A biochemical mechanism for Stu2/XMAP215-family microtubule polymerases

Technical Summary: A Biochemical Mechanism for Stu2/XMAP215-Family Microtubule Polymerases

Problem Statement
Microtubule (MT) dynamics are regulated by Stu2/XMAP215-family polymerases, which utilize multiple tubulin-binding TOG (Tumor Overexpressed Gene) domains to catalyze MT growth. While the structural diversity of these polymerases is known—ranging from fungal homodimers with two TOG domains per monomer to metazoan monomers with five or six TOG domains—a quantitative biochemical mechanism explaining how polymerase activity arises from the number and binding properties of these TOG domains has remained undefined. Existing models, such as the "concentrating reactants" model and the "polarized unfurling" model, offer qualitative explanations but fail to quantitatively link polymerase activity to TOG domain count and tubulin-binding kinetics. Furthermore, it is unclear if the mechanism is conserved across different oligomerization states (monomeric vs. dimeric) and taxa.

Methodology
The authors tested whether an enzyme-like biochemical model, previously established for the unrelated actin polymerase Ena/VASP, could quantitatively describe Stu2 microtubule polymerase activity. The study employed a multi-faceted experimental approach:

1. In Vitro Microtubule Dynamics Assays: Using Total Internal Reflection Fluorescence (TIRF) and Differential Interference Contrast (DIC) microscopy, the authors measured MT growth rates in the presence of constant concentrations of Stu2 variants (100 nM) across a range of yeast tubulin concentrations (0.4–1.4 µM).
   • Constructs: They utilized wild-type Stu2 (TOG1-TOG2), a variant with two identical TOG2 domains (TOG2-TOG2), and a variant with one inactivated TOG1 domain (TOG1*-TOG2, R200A mutation) to vary the number of functional binding domains.
2. Quantification of End-Resident Polymerase: To validate the enzyme-like assumption that the amount of polymerase at the MT end is independent of growth rate, the authors quantified the fluorescence intensity of GFP-tagged Stu2 at the MT tip using TIRF microscopy at low (0.6 µM) and high (1.4 µM) tubulin concentrations.
3. Bio-layer Interferometry (BLI): To determine the kinetic parameters of TOG:tubulin interactions under conditions mimicking the polymerase assay (pH 6.9, 10 mM NaCl), the authors performed BLI measurements. They used sortase-mediated ligation to site-specifically biotinylate the C-terminus of $\beta$-tubulin, immobilized it on streptavidin biosensors, and measured binding kinetics with the TOG2 domain.
4. Mathematical Modeling: The authors adapted the Ena/VASP biochemical model (Breitsprecher et al., 2011) to Stu2. This model treats the polymerase as an enzyme acting on the polymer end, where unpolymerized tubulin is the substrate. The model relates the growth rate ($V$) to the number of TOG domains ($N$), association/dissociation rate constants ($k_{on}$, $k_{off}$, $K_D$), and a transfer rate constant ($k_t$) describing the delivery of bound tubulin to the MT end. The apparent Michaelis constant ($K_M$) is defined as $K_D + k_t/k_{on}$, and maximal velocity ($V_{max}$) is defined as $k_t \times N$.

Key Results

• Enzyme-Like Kinetics: Stu2-stimulated MT growth rates exhibited a hyperbolic dependence on tubulin concentration, consistent with Michaelis-Menten kinetics. The $V_{max}$ and $K_M$ values were derived for different Stu2 constructs. Notably, the construct with two active TOG domains (TOG1*-TOG2) showed roughly half the maximal activity of constructs with four active TOG domains, supporting a linear relationship between active TOG count and activity.
• Constant Polymerase Occupancy: Fluorescence intensity measurements revealed that the amount of Stu2 residing at the MT end did not vary significantly between low and high tubulin concentrations (and thus different growth rates). This confirms that the observed changes in growth rate are due to kinetic effects rather than differential accumulation of the polymerase on the MT end.
• High-Affinity, Slow Dissociation Binding: BLI measurements under assay-relevant conditions revealed a high affinity of TOG2 for tubulin ($K_D \approx 10$ nM), which is significantly tighter than previous measurements made at higher ionic strength. The dissociation rate ($k_{off}$) was slow ($0.03 s^{-1}$), while the association rate ($k_{on}$) was $3.2 \times 10^6 M^{-1}s^{-1}$.
• Unification of Data: By integrating the polymerase activity data with the binding kinetics, the authors solved for the unmeasured transfer rate constant ($k_t$), finding it to be approximately $5.0 s^{-1}$. This rate is comparable to the rate of uncatalyzed tubulin incorporation, suggesting that TOG-mediated delivery can substantially exceed background rates.
• Stoichiometry Estimation: Using the derived $k_t$ and measured $V_{max}$, the authors calculated the number of TOG domains active at the MT end. For the TOG2-TOG2 dimer, this corresponded to approximately 12 total TOG domains (implying 3 dimers at the end), and for the TOG1*-TOG2 variant, ~8 total TOG domains (4 dimers). These estimates are consistent with prior observations of polymerase occupancy.

Significance and Claims
The paper claims to provide the first quantitative biochemical mechanism for TOG-based polymerases, unifying polymerase activity with the number of TOG domains and their tubulin-binding kinetics. The authors conclude that:

1. Mechanism: Stu2 functions as a "tubulin-shuttling antenna" that operates with high efficiency. The rate-limiting step is the association of unpolymerized tubulin with an empty TOG domain ($k_{on}$). Once bound, the tubulin is almost invariably delivered to the MT end because the transfer rate ($k_t$) is much faster than the dissociation rate ($k_{off}$).
2. Convergent Evolution: The finding that Stu2 (microtubule) and Ena/VASP (actin) polymerases share the same enzyme-like mechanism—despite being unrelated and acting on different cytoskeletal polymers—provides an example of convergent evolution. The common design principle is the localization of flexibly linked subunit-binding domains to the growing polymer end.
3. Model Validation: The data support the "concentrating reactants" model and challenge the "polarized unfurling" model, which lacks quantitative predictions regarding kinetics and appears inconsistent with observations of polymerase variants.
4. Generality: The proposed mechanism does not depend on a specific oligomerization state and should be broadly applicable to polymerases across taxa, including monomeric metazoan XMAP215/chTOG and dimeric fungal Stu2/Alp14, provided the TOG:tubulin affinity is sufficiently high relative to the $K_M$.

The authors maintain a modest tone, noting that while their data supports the concentrating reactants model, they do not definitively invalidate the polarized unfurling model, though they argue the latter currently lacks the necessary quantitative framework and biochemical consistency.