Proteins in a shear flow

The conformational dynamics of a single protein molecule in a shear flow is investigated using Brownian dynamics simulations. A structure-based coarse grained model of a protein is used. We consider two proteins, ubiquitin and integrin, and find that at moderate shear rates they unfold through a sequence of metastable states - a pattern which is distinct from a smooth unraveling found in homopolymers. Full unfolding occurs only at very large shear rates. Furthermore, the hydrodynamic interactions between the amino acids are shown to hinder the shear flow unfolding. The characteristics of the unfolding process depend on whether a protein is anchored or not, and if it is, on the choice of an anchoring point.

💡 Research Summary

This study investigates how a single protein molecule responds to shear flow, using Brownian dynamics simulations combined with a structure‑based coarse‑grained (CG) model. The authors focus on two representative proteins: ubiquitin, a small single‑domain protein (76 residues), and integrin, a larger multi‑domain protein (~300 residues). Each amino acid is represented by a spherical bead, and native contacts derived from the crystal structure are encoded as attractive potentials, preserving the protein’s topology while allowing large‑scale conformational changes.

Shear flow is imposed as a linear velocity gradient (γ̇) that generates a Stokes drag on each bead and a rotational component consistent with low‑Reynolds‑number hydrodynamics. Two key variables are explored: (1) the inclusion or exclusion of hydrodynamic interactions (HI) between beads, modeled via the Rotne‑Prager‑Yamakawa tensor, and (2) the anchoring condition of the protein (free, N‑terminal fixed, C‑terminal fixed, or fixed at an internal loop).

The main findings can be summarized as follows:

1. Metastable unfolding pathway – At moderate shear rates (γ̇ ≈ 10⁶ s⁻¹), both proteins do not unwind smoothly. Instead, they pass through a series of distinct metastable states, each characterized by a subset of native contacts that remain intact. These states correspond to partial unfolding of specific secondary‑structure elements (α‑helices or β‑sheets) or the separation of individual domains. The transition from one metastable state to the next is triggered by the progressive rupture of contacts that experience the highest local shear stress. This stepwise behavior contrasts sharply with the continuous, “smooth unraveling” observed for homopolymers under the same flow conditions, highlighting the importance of the protein’s native contact network.

2. Hydrodynamic interactions hinder unfolding – When HI are included, the effective drag on each bead is reduced because the flow field generated by one bead partially shields its neighbors. Consequently, the relative velocity between beads is smaller, and a higher shear rate is required to break the same contacts. Simulations without HI show unfolding at shear rates roughly an order of magnitude lower. This result suggests that in realistic, viscous cellular environments where HI are significant, proteins are mechanically more robust than predictions based on free‑draining models.

3. Effect of anchoring point – The location where the protein is tethered dramatically reshapes the unfolding pathway. An N‑terminal anchor distributes the shear force more uniformly along the chain, leading to a relatively symmetric sequence of metastable states. A C‑terminal anchor concentrates stress near the opposite end, causing early rupture of contacts in the distal domain and a faster loss of that region’s structure. Anchoring at an internal loop creates a mixed scenario: one side of the protein experiences high tension and unfolds quickly, while the other side remains partially folded for a longer time. The anchoring condition also influences the critical shear rate required for complete unfolding.

4. Complete unfolding at extreme shear – Only when the shear rate is pushed to very high values (γ̇ > 10⁸ s⁻¹) do the proteins bypass all intermediate metastable states and become fully extended. At these rates, the mechanical work supplied by the flow overwhelms all stabilizing native contacts, regardless of HI or anchoring. This defines an upper bound for shear‑induced denaturation in the model.

5. Protein‑specific differences – Ubiquitin, being small and monomeric, exhibits fewer metastable intermediates and reaches full extension at lower shear rates than integrin. Integrin’s multi‑domain architecture generates a richer landscape of intermediate states and a higher resistance to shear, reflecting the role of domain‑domain contacts in mechanical stability.

Overall, the paper provides a quantitative framework for understanding shear‑induced protein unfolding. By systematically varying HI and anchoring, the authors demonstrate that both long‑range hydrodynamic coupling and the mechanical boundary conditions are crucial determinants of the unfolding pathway. The work has implications for interpreting experiments in microfluidic devices, for assessing protein stability under physiological flows (e.g., blood circulation), and for designing shear‑resistant biomaterials. It also underscores that protein unfolding under flow cannot be extrapolated from polymer physics alone; the specific native topology and interaction network must be taken into account.