betaFeedback

WorksPhysics and Astronomy

Vortex Dynamics in Tubular Fluid Membranes

Abstract

Thin cylindrical membranes arise in a wide variety of biological systems ranging from tubular structures on and within cell membranes to in-vitro experiments on artificial vesicles. Motor proteins embedded in such fluidic membranes often induce vortex-like flows. In this work, we construct a class of 2D vortex flow in a thin tubular membrane, coupled to 3D external embedding fluids. The cylinder topology enforces the creation of an additional saddle in the flow field, consistent with Poincaré Index Theorem. In this setup, the incompressibility of the membrane fluid can be utilized to cast the dynamics of a multi-vortex system in the form of a Hamiltonian, This Hamiltonian also incorporates the specific couplings of the 2D membrane flow with the 3D external fluids. The cylinder geometry breaks the in-plane rotational symmetry of the membrane and leads to several interesting features in the multi-vortex dynamics, such as orbit pinching, For a two-vortex system of same circulation, we observe closed orbits with the inter-vortex separation oscillating in time, unlike flat and spherical fluid membranes, where the separation remains constant. Vortex pairs (vortices with opposite circulation) move together along helical geodesics in accordance with a conjecture by Kimura, Proceedings of the Royal Society A, Vol 455 (1999), now extended to tubular geometries. We also explore relative equilibria of multi-vortex systems in this setup and demonstrate vortex leapfrogging via numerical simulations. Our results will be interesting in the context of microfluidic flows arising in nature as well as experimental studies in membrane tubes similar to PNAS 108 (31) 12605-12610 (2011).

1

(2011).

Introduction

The collective dynamics of biological nanomachines at fluid interfaces and membranes is currently an active arena for studying fascinating aspects of dynamical systems. These self-driven motors are capable of converting chemical or other forms of energy into mechanical motion. Moreover, mutual hydrodynamic interactions can often drive the system into an organized state. Biological membranes are particularly interesting in this context because of their fluidic nature [3], which is vital for many living processes. Such a membrane can be well approximated as a thin 2D sheet of viscous fluid with associated flows governed by low Reynolds hydrodynamics [4]. Moreover, the membrane fluid can exchange momentum with the external embedding fluids in the ambient space, thus they are essentially quasi-2D in nature. The early works of Saffman and Delbrück Ref. [5, 6] explored the low Reynolds hydrodynamics of inclusions embedded in such quasi-2D, flat membranes, followed by several advancements Ref. [7, 8, 9, 10, 11, 12, 13, 14, 15]. The quasi-2D nature of the membrane introduces an additional length scale (Saffman length) which is the ratio of the membrane viscosity (η2D\eta_{2D}) and the viscosity of the external fluid (η\eta). Beyond the Saffman length, the in-plane flows are governed by the traction stress from the external embedding fluids. In addition to this interesting feature, membranes found in nature typically have a confined geometry with non-zero curvature. Moreover, they host a large number of inclusions. The impact of membrane topology and curvature on dynamics of such inclusions has been studied in many recent works, Ref. [2, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31] and also being probed in recent experiments, Ref. [2, 17, 20]. In particular, good agreement with Saffman Delbrück theory has been demonstrated in membrane tubes Ref. [2].

The physics of rotating motors in fluid interfaces has recently been the subject of many theoretical and experimental investigations, Ref. [32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44]. In this paper, we focus on the hydrodynamic interactions of rotating inclusions embedded in tubular fluid membranes. The inclusions considered in this work may be considered as driven-to-rotate entities which induce vortex-like flows in membranes. Such vortices are known to self-organize into crystallized states under suitable conditions and exhibit hyperuniform order [45, 46, 47, 48] in flat membranes. The dynamics of such rotating units in spherical fluid membranes is being explored as well, see Ref. [27]. However, analogous studies in membranes of tubular geometries have been relatively less explored. The formation of membrane tubes is an active area of research, Ref. [49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59], arising in a wide variety of biological systems due to polymerization of actin fibers Ref. [60] and microtubules Ref. [61]. and hosting motor proteins Ref. [62, 63]. Such tubular structures exist within and on cell membranes, often bridging two nearby membranes. Moreover, forces exerted on vesicles also lead to tubule formation, as observed in vitro Ref. [57, 63, 64, 65, 66, 67, 68, 69, 70]. For example, such tubular membrane geometries of varying radii can be created by pulling a membrane-tethered bead from the surface of a cell or vesicle using a laser trap. Biological membranes in the Endoplasmic Reticulum and the Golgi apparatus can also form tubular structures Ref. [71, 72, 73, 74, 75, 76, 77, 78, 79], playing vital role in inter-membrane transport Ref. [80, 81]. Such membranes can host a large class of proteins, some of which may be rotating, such as ATP synthase arising in mitochondrial cristae and thylakoids in chloroplasts, Ref. [82, 83, 84].

The tubular membrane geometry that we explore in this paper is distinct from flat and spherical membranes, since the in-plane rotational symmetry is broken. We find that this breakdown of rotational symmetry, along with the confined cylinder topology (in the angular direction) leads to several interesting features in the dynamics of microvortex assemblies in membrane tubes. First, we show that the cylinder topology enforces the creation of an additional saddle in the flow field, such that a global flow constraint emerging from topological considerations is satisfied (Poincaré Index Theorem). Next, utilizing the incompressibility condition of the membrane fluid flow, we cast the dynamics of the multi-vortex system in terms of a Hamiltonian on the cylinder, similar to the approach followed in flat membranes [47, 48]. The Hamiltonian depends on the system parameters, the Saffman length λ\lambda, the tube radius R and the circulation strength of the vortices. This offers experimentally accessible tuning parameters to control the dynamics of the multi-vortex system. Focusing on a simple system of two vortices of same circulation strength τ\tau, we show that the breaking of rotational symmetry leads to non-conservation of the inter-vortex separation, unlike flat and spherical fluid membranes. The vortices generically move in closed orbits, which can be severely distorted by tuning the initial inter-vortex separation as well as the parameter λR\frac{\lambda}{R}. For example, upon reducing the value of λR\frac{\lambda}{R} we find that the single orbit breaks into two smaller orbits. Also, when they are initially sufficiently separated along the longitudinal (z-axis) direction of the tube, we observe that the vortices prefer two distinct orbits that cover the entire circumference of the tube. Two vortices with opposite circulation generically move together along helical trajectories, in accordance with a conjecture by Kimura [1]. We also explore relative equilibria and relative periodic orbits including vortex leapfrogging. A mathematical understanding of confined versus unconfined dynamics of vortices as well as the pinching of vortex orbits emerges from the associated conservation laws of the multivortex system.

Besides being of experimental relevance Ref. [2], we believe our results will be interesting for understanding the collective dynamics of vortices in tubular fluid interfaces and membranes. Moreover, many of our results will be applicable to more setups involving rotating motors (both living and non-living) in tubular fluid interfaces.

It’s worth commenting on how some of our results compare with spherical (Ref. [27]) and flat membranes. The in-plane fluid flows in both spherical and tubular membranes depart from flat membranes in the regime λR≫1\frac{\lambda}{R} \gg1. For the spherical membrane, modifications to fluid flow arise from the intrinsic Gaussian curvature as well as the non-local interactions mediated via the external 3D fluids (related to extrinsic geometry), leading to a global rotation (see Ref. [18] for details) of the entire membrane fluid (along with internal fluids). However, the in-plane rotational symmetry ensures that many features of fluid flow and vortex dynamics in flat membranes continue to hold in spherical membranes, see Ref. [27], even in the high curvature regime. On the other hand, for cylindrical membranes, although the intrinsic Gaussian curvature is zero, the extrinsic geometry of the cylinder comes into play via the couplings to the external 3D fluids, leading to distinct features in flows along the transverse and longitudinal directions. The cylindrical geometry breaks the in-plane rotational symmetry and leads to new phenomenon, such as non-conservation of inter-vortex separation and orbit pinching of same sign vortices, helical trajectories of vortex dipoles, leapfrogging in multi-vortex systems, which are the main results of the paper.

The paper is organized as follows: In Sec. II we provide a brief review of hydrodynamics of generic curved membranes, following [18] and [27, 29, 30]. Specializing to membrane tubes, we construct the point vortex solution, explore topological aspects of the flow and provide the multi-vortex Hamiltonian in Sec. III. The dynamics governed by the Hamiltonian is next explored in Sec. IV, initially focusing on two vortices and then extending to multi-vortex systems, illustrating the conserved quantities. Investigations on vortex equilibria, relative periodic orbits and vortex leapfrogging in the tubular geometry are also provided. We finally conclude with a future outlook in Sec. V.

Membrane Hydrodynamics

In this section, we present a brief review of viscous hydrodynamics in membranes of fixed geometries, closely following Ref. [18] and Ref.[27, 29]. We will describe the setup for arbitrarily curved but fixed membrane geometry and later specialize to cylindrical membranes. In the fluid approximation, the membrane is considered as a two-dimensional, curved, incompressible monolayer of viscosity η2d\eta_{2d} surrounded by external embedding fluids of different viscosities, both outside and inside the membrane. In order to simplify the analysis, we consider situations where the membrane geometry is fixed in time and only consider in-plane shear flows with no normal component. The membrane fluid is coupled to the external embedding fluids, and hence the membrane is essentially quasi 2D. The key hydrodynamic equations appropriate for this setup are as follows:

Dαvα=0(1)D^{\alpha}v_{\alpha}=0 \tag*{(1)}
σαext=−η2D(K(x⃗)vα+DμDμvα)+Dαp+Tα(2)\sigma_{\alpha}^{ext}=-\eta_{2D}\left(K(\vec{x})v_{\alpha}+D^{\mu}D_{\mu}v_{\alpha}\right)+D_{\alpha}p+T_{\alpha} \tag*{(2)}
∇⋅v±=0,η±∇2v±=∇±p±(3)\nabla\cdot\mathbf{v}_{\pm}=0,\qquad\eta_{\pm}\nabla^{2}\mathbf{v}_{\pm}=\nabla_{\pm}p^{\pm} \tag*{(3)}
Tα=σαr−∣r=R−σαr+∣r=R,σij±=η±(Divj±+Djvi±)−gijp±(4)T_{\alpha}=\left.\sigma_{\alpha r}^{-}\right|_{r=R}-\left.\sigma_{\alpha r}^{+}\right|_{r=R},\qquad\sigma_{ij}^{\pm}=\eta_{\pm}\left(D_{i}v_{j}^{\pm}+D_{j}v_{i}^{\pm}\right)-g_{ij}p_{\pm} \tag*{(4)}
v±∣r=R=v.(5)\left.\mathbf{v}_{\pm}\right|_{r=R}=\mathbf{v}. \tag*{(5)}

