from the G₁ phase to the S phase and Cyclin A is critical for progression through the S phase of the cell division cycle [2,3,4].

Mutations in CDKs and/or their inhibitors are associated with overexpression and amplification, leading to uncontrolled cell growth: tumors. As CDKs play such a huge role in the cell division cycle, they have become important targets in drug-discovery programs for cancer, diabetes, and immune diseases. Certain macrocyclic compounds have been found to be inhibitors to the Cyclin A and E binding sites to CDK2 and therefore show great potential as anti-tumor agents [4].

II. CDK2 General Structure

CDK2 is a monomer comprised of a polypeptide chain of 298 amino acid residues. Structurally, CDK2 is made up of mainly α-helix elements as well as a β-sheet terminus [5].

CDK2 is composed of two lobes: a smaller amino-terminal lobe that contains primarily β-sheets and the PSTAIRE helix , and a large carboxy-terminal lobe made up of α-helices.

Two α-helices, the PSTAIRE helix and the L12 helix , are important for cyclin binding and the conformation changes which occur in CDK2 after cyclin is bound. The T-loop , also called the activation loop, is located near the kinase active site. It provides selectivity for substrate binding by blocking the binding of protein substrates at the entrance of the active-site cleft [5]. Thiazolidinone showed potential in vitro as a CDK2 inhibitor. This inhibitor binds to the hinge region of CDK2, which contains the T-loop, PSTAIRE helix, and L12 helix [6].

III. Cyclin A and E General Structure

A basic knowledge of the structure of the cyclin forms that bind to CDK2 is useful to fully understand how synthetic inhibitors act to deactivate CDK2. Both Cyclin A and Cyclin E are structurally similar but have slight sequential differences, particularly in the C-terminal regions of the two forms, which affect the strength of bonding at the cyclin/CDK2 interface [2,3].

Cyclin E is 395 amino acid residues long and consists of two five-helical domains with additional helices at the N and C terminals. For a schematic drawing clarifying the names of the helices of Cyclin E, please click here . The α3 helix forms a hydrophobic core which acts as a point of organization for the α1, α2, α4, and α5 helices arranged around this core. Ala154 and Ala184 interact to allow close packing of the α2 and α3 helices [2] .

Cyclin A is 431 amino acid residues long and consists of two helical domains with identical chain topology. Cyclin A is a globular structure of 12 α-helices. For a schematic drawing clarifying the names of the helices of Cyclin A, please click here .

Each helical domain consists of a right-handed 3-helix bundle with two extra helices packed against the bundle’s side. The two domains are connected by a linker of five amino acids. One domain contains the cyclin box , forming the binding site for the PSTAIRE helix and making contacts with the T-loop and N-terminal β-sheet of CDK2.

Similar to Cyclin E, the hydrophobic core of the 5-helix motif is formed by the α3 helix. The α1 and α2 helices pack against the α3 helix and help form the hydrophobic core while participating in inter-repeat packing with α1’ and α2’ helices . α4 and α5 helices pack against the α3 helix and interact with CDK2. Alanine residues allow close packing interactions between helices, the most significant being Ala 235 and Ala 264 from the α2 and α3 helices.

IV. Cyclin Binding

Understanding how cyclin binding occurs is essential for successful design of a CDK2 synthetic inhibitor.

The cyclin dependent kinase family, which includes CDK2, is inactive until complexed with a regulatory protein called cyclin [5]. Cyclins share a homologous region of about 100 amino acids termed the cyclin box but differ in other regions giving them a unique structure [3]. Therefore, different kinases bind different forms of cyclin. CDK2 interacts with the A and E forms only. Cyclin E associates with CDK2 to drive cells from the G₁ to the S phase through phosphorylation of certain targets. Cyclin E is not required for cell cycling since Cyclin A is structurally similar and can perform the same cell cycle functions. However, Cyclin E is is required for re-entry of a cell into the cell cycle from the G₀ phase, which is why the presence of both cyclin forms is necessary in the cell. On entry into S phase, Cyclin E is abruptly destroyed by the proteasome that it targets. Cyclin A, expressed in response to the CDK2/Cyclin E activities, then associates with CDK2 to drive cells through S phase.

Cyclin E is found in high concentrations in most tumors and almost all aggressive cancers. This fact further stresses the importance of finding an inhibitor to act as an anti-tumor agent, which will selectively bind CDK2 and prevent binding with Cyclin E [2].

Binding of cyclin induces a conformational change that alters the structure around the active site. The complex shown to the left is a view of CDK2 unbound to cyclin.

The PSTAIRE and L12 helices are involved in the conformational changes in CDK2 after it has bound a form of cyclin. The L12 helix prevents motion of the PSTAIRE helix in free CDK2. The binding of Cyclin A to CDK2 melts the L12 helix into a β-sheet, allowing the PSTAIRE helix to rotate nearly 90 degrees about its helical axis, thereby reconfiguring the active site.

The PSTAIRE helix interacts directly with cyclin by forming a hydrogen-bond between the Glu51 residue and the Lys33 residue on Cyclin E. The PSTAIRE helix is virtually parallel to the α5 Cyclin E helix and almost perpendicular to the C-terminal end of the α3 helix.

The T-loop is also extremely important for correct cyclin binding since accurate orientation of this structural element is required for selective recognition of a sequence motif present on the Cyclin A and E forms. It associates with the C-terminal lobe of cyclin. Phospho-Thr160 acts as an organizing center and makes contacts with three Arginine residues (Arg 50, 126 and 150) also on the CDK2 molecule. Arg50 and Arg150 form hydrogen-bonds with carbonyl groups on Leu187 and Glu188 on Cyclin E.

