SummaryGrowing microtubule ends serve as transient binding platforms for essential proteins that regulate microtubule dynamics and their interactions with cellular substructures. End-binding proteins (EBs) autonomously recognize an extended region at growing microtubule ends with unknown structural characteristics and then recruit other factors to the dynamic end structure. Using cryo-electron microscopy, subnanometer single-particle reconstruction, and fluorescence imaging, we present a pseudoatomic model of how the calponin homology (CH) domain of the fission yeast EB Mal3 binds to the end regions of growing microtubules. The Mal3 CH domain bridges protofilaments except at the microtubule seam. By binding close to the exchangeable GTP-binding site, the CH domain is ideally positioned to sense the microtubule's nucleotide state. The same microtubule-end region is also a stabilizing structural cap protecting the microtubule from depolymerization. This insight supports a common structural link between two important biological phenomena, microtubule dynamic instability and end tracking.

SummaryBackgroundThe dynamic properties of microtubules depend on complex nanoscale structural rearrangements in their end regions. Members of the EB1 and XMAP215 protein families interact autonomously with microtubule ends. EB1 recruits several other proteins to growing microtubule ends and has seemingly antagonistic effects on microtubule dynamics: it induces catastrophes, and it increases growth velocity, as does the polymerase XMAP215.ResultsUsing a combination of in vitro reconstitution, time-lapse fluorescence microscopy, and subpixel-precision image analysis and convolved model fitting, we have studied the effects of EB1 on conformational transitions in growing microtubule ends and on the time course of catastrophes. EB1 density distributions at growing microtubule ends reveal two consecutive conformational transitions in the microtubule end region, which have growth-velocity-independent kinetics. EB1 binds to the microtubule after the first and before the second conformational transition has occurred, positioning it several tens of nanometers behind XMAP215, which binds to the extreme microtubule end. EB1 binding accelerates conformational maturation in the microtubule, most likely by promoting lateral protofilament interactions and by accelerating reactions of the guanosine triphosphate (GTP) hydrolysis cycle. The microtubule maturation time is directly linked to the duration of a growth pause just before microtubule depolymerization, indicating an important role of the maturation time for the control of dynamic instability.ConclusionsThese activities establish EB1 as a microtubule maturation factor and provide a mechanistic explanation for its effects on microtubule growth and catastrophe frequency, which cause microtubules to be more dynamic.

Machine learning (ML), artificial intelligence (AI) and other modern statistical methods are providing new opportunities to operationalize previously untapped and rapidly growing sources of data for patient benefit. Whilst there is a lot of promising research currently being undertaken, the literature as a whole lacks: transparency; clear reporting to facilitate replicability; exploration for potential ethical concerns; and, clear demonstrations of effectiveness. There are many reasons for why these issues exist, but one of the most important that we provide a preliminary solution for here is the current lack of ML/AIspecific best practice guidance. Although there is no consensus on what best practice looks in this field, we believe that interdisciplinary groups pursuing research and impact projects in the ML/AI for health domain would benefit from answering a series of questions based on the important issues that exist when undertaking work of this nature. Here we present 20 questions that span the entire project life cycle, from inception, data analysis, and model evaluation, to implementation, as a means to facilitate project planning and post-hoc (structured) independent evaluation. By beginning to answer these questions in different settings, we can start to understand what constitutes a good answer, and we expect that the resulting discussion will be central to developing an international consensus framework for transparent, replicable, ethical and effective research in artificial intelligence (AI-TREE) for health.