We refer to Appendix A of Ref. [27] by one of the authors for a detailed derivation of the equations. We use Greek indices to represent the in-plane 2D membrane coordinates. Latin indices are used for 3D coordinates to describe flows in the ambient 3D embedding fluids surrounding the 2D membrane. Eq. (1) is the incompressibility condition for the 2D membrane fluid velocity field vαv_{\alpha}. DD is the metric compatible covariant derivative. Note that on a curved manifold, incompressibility implies that the divergence is taken with respect to the covariant derivative. Eq. (2) is the boundary condition for tangential stress balance at the membrane surface. σαext\sigma_{\alpha}^{ext} represents the local stress exerted by the rotating inclusions embedded in the membrane (this will be a point torque for the following discussions, see Ref.[27] for details ) and is balanced by the in-plane membrane stress and the tangential traction stress from the external embedding fluids (denoted by TαT_{\alpha}) in the equation. Generic membrane surface coordinates are denoted by “x”. Notably, in the low Reynolds regime, inertia terms are absent and fluid flows are dictated by Stokes equations, with viscosity η2D\eta_{2D} being the viscosity of the membrane fluid. K(x⃗)K(\vec{x}) is the local Gaussian curvature and pp represents the 2D membrane pressure. The membrane fluid is surrounded by external 3D embedding fluids with viscosities η±\eta_{\pm} and pressure p±p_{\pm} where “±\pm” denotes the outer/inner fluid respectively. Eq. (3) are the relevant Stokes equations for the external incompressible embedding fluids in 3D. Eq. (4) gives the expression for the traction vector TαT_{\alpha}. The coordinate rr represents a generic coordinate along the membrane normal. Finally, Eq. (5) is the no-slip boundary condition. We also define two Saffman length scales

λ+=η2dη+,λ−=η2dη−.\lambda_{+}=\frac{\eta_{2d}}{\eta_{+}},\qquad\lambda_{-}=\frac{\eta_{2d}}{\eta_{-}}.

We refer to Appendix A and B of Ref. [27] for more details on the appearance of Gaussian curvature and external torque in the stress balance condition Eq. (2). In general, the viscosities η+\eta_{+} and η−\eta_{-} of the external fluids give rise to two Saffman lengths λ+\lambda_{+} and λ−\lambda_{-}. However, the geometric asymmetry between the two embedding fluids for the cylindrical case is not much, since both the internal and external fluids are unbounded, unlike spherical membranes where the internal fluid is bounded. As a result, the effects of asymmetry in the viscosities are much weaker for cylindrical membranes [18]. In what follows, we will choose η≡η+=η−\eta\equiv\eta_{+} = \eta_{-} and thus restrict to a unique Saffman length λ≡η2Dη\lambda\equiv\frac{\eta_{2D}}{\eta}, with full expressions for asymmetric viscosities in Appendix A. Higher values of λR\frac{\lambda}{R} indicate the high curvature regime (thin membrane tube). Typical values of λ\lambda range from 0.1 to 10 micrometers while typical membrane tethers have radius of the order of few nanometers, justifying the thin tube approximation i.e. the high curvature regime (λR≫1)\left(\frac{\lambda}{R} \gg1\right) of this work. Let us note that both these assumptions can be relaxed if one uses the full vortex solution constructed in Appendix A.

We summarize the notations in a table:

SymbolMeaning
xxGeneric surface coordinates
vαv_{\alpha}2D velocity of the membrane fluid
α\alphaSurface coordinate index for the 2D membrane
σαext\sigma_{\alpha}^{\mathrm{ext}}Local stress exerted by the inclusions embedded in the membrane
η2D\eta_{2D}Viscosity of the 2D membrane fluid
DDCovariant derivative compatible with the metric
K(x⃗)K(\vec{x})The local Gaussian curvature of the membrane
ppLocal 2D membrane pressure
σ3D\sigma^{3D}Stress tensor of the external fluids
zzA generalized coordinate normal to the membrane surface
λ\lambdaSaffman length
τ\tauVortex circulation strength

Table 7.

Vortex Solution and Multi-vortex Hamiltonian

In this section, we specialize to membranes of cylindrical geometry which have zero Gaussian curvature i.e. K(x)=0K(x) = 0 in Eq. (2). This greatly simplifies the coupled system of

Figure 11 (PDF p. 8). (a) Case A; (b) Case B; (c) Case C; (d) Case D; (e) Case E

equations and one can arrive at an analytic vortex solution in this setup. Note that even in the absence of Gaussian curvature, the extrinsic geometry of the cylinder modifies the momentum exchange between the membrane fluid and the external embedding fluids and has significant impact on the multi-vortex dynamics we are going to study. We refer to Appendix Sec. A for a detailed computation of the single vortex solution and here summarize the key results. The incompressible nature of the membrane fluid can be utilized to express the membrane velocity field confined to the cylinder surface in terms of a scalar stream function ψ\psi as follows (here θ\theta and zz are the usual angular and axial coordinates along the cylinder surface of fixed radius RR):

vθ=∂zψvz=−1R∂θψ(6)v_{\theta}=\partial_{z}\psi\qquad v_{z}=-\frac{1}{R}\partial_{\theta}\psi \tag*{(6)}

As shown in detail in Appendix Sec. A, in the limit of thin tubular membranes, the stream function ψ\psi for a point vortex situated at the origin (θ,z)=(0,0)(\theta,z)=(0,0) takes the following simple form

ψ(θ,z)=τ4πη2D(−log⁡[1−2e−∣z∣Rcos⁡θ+e−2∣z∣R]+λ2Re−2∣z∣λR).(7)\psi(\theta,z)=\frac{\tau}{4\pi\eta_{2D}}\left(-\log\left[1-2e^{-\frac{\lvert z\rvert}{R}}\cos\theta+e^{-\frac{2\lvert z\rvert}{R}}\right]+\sqrt{\frac{\lambda}{2R}}e^{-\frac{\sqrt{2}\lvert z\rvert}{\sqrt{\lambda R}}}\right). \tag*{(7)}

where τ\tau is the strength of the circulation. The logarithmic term in the stream function is expected from 2D vortex dynamics Ref. [85], while the additional exponential correction arises from the exchange of momentum between the 2D membrane fluid and external 3D fluids, illustrating the quasi-2D nature of the system. The exponential correction term originates from the zero mode in the compact θ\theta direction of the tube (and features in the θ\theta component of the vortex velocity field computed afterwards). The associated decay length in the exponential term is given by λR\sqrt{\lambda R} which is larger than the tube radius RR. For a vortex located at (θ0,z0)(\theta_0,z_0) on the membrane tube, the velocity field at any location (θ,z)(\theta,z) is given by

vθ[θ,z,θ0,z0]=τ(z−z0)4πRη2D∣z−z0∣(2−2cos⁡(θ−θ0)e∣z−z0∣R1+e2∣z−z0∣R−2cos⁡(θ−θ0)e∣z−z0∣R−e−Rλ∣z−z0∣)v_\theta[\theta,z,\theta_0,z_0]=\frac{\tau(z-z_0)}{4\pi R\eta_{2D}|z-z_0|}\left(\frac{2-2\cos(\theta-\theta_0)e^{\frac{|z-z_0|}{R}}}{1+e^{\frac{2|z-z_0|}{R}}-2\cos(\theta-\theta_0)e^{\frac{|z-z_0|}{R}}}-e^{-\sqrt{\frac{R}{\lambda}}|z-z_0|}\right)
vz[θ,z,θ0,z0]=−τ4πη2DRsin⁡(θ−θ0)cos⁡(θ−θ0)−cosh⁡z−z0R(8)v_z[\theta,z,\theta_0,z_0]=-\frac{\tau}{4\pi\eta_{2D}R}\frac{\sin(\theta-\theta_0)}{\cos(\theta-\theta_0)-\cosh\frac{z-z_0}{R}} \tag*{(8)}

As explained in Appendix A, the above velocity field is constructed out of the stream function ψ\psi via vα=ϵαβDβψv_\alpha=\epsilon_{\alpha\beta}D^\beta\psi, such that the incompressibility constraint Dαvα=0D^\alpha v_\alpha=0 is automatically satisfied. Moreover, the velocity fields vanish at large separation in the zz direction. The vortex solution is presented in Figure 1. As evident from the streamlines, the cylinder topology enforces the creation of an additional saddle defect in the flow such that the net index is zero, consistent with Poincaré Index theorem. We now cast the dynamics of the multi-vortex system such that a vortex is simply advected by the local fluid flow generated by

Plot of the vortex solution showing the vortex core and saddle

Figure 1. Plot of the vortex solution Eq. (8) with the blue and red dot representing the vortex core and saddle respectively.

Figure 6. Vortex leapfrogging for λR=100\frac{\lambda}{R}=100 in a 4 vortex system. Two vortices with positive circulation τ=+1/2\tau=+1/2 each are marked red, while two vortices of circulation τ=−12\tau=-\frac{1}{2} each are shown in blue, with the same color code for trajectories. Numerical errors in L2L^2, CθC_\theta, CzC_z are also shown.

the remaining vortices. Following standard treatment in vortex literature, such dynamics of the many-vortex system can be encapsulated in the form of a Hamiltonian on the cylinder, utilizing the cylinder stream function Eq. (7)

H=∑i≠jτiτj4πη2D(−log⁡[1−2e−∣zi−zj∣Rcos⁡(θi−θj)+e−2∣zi−zj∣R]+λ2Re−2∣zi−zj∣λR)(9)H=\sum_{i\ne j}\frac{\tau_i\tau_j}{4\pi\eta_{2D}}\left(-\log\left[1-2e^{-\frac{|z_i-z_j|}{R}}\cos(\theta_i-\theta_j)+e^{-\frac{2|z_i-z_j|}{R}}\right]+\sqrt{\frac{\lambda}{2R}}e^{-\frac{\sqrt{2}|z_i-z_j|}{\sqrt{\lambda R}}}\right) \tag*{(9)}

where Qi=∣τi∣RθiQ_i=\sqrt{|\tau_i|}R\theta_i and its associated conjugate momentum is Pi=∣τi∣ziP_i=\sqrt{|\tau_i|}z_i, in terms of which the dynamics is governed by the usual Hamilton’s equations Q˙i=∂PiH\dot{Q}_i=\partial_{P_i}H, P˙i=−∂QiH\dot{P}_i=-\partial_{Q_i}H. Just like the vortex stream function, the Hamiltonian features logarithmic interactions along with an extra interaction term arising from the zero mode in the compact angular direction of the cylinder. For spherical membranes, the Hamiltonian also receives a dominant contribution from a similar zero mode that gives rise to global rotation Ref. [27]. In the usual coordinates (θ,z)(\theta,z), the dynamical equations for the multi-vortex system thus reads