Cyclin A binds to one side of the catalytic cleft, interacting with both lobes of CDK2 to form a large, continuous protein-protein interface. This Cyclin A-CDK2 interface is formed from the interlocking of the PSTAIRE helix , the T-loop , portions of the N-terminal sheet and C-terminal lobe from CDK2, and helices α3 , α4 , and α5 from the first repeat as well as the N-terminal helix from Cyclin A.

The L12 helix of CDK2 is a crucial component of this interface. In the Cyclin A-CDK2 complex, this helix rotates and moves inward toward the catalytic cleft. The cyclin box surrounds the middle part of the PSTAIRE helix so that the α5 helix lies parallel to the PSTAIRE helix on one side and the C terminus of the α3 helix contacts the other side perpendicularly. The PSTAIRE helix is bound by an extended region of hydrophobic interactions, whereas the N- and C-terminal regions are bound by networks of hydrogen-bonds. Ile 49 , in particular, fits tightly into a hydrophobic pocket lined with side chains of Leu 263, Phe 267 , Leu 299 , Leu 306 , and an aliphatic region of Lys 266 from Cyclin A.

Van der Waals interactions, such as those between Ile 52 and Phe 304 of Cyclin A, are also important to the Cyclin A-CDK2 complex. The N-terminal portion of the T-loop is bound by helices α1 , α2 , and α3 from the first repeat of Cyclin A, as well as its N-terminal helix . Van der Waals interactions occur between Ala 151, Phe 152 , and Tyr 159 of CDK2 and Phe 267 , Ile 182 , and Ile 270 of Cyclin A. Hydrogen-bonds link Arg 150 to the backbone carbonyl groups Glu 269 and Ile 270 of Cyclin A.

Cyclin binding occurs mainly at the PSTAIRE helix and the T-loop on CDK2. The CDK2/cyclin interface area is 3252 A°2 in Cyclin E and 2839 A°2 in Cyclin A. Cyclin E has a larger area available for interactions between the two complexes and therefore the potential for stronger binding. This increased binding affinity of CDK2 for Cyclin E has been confirmed experimentally [2,3].

V. Synthetic Inhibitor Binding

Inhibitors of CDK2 show exciting potential as tumor suppressors. The majority of kinase inhibitors interact with the kinase backbone motif which is part of the cyclin binding site. Therefore, the inhibitors act by sterically inhibiting the binding of the activation molecule that the kinases need in order to have a physiologically significant activity [4].

Thiazolidinone inhibitors are one example of a compound that acts on CDK2 by the method described above. It forms hydrogen-bonds with the Glu81 and Leu83 residues which anchor the conformationally free inhibitor molecule to the kinase.

The sulfonate group on the ligand also interacts with the NH group on the backbone of Asp86 and the carbonyl group on the backbone of Leu10.

The thiazolidinone inhibitor was also designed to selectively inhibit the CDK2 form over other members of the kinase family. This target specificity makes thiazolidinone inhibitors an even more promising discovery in terms of medical applications [6].

Macrocyclic aminopyrimidines are another class of molecules which act to inhibit CDK2. These compounds showed potential both in-vitro and in-vivo as inhibitors and they interact with the same general region of CDK2 that the thiazolidinone inhibitors do. However, structural differences between the two inhibitors result in macrocyclic aminopyrimidines interacting with slightly different amino acid residues. The aminopyrimidine recognition site was shown to bind to the hinge region of CDK2 through two hydrogen-bonds . An oxygen atom from the meta-sulfonamide forms two hydrogen-bonds to the backbone nitrogen and to a side chain oxygen of Asp86 . The amide nitrogen atom from this group interacts with the backbone carbonyl group of Ile10 [4].

VI. References

**[1] Card, G.L, P. Knowles, H. Laman, N. Jones, and N.Q McDonald. 2000. Crystal structure of γ-herpesvirus cyclin-CDK complex. The EMBO Journal 19 (12): 2877-2888.**

[2] Honda, R., E.D. Lowe, E. Dubinina, V. Skamnaki, A. Cook, N.R. Brown, and L.N. Johnson. 2005. The structure of cyclin E1/CDK2: implications for CDK2 activation and CDK2-independent roles. The EMBO Journal 24: 452-263.

[3] Jeffrey, P.D., A.A. Russo, K. Polyak, E. Gibbs, J. Hurwitz, J. Massague, and N.P. Pavletich. 1995. Mechanism of CDK activation revealed by the structure of a cyclinA-CDK2 complex. Nature 376: 313-320..

[4] Lucking, U., G. Siemeister, M. Schafer, H. Briem, Martin Kruger, P. Lienau, and R. Jautelat. 2007. Macrocyclic aminopyrimidines as multitarget CDK and VEGF-R inhibitors with potent antiproliferative activities. ChemMedCHem 2: 63-77.

**[5] Morgan, D. O. “Cyclin-Dependent Kinases”. The Cell Cycle: Principles of Control. Online: New Science Press Ltd. http://www.new-science-press.com /content/pdf/nsp-cellcycle-3-1.pdf**

[6] Richardson, C. M., C. L. Nunns, D. S. Williamson, M. J. Parratt, P. Dokurno, R. Howes, J. Borgognoni, M. J. Drysdale, H. Finch, R. E. Hubbard, P. S. Jackson, P. Kierstan, G. Lentzen, J. D. Moore, J. B. Murray, H. Simmonite, A. E. Surgenor, and C. J. Torrance. 2007. Discovery of a potent CDK2 inhibitor with a novel binding mode, using virtual screening and initial, structure-guided lead scoping. Bioorganic & Medicinal Chemistry Letters 17: 3880-3885.

**[7] Watson, J.D., T. Baker, S. P. Bell, A. Gann, M. Levine, and R. Losick. Molecular Biology of the Gene, 6th Ed. New York: Cold Spring Harbor Laboratory Press, 2008.

Back to Top**