MARK2
Microtubule associated regulating kinase 2
Unless otherwise stated, the structures are wild-type versions of the MARK2 protein obtained from Rattus norvegicus, as investigated by A. Marx et al. (2006) [1].
All images were rendered using Pymol.
Investigate the structure of MARK2 using the links below
"AN OUNCE OF SAUCE
COVERS A MULTITUDE OF SINS."
Anthony Bourdain
MARK2 as a dimer
The original work conducted by A. Marx et al. (2006) involved the crystallisation of the protein from the species Rattus norvegicus. The crystal they obtained was a homodimer, containing two molecules (A+B) of MARK2 per asymmetric unit. Each monomeric unit interacts very tightly by the kinase domains.
The interaction between a region on the large lobe of one monomeric unit and the small lobe of the other monomeric unit provides a cleft that helix G fits into.
A lot of the interaction is governed by the interaction of a highly disordered region called the activation loop. However, due to the degree of disorder, this region is unidentified in the structure. The group investigated the dimerisation of a double mutant (T208A/S212A), which contained less disordered. This allowed identification of a disulphide bond between C210 of each monomeric unit. This is shown in figure two.
Figure 1. MARK2 in its crystallised, dimer form.
Figure 2. The disulphide bond within the dimer interface
Figure 3. The dimer interface. For better resolution, click the HD button.
The rendered structures on the left shows one monomeric unit of MARK2. The MARK2 molecule obtained from Rattus norvegicus consists of two domains: the kinase domain and the UBA domain. A large linker connects these two domains. In addition to these domains is a KA1 domain, responsible for binding to lipids, however this structure was not crystallised.
The structure on the right shows the secondary structure of the protein (red: alpha helical; yellow: beta strands; green: loops/turns). The molecule is predominantly alpha helical, although five beta strands exist within the kinase domain.
MARK2 as a monomer
Figure 4. Two structures showing MARK2 as a monomeric unit. Left=the different domains; right=secondard structure.
A Ramachandran plot of the favourable torsion angles. Determined using RAMPAGE [2]. Note that there are multiple points in the bottom left quadrant, suggesting a large helical propensity.
A Ramachandran plot of the favourable torsion angles. Determined using RAMPAGE [2]. Note that there are multiple points in the bottom left quadrant, suggesting a large helical propensity.
The kinase domain
Homologous to other kinase domains is the kinase domain of MARK2. It exists as a bilobed structure, containing an N-terminal small lobe, and a C-terminal large lobe (that connects to the UBA domain via a linker of ~21 residues).
The small lobe has two key components that have been highlighted in the image. There is the P-loop highlighted in cyan blue, responsible for recognising the phosphorylated site. More importantly, a mutation of K82R showed that this residue is essential for catalysis. This is the invariant lysine residue that is responsible for bonding to the ATP substrate used in phosphorylation, as its orientation into the cleft would suggest.
The large lobe similarly has two key regions. The activation loop in the wild-type structure was extremely disordered, however the double mutant T208A/S212A revealed 5 more residues than the wild type. These residues are part of the P+1 loop. Within the activation loop is also the primary phosphorylation site, T208, however this was disordered in the structural rendering. pT208 is responsible for engaging in ion pair interaction with R147 in the catalytic loop, coloured in yellow in the structure. This interaction orientates D175, adjacent to R147, ideally for attacking the OH group of the substrate. The disordered state of the loop structures often causes misunderstanding of the structure in three-dimensional; in vivo these loops have flexibility that a crystal model does not predict. However, molecular dynamic simulations can help increase the understanding of the disorder.
Connecting the two lobes is a linker region composed of six residues. Furthermore, there are hydrogen bonds made between the small and large lobe; these are restricted to a narrow region allowing the small lobe to be freed from the large lobe, revealing the catalytic cleft.
Figure 6. A detailed look at the kinase domain of MARK2 T208A/S212A.
The UBA domain
The UBA domain of MARK2 consists of three small helices, with two helices being roughly antiparallel to each other. This is inconsistent with all known structures of the UBA domains whereby the two helices are parallel. Figure 8 shows the alignment of MARK2 UBA with RUH-074 UBA, a typical UBA domain [3].
The UBA domain binds to the small lobe of the kinase domain at the back end of the cleft. The majority of interactions are hydrophobic interactions and some hydrogen bonding. It is believed that the unexpected, antiparallel helices arise due to the interaction with the N-lobe. Alternatively, antiparallel helices are just a specific feature of the MARK2 kinase.
The role of UBA in MARK2 is not completely understood. The presence of UBA would suggest a ubiquitin binding function. In the structure present, there is no open space that would allow ubiquitin to bind; a MARK2-ubiquitin complex has not been identified. Another idea is that this domain has an autoregulatory role whereby is binds into the catalytic cleft of the kinase domain, thus inactivating the protein.
Connecting the UBA to the kinase domain is a linker composed of ~20 residues. The first half of this linker has a motif similar to the common docking motif in MAP kinases that is responsible for binding upstream/downstream regulators. The second half has an extended conformation distant from the kinase domain, which could suggest that the UBA domain is capable of moving away from the kinase domain, providing different regulatory states.
Figure 7. A detailed look at the UBA domain and the linker that connects it to the kinase domain.
Figure 8. An alignment of MARK2 UBA domain (blue) with the UBA domain of RUH-074 (teal). Arrows are drawn to show the directions of the helices.
References:
[1] MARX, A., NUGOOR, C., MULLER, J., PANNEERSELVAM, S., TIMM, T., BILANG, M., MYLONAS, E., SVERGUN, D. I., MANDELKOW, E. M. & MANDELKOW, E. 2006. Structural variations in the catalytic and ubiquitin-associated domains of microtubule-associated protein/microtubule affinity regulating kinase (MARK) 1 and MARK2. J Biol Chem, 281, 27586-99.
[2] http://mordred.bioc.cam.ac.uk/~rapper/rampage.php
[3] KITASAKA, S., RUHUL MOMEN, A.Z.M., HIROTA, H., MUTO, Y., YOKOYAMA, S. 2007. Solution structure of RUH-074, a human UBA domain. RCSB PDB.