Rθ˙i=∑j≠iNvθ[θi,zi,θj,zj]R\dot{\theta}_i=\sum_{j\ne i}^{N}v_\theta[\theta_i,z_i,\theta_j,z_j]
z˙i=∑j≠iNvz[θi,zi,θj,zj](10)\dot{z}_i=\sum_{j\ne i}^{N}v_z[\theta_i,z_i,\theta_j,z_j] \tag*{(10)}

where the velocity field (vθ,vz)(v_\theta, v_z) is given by Eq. (8). Written explicitly, the dynamical equations take the following form

Rθ˙i=∑j≠iNτj(zi−zj)4πη2D∣zi−zj∣(2−2cos⁡(θi−θj)1+e2∣zi−zj∣R−2cos⁡(θi−θj)e∣zi−zj∣R−e−2Rλ∣zi−zj∣)R\dot{\theta}_i=\sum_{j\ne i}^{N}\frac{\tau_j(z_i-z_j)}{4\pi\eta_{2D}|z_i-z_j|}\left(\frac{2-2\cos(\theta_i-\theta_j)}{1+e^{\frac{2|z_i-z_j|}{R}}-2\cos(\theta_i-\theta_j)}e^{\frac{|z_i-z_j|}{R}}-e^{-\sqrt{\frac{2}{R\lambda}}|z_i-z_j|}\right)
z˙i=−∑j≠iNτj4πη2DRsin⁡(θi−θj)cos⁡(θi−θj)−cosh⁡zi−zjR.(11)\dot{z}_i=-\sum_{j\ne i}^{N}\frac{\tau_j}{4\pi\eta_{2D}R}\frac{\sin(\theta_i-\theta_j)}{\cos(\theta_i-\theta_j)-\cosh\frac{z_i-z_j}{R}}. \tag*{(11)}

The translational symmetry of the Hamiltonian leads to the conservation of the following quantities

Cθ≡∑iτiθiCz=∑iτizi(12)C_\theta\equiv\sum_i\tau_i\theta_i \qquad C_z=\sum_i\tau_i z_i \tag*{(12)}

which is ensured in our multi-vortex simulations to a good accuracy (Numerical errors of the order of 10−1410^{-14}). Note that there are fewer conserved quantities arising from the symmetries of the Hamiltonian and thus in general, one expects non-integrable dynamics for multi-vortex systems, as will be evident in the simulations carried out in the next section. We further define a quantity

L2≡12∣∑i≠jτiτjlij2∣(13)L^2\equiv\frac{1}{2}\left|\sum_{i\ne j}\tau_i\tau_j l_{ij}^2\right| \tag*{(13)}

where lij2=R2Min⁡((θi−θj)2,(2π−(θi−θj))2)+(zi−zj)2l_{ij}^2=R^2\operatorname{Min}\left((\theta_i-\theta_j)^2,\left(2\pi-(\theta_i-\theta_j)\right)^2\right)+(z_i-z_j)^2 is the distance function appropriate for the cylinder geometry. For the two vortex systems, we will denote the quantity in Eq. (13) by L122L_{12}^2 and reserve the notation L2L^2 for multi-vortex systems. The quantity L2L^2 is conserved in flat and spherical membranes but violated in membrane tubes due to loss of in-plane rotational symmetry, as will be explicitly demonstrated in the simulations presented in § 0.4.

Vortex dynamics in membrane tubes

In this section, we numerically simulate the vortex dynamics on the surface of the membrane tube using the dynamical equations Eq. (10). The vortex trajectories are computed using a fourth-order Runge–Kutta (RK4) integration scheme with an adaptive time step. At each iteration, the time step is adjusted to ensure that the relative decrease in distance between any two approaching vortices does not exceed a specified threshold (0.05 or 0.1 in the present simulations). Additionally, the time step is constrained by a fixed upper bound, ranging from 0.01 to 0.001, depending on the simulation. This upper bound in timesteps maintains the maximum local integration error per time step in the order of 1.0×e−101.0 \times e^{-10} to 1.0×e−151.0 \times e^{-15}. The simulations are implemented in Python 3, with minor performance optimizations carried out using the Numba library. In all cases, we include a 3D plot showing the vortex trajectories on the surface of the membrane tube. Trajectories of vortices with positive circulation (anti-clockwise) are colored blue, while vortices of negative circulation (clockwise) are colored red. The initial locations of vortices are marked by a colored dot, following the same color code as the trajectories. We first discuss the two vortex case for λR=100\frac{\lambda}{R}=100 in Figure 2a-Figure 2f) for equal (or opposite) vortex strength. This typically results in closed orbits for vortex pairs of same circulation and helical geodesics for vortex pairs of opposite circulation. Next, in Figure 3a we show the trajectories for two vortices of unequal strength but same sign, giving rise to concentric orbits. Figure 3b illustrates interesting effects resulting from tuning λR\frac{\lambda}{R}. Multi-vortex systems are explored in Figure 4a-Figure 4d). Vortex leapfrogging and relative periodic orbits are illustrated in various situations Figure 5-Figure 7 along with the associated numerical errors in conservation of CθC_\theta and CzC_z. Finally, we study the stability of vortex rings in Figure 10.

Figure 2. (f) Case F. Two vortex dynamics in the membrane tube with λR=100\frac{\lambda}{R}=100. We show variations in vortex locations (θ,z)(\theta,z) and the evolution of L2L^2 defined in Eq. (13) with respect to time. For the 3D cylinder plot appearing on the extreme right, blue dots indicate initial location of vortices with positive circulation, while red dots indicate vortices with negative circulation. The same color code is used for the vortex trajectories.

Two rows of vortex dynamics plots and cylinder trajectories

Figure 3. (a) Same configuration as Figure 2a but with unequal vortex strengths of same sign. (b) Same configuration as Figure 2b but with λR=5\frac{\lambda}{R}=5. Two vortex dynamics with tuning of parameters: In each row, we show the vortex locations (θ,z)(\theta,z) and the evolution of L2L^2 defined in Eq. (13) with respect to time. The 3D cylinder plot appearing on the extreme right,shows the initial vortex locations (blue dots) as well as the vortex trajectories.

Figure 4. Multi-vortex dynamics in the membrane tube : In each row, we show the locations (θ,z)(\theta,z) of all 4 vortices and the evolution of L2L^2 defined in Eq. (13) with respect to time. The 3D cylinder plot appearing on the extreme right, shows the initial vortex locations as well as the vortex trajectories. Blue dots indicate initial location of vortices with positive circulation, while red dots indicate vortices with negative circulation. The same color code is used for the vortex trajectories.

Figure 5. Vortex leapfrogging for λR=100\frac{\lambda}{R}=100 in a 3 vortex system. The vortex with positive circulation τ=+1\tau=+1 is marked red, while two vortices of circulation τ=−12\tau=-\frac{1}{2} each is shown in blue, with the same color code for trajectories. Numerical errors in L2L^2, CθC_\theta, CzC_z are also shown.

Figure 7. Vortex leapfrogging for λR=100\frac{\lambda}{R}=100 in a 10 vortex system. Five vortices with random splittings of vortex strengths adding up to a net positive circulation ∑iτi=+1\sum_i \tau_i=+1 are marked red, while remain vortices with random splittings of vortex strengths adding up to a net negative circulation ∑iτi=−1\sum_i \tau_i=-1 are shown in blue, with the same color code for trajectories. Numerical errors in L2L^2, CθC_\theta, CzC_z are also shown.

Figure 10. Left: An equilibria of 4 vortices arranged in a ring, with intervortex separation πR2\frac{\pi R}{2}, Right: Mild perturbation along θ\theta or zz or both renders the system unstable.

The two-vortex system

We first consider a two-vortex system of the same strength τ\tau with λ/R=100\lambda/R=100. In Figure 2a, we show the variation of vortex locations (θ,z)(\theta,z) and L122L_{12}^{2} defined in Eq. (13) with time and the rightmost plot shows a trace of the vortex trajectories. From the plots, it is clear that a two vortex system which is initially closely spaced will orbit each other in a closed orbit, with the inter-vortex distance oscillating in time. The closed orbits indicate a near-equilibrium configuration modulated by the delicate balance provided by the Hamiltonian. This is in contrast to flat and spherical fluid membranes, where L122L_{12}^{2} is conserved and arises due to the breaking of in-plane rotational symmetry in tubular membranes. However, the orbit shape gets remarkably distorted for vortices, which are initially well separated. This is illustrated in Figure 2b where the vortices are initially well separated along the angular θ^\hat{\theta} direction. This can also lead to pinching of a single closed orbit into two separate orbits. For example, two vortices initially well separated along the longitudinal z^\hat{z} direction of the tube prefer to go in separate orbits, encircling the tube, see Figure 2c.

Comments on pinching of orbits: A mathematical understanding of these results emerges easily from the associated conserved quantities Eq. (12) and the conservation of the Hamiltonian Eq. (9). Using Eq. (12) one can eliminate say (θ2,z2)(\theta_2,z_2) in favor of (θ1,z1)(\theta_1,z_1), thus the Hamiltonian Eq. (9) for two vortices becomes a function of only one of the vortex co-ordinates i.e., (θ1,z1)(\theta_1,z_1) and the constants of motion i.e. Cθ,CzC_\theta,C_z. Analyzing the level curves of this Hamiltonian in the (θ1,z1)(\theta_1,z_1) plane, one readily observes that for the situations described in Figure 2a and Figure 2b the contour plots are closed loops, while for Figure 2c, they traverse the entire angular direction of the cylinder, The level curves of the Hamiltonian in the (θ1,z1)(\theta_1,z_1) plane is displayed in the left panel of Figure 8 for generic choice of constants of motion CθC_\theta and CzC_z. In the right panel, we show the level curves of the Hamiltonian for 4 different values Cz=0.5,1,1.5,2C_z = 0.5, 1, 1.5, 2 at a fixed value of Cθ=0C_\theta= 0. This is achieved by tuning the initial separation along z-direction for the two vortex system, with θ1=θ2=0\theta_1 = \theta_2 = 0. We observe that for lower values of CzC_z, i.e. less initial separation in z, the level curves are closed loops, while for higher values of z-separation, the level curves traverse the entire angular direction of the cylinder leading to the pinching of orbits.

