This presentation is accessible only to the Illinois State University community.

  • Off-Campus ISU Users: To download this item, click the "Off-Campus Download" button below. You will be prompted to log in with your ISU ULID and password.



Download Presentation (452 KB)

Publication Date


Document Type

Restricted Presentation

Presentation Type


Degree Type



Biological Sciences


Martin Engelke

Mentor Department

Biological Sciences


Kinesin motor proteins are vitally important for many cellular functions such as mitosis, ciliogenesis, and transporting cellular cargo. Kinesins move various cargo along microtubules unidirectionally from the minus- to the plus-end of a microtubule. There are three structural features that make up a kinesin: the stalk, the motor, and tail domains. The motor domains directly bind to and walk along microtubules while the tail domains interact with the cargo. The protein’s stalk portion comprises coiled-coil segments that allow the motor to dimerize via alpha-helices interacting through hydrophobic forces. These rigid segments are interrupted by flexible hinges that allow the kinesin molecule to fold back on itself in the absence of cargo. In this state, parts of the stalk and the tail domains bind to and inhibit the motor domains in a process called autoinhibition. This prevents the motor from binding and moving along microtubules when no cargo is present to conserve energy. In our lab, we study kinesin-2 family motors, specifically, KIF3A/KIF3B. Some motors of the kinesin-2 family form heterodimers, which is a unique feature compared to all other motor proteins found in mammals which form homodimers. In this work, we aim to elucidate the function of individual coiled-coil domains in the stalk during autoinhibition. The precise location of the coil domains within the stalk is, however, not known. Available coiled-coil prediction programs such as DeepCoil and COILS Server are programmed to predict coiled-coil formation in homomultimers, complicating the prediction of coil domains in the heterodimeric KIF3A/KIF3B motor. To overcome this hurdle, we took a rational prediction approach. We used DeepCoil to obtain the probability of individual amino acids being involved in the coiled-coil formation, and mapped these probabilities onto a 3D atomic model of the motor stalk generated by the MODELLER algorithm. We then used rational matching by looking at the hydrophobicity of each amino acid in relation to the probability predicted by DeepCoil and used this information to assign the amino acids sequences to the most probable location for each coilcoiled segment. In agreement with the reported coil domain architecture of the closely related homodimeric kinesin KIF17, we found that KIF3A/KIF3B likely also forms three distinct coiledcoil stretches in its stalk. In future work, we plan to generate mutant kinesins in which each of the predicted coil domains is deleted separately and then use motor activity assays to assess the function of each coil domains in KIF3A/KIF3B autoinhibition.


Authors: Samuel Irwin and Martin F. Engelke

Characterization Of The Coiled-Coil Regions In The Heterodimeric Kif3a/Kif3b Motor Protein
Off-Campus Download COinS