Figure 8. Left: Level curves of the Hamiltonian in the (θ1,z1)(\theta_1,z_1) plane for λ/R=100\lambda/R = 100 and generic values of CθC_\theta and CzC_z, Right: Level curves of the Hamiltonian in the (θ1,z1)(\theta_1,z_1) plane for Cz=0.5C_z = 0.5 (blue), 1 (orange), 1.5 (green) and 2 (red) at a fixed value of Cθ=0C_\theta= 0. This is achieved by tuning the initial separation along z-direction for the two vortex system, with θ1=θ2=0\theta_1 = \theta_2 = 0. Notice that the red curve for Cz=2C_z = 2 traverses the entire angular direction of the cylinder.

For the two-vortex system of opposite strength (also known as vortex pair), we observe vastly different interactions, see Figure 2d-Figure 2f). A vortex pair separated along the angular (transverse) direction of the tube move in vertical parallel trajectories (Figure 2d) while vortex pairs separated along the longitudinal direction of the tube move together along the transverse direction in separate closed orbits (Figure 2e). Arbitrarily situated vortex pairs move together along helical geodesics, in accordance with a conjecture by Kimura Ref. [1]. Note that in all the situations with vortex pairs, the net circulation is zero and L2L^2 is conserved in time. These results are evident from Eq. (11) where for equal and opposite strength τ1=−τ2\tau_1=-\tau_2 we find that θ˙1=θ˙2\dot{\theta}_1=\dot{\theta}_2 and z˙1=z˙2\dot{z}_1=\dot{z}_2. Using this fact, it is straightforward to compute the pitch of the helix zPz_P. The time period T for one complete rotation around the membrane tube of radius R for the helical motion in Figure 2f for a vortex pair with τ1=−τ2≡τ\tau_1=-\tau_2\equiv\tau can be easily constructed from the fact that in this situation θ˙1=θ˙2\dot{\theta}_1=\dot{\theta}_2 and z˙1=z˙2\dot{z}_1=\dot{z}_2, which implies that the θ\theta and zz separation between the vortices remain constant in time. Thus, we can utilize the velocity expression in Eq. (11) to find the time period T (note that the velocities depend on the separation in θ\theta and zz and hence are constants as well)

2πRT=∣τ(z1i−z2i)4πRη2D∣z1i−z2i∣((2−2cos⁡(θ1i−θ2i))e∣z1i−z2i∣R1+e2∣z1i−z2i∣R−2cos⁡(θ1i−θ2i)e∣z1i−z2i∣R−e−2R∣z1i−z2i∣)∣\frac{2\pi R}{T}=\left|\frac{\tau(z_1^i-z_2^i)}{4\pi R\eta_{2D}|z_1^i-z_2^i|}\left(\frac{(2-2\cos(\theta_1^i-\theta_2^i))e^{\frac{|z_1^i-z_2^i|}{R}}}{1+e^{\frac{2|z_1^i-z_2^i|}{R}}-2\cos(\theta_1^i-\theta_2^i)e^{\frac{|z_1^i-z_2^i|}{R}}}-e^{-\sqrt{\frac{2}{R}}|z_1^i-z_2^i|}\right)\right|

where (θ1i,z1i)(\theta_1^i,z_1^i) and (θ2i,z2i)(\theta_2^i,z_2^i) are the initial vortex locations. Thus, the pitch of the helix zPz_P can be computed from the z˙1=z˙2\dot{z}_1=\dot{z}_2 expression in Eq. (11) and given by

zP=2πRsin⁡(θ1i−θ2i)cos⁡(θ1i−θ2i)−cosh⁡z1i−z2iR2−2cos⁡(θ1i−θ2i)e∣z1i−z2i∣R1+e2∣z1i−z2i∣R−2cos⁡(θ1i−θ2i)e∣z1i−z2i∣R−e−2R∣z1i−z2i∣(14)z_P=\frac{\frac{2\pi R\sin(\theta_1^i-\theta_2^i)}{\cos(\theta_1^i-\theta_2^i)-\cosh\frac{z_1^i-z_2^i}{R}}}{\frac{2-2\cos(\theta_1^i-\theta_2^i)e^{\frac{|z_1^i-z_2^i|}{R}}}{1+e^{\frac{2|z_1^i-z_2^i|}{R}}-2\cos(\theta_1^i-\theta_2^i)e^{\frac{|z_1^i-z_2^i|}{R}}}-e^{-\sqrt{\frac{2}{R}}|z_1^i-z_2^i|}} \tag*{(14)}

where (θ1i,z1i)(\theta_1^i,z_1^i) and (θ2i,z2i)(\theta_2^i,z_2^i) are the initial locations of the vortex pair.

Comments on confined and unconfined dynamics of vortices: The confined dynamics of vortices of same circulation (or equivalently, the unconfined dynamics of vortices of opposite circulation) can be easily understood from the conservation of the vortex Hamiltonian Eq. (9) and the conservation of CzC_z, Eq. (12). For vortices of same circulation τ1=τ2≡τ\tau_1=\tau_2\equiv\tau, z1=Czτ−z2z_1=\frac{C_z}{\tau}-z_2 which implies ∣z1−z2∣=∣Czτ−2z2∣|z_1-z_2|=\left|\frac{C_z}{\tau}-2z_2\right|. Thus, the Hamiltonian decays exponentially as z2→∞z_2\to\infty, which is forbidden by energy conservation. This in turn explains the bounded orbits for vortices of same strength. On the other hand, for vortices of equal and opposite strength, then ∣z1−z2∣=∣Czτ∣\lvert z_1-z_2\rvert= \left\lvert\frac{C_z}{\tau} \right\rvert which is a constant. Thus, the Hamiltonian conservation allows the vortices to drift together to large zz leading to unconfined dynamics in this situation.

Additionally, we note that the ratio of the vortex strengths τ1/τ2\tau_1/\tau_2 and λ/R\lambda/R provide additional tuning parameters that can be used to control the vortex trajectories, as illustrated in Figure 3a and Figure 3b respectively. Figure 3a shows two concentric orbits arising from unequal vortex strengths, whereas in Figure 3b we simulate the same initial configuration as Figure 2b but with λ/R=5\lambda/R=5. In this situation, the single closed orbit of Figure 2b breaks into two separate local closed orbits, as seen in Figure 3b. This curvature mediated pinching of orbits can also be understood in terms of the level curves of the Hamiltonian Eq. (9). In this situation, translational symmetry can be used to express the Hamiltonian of the two-vortex system in terms of separation Δθ\Delta\theta and Δz\Delta z. We investigate the level curves of the Hamiltonian in the (Δθ,Δz)(\Delta\theta,\Delta z) plane as shown in Figure 9 for large and small values of λ/R\lambda/R. The variations in (Δθ,Δz)(\Delta\theta,\Delta z) gets localized upon reducing λ/R\lambda/R, consistent with Figure 3b.

Figure 9. Level curves of the Hamiltonian in the (Δθ,Δz)(\Delta\theta,\Delta z) plane for Left: λ/R=100\lambda/R=100 and Right: for λ/R=1\lambda/R=1,

Essential features of the multi-vortex system

Moving on to multi-vortex assemblies Figure 4a-Figure 4d), we first simulate a set of 4 vortices arranged along the transverse (θ^\hat{\theta}) direction of the membrane tube, of the same (Figure 4a) and alternating (Figure 4b) circulations. All simulations are performed with λ/R=100\lambda/R=100. A distinct feature of multi-vortex assemblies of the same strength (Figure 4a) is that their dynamics remains confined in a region determined by the initial distribution, with L2L^2 oscillating in time. This can be again explained from the conservation of the Hamiltonian Eq. (9) like the two vortex system. On the other hand, the multi-vortex systems of alternating circulation (with zero net circulation) show unconfined dynamics, yet L2L^2 remains conserved in time, see Figure 4b. Vortices initially situated along the longitudinal direction of the cylinder Figure 4c and Figure 4d share the same features as their transverse counterparts, ie. Figure 4a and Figure 4b, except the fact that the compact cylinder topology enforces the vortices of alternating circulations to exhibit closed orbits wrapping the entire membrane tube, see Figure 4d. In Figure 5-Figure 7 we demonstrate the interesting phenomenon of vortex leapfrogging in the tubular fluid membrane. In such situations, the vortices move in periodic or quasi-periodic orbits along with translation along the tube surface. They are fairly easy to design using the results from two vortex systems. As observed in Figure 2d-Figure 2f), vortex pairs (of opposite circulation) translate together while vortices of same circulation perform periodic orbits with no net translation. In order to achieve relative periodic orbits (leapfrogging), one can take a vortex pair (of circulation τ1=+1\tau_1 = +1 and τ2=−1\tau_2 = -1) and divide them into two groups A and B, such that the circulations in group A add up to +1 while that of Group B add up to -1 (this is also known as “Vortex splitting”). The vortices in Group A and Group B will translate together, while within each individual group we will have periodic or quasi periodic orbits of the split vortices. This is illustrated in several leapfrogging scenarios in Figure 5-Figure 7. Finally, we note that equilibria of vortex rings can be constructed for NN vortices arranged along the transverse direction of the tube with intervortex separation 2πRN\frac{2\pi R}{N}. Mild perturbation along θ\theta or zz (or both) in any one of the vortices renders the system unstable. This is illustrated in Figure 10 for N=4N = 4. A more systematic perturbative analysis of vortex rings and crystals in our setup is left for upcoming works.

Conclusion

In this paper, we focus on the hydrodynamic interactions of vortex assemblies embedded in tubular fluid membranes. The tubular membrane geometry is distinct from flat and spherical membranes, due to loss of in-plane rotational symmetry. We find that the confined cylinder topology (in the angular direction) leads to several interesting features in the dynamics of vortices in membrane tubes. We first construct an analytic vortex solution in this tubular fluid membrane setup and show that the cylinder topology enforces the creation of an additional saddle defect in the flow, consistent with Poincaré Index theorem. Next, the incompressibility condition of the membrane fluid flow is utilized to construct a Hamiltonian on the cylinder that governs the dynamics of the multi-vortex system. The Hamiltonian depends on the system parameters, the Saffman length λ\lambda, the tube radius RR and the circulation strength τ\tau of the vortices. This offers experimentally accessible tuning parameters to control the dynamics of the multi-vortex system. Focusing on a simple system of two vortices of same circulation strength τ\tau, we show that the breaking of rotational symmetry leads to non-conservation of the inter-vortex separation, unlike flat and spherical fluid membranes. The vortices generically move in closed orbits, which can be severely distorted by tuning the initial inter-vortex separation as well as the parameter λR\frac{\lambda}{R}. In particular, we observe several situations where a single closed orbit breaks into two upon tuning system parameters. We also illustrate this pinching from a mathematical point of view by considering the level sets of the associated vortex Hamiltonian. Two vortices with opposite circulation generically move together along helical trajectories, thus confirming a conjecture by Kimura on the dynamics of vortex pairs Ref. [1]. We also explore relative periodic orbits and relative equilibria including vortex leapfrogging.

The work needs extension in several directions: first, although we performed the analysis of vortex dynamics in the limit of thin membranes for ease of numerical implementation, the full vortex solution constructed in Appendix § 0.4.1 can be utilized to perform vortex simulations in situations where the Saffman length is comparable to membrane radius as well as asymmetric viscosities of external fluids coupled to the membrane. Addition of short range repulsive interactions and thermal noise will be interesting and important for studying diffusion of motor proteins in membrane tubes, since the diffusion constant can be related to the hydrodynamic mobility via the Stokes-Einstein relation. Secondly, although the current model deals with membranes of fixed geometry, it will be worth pursuing similar studies of vortex dynamics in membranes of dynamical geometry, together with corrections from inertia terms in the Navier-Stokes equation. Besides being of experimental relevance (Ref. [2]), we believe our results will be interesting for understanding the collective dynamics of vortices and rotors in tubular fluid interfaces and membranes. We believe many of our results will be applicable to more general setups involving rotating motors (both living and non-living) in tubular fluid interfaces, analogous to the studies in flat membranes Ref. [45, 46, 47, 48] and vortex simulations in superconductors, Ref. [86]. Dynamics of rotating matter in fluid interfaces is currently an active area of research Ref.[32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44], both from an experimental and theoretical perspective. It will also be interesting to connect our results to vortex and mass dynamics in inviscid cylinder geometries, Ref.[87, 88]. Let us note that the vortex dynamics studied in this work bears many resemblances to the study of vortex dynamics in inviscid fluids, Ref. [87], as expected from the symmetry considerations. We leave a more detailed comparison of our results with existing numerical simulations and experiments for upcoming works. An analysis of thermal fluctuations in this setup, along the lines of the work of Sokolov and Diamant Ref. [89] is also left for the future.

Acknowledgments

We are very thankful to Naomi Oppenheimer, Haim Diamant, Mark Henle, Sarthak Bagaria, Samyak Jain, Prasad Perlekar and Mustansir Barma. R.S is supported by DST INSPIRE Faculty fellowship, India (Grant No.IFA19-PH231), NFSG and OPERA Research Grant from Birla Institute of Technology and Science, Pilani (Hyderabad Campus). The work is dedicated to the memory of Professor Alex J. Levine, UCLA.

Construction of the Vortex solution in the Tubular Fluid Membrane

In this Appendix, we will sketch the construction of the flow sourced by a point vortex embedded in a tubular fluid membrane, extending Ref.[18, 27, 29]. We will first provide the vortex solution for cylindrical fluid membranes of arbitrary radius and finally consider the limit of thin tubular membrane, which is of relevance to our study in this paper. The cylinder surface metric in the usual (θ,z)(\theta, z) coordinates is

ds2=R2dθ2+dz2(15)ds^2 = R^2 d\theta^2 + dz^2 \tag*{(15)}

and

εαβ=(0R−R0)(16)\varepsilon_{\alpha\beta} = \begin{pmatrix} 0 & R \\ -R & 0 \end{pmatrix} \tag*{(16)}

is the corresponding Levi Civita tensor. The incompressibility condition of the membrane fluid (Eq. (1) of main text) allows us to write the 2D velocity field on the cylinder surface in terms of a scalar stream function ψ\psi as follows:

vα=εαβDβψ.v_\alpha= \varepsilon_{\alpha\beta}D^\beta\psi.

Since covariant derivatives on a scalar function commute, this automatically ensures that the incompressibility condition is met. Decomposing ψ\psi in terms of eigenmodes of the Laplace Beltrami operator for the cylinder surface, we get

vα=εαβDβ(∫dΛ A(Λ)eiΛμxμ)(17)v_\alpha= \varepsilon_{\alpha\beta}D^\beta\left(\int d\Lambda\ A(\Lambda)e^{i\Lambda_\mu x^\mu}\right) \tag*{(17)}

where Λα≡(n,q≡kR)\Lambda_\alpha\equiv(n,q\equiv\frac{k}{R}) and xα≡(θ,z)x^\alpha\equiv(\theta,z) such that Λμxμ=nθ+kRz\Lambda_\mu x^\mu=n\theta+\frac{k}{R}z. The discrete index “n” appears in the mode decomposition due to the periodicity along the compact θ\theta direction of the membrane geometry. Note that ∫dΛ\int d\Lambda is shorthand for ∑n=−∞∞∫−∞∞dq\sum_{n=-\infty}^{\infty}\int_{-\infty}^{\infty}dq. The unknown coefficients A(Λ)A(\Lambda) will be solved below using the stress-balance condition Eq. (2) of main text. In the stress balance condition, we insert the external point torque σαext=τεαγDγδ2(θ−θ0,z−z0)\sigma^{\mathrm{ext}}_\alpha=\tau\varepsilon_{\alpha\gamma}D^\gamma\delta^2(\theta-\theta_0,z-z_0) and also take an anti-symmetric derivative on both sides of the equation to eliminate the membrane pressure p. The expression for the traction vector TαT_\alpha appearing on the RHS of the stress balance condition in Eq. (2) requires a straightforward but somewhat long computation, which we just summarize below.

Computation of TαT_\alpha: In order to compute the in-plane components of the traction vector TαT_\alpha we utilize the known solution for Stokes flow in 3D (Eq. (3) of main text) in cylindrical coordinates following Happel and Brenner Ref.[90]:

v⃗±(r,θ,z)=∇⃗f±(r,θ,z)+∇⃗×[g±(r,θ,z)z^]+r∂r[∇⃗h±(r,θ,z)]+∂zh±(r,θ,z)z^,\vec{v}^{\pm}(r,\theta,z)=\vec{\nabla}f^{\pm}(r,\theta,z)+\vec{\nabla}\times\left[g^{\pm}(r,\theta,z)\hat{z}\right]+r\partial_r\left[\vec{\nabla}h^{\pm}(r,\theta,z)\right]+\partial_z h^{\pm}(r,\theta,z)\hat{z},
p±(r,θ,z)=−2η±∂z2h±(r,θ,z).p^{\pm}(r,\theta,z)=-2\eta_{\pm}\partial_z^2h^{\pm}(r,\theta,z).

where f±f^{\pm}, g±g^{\pm}, h±h^{\pm} are harmonic functions of the Laplacian operator in 3D in cylindrical coordinates. Such harmonic functions on the cylinder admit a decomposition.

f±(r,θ,z)=∑n=−∞∞∫−∞∞dq eiΛμxμF±(Λ) ξn±(qr)f^{\pm}(r,\theta,z)=\sum_{n=-\infty}^{\infty}\int_{-\infty}^{\infty}dq\,e^{i\Lambda_{\mu}x^{\mu}}F^{\pm}(\Lambda)\,\xi_{n}^{\pm}(qr)

and likewise for g±g^{\pm}, h±h^{\pm}. In the above expression,

ξn+(qr)≡Kn(∣q∣r),ξn−(qr)≡In(∣q∣r)\xi_{n}^{+}(qr)\equiv K_{n}(|q|r),\qquad\xi_{n}^{-}(qr)\equiv I_{n}(|q|r)

where KnK_{n} and InI_{n} are modified Bessel functions of order nn of first and second kind, respectively. The covariant components of the external fluid velocities thus read as follows:

vr±(Λ,r)=∣q∣ξ~n±(qr)F±(Λ)+inrξn±(qr)G±(Λ)+H±(Λ)[−∣q∣ξ~n±(qr)+r(q2+n2r2)ξn±(qr)]v_{r}^{\pm}(\Lambda,r)=|q|\widetilde{\xi}_{n}^{\pm}(qr)F^{\pm}(\Lambda)+\frac{in}{r}\xi_{n}^{\pm}(qr)G^{\pm}(\Lambda)+H^{\pm}(\Lambda)\left[-|q|\widetilde{\xi}_{n}^{\pm}(qr)+r\left(q^{2}+\frac{n^{2}}{r^{2}}\right)\xi_{n}^{\pm}(qr)\right]
vθ±(Λ,r)=imξn±(qr)F±(Λ)−∣q∣rξn±(qr)G±(Λ)+inH±(Λ)[∣q∣rξ~n±(qr)−ξn±(qr)]v_{\theta}^{\pm}(\Lambda,r)=im\xi_{n}^{\pm}(qr)F^{\pm}(\Lambda)-|q|r\xi_{n}^{\pm}(qr)G^{\pm}(\Lambda)+inH^{\pm}(\Lambda)\left[|q|r\widetilde{\xi}_{n}^{\pm}(qr)-\xi_{n}^{\pm}(qr)\right]
vz±(Λ,r)=iqξn±(qr)F±(Λ)+iqH±(Λ)[∣q∣rξ~n±(qr)+ξn±(qr)]v_{z}^{\pm}(\Lambda,r)=iq\xi_{n}^{\pm}(qr)F^{\pm}(\Lambda)+iqH^{\pm}(\Lambda)\left[|q|r\widetilde{\xi}_{n}^{\pm}(qr)+\xi_{n}^{\pm}(qr)\right]

where

ξ~+(qr)≡dKn(u)du∣u=∣q∣r,ξ~−(qr)≡dIn(u)du∣u=∣q∣r\widetilde{\xi}^{+}(qr)\equiv\left.\frac{dK_{n}(u)}{du}\right|_{u=|q|r},\qquad\widetilde{\xi}^{-}(qr)\equiv\left.\frac{dI_{n}(u)}{du}\right|_{u=|q|r}

The 6 unknown coefficients F±F^{\pm}, G±G^{\pm}, H±H^{\pm} can be solved in terms of membrane mode coefficient AA using the following 6 equations derived from the no-slip boundary condition (Eq. (5) of main text):

vr±(Λ,R)=0,Λαvα±(Λ,R)=0v_{r}^{\pm}(\Lambda,R)=0,\qquad\Lambda^{\alpha}v_{\alpha}^{\pm}(\Lambda,R)=0
iϵαγΛγvα±(Λ,R)=−ΛβΛβA(Λ)i\epsilon^{\alpha\gamma}\Lambda_{\gamma}v_{\alpha}^{\pm}(\Lambda,R)=-\Lambda_{\beta}\Lambda^{\beta}A(\Lambda)

The solution for the 6 coefficients F±F^{\pm}, G±G^{\pm}, H±H^{\pm} is next obtained in Mathematica. Plugging in these solutions, the required antisymmetric derivative of the traction vector TαT_{\alpha} is given as follows:

ϵαγDγσrα±∣r=R=η±∫DΛ eiΛμxμA(Λ)R3C±(Λ)(18)\epsilon^{\alpha\gamma}D_{\gamma}\sigma_{r\alpha}^{\pm}\big|_{r=R}=\eta_{\pm}\int D\Lambda\,e^{i\Lambda_{\mu}x^{\mu}}\frac{A(\Lambda)}{R^{3}}C^{\pm}(\Lambda) \tag*{(18)}

where

C±(n,k)=2n2[ρ±(n,k)]3+(n2+k2)2[ρ±(n,k)]2+2ρ±(n,k)(k4−n4)−(k2+n2)3ρ±(n,k)k2−[ρ±(n,k)−n][ρ±(n,k)+n][ρ±(n,k)+2](19)C^{\pm}(n,k)=\frac{2n^{2}\left[\rho^{\pm}(n,k)\right]^{3}+\left(n^{2}+k^{2}\right)^{2}\left[\rho^{\pm}(n,k)\right]^{2}+2\rho^{\pm}(n,k)\left(k^{4}-n^{4}\right)-\left(k^{2}+n^{2}\right)^{3}}{\rho^{\pm}(n,k)k^{2}-\left[\rho^{\pm}(n,k)-n\right]\left[\rho^{\pm}(n,k)+n\right]\left[\rho^{\pm}(n,k)+2\right]} \tag*{(19)}

and

k=qR.ρ±(n,k)≡∣k∣ξ~n±(k)ξn±(k).(20)k = qR. \qquad\rho^{\pm}(n,k) \equiv\frac{\lvert k\rvert\tilde{\xi}^{\pm}_{n}(k)}{\xi^{\pm}_{n}(k)}. \tag*{(20)}

More explicitly,

ρ+[n,k]=∣k∣dKn[u]du∣u=∣k∣Kn[∣k∣],ρ−[n,k]=∣k∣dIn[u]du∣u=∣k∣In[∣k∣].(21)\rho_{+}[n,k] = \frac{\left.\lvert k\rvert\frac{dK_n[u]}{du}\right|_{u=\lvert k\rvert}}{K_n[\lvert k\rvert]}, \qquad\rho_{-}[n,k] = \frac{\left.\lvert k\rvert\frac{dI_n[u]}{du}\right|_{u=\lvert k\rvert}}{I_n[\lvert k\rvert]}. \tag*{(21)}

The fundamental equation of stress balance on the cylindrical membrane (after taking antisymmetric derivative of both sides of (2) of main text) thus reads as

−ϵαβDβ[τϵαγDγ(14π2R∫dΛ eiΛβ(x−x0)β⏟δ2(θ−θ0,z−z0))]=∫dΛ η2DA(Λ)R4Cn(k)⏟membrane stress + TractioneiΛαxα-\epsilon^{\alpha\beta}D_\beta\left[\tau\epsilon_{\alpha\gamma}D^\gamma\left(\underbrace{\frac{1}{4\pi^2R}\int d\Lambda\,e^{i\Lambda_\beta(x-x_0)^\beta}}_{\delta^2(\theta-\theta_0,z-z_0)}\right)\right]=\int d\Lambda\,\frac{\eta_{2D}A(\Lambda)}{R^4}\underbrace{C_n(k)}_{\text{membrane stress + Traction}}e^{i\Lambda_\alpha x^\alpha}

with

cn(k)=(n2+k2)2−Rλ−C−(n,k)+Rλ+C+(n,k)(22)c_n(k)=(n^2+k^2)^2-\frac{R}{\lambda_-}C^-(n,k)+\frac{R}{\lambda_+}C^+(n,k) \tag*{(22)}

We solve for A(Λ)A(\Lambda) from the above equation as

A(Λ)=τR4π2η2Dn2+k2cn(k)e−iΛαx0α(23)A(\Lambda)=\frac{\tau R}{4\pi^2\eta_{2D}}\frac{n^2+k^2}{c_n(k)}e^{-i\Lambda_\alpha x_0^\alpha} \tag*{(23)}

Plugging this solution of A(Λ)A(\Lambda) into (17), we get

vα=ϵαβDβ(∫dΛ τR4π2η2Dn2+k2cn(k)eiΛ⋅(x−x0)⏟≡ψ)(24)v_\alpha=\epsilon_{\alpha\beta}D^\beta\left(\underbrace{\int d\Lambda\,\frac{\tau R}{4\pi^2\eta_{2D}}\frac{n^2+k^2}{c_n(k)}e^{i\boldsymbol{\Lambda}\cdot(x-x_0)}}_{\equiv\psi}\right) \tag*{(24)}

where we can now identify the stream function of the flow field at (θ,z)(\theta,z) sourced by a vortex of strength τ\tau situated at the origin

ψ(θ,z)=τ4π2η2D∑n=−∞∞∫−∞∞dk n2+k2cn(k)ei(nθ+kRz).(25)\psi(\theta,z)=\frac{\tau}{4\pi^2\eta_{2D}}\sum_{n=-\infty}^{\infty}\int_{-\infty}^{\infty}dk\,\frac{n^2+k^2}{c_n(k)}e^{i\left(n\theta+\frac{k}{R}z\right)}. \tag*{(25)}

The vortex flow velocity is given in terms of ψ\psi as

vθ=∂zψv_\theta=\partial_z\psi
vz=−1R∂θψ(26)v_z=-\frac{1}{R}\partial_\theta\psi \tag*{(26)}

References

In the high curvature limit and assuming η−=η+≡η\eta_- = \eta_+ \equiv\eta,

lim⁡R/λ→0cn(k)={k4+2k2Rλn=0(k2+n2)2n≥1(27)\lim_{R/\lambda\to0} c_n(k) = \begin{cases} k^4 + 2k^2 \dfrac{R}{\lambda} & n=0 \\ \left(k^2+n^2\right)^2 & n\geq1 \end{cases} \tag*{(27)}

In this limit, we can perform the integrals and sums analytically, leading to the following stream function for a vortex of strength τ\tau located at the origin.

ψ(θ,z)=τ4πη2D(−log⁡[1−2e−∣z∣Rcos⁡θ+e−2∣z∣R]+λ2Re−2λR∣z∣)(28)\psi(\theta,z)=\frac{\tau}{4\pi\eta_{2D}}\left(-\log\left[1-2e^{-\frac{|z|}{R}}\cos\theta+e^{-\frac{2|z|}{R}}\right]+\sqrt{\frac{\lambda}{2R}}e^{-\sqrt{\frac{2}{\lambda R}}|z|}\right) \tag*{(28)}

which is the Eq. (7) of main text.

The corresponding vortex flow at the location (θ,z)(\theta,z) sourced by a vortex of strength τ\tau, situated at (θ0,z0)(\theta_0,z_0) is given by

vθ=τ4πη2D∣z−z0∣(2−2cos⁡(θ−θ0)e∣z−z0∣R1+2e∣z−z0∣R−2cos⁡(θ−θ0)e∣z−z0∣R−e−2Rλ∣z−z0∣(z−z0)R)v_\theta=\frac{\tau}{4\pi\eta_{2D}|z-z_0|}\left(\frac{2-2\cos(\theta-\theta_0)e^{\frac{|z-z_0|}{R}}}{1+2e^{\frac{|z-z_0|}{R}}-2\cos(\theta-\theta_0)e^{\frac{|z-z_0|}{R}}}-\frac{e^{-\sqrt{\frac{2}{R\lambda}}|z-z_0|}(z-z_0)}{R}\right)
vz=−τ4πη2DRsin⁡(θ−θ0)cos⁡(θ−θ0)−cosh⁡z−z0R(29)v_z=-\frac{\tau}{4\pi\eta_{2D}R}\frac{\sin(\theta-\theta_0)}{\cos(\theta-\theta_0)-\cosh\frac{z-z_0}{R}} \tag*{(29)}

which is the Eq. (8) of main text. Note that in the limit z→z0z\to z_0, the vortex flow Eq. (29) simplifies to

lim⁡z→z0vθ=0\lim_{z\to z_0}v_\theta=0
lim⁡z→z0vz=τcot⁡(θ−θ02)4πRη2D(30)\lim_{z\to z_0}v_z=\frac{\tau\cot\left(\frac{\theta-\theta_0}{2}\right)}{4\pi R\eta_{2D}} \tag*{(30)}

and in the limit θ→θ0\theta\to\theta_0, the vortex flow Eq. (29) simplifies to

lim⁡θ→θ0vθ=−τ(z−z0)4πRη2D∣z−z0∣(e−2Rλ∣z−z0∣+2−1+e∣z−z0∣R)\lim_{\theta\to\theta_0}v_\theta=-\frac{\tau(z-z_0)}{4\pi R\eta_{2D}|z-z_0|}\left(e^{-\sqrt{\frac{2}{R\lambda}}|z-z_0|}+\frac{2}{-1+e^{\frac{|z-z_0|}{R}}}\right)
lim⁡θ→θ0vz=0(31)\lim_{\theta\to\theta_0}v_z=0 \tag*{(31)}

Data Availability The data that supports the findings of this study are available within the article and its supplementary material.

References

  1. [2]Yegor A. Domanov, Sophie Aimon, Gilman E. S. Toombes, Marianne Renner, François Quemeneur, Antoine Triller, Matthew S. Turner, and Patricia Bassereau. Mobility in geometrically confined membranes. Proceedings of the National Academy of Sciences, 108(31):12605–12610, 2011.DOI
  2. [3]S. J. Singer and Garth L. Nicolson. The fluid mosaic model of the structure of cell membranes. Science, 175(4023):720–731, 1972.pubmed.ncbi.nlm.nih.gov/4333397
  3. [4]E. M. Purcell. Life at low reynolds number. American Journal of Physics, 45(1):3–11, 01 1977.
  4. [5]P. G. Saffman. Brownian motion in thin sheets of viscous fluid. Journal of Fluid Mechanics, 73(4):593–602, 1976.DOI
  5. [6]P G Saffman and M Delbrück. Brownian motion in biological membranes. Proceedings of the National Academy of Sciences, 72(8):3111–3113, 1975.pubmed.ncbi.nlm.nih.gov/1059096
  6. [7]B. D. Hughes, B. A. Pailthorpe, and L. R. White. The translational and rotational drag on a cylinder moving in a membrane. Journal of Fluid Mechanics, 110:349–372, 1981.DOI
  7. [8]Evan Evans and Erich Sackmann. Translational and rotational drag coefficients for a disk moving in a liquid membrane associated with a rigid substrate. Journal of Fluid Mechanics, 194:553–561, 1988.DOI
  8. [9]David K. Lubensky and Raymond E. Goldstein. Hydrodynamics of monolayer domains at the air–water interface. Physics of Fluids, 8(4):843–854, 04 1996.DOI
  9. [10]Howard A. Stone and Armand Ajdari. Hydrodynamics of particles embedded in a flat surfactant layer overlying a subphase of finite depth. Journal of Fluid Mechanics, 369:151–173, 1998.DOI
  10. [11]TH. M. FISCHER. The drag on needles moving in a langmuir monolayer. Journal of Fluid Mechanics, 498:123–137, 2004.DOI
  11. [12]Ali Naji, Alex J Levine, and PA Pincus. Corrections to the saffman-delbruck mobility for membrane bound proteins. Biophysical journal, 93(11):L49–51, December 2007.DOI
  12. [13]Naomi Oppenheimer and Haim Diamant. Correlated diffusion of membrane proteins and their effect on membrane viscosity. Biophysical Journal, 96(8):3041–3049, 2009.DOI
  13. [14]Naomi Oppenheimer and Haim Diamant. Correlated dynamics of inclusions in a supported membrane. Phys. Rev. E, 82:041912, Oct 2010.DOI
  14. [15]Naomi Oppenheimer and Haim Diamant. In-plane dynamics of membranes with immobile inclusions. Phys. Rev. Lett., 107:258102, Dec 2011.
  15. [16]D. R. Daniels and M. S. Turner. Diffusion on membrane tubes: A highly discriminatory test of the saffman-delbrueck theory. Langmuir, 23(12):6667–6670, 2007. PMID: 17489611.pubmed.ncbi.nlm.nih.gov/17489611
  16. [17]M. L. Henle, R. McGorty, A. B. Schofield, A. D. Dinsmore, and A. J. Levine. The effect of curvature and topology on membrane hydrodynamics. Europhysics Letters, 84(4):48001, nov 2008.DOI
  17. [18]Mark L. Henle and Alex J. Levine. Hydrodynamics in curved membranes: The effect of geometry on particulate mobility. Phys. Rev. E, 81:011905, Jan 2010.pubmed.ncbi.nlm.nih.gov/20365397
  18. [19]Francis G. Woodhouse and Raymond E. Goldstein. Shear-driven circulation patterns in lipid membrane vesicles. Journal of Fluid Mechanics, 705:165–175, 2012.
  19. [20]Aurelia R. Honerkamp-Smith, Francis G. Woodhouse, Vasily Kantsler, and Raymond E. Goldstein. Membrane viscosity determined from shear-driven flow in giant vesicles. Phys. Rev. Lett., 111:038103, Jul 2013.
  20. [21]François Quemeneur, Jon K. Sigurdsson, Marianne Renner, Paul J. Atzberger, Patricia Bassereau, and David Lacoste. Shape matters in protein mobility within membranes. Proceedings of the National Academy of Sciences, 111(14):5083–5087, 2014.DOI
  21. [22]D. Daniels. Curvature correction to the mobility of fluid membrane inclusions. The European Physical Journal E, 39(10):96, 10 2016.pubmed.ncbi.nlm.nih.gov/27771858
  22. [23]J. K. Sigurdsson and P. J. Atzberger. Hydrodynamic coupling of particle inclusions embedded in curved lipid bilayer membranes. Soft Matter, 12:6685–6707, 2016.DOI
  23. [24]B.J. Gross and P.J. Atzberger. Hydrodynamic flows on curved surfaces: Spectral numerical methods for radial manifold shapes. Journal of Computational Physics, 371:663–689, 2018.DOI
  24. [25]David A. Rower, Misha Padidar, and Paul J. Atzberger. Surface fluctuating hydrodynamics methods for the drift-diffusion dynamics of particles and microstructures within curved fluid interfaces. Journal of Computational Physics, 455:110994, 2022.DOI
  25. [26]Amaresh Sahu, Yannick A.D. Omar, Roger A. Sauer, and Kranthi K. Mandadapu. Arbitrary lagrangian–eulerian finite element method for curved and deforming surfaces: I. general theory and application to fluid interfaces. Journal of Computational Physics, 407:109253, 2020.DOI
  26. [27]R. Samanta and N. Oppenheimer. Vortex flows and streamline topology in curved biological membranes. Physics of Fluids, 33(5):051906, 05 2021.DOI
  27. [28]Patrick D. Tran, Thomas A. Blanpied, and Paul J. Atzberger. Protein drift-diffusion dy- namics and phase separation in curved cell membranes and dendritic spines: Hybrid discrete-continuum methods. Phys. Rev. E, 106:044402, Oct 2022.
  28. [29]Sarthak Bagaria and Rickmoy Samanta. Dynamics of force dipoles in curved fluid membranes. Phys. Rev. Fluids, 7:093101, Sep 2022.DOI
  29. [30]Samyak Jain and Rickmoy Samanta. Force dipole interactions in tubular fluid membranes. Physics of Fluids, 35(7):071901, 07 2023.DOI
  30. [31]Wenzheng Shi, Moslem Moradi, and Ehssan Nazockdast. Hydrodynamics of a single filament moving in a spherical membrane. Phys. Rev. Fluids, 7:084004, Aug 2022.DOI
  31. [32]Eric Climent, Kyongmin Yeo, Martin R. Maxey, and George E. Karniadakis. Dynamic self-assembly of spinning particles. Journal of Fluids Engineering, 1(4):379–387, 2007.DOI
  32. [33]Debasish Das and David Saintillan. Electrohydrodynamic interaction of spherical particles under quincke rotation. Phys. Rev. E, 87:043014, Apr 2013.pubmed.ncbi.nlm.nih.gov/23679520
  33. [34]Yaouen Fily, Aparna Baskaran, and M. Cristina Marchetti. Cooperative self-propulsion of active and passive rotors. Soft Matter, 8(10):3002–3009, 2012.DOI
  34. [35]M. E. J. Friese, T. A. Nieminen, N. R. Heckenberg, and H. Rubinsztein-Dunlop. Optical alignment and spinning of laser-trapped microscopic particles. Nature, 394(6691):348–350, July 1998.DOI
  35. [36]S. Fürthauer, M. Strempel, S. W. Grill, and F. Jülicher. Active chiral processes in thin films. Phys. Rev. Lett., 110:048103, Jan 2013.pubmed.ncbi.nlm.nih.gov/25166204
  36. [37]Bartosz A. Grzybowski, Howard A. Stone, and George M. Whitesides. Dynamic self-assembly of magnetized, millimetre-sized objects rotating at a liquid-air interface. Nature, 405(6790):1033–1036, Jun 2000.pubmed.ncbi.nlm.nih.gov/10890439
  37. [38]Bartosz A. Grzybowski, Xingyu Jiang, Howard A. Stone, and George M. Whitesides. Dynamic, self-assembled aggregates of magnetized, millimeter-sized objects rotating at the liquid-air interface: Macroscopic, two-dimensional classical artificial atoms and molecules. Phys. Rev. E, 64:011603, Jun 2001.pubmed.ncbi.nlm.nih.gov/11461266
  38. [39]Bartosz A. Grzybowski and George M. Whitesides. Dynamic aggregation of chiral spinners. Science, 296(5568):718–721, 2002.pubmed.ncbi.nlm.nih.gov/11976449
  39. [40]M. Leoni and T. B. Liverpool. Dynamics and interactions of active rotors. Europhysics Letters, 92(6):64004, jan 2011.
  40. [41]I. Llopis and I. Pagonabarraga. Hydrodynamic regimes of active rotors at fluid interfaces. The European Physical Journal E, 26(1):103–113, May 2008.
  41. [42]Enkeleida Lushi and Petia M. Vlahovska. Periodic and chaotic orbits of plane-confined micro-rotors in creeping flows. Journal of Nonlinear Science, 25(5):1111–1123, April 2015.DOI
  42. [43]Kyongmin Yeo, Enkeleida Lushi, and Petia M. Vlahovska. Collective dynamics in a binary mixture of hydrodynamically coupled microrotors. Phys. Rev. Lett., 114:188301, May 2015.pubmed.ncbi.nlm.nih.gov/26001020
  43. [44]Alvin Modin, Matan Yah Ben Zion, and Paul M. Chaikin. Hydrodynamic spin-orbit coupling in asynchronous optically driven micro-rotors. Nature Communications, 14(1):4114, Jul 2023.DOI
  44. [45]Peter Lenz, Jean-François Joanny, Frank Jülicher, and Jacques Prost. Membranes with rotating motors. Phys. Rev. Lett., 91:108104, Sep 2003.
  45. [46]P. Lenz, J.-F. Joanny, F. Jülicher, and J. Prost. Membranes with rotating motors: Microvortex assemblies. The European Physical Journal E, 13(4):379–390, Apr 2004.pubmed.ncbi.nlm.nih.gov/15170537
  46. [47]Naomi Oppenheimer, David B. Stein, and Michael J. Shelley. Rotating membrane inclusions crystallize through hydrodynamic and steric interactions. Phys. Rev. Lett., 123:148101, Oct 2019.DOI
  47. [48]Naomi Oppenheimer, David B. Stein, Matan Yah Ben Zion, and Michael J. Shelley. Hyperuniformity and phase enrichment in vortex and rotor assemblies. Nature Communications, 13(1), February 2022.DOI
  48. [49]Imre Derényi, Frank Jülicher, and Jacques Prost. Formation and interaction of membrane tubes. Phys. Rev. Lett., 88:238101, May 2002.DOI
  49. [50]B. Božić, V. Heinrich, S. Svetina, and B. Žekš. Shapes of nearly cylindrical, axisymmetric bilayer membranes. Eur. Phys. J. E, 6(1):91–98, 2001.
  50. [51]Dirk Jan Bukman, Jian Hua Yao, and Michael Wortis. Stability of cylindrical vesicles under axial tension. Phys. Rev. E, 54:5463–5468, Nov 1996.pubmed.ncbi.nlm.nih.gov/9965732
  51. [52]Volkmar Heinrich, Bojan Božić, Saša Svetina, and Boštjan Žekš. Vesicle deformation by an axial load: From elongated shapes to tethered vesicles. Biophysical Journal, 76(4):2056–2071, 1999.pubmed.ncbi.nlm.nih.gov/10096901
  52. [53]Shao-Guang Zhang. Instability of circular cylinders and polygonal solutions for cylindrical vesicles. Journal of the Physical Society of Japan, 68(11):3603–3607, 1999.DOI
  53. [54]Ana-Sun Ćana Smith, Erich Sackmann, and Udo Seifert. Pulling tethers from adhered vesicles. Phys. Rev. Lett., 92:208101, May 2004.pubmed.ncbi.nlm.nih.gov/15169384
  54. [55]Thomas R. Powers, Greg Huber, and Raymond E. Goldstein. Fluid-membrane tethers: Min- imal surfaces and elastic boundary layers. Phys. Rev. E, 65:041901, Mar 2002.DOI
  55. [56]Drazen Raucher and Michael P. Sheetz. Characteristics of a membrane reservoir buffering membrane tension. Biophysical Journal, 77(4):1992–2002, 1999.pubmed.ncbi.nlm.nih.gov/10512819
  56. [57]Deborah Kuchnir Fygenson, John F. Marko, and Albert Libchaber. Mechanics of microtubule-based membrane extension. Phys. Rev. Lett., 79:4497–4500, Dec 1997.DOI
  57. [58]D. R. Daniels and M. S. Turner. Spicules and the effect of rigid rods on enclosing membrane tubes. Phys. Rev. Lett., 95:238101, Dec 2005.DOI
  58. [59]F. Brochard-Wyart, N. Borghi, D. Cuvellier, and P. Nassoy. Hydrodynamic narrowing of tubes extruded from cells. Proceedings of the National Academy of Sciences, 103(20):7660–7663, 2006.
  59. [60]David Boal. Mechanics of the Cell. Cambridge University Press, 2001.
  60. [61]Bruce Alberts, Alexander Johnson, Julian Lewis, Martin Raff, Keith Roberts, and Peter Walter. Molecular Biology of the Cell, 4th edition. Garland Science, 2002.
  61. [62]D.A. Berk and R.M. Hochmuth. Lateral mobility of integral proteins in red blood cell tethers. Biophysical Journal, 61(1):9–18, 1992.pubmed.ncbi.nlm.nih.gov/1540701
  62. [63]Aurélien Roux, Damien Cuvelier, Pierre Nassoy, Jacques Prost, Patricia Bassereau, and Bruno Goud. Role of curvature and phase transition in lipid sorting and fission of membrane tubules. The EMBO Journal, 24(8):1537–1545, 2005.pubmed.ncbi.nlm.nih.gov/15791208
  63. [64]Aurélien Roux, Giovanni Cappello, Jean Cartaud, Jacques Prost, Bruno Goud, and Patricia Bassereau. A minimal system allowing tubulation with molecular motors pulling on giant liposomes. Proceedings of the National Academy of Sciences, 99(8):5394–5399, 2002.pubmed.ncbi.nlm.nih.gov/11959994
  64. [65]Gerbrand Koster, Martijn VanDuijn, Bas Hofs, and Marileen Dogterom. Membrane tube formation from giant vesicles by dynamic association of motor proteins. Proceedings of the National Academy of Sciences, 100(26):15583–15588, 2003.pubmed.ncbi.nlm.nih.gov/14663143
  65. [66]Joseph E. Reiner, Jeffrey M. Wells, Rani B. Kishore, Candace Pfefferkorn, and Kristian Helmerson. Stable and robust polymer nanotubes stretched from polymersomes. Proceedings of the National Academy of Sciences, 103(5):1173–1177, 2006.pubmed.ncbi.nlm.nih.gov/16432242
  66. [67]Amin Rustom, Rainer Saffrich, Ivanka Markovic, Paul Walther, and Hans-Hermann Gerdes. Nanotubular highways for intercellular organelle transport. Science, 303(5660):1007–1010, 2004.pubmed.ncbi.nlm.nih.gov/14963329
  67. [68]Joel M. Schnur. Lipid tubules: A paradigm for molecularly engineered structures. Science, 262(5140):1669–1676, 1993.pubmed.ncbi.nlm.nih.gov/17781785
  68. [69]Evan Evans, Howard Bowman, Andrew Leung, David Needham, and David Tirrell. Biomembrane templates for nanoscale conduits and networks. Science, 273(5277):933–935, 1996.pubmed.ncbi.nlm.nih.gov/8688071
  69. [70]T. Roopa and G. V. Shivashankar. Nanomechanics of membrane tubulation and dna assembly. Applied Physics Letters, 82(10):1631–1633, 03 2003.DOI
  70. [71]Arpita Upadhyaya and Michael P. Sheetz. Tension in tubulovesicular networks of golgi and endoplasmic reticulum membranes. Biophysical Journal, 86(5):2923–2928, May 2004.pubmed.ncbi.nlm.nih.gov/15111408
  71. [72]Tom Shemesh, Alberto Luini, Vivek Malhotra, Koert N.J. Burger, and Michael M. Kozlov. Prefission constriction of golgi tubular carriers driven by local lipid metabolism: A theoretical model. Biophysical Journal, 85(6):3813–3827, Dec 2003.
  72. [73]Noah Sciaky, John Presley, Carolyn Smith, Kristien J.M. Zaal, Nelson Cole, Jorge E. Moreira, Mark Terasaki, Eric Siggia, and Jennifer Lippincott-Schwartz. Golgi tubule traffic and the effects of brefeldin a visualized in living cells. Journal of Cell Biology, 139(5):1137–1155, 12 1997.pubmed.ncbi.nlm.nih.gov/9382862
  73. [74]Peter Watson and David J. Stephens. Er-to-golgi transport: Form and formation of vesicular and tubular carriers. Biochimica et Biophysica Acta (BBA) - Molecular Cell Research, 1744(3):304–315, 2005. The Golgi Complex.pubmed.ncbi.nlm.nih.gov/15979504
  74. [75]Mark S. Ladinsky, David N. Mastronarde, J. Richard McIntosh, Kathryn E. Howell, and L. Andrew Staehelin. Golgi structure in three dimensions: Functional insights from the normal rat kidney cell. Journal of Cell Biology, 144(6):1135–1149, 03 1999.pubmed.ncbi.nlm.nih.gov/10087259
  75. [76]Nelson B. Cole, Carolyn L. Smith, Noah Sciaky, Mark Terasaki, Michael Edidin, and Jennifer Lippincott-Schwartz. Diffusional mobility of golgi proteins in membranes of living cells. Science, 273(5276):797–801, 1996.pubmed.ncbi.nlm.nih.gov/8670420
  76. [77]Matthias Weiss, Hitoshi Hashimoto, and Tommy Nilsson. Anomalous protein diffusion in living cells as seen by fluorescence correlation spectroscopy. Biophysical Journal, 84(6):4043–4052, Jun 2003.pubmed.ncbi.nlm.nih.gov/12770908
  77. [78]Ivo F. Sbalzarini, Anna Mezzacasa, Ari Helenius, and Petros Koumoutsakos. Effects of organelle shape on fluorescence recovery after photobleaching. Biophysical Journal, 89(3):1482–1492, Sep 2005.pubmed.ncbi.nlm.nih.gov/15951382
  78. [79]Matthias Weiss and Tommy Nilsson. In a mirror dimly: tracing the movements of molecules in living cells. Trends in Cell Biology, 14(5):267–273, May 2004.pubmed.ncbi.nlm.nih.gov/15130583
  79. [80]Fabio Demontis. Nanotubes make big science. PLOS Biology, 2(7), 07 2004.pubmed.ncbi.nlm.nih.gov/15252458
  80. [81]Jan Ellenberg, Eric D. Siggia, Jorge E. Moreira, Carolyn L. Smith, John F. Presley, Howard J. Worman, and Jennifer Lippincott-Schwartz. Nuclear membrane dynamics and reassembly in living cells: Targeting of an inner nuclear membrane protein in interphase and mitosis. Journal of Cell Biology, 138(6):1193–1206, 09 1997.pubmed.ncbi.nlm.nih.gov/9298976
  81. [82]Alexander Hahn, Janet Vonck, Deryck J. Mills, Thomas Meier, and Werner Kühlbrandt. Structure, mechanism, and regulation of the chloroplast atp synthase. Science, 360(6389):eaat4318, 2018.
  82. [83]Jan Pieter Abrahams, Andrew G. W. Leslie, René Lutter, and John E. Walker. Structure at 2.8 Å resolution of f1-atpase from bovine heart mitochondria. Nature, 370(6491):621–628, Aug 1994.pubmed.ncbi.nlm.nih.gov/8065448
  83. [84]Daniela Stock, Clyde Gibbons, Ignacio Arechaga, Andrew GW Leslie, and John E Walker. The rotary mechanism of atp synthase. Current Opinion in Structural Biology, 10(6):672–679, 2000.pubmed.ncbi.nlm.nih.gov/11114504
  84. [85]Paul K. Newton. The N-Vortex Problem - Analytical Techniques. Springer New York, 2001.DOI
  85. [86]Q. Le Thien, D. McDermott, C. J. O. Reichhardt, and C. Reichhardt. Enhanced pinning for vortices in hyperuniform pinning arrays and emergent hyperuniform vortex configurations with quenched disorder. Physical Review B, 96(9), September 2017.DOI
  86. [87]James Montaldi, Anik Soulière, and Tadashi Tokieda. Vortex dynamics on a cylinder. SIAM Journal on Applied Dynamical Systems, 2(3):417–430, January 2003.DOI
  87. [88]Jaime Andrade, Stefanella Boatto, Thierry Combot, Gladston Duarte, and Teresinha J. Stuchi. N-body dynamics on an infinite cylinder: the topological signature in the dynamics. Regular and Chaotic Dynamics, 25(1):78–110, Jan 2020.DOI
  88. [89]Yulia Sokolov and Haim Diamant. Many-particle mobility and diffusion tensors for objects in viscous sheets. The Journal of Chemical Physics, 149(3):034901, 07 2018.DOI
  89. [90]John Happel and Howard Brenner. Low Reynolds number hydrodynamics with special applications to particulate media. Springer Dordrecht, 1983.DOI

Paper details

vlnc:W_xnn3fbdw
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 8

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents