Unlike inanimate physical and chemical media, living systems feature a particular degree of complexity, hierarchical organization, and evolutionary structure. This poses unique challenges to the way in which we rationalize and systematically phrase our understanding of these systems, that is, how we set up models (in the broadest sense) of these complex living systems. Ideally, we would like such models to represent mechanistic reasoning (explanation) of the systems in a way that allows us to experimentally test predictions about the system’s behavior. For many biological questions, qualitative flowchart-type descriptions suffice (if gene X is knocked out, protein Y is not available any longer, which has this or that effect). However, in many situations, observations cannot be addressed by verbal arguments or cartoon-type approaches alone, and then, mathematical models are needed (Baker et al., 2008; Segel and Edelstein-Keshet, 2013; Ganusov, 2016; Fang et al., 2019). This applies not only to most quantitative questions but also to many qualitative features because typically the responses of biological systems to external changes or perturbations do not occur proportionally (linearly) but rather exhibit a more complex (nonlinear) dependence that is often unexpected.

Prominent examples are collective effects, where ordered large-scale patterns result from interactions and correlations of large numbers of molecular or nano-scale components of the system. The key characteristics of such structures are emergent properties in the sense that they do not reflect the properties of the individual building blocks in a linear, additive sense, but they are genuine ensemble properties that arise from the collective interactions of many constituents, that is, they only exist when large numbers of building blocks come together (Murray, 2001; Baker et al., 2008; Champneys et al., 2021). This may be seen in multicellular systems but also on the subcellular level (Edelstein-Keshet, 2005; Alon, 2006; Keener and Sneyd, 2009; Beta and Kruse, 2017; Halatek et al., 2018). Paradigmatic examples that have attracted much attention over the past decade are intracellular patterns that can be observed in the actin cytoskeleton of eukaryotic cells. From the collective action of large numbers of polymerizing actin filaments, motor proteins, and regulatory signaling components (Machesky and Insall, 1999), different macro-scale functional structures emerge that drive essential processes of cellular life (Kabsch and Vandekerckhove, 1992), such as motility, division, and nutrient uptake. Among them, wave patterns in the cell cortex have been observed as a common motif across many different cell types, such as human neutrophils, Xenopus laevis oocytes, and Dictyostelium discoideum. They have been intensely studied and became the target of many modeling efforts (Carlsson, 2010; Allard and Mogilner, 2013; Inagaki and Katsuno, 2017; Beta et al., 2020; Michaud et al., 2021; Povarova et al., 2023). Nevertheless, key questions remain open. For example, it is unclear to what extent the underlying mechanisms and the biological functions of actin waves are similar across different cell types, which of the wave characteristics are essential, and which are merely side effects of other, more essential, processes.

These questions pose many challenges to experimental research that can only be addressed together with theoretical methodologies. Due to the inherent complexity, models of actin waves typically lean on a reductionist approach (Carlsson, 2010), where the key players and their interactions are determined and cast into equations that recover the essential phenomena. However, as for many biological systems, it is not straightforward to identify a reduced model of actin waves. Their dynamics rely on dozens of molecular players (Kabsch and Vandekerckhove, 1992; Dominguez and Holmes, 2011), connected in a complex interaction network that is only partially known, it includes both chemical and mechanical factors, and features multiple feedback loops. Identifying the core molecular components that are essential for the formation of actin waves is a major outstanding challenge. Consequently, it often remains uncertain whether models of actin waves correctly represent the underlying mechanism, even if numerical simulations of the model reproduce the experimental behaviors.

Here, mathematics and, in particular, bifurcation theory can provide useful means to design and validate models, especially after simplifications have been introduced. Similar to the laws of thermodynamics which all physical systems have to obey, bifurcation theory imposes constraints on the form of a dynamic system that exhibits specific qualitative properties, such as excitability or oscillations, and explains their emergence, evolution, and persistence. It defines the different options and minimal mathematical features required to recover these properties, independent of the specific details of the model. The most notable example is the simplification of the Hodgkin–Huxley model of neuronal excitation to the phenomenological FitzHugh–Nagumo model (Ermentrout and Terman, 2010). Hodgkin and Huxley, 1952 proposed a seminal four-variable model based on voltage-gated ion channels to describe the propagation of an action potential along the membrane. But while the model captures realistic kinetics, the analysis and, thus, understanding its robustness is difficult. Instead, our understanding of many of the qualitative dynamic properties has been advanced through the analysis of the two-variable FitzHugh–Nagumo model (Fitzhugh, 1961; Nagumo et al., 1962), where one of the variables behaves as a switch (or excitable element) activating a second variable that damps the excitation via a slow negative feedback. While oversimplified from a physiological standpoint, the FitzHugh–Nagumo model is amenable to analysis and has proven valuable to understand the basic mechanisms of excitable and oscillatory dynamics, in general (Tyson and Keener, 1988; Meron, 1992). For that reason, models such as the FitzHugh–Nagumo model are frequently denoted as ‘toy models.’ A theoretical approach relying on reduced models is important for mapping out the range of possible behaviors, and for elucidating the dependence of simulated patterns on the chosen assumptions, details, and parameter regimes of a given model; in other words, for obtaining reliable conclusions and predictions that go beyond model-specific statements (Cross and Hohenberg, 1993; Champneys et al., 2021). It allows us to uncover model-independent ‘anchors’ around which specific model descriptions can be developed and, as such, provide important quality control and guideline for a reductionist model design.

In this review, we focus on actin waves to exemplify this approach. An analogy to electrophysiological waves in neurons and cardiac cells suggests that universal principles may also be inherent in self-organized cortical actin waves (Karma, 2013; Alonso et al., 2016). In addition, an increasing body of experimental results became available in recent years, which has prompted different modeling groups to focus on different specific experimental observations. We first summarize the current state of experimental research on intracellular actin wave patterns and then review the different modeling approaches that have been proposed to describe these phenomena. At the end, we provide a brief, non-technical introduction to the ideas of bifurcation theory and illustrate how these mathematical tools may be beneficial for guiding future modeling in this field.

Actin waves are micron-scale cytoskeletal regions of increased filamentous actin density that propagate in a wave-like fashion in the actin cortex (Carlsson, 2010). They are typically also enriched in other actin-related proteins and regulatory components and may be associated with localized signaling processes at the adjacent plasma membrane. Actin waves have been observed in many different cell types, including human neutrophils (Weiner et al., 2007), dendritic cells (Stankevicins et al., 2020), neurons (Winans et al., 2016), T cells (Lam Hui et al., 2014), and breast cancer cells (Marchesin et al., 2015; Zhan et al., 2020). Besides examples from Caenorhabditis elegans (Michaux et al., 2018), actin waves have been particularly well studied in D. discoideum (Vicker et al., 1997; Gerisch et al., 2004; Arai et al., 2010) and X. laevis oocytes (Bement et al., 2015; Michaud et al., 2021; see also Figure 1 for examples from various species). Their size, morphology, speed of propagation, and biochemical composition may vary between different cell types. Yet, despite these diverse appearances, it is not clear how many truly distinct mechanisms give rise to actin waves in various cell types, nor how many variations of each underlying mechanism exist. The emergence of actin waves has been related to different cellular functions, such as motility, division, and phagocytosis (Carlsson, 2010; Inagaki and Katsuno, 2017). But also here, functional roles may vary between cell types. In some cases, it is unclear whether the wave patterns necessarily have a specific function or whether the waves are by-products of signaling dynamics or ever-changing cell morphology.

Figure 1

Download asset Open asset

Wave-like dynamics of actin structures were reported already several decades ago, starting with circular dorsal ruffles (CDRs) in the 1970s (Abercrombie et al., 1970) in various cell types, such as fibroblasts and glial cells, in response to receptor-tyrosine-kinase growth factors (Buccione et al., 2004). CDRs are transient actin-based structures that consist of a dynamic ring-shaped ridge expanding and contracting across the dorsal plasma membrane in a meandering, wave-like fashion (Chhabra and Higgs, 2007; Swanson, 2008). A different but potentially related wave phenomenon was observed in the mid-1990s, where protrusions of the cell border were reported that move along the periphery of D. discoideum cells and human epidermal keratinocytes (Killich et al., 1993; Alt et al., 1995). With the rapid improvements in live-cell fluorescence microscopy, these findings were soon complemented with real-time observations of wave dynamics at the cytoskeletal level relying on fluorescent markers of actin and actin-related proteins in D. discoideum (Vicker et al., 1997; Vicker, 2002b; Gerisch et al., 2004; Bretschneider et al., 2004) and in other cell types (Vicker, 2002a; Weiner et al., 2007). They triggered growing interest in the phenomenon of actin waves, resulting in a boost of experimental and theoretical studies over the following decades.

A strong focus of the earlier work in this field was on periodic and wave-like morphodynamics of the cell outline (Verkhovsky, 2015). In spreading mouse embryonic fibroblasts, periodic contractions of lamellipodia were found to correlate with rearward moving actin waves (Giannone et al., 2004; Giannone et al., 2007). Lamellipodial protrusions and retractions may also travel laterally along the membrane (Döbereiner et al., 2006; Machacek and Danuser, 2006), as observed in fish epithelial keratocytes (Barnhart et al., 2011; Lou et al., 2015; Barnhart et al., 2017), D. discoideum (Driscoll et al., 2012; Driscoll et al., 2015), and Xenopus carcass fibroblasts (Ryan et al., 2012a). Besides wave dynamics at the leading edge of spreading and migrating cells, wave-like structures were also reported from cultured neurons that may exhibit actin-based, fin-like membrane protrusions traveling along their thin axonal extensions (Ruthel and Banker, 1998; Toriyama et al., 2006; Tomba et al., 2017). However, fin-like actin waves are not restricted to neural cells but may also occur in other cell types when cultured on thin fibers (Guetta-Terrier et al., 2015).

Wave dynamics is not only observed at the cell border. Traveling actin waves also emerge at the ventral and dorsal cell surfaces. For example, in human neutrophils, propagating waves of Hem-1, a member of the WAVE complex that regulates Arp2/3 activity, together with actin assembly were observed (Weiner et al., 2007; Millius et al., 2009). These waves show signatures of excitable behavior and organize the protrusion dynamics at the leading edge of migrating neutrophils. Similarly, the amoeboid motion of dendritic cells that mediate immune responses to various pathogens is affected by the presence of actin waves that can induce switches between diffusive and persistent states of motion (Stankevicins et al., 2020). Ventral actin waves can also associate with integrin-mediated adhesions to form so-called adhesive F-actin waves, as was observed in several mammalian cell lines (Case and Waterman, 2011). However, in many cases it is not clear whether the waves are generated at the cytoskeletal level or merely reflect wave dynamics that emerges in the upstream signaling network (see Box 1 for an overview of some of the major signaling components regulating actin assembly). In rat basophilic leukemia mast cells, for example, calcium oscillations may change the dynamic character of the actin wave patterns (Wu et al., 2013) and distinct phosphoinositides control the frequency and amplitude of cortical oscillations (Xiong et al., 2016). In oocytes and embryonic cells of frogs and echinoderms, Rho signaling activation together with inactivation by the GAP RGA3/4 (a GTPase-activating protein recruited by F-actin) result in excitable waves (Bement et al., 2015) that were successfully reconstituted also in an in vitro system (Landino et al., 2021). Important functions during cell division, such as division plane placement, have been attributed to this form of ‘cortical excitability’ (Michaud et al., 2021) and to the onset of excitable behavior (Swider et al., 2022; Michaud et al., 2022).

Box 1

Common regulators of actin assembly

Box 1—figure 1 represents some of the most common regulators of actin polymerization (found in many eukaryotic cells) and a few of the feedback interactions identified experimentally in various species.

The assembly of filamentous actin (F-actin, branched structures) takes place close to the cell membrane. Membrane-bound complexes (WASP or WAVE, yellow ovals) activate a protein that forms the actin branch-points (Arp2/3, blue dots; Svitkina and Borisy, 1999; Blanchoin et al., 2000). Other membrane binding proteins such as formins (e.g., mDia, not shown) also lead to growth of F-actin strands. The actin filament tips (‘barbed ends’) exert force on the cell membrane (Mogilner and Oster, 2003), leading to protrusion of the cell edge. F-actin assembly is promoted by GTPases such as Rac and Cdc42 (which activate WASP, WAVE) and by Rho (which activates mDia; Watanabe et al., 1997).

In D. discoideum and neutrophils, branching nucleation dominates at the cell front (Glogauer et al., 2000; Welch and Mullins, 2002). Various regulators, including Hem-1 (a membrane-resident part of the WAVE2 complex) in neutrophils (Weiner et al., 2007), participate in feedback loops activating Arp2/3. In many mammalian cells, GTPases such as Rac, Cdc42, and Rho are central to cell polarity, motility, and F-actin dynamics (Nobes and Hall, 1995; Etienne-Manneville and Hall, 2002). In D. discoideum, the GTPase Ras plays a similar role (Kortholt and van Haastert, 2008). GTPases are molecular switches, with an active, membrane-bound ‘ON’ form (dark green octagon), and an inactive cytosolic ‘OFF’ form (light green). Only the active form has downstream action on other effectors. GTPases are activated by guanine nucleotide exchange factors (GEFs) and inactivated by GTPase-activating proteins (GAPs) (Hodge and Ridley, 2016). In some systems, including embryos of C. elegans, Xenopus, and starfish, F-actin is assembled by mDia, which is activated by Rho (Bement et al., 2015; Michaux et al., 2018; Michaud et al., 2021; Yao et al., 2022).

Phosphoinositides (green quadrilaterals) are membrane lipids involved in signaling to actin. PIP₃, the triply phosphorylated version (dark green), greatly accelerates F-actin assembly and is correlated with Rac activity (Van Keymeulen et al., 2006). PIP₃ is generated from its less active form, PIP₂ (light green) by the kinase Phosphoinositide 3-kinase (PI3K, adds a phosphate group), and converted back to PIP₂ by the Phosphatase and Tensin homolog (PTEN) (Kölsch et al., 2008). Since PI3K and PTEN spatially exclude each other, they may be mutually antagonistic (Jiang and Liu, 2008). PI3K and PTEN have active membrane-bound forms and inactive cytosolic forms, with rapid exchange rates (secs) that may be affected by feedback from PIP₂ and PIP₃ (Hilgemann, 2007). On the timescale of actin waves (seconds, minutes), the total amounts of PI3K, and of a given GTPase is roughly constant, since rates of synthesis or degradation of such proteins are much slower (hours) (Spiering and Hodgson, 2011), justifying a common modeling assumption of conservation.

Box 1—figure 1

Download asset Open asset

It is accepted that myosin also plays a role in excitability, contractility of the cell cortex, and/or oscillatory behavior. For example, cycles of periodic lamellipodial contraction critically depend on myosin II activity (Giannone et al., 2007). Furthermore, the works of Michaud et al., 2021 and those leading up to it implicitly connect Rho signaling to myosin contractility, though in many studies myosin is lumped with F-actin in what is denoted as ‘actomyosin’ (e.g., see Driscoll et al., 2015). More explicit experimental probing of the spatial waves of Rho and myosin in cells is shown in Graessl et al., 2017. The role of feedback from myosin back to its signaling regulators has also been explored in cells confined to 1D by Hadjitheodorou et al., 2021. Here, the behavior is reduced to front-back cell repolarization under specific conditions (a geometrically simple form of a ‘1D actomyosin wave’). While important, in this review we limit our main focus to ‘actin waves,’ and only occasionally mention myosin as a separate player.

One of the most thoroughly studied examples of ventral actin waves are those observed in D. discoideum (Vicker, 2002b; Bretschneider et al., 2004; Gerisch et al., 2004). These waves are composed of large ring-shaped actin structures that deform and meander across the substrate-attached bottom membrane and extend 1–2 µm into the cytosolic region (Bretschneider et al., 2009). The F-actin ring separates an inner region enriched in F-actin and Arp2/3 from the outer part of the cortex characterized by increased levels of cortexilin I, myosin II, and different formins (Schroth-Diez et al., 2009; Ecke et al., 2020). In addition, both regions exhibit distinct network architectures (Jasnin et al., 2019) and surface charges (Banerjee et al., 2022). The different class I myosins colocalize with the wave or the inner region and may inhibit wave formation in a lipid-binding dependent manner (Brzeska et al., 2014; Brzeska et al., 2020). When traveling over larger cortical areas of oversized cells, waves maintain a preferred size and may annihilate upon collision (Gerhardt et al., 2014a; Miao et al., 2019). This cytoskeletal organization is coupled to wave patterns in the upstream signaling system, such that the inner patch enclosed by the actin wave is enriched in activated Ras and PIP₃, whereas PTEN and PIP₂ are found outside the ring (Asano et al., 2008; Gerisch et al., 2011). A rich variety of dynamic modes was observed for these patches, including alternating oscillations, rotational movement, and periodic switching between the ventral and dorsal membranes (Gerisch et al., 2011; Gerisch et al., 2012; Helenius et al., 2018). Many of these dynamic patterns are preserved in the signaling systems even if cytoskeletal activity is suppressed by treatment with inhibitors of actin polymerization (Arai et al., 2010; Fukushima et al., 2019; Hörning and Shibata, 2019).

Actin waves in D. discoideum have been associated with various cellular functions. When colliding with the cell border, they push the membrane outward (Vicker, 2000; Bretschneider et al., 2009; Taniguchi et al., 2013). Even though not strictly required for cell motility, waves may also affect the movement by triggering transitions between the amoeboid and a more persistent, keratocyte-like mode of locomotion (so-called fan-shaped cells) (Asano et al., 2008; Miao et al., 2017; Cao et al., 2019b; Moreno et al., 2020; Ghabache et al., 2021; Moldenhawer et al., 2022). Moreover, waves may also drive the division of cells (Flemming et al., 2020). While initiation and guidance of waves by chemoattractants was not observed (Gerhardt et al., 2014b; Lange et al., 2016), electrical fields affect the wave propagation and result in more abundant formation of protrusions in the field’s direction (Yang et al., 2022). Similarly, nano- and micrometer-sized surface ridges guide the propagation of actin waves (Driscoll et al., 2014; Sun et al., 2015; Honda et al., 2021). The ring-shaped actin waves in Dictyostelium furthermore show structural similarities with phagocytic cups (Gerisch et al., 2009; Gerisch, 2010), an observation that has also been reported for macrophages (Masters et al., 2016). These waves may thus serve as precursors of macropinocytic cups (Veltman et al., 2016; Lutton et al., 2022), which is also supported by the vertical orientation of the actin filaments inside the ventral ring-shaped structures, indicating a force generating machinery that may drive the formation of an outward protruding cup (Jasnin et al., 2019).

The broad range of actin wave patterns has stimulated a wide spectrum of modeling approaches, varying greatly in their complexity and level of detail (Ryan et al., 2012b; Allard and Mogilner, 2013; Sept and Carlsson, 2014; Beta and Kruse, 2017; Beta et al., 2020). Some works start with detailed signaling circuits of interacting components (Khamviwath et al., 2013; Avila Ponce de León et al., 2021), resulting in models that consist of large sets of coupled equations. In most situations, the exact functional forms and parameter values cannot be uniquely determined from experimental observations and are somewhat arbitrary. The analysis of such models mainly relies on numerical simulations (numerical integration of the model equations), often yielding results that quantitatively match specific experimental data after adjusting parameters. However, it is not always clear how robust these findings are with respect to model details and whether the agreement with the experimental data provides valid intuition about the underlying mechanism.

Along with these detailed approaches, there are many attempts to formulate models that reduce the complexity by focusing on a specific part of the signaling or cytoskeletal system. These models contain a lower number of equations, while still maintaining an explicit biological interpretation of the model variables; see, for example, Table 1. Identifying the core wave mechanisms in a given circuit implies that layers of complexity are removed to find the essential underlying motif that accounts for the observed patterns. Here, the role of actin in the generation of waves is a central aspect to link model predictions to experimental findings. Largely, this involves experimental observations of waves in systems where mutations, knockouts, or drug treatments are used to eliminate specific components of the signaling circuit. We will provide examples of this approach in the context of D. discoideum below. However, whether actin plays similar roles for wave generation in different cell types is an open question. Note also that some of the reduced models are even more simplified. They are designed to capture the observed macroscopic patterns with the most minimal mathematical ingredients. Here, the choice of model equations is mainly driven by mathematical intuition and only to a lesser extent by biological mechanisms and signaling motifs. Consequently, the model variables only maintain a vague biological interpretation and are mostly seen as effective, lumped quantities that represent the joint action of many involved players.

In this section, we will summarize the large body of work on actin wave models that are of reduced complexity in the sense explained above. In general, they all rely on a set of interacting species that are connected via positive and negative feedback loops, resulting in a nonlinear temporal dynamics at each location in the cell cortex (e.g., switch-like or time-periodic behavior). In addition, the local dynamics is spatially coupled to adjacent locations in the cortex, for example, via diffusive transport of some of the involved components or as a consequence of nonlocal changes in the mechanical properties. Local changes will thus affect the dynamics at neighboring points in the cortex, which may result in the formation of propagating waves, see Figure 2A–C for a schematic representation of this generic principle. Mathematically speaking, most of these models belong to the so-called reaction-diffusion-type, where spatial coupling is established by diffusive transport. However, extensions and alternative approaches will be also mentioned at the end of this section. Note, however, that we focus on descriptions of actin waves under spatially isotropic, homogeneous conditions and do not elaborate on the impact of externally applied chemical gradients, electric fields, or other directional cues.

Figure 2

Download asset Open asset

Reaction-diffusion-type models

Most actin wave models describe the interactions of molecular players using traditional chemical reaction kinetics localized at the cell cortex or/and the membrane (local dynamics, see Figure 2A). In addition, spatial coupling is assumed, typically based on diffusive transport of the species involved (see Figure 2B). Such models can be mathematically expressed as reaction-diffusion (RD) systems, composed of reaction terms that represent the local intracellular kinetics and a diffusion term, accounting for the diffusive transport of all or some of the species. The reaction terms involve nonlinear interactions between the reacting species that are often assumed to be of Michaelis–Menten type (saturating) or switch-like (sigmoidal, a.k.a. ‘Hill function’ type). While the number of interacting species can be large, it is often possible, under certain assumptions, to reduce the number of variables in an RD model to three, two, or even a single dynamic variable in the case of minimal phenomenological models, as briefly discussed in Box 2. Due to the large body of such works, here we survey them according to different cell types (species).

Box 2

Methods of reduction for systems of differential equations

The complexity of cellular systems tends to obscure the identities of elements that act as ‘central organizers.’ Mathematical methods can offer significant model simplification (reduction) in whittling away the peripheral vs. central components. In what follows, we highlight some common methods.

Dimensional analysis

Rescaling variables in terms of ‘baseline values’ (for concentrations, fluorescence levels, etc.), relevant timescales (reciprocals of typical decay rates), or typical lengthscales (e.g., cell diameter or diffusion lengthscale) can reduce the number of free parameters (Lin and Segel, 1988; Murray, 2002; Breña–Medina et al., 2014; Holmes et al., 2015; Bernitt et al., 2017; Verschueren and Champneys, 2017). In some cases, for example, Flyvbjerg et al., 1996; Bailey et al., 2011, this step, together with data that follows some ‘qualitative trends,’ can already reveal precise underlying molecular mechanisms that are unknown ab initio (i.e., from first principles).

Scale separation analysis

Rescaling (i.e., reducing a model to a dimensionless formulation as mentioned above) can reveal additional relative rates of underlying processes, such as rate-limiting steps for chemical reactions. Then, it is possible to make simplifying assumptions (replace some variables by ‘slowly changing parameters’) in a process denoted quasi steady-state analysis (QSS). This can also help to reduce a large and complicated model into a hierarchy of submodels operating at fast, medium, and slow timescales. Each can be analyzed in turn in a more complete way, for example, see Keener and Sneyd, 2009; Desroches et al., 2022; Tran et al., 2009; Newby and Bressloff, 2010; Plazen et al., 2023; Plazen and Khadra, 2023. Rescaling can also identify submodels or processes that are governed by a ‘small (dimensionless) parameter’ (e.g., ratios of rates of diffusion of the membrane-bound and cytosolic forms of regulatory proteins). Then, asymptotic analysis (a.k.a. singular perturbation theory) can help to decipher how system behavior depends on such quantities.

Piecewise-linear approximations

Many models for biochemical interactions exploit Hill functions to depict switching between states (e.g., self-amplified rates of activation). Approximating such terms by ‘sharp switches,’ that is, step functions or piecewise-linear terms (Gebert et al., 2007), can vastly simplify the process of solving for steady states explicitly (hence bypassing the need for numerical analysis). In some cases, for example, Holmes and Edelstein-Keshet, 2016; Holmes et al., 2016, this provides a reasonable first step in mapping out the parameter dependence of various regimes.

Reducing PDEs to ODEs

Approximating or rewriting PDEs by ODEs can also vastly increase the availability of methods of analysis. Local perturbation analysis (LPA) exploits the differences between slow and rapidly diffusing components, approximating these by zero and infinite rates of diffusion, thereby defining ‘local’ and ‘global’ variables. LPA can then be used in conjunction with (simpler) ODE tools to find parameter regimes in which a localized stimulus is expected to generate ‘interesting behavior,’ examples include Marée et al., 2006; Holmes et al., 2012; Holmes et al., 2015; Mata et al., 2013; Nakamura and Shibata, 2015. Another form of reduction is by transformation based on a conserved property, for example, for steady propagating patterns it is convenient to use Galilean transformation ${\displaystyle (x,t)\to \xi =x\pm vt}$ (a.k.a. a co-moving coordinate frame transformation), where ${\displaystyle v}$ is a propagation speed. Examples in the context of actin waves include Yochelis et al., 2020; Champneys et al., 2021; Yochelis et al., 2022.

All the above methods contribute to the ability to analyze the model equations and facilitate the use of bifurcation theory to uncover the origin of wave mechanisms, as elaborated in the text.

Models of actin waves in Dictyostelium

Reduced mechanistic models of actin waves critically depend on identifying the core wave generator of the system. Here, we will present examples from D. discoideum to demonstrate how the list of key components can be narrowed down by combining experimental and modeling studies. In 2010, traveling waves of PIP₃ and PTEN were reported from cells treated with Latrunculin A, a drug that inhibits actin polymerization (Arai et al., 2010), demonstrating that neither F-actin nor the chemoattractant cAMP are needed for wave generation (only later it was argued that F-actin could play a stabilizing role in wave dynamics; see Nishikawa et al., 2014). These observations inspired a model in which the key dynamics is maintained via PIP₂/PIP₃ signaling (see Figure 2D–F; Arai et al., 2010; Shibata et al., 2012). The model is based on assuming that PIP₃ negatively regulates membrane association of PTEN, while PIP₂ helps to activate PTEN, thus effectively providing a positive feedback from PTEN to itself. PI3K was taken to be independent of PIP₂ and PIP₃, and constant in the absence of chemoattractants. PTEN was assumed to be well-mixed in the cytosol and non-diffusive on the membrane originally, but diffusion was included in a later variant of the model (Shibata et al., 2013). Numerical simulations reproduced the experimentally observed signaling waves and some analysis was also performed (Nakamura and Shibata, 2015), including a ‘local perturbation analysis’ (LPA) (Holmes, 2014).

Subsequent experiments demonstrated that excitable behavior can also be present (Huang et al., 2013; Nishikawa et al., 2014). It was observed that Ras activation and increased levels of PI3K are associated with the wave patterns (Fukushima et al., 2019). Moreover, based on more recent fluorescence recordings, Ras appears as the central player of the wave generator, in that Ras waves affect PI3K and PIP₃ but can persist independently of those intermediates (Fukushima et al., 2019). In light of these newer findings, an extended model was proposed that includes active and inactive forms of Ras, as well as an unspecified Ras GAP, while the kinetics of PIP₂ and PIP₃ are included in a simplified form (Fukushima et al., 2019). Important features of the model are that PIP₃ helps to recruit Ras to the membrane in a positive feedback loop, and both the active and the inactive Ras affect the Ras GAP. Besides models that exclusively focus on a wave generator in the signaling system, wave dynamics in D. discoideum has also been addressed by so-called ‘excitable network models’ (Xiong et al., 2010; Pal et al., 2019). Here, two excitable modules are coupled together that represent the signaling and cytoskeletal dynamics (Huang et al., 2013), respectively. They have been successfully adapted to reproduce many experimental observations of actin wave patterns under a wide range of different conditions (Miao et al., 2017; Miao et al., 2019).

Reaction-diffusion models of intracellular waves in D. discoideum were also studied inside deforming model cells. The dynamic phase field approach represents cell shape by a function that takes on distinct values (e.g., 1 or 0) inside versus outside of the cell with a smooth interface connecting these values along the cell boundary (e.g., see Shao et al., 2010; Camley et al., 2017). This approach has been used, for example, in a study focusing on the dynamics of phase singularities in a variant of the PIP₂/PIP₃ model (Taniguchi et al., 2013). A phase field model for the dynamics of PIP₃ and PTEN has also been extended by a noisy excitable module to account for the observations of transient ‘holes’ in the PTEN distribution (Knoch et al., 2014). In Ghabache et al., 2021, different distributions of actin and myosin were prescribed within the phase field domain to account for different motility modes. Finally, with further reduction, generic wave generators (a.k.a. ‘toy’ models) have been combined with a dynamic phase field to qualitatively address a wide range of experimental observations in D. discoideum. Even though the relation of these models to individual molecular players in the cell often remains vague, they successfully captured observations of cell-to-cell variability (Alonso et al., 2018), of different motility phenotypes (Cao et al., 2019b; Moreno et al., 2020; Moldenhawer et al., 2022), and of wave-mediated cell division events (Flemming et al., 2020). In combination with an F-actin orientation field, a noisy FitzHugh-Nagumo-type model also explains how actin spots become mobile and form traveling actin waves (Whitelam et al., 2009).

Models of actin waves in oocytes

In cell division, signals from the mitotic spindle set up waves of Rho and F-actin along the cortex, which are funneled to the cell’s equator, where the contractile ring eventually splits the mother cell into its two daughter cells. In immature oocytes of frogs (X. laevis) and starfish (Patiria miniata), where the cortex is quiescent, vibrant dynamic wave patterns of Rho and F-actin can be induced by expressing two regulators of Rho (the GEF Ect2 and the GAP RGA3/4) at various ratios (Michaud et al., 2021; Michaud et al., 2022). Here, waves of F-actin are observed to follow closely behind waves of activity of the GTPase Rho. A model of this system was proposed taking into account diffusively mobile active and inactive Rho, stationary F-actin, an autocatalytic coupling of Rho to itself via the Rho-GEF Ect2, as well as inactivation of Rho by F-actin, see Figure 2G–I (Goryachev et al., 2016; Michaud et al., 2021). Recent experiments have confirmed that the core circuit of F-actin, Ect2, and the Rho-GAP RGA-3/4, which was identified as the negative regulator correlated with F-actin feedback, controls the dynamics of patterns in the cortex of oocytes (Michaud et al., 2022). Based on extracts of Xenopus oocytes and artificial lipid membranes, the Rho and F-actin wave patterns could even be reconstituted in vitro (Landino et al., 2021). Traveling waves were observed that emerge from random foci and expand as target waves that annihilate upon collision. Surprisingly, even diluted extracts displayed similar robust dynamics. Inhibiting Rho or actin polymerization destroyed the dynamics, proving that both components are essential for the waves. In the abovementioned models of actin waves in X. oocytes, the role of myosin in cortical contraction is not modeled explicitly.

Disordered wave patterns in the starfish oocyte system have also been analyzed from the viewpoint of spiral wave turbulence (Tan et al., 2020). Comparing the experimental observations to the complex Ginzburg–Landau (CGL) equation, which describes the dynamics of any reaction-diffusion system close to the onset of oscillations, suggests that the observed spiral wave properties are generic (model independent) features. However, while this model is based on the creation/annihilation dynamics of topological defects (intrinsically arising dislocations in the phase field) at the cores of the Rho spiral vortices, it is only indirectly related to the specific underlying biological mechanisms.

Models of actin waves in various other species

In neutrophils, the dynamics of the Hem-1 component of the SCAR/WAVE complex gives rise to actin waves that correlate with cell edge protrusions (Weiner et al., 2007). As demonstrated in this pioneering work, Hem-1 can self-activate and is negatively regulated by F-actin (via an unknown mechanism), as shown in Figure 3A. Assuming a constant total pool of actin that polymerizes in a Hem-1-dependent fashion, a simple model that includes only Hem-1 and F-actin was proposed that produced waves resembling the experimental observations (Weiner et al., 2007). Rather than local coupling by diffusion (associated with nearest-neighbor interactions), the authors used a nonlocal kernel (integral convolution with a Gaussian) to model actin-dependent autoactivation of Hem-1 over some spatial neighborhood.

In early C. elegans embryos, control cells show oscillating focal pulses of Rho activity, rather than oscillatory waves (Michaux et al., 2018; Yao et al., 2022). The GTPase Rho leads to actin assembly by formins of the Dia family, a route that is parallel to the ‘common’ WAVE/WASP activation of Arp2/3-dependent actin branching. Here, negative feedback stems from the Rho GAP (RGA-3/4) that gets recruited by F-actin and locally inactivates Rho, as shown in Figure 3C. Recently, it was demonstrated that well-controlled perturbations to anillin or formin affect the nature of the observed patterns (see also Bolado-Carrancio et al., 2020 for a study that combines experiments on breast cancer cells and a model that explains propagating waves of Rac and Rho in cell migration). Rho is also known to activate myosin contraction, a feature that many of the above models do not address. One exception is the recent model of Staddon et al., 2022, where actomyosin leads to contraction and also creates stress that feeds back to Rho activation. This is, however, a time-dependent model with as yet no spatial degrees of freedom.

In a different approach, several models assign autocatalysis to F-actin itself, for example, mediated by Arp2/3 branching, and propose or identify an additional inhibitor of assembly or a promoter of disassembly, as shown in Figure 3D. A simple example of this type includes only monomeric and polymerized actin as well as an inhibitor of actin polymerization (Yochelis et al., 2020). This model was derived from a more detailed model of CDRs in fibroblasts (Bernitt et al., 2017). Its simplicity allows for in-depth mathematical analysis (Yochelis et al., 2022). A more complex variant of this approach has been proposed by Wasnik and Mukhopadhyay, 2014, where different forms of actin monomers and autocatalytic dendritic actin growth are included, as well as negative feedback via coronin that binds to F-actin and promotes its disassembly.

In summary, Figure 3 displays several circuits associated with actin waves in various species including (A) neutrophils, (C) C elegans, Xenopus, and starfish oocytes, and (E) D. discoideum. Notably, common features of these circuits, shown in (F), include the following: (i) the main variable is associated with F-actin assembly or F-actin itself. (ii) The same species creates autocatalytic positive feedback (curved green arrow), enhancing its own activation or assembly from the reservoir. (iii) The main variable promotes some effector that creates negative feedback, generally on a slower timescale. It is well known that such positive and negative feedbacks can give rise to propagating waves, though the details of the patterns and their regimes of existence depend on the specific model. Note, however, that the key motif of these networks is quite general and robustly persists also in the simpler toy models, namely the autoactivation of a key player coupled to slow damping via negative feedback from its product (see also Figure 2A–C).

Figure 3

Download asset Open asset

As we demonstrated in the previous sections, many models implement similar ideas with very different levels of detail. While most models of actin waves have been designed to exhibit oscillatory or excitable dynamics, their fidelity to specific experimental observations varies significantly, from phenomenological to realistic aspects, as also indicated in Table 1. Beyond biochemical reaction networks, which are typically described by rate equations, a.k.a. coupled ordinary differential equations (ODEs), transport and mechanical coupling elevate the level of complexity due to space dependence that increases the number of degrees of freedom (Cross and Hohenberg, 1993). In cases where space is of interest, such as for propagating actin waves, the ODE formulation is extended to partial differential equations (PDEs) (Murray, 2001; Baker et al., 2008).

In general, it is impractical, if not impossible, to specify all the processes involved or to obtain a complete set of measurements of rates, concentrations, and interactions within the cell. Alternatively, we seek to discover which of the cell components play critical roles in the observed phenomena. The challenge is finding a compromise between models that have enough ‘realism’ to capture actual biology (e.g., can be related to known cell components) while being ‘simple enough’ to permit in-depth mathematical analysis. This wide spectrum of possible model designs raises fundamental questions to which mathematical theory can contribute: What aspects of the emergent behavior depend on details of a proposed model and what is generic, that is, independent of model details? And is it possible to develop a general theory of intracellular wave dynamics that is still useful to biology? Essentially, any model represents an approximate description of a set of observations, which is often a system’s response upon variation of a controlled quantity. However, the validity of a model is not solely determined by its fidelity (fitting) to the data but also by the ability to predict novel features of the system.

The advantage of simple models with few core building blocks is that they are amenable to mathematical analysis that can reveal the wave generating mechanism. The analysis can then give insights into basic (model-independent) mechanisms of pattern formation even though such models are often far from specific biological interpretations. It is the role of mathematical theory to uncover these core building blocks also in more complex, realistic models that are derived from mechanistic knowledge about the system. Moreover, such core building blocks often represent the more general and robust properties of the system that, in some cases, may be universally valid for a class of similar phenomena, such as actin waves in different biological species.

Previously, we reviewed different modeling approaches. In what follows, our aim is to introduce methods from the so-called bifurcation theory of nonlinear PDEs that provide a framework to identify the core building blocks of complex biological models. Through examples from the field of actin waves, we also demonstrate the methodology’s strength. We emphasize that the purpose is to provide a descriptive picture rather than a rigorous mathematical review, while organizing the more technical details, for math-inclined readers, in focused boxes.

Mathematical methodology for actin waves

In the 18th century, Laplace was convinced that Newtonian mechanics is a sufficient methodology to uncover the ‘secrets’ of the universe, “Give me the positions and velocities of all the particles in the universe, and I will predict the future.” It took, however, another century until Poincaré demonstrated the inherent limitation of Newtonian (linear) mechanics arising already in the (nonlinear) motion of three interacting particles. In contrast to Laplace’s philosophy, Poincaré showed that such motion cannot be predicted (Poincaré, 1885), laying the foundations of what was to become the so-called chaotic dynamics as well as bifurcation theory, in general.

The essential difference between linear and nonlinear systems is the number of possible solutions: While linear models have only one solution, nonlinear models may admit coexisting solutions, which can be uniform or may vary in space and time. Furthermore, the stability properties of solutions are important as they affect sensitivity to initial conditions and to perturbations as the system evolves in time. However, uncovering the evolution of multivariable nonlinear PDEs is a paramount and ongoing mathematical challenge. Thus, in many cases, we rely on in silico simulations to describe complex systems (e.g., biological, chemical, and ecological). These simulations are numerical experiments rather than theoretical frameworks. In what follows, we briefly survey and exemplify the bifurcation methodology that is, to date, the most powerful approach to analyze the solutions of nonlinear PDE systems.

Fundamentals of bifurcation theory at a glance

Bifurcation theory has been developed to analyze dynamical systems described by ODEs (Strogatz, 2018). The term ‘bifurcation’ stems from Latin and describes a branching or splitting. In the context of dynamical systems, it was introduced by Henri Poincaré and refers to the branching of solutions (Poincaré, 1885). Mathematically, a bifurcation point corresponds to a sudden qualitative change in the system’s behavior that is related to a transition between different states of the system or to an exchange of a state’s stability, as a control parameter is varied.

Conventionally, bifurcations of solution branches are described in diagrams that show the effect of changes in control parameters on the existence of solutions and their stability. (For example, a one-parameter bifurcation diagram resembles a tree-like structure, as shown in the right panel of Figure 4A.) Stable (unstable) solutions are typically depicted by solid (dashed or dotted) curves. A bifurcation diagram can be thought of as a ‘landscape’ on which the system evolves, similar to a ‘potential surface’ in physics or the ‘Waddington landscape’ in developmental biology. In Figure 4, we show a schematic example of many possible behaviors by a mathematical model, the dependence of that behavior on a control parameter, and the values of the parameter at which the behavior changes. The main ‘branches’ of this diagram represent the system’s behavior after transients are gone, and the system has settled into one of its stable regimes. We elaborate further on the concepts of stability and classifications in Box 3.

The appearance and/or stability of these solutions obey specific mathematical rules, for example, past a bifurcation point, an increase in the number of solutions is always by pairs (one stable and one unstable), while if the number of solutions remains the same, then their stability is exchanged (i.e., the stable becomes unstable and vice versa). Different bifurcation scenarios can be distinguished. They represent model-independent behaviors near the bifurcation onset. In Figure 4, we describe two representative cases of a ‘pitchfork’ and a ‘Hopf’ bifurcation. A pitchfork bifurcation represents the formation of two coexisting stable states (a regime of bistability) out of one stable state, as a function of a control parameter. A bistabile regime is observed, for example, when reactants undergo a chemical reaction that results in a mixture of two chiral molecules (enantiomers). Both enantiomers are stable and may coexist, while the initial non-chiral reactants are unstable. Also in many biological systems, multistability is observed, for example, when cells differentiate into different coexisting types of tissue or ecosystems converge to different compositions of stable coexisting populations or vegetation types. A Hopf bifurcation describes an onset of persistent time-periodic oscillations (often referred to as ‘limit cycle’ solutions). They are ubiquitously observed in biological systems, for example, in the regular pacing activity of the sinoatrial node in the heart, the circadian rhythms, the Rho activity in early C. elegans embryos, the genetic clock in somitogenesis, or cyclic adenosine monophosphate (cAMP) dynamics in developing D. discoideum cells. Such cyclic dynamics require at least one activating species with autocatalytic or enzymatic steps and an inhibiting species that evolves at a slower timescale.

Basic bifurcations, such as the pitchfork and Hopf bifurcations in Figure 4, are governed by variation of a single control parameter and are termed co-dimension one bifurcations, while more complex bifurcation scenarios may require the tuning of several control parameters and are, thus, termed as higher co-dimension bifurcations. In general, the higher the co-dimension of a bifurcation, the richer the qualitative accompanied dynamics as well as the sensitivity to initial conditions. Consequently, studying high co-dimension bifurcations may reveal the core building blocks of a model and allow the design of specific tests to differentiate between different dynamic behaviors in experiments. For further details, we refer the reader to introductory (Strogatz, 2018) and to more advanced texts (Crawford, 1991; Kuznetsov, 2004; Guckenheimer and Holmes, 2013).

Figure 4

Download asset Open asset

Box 3

Linear vs. nonlinear bifurcations

In this box, we briefly overview the concept of stability. For illustrative purposes, consider a cell that could be rounded (state A) or spread out (state B). A state is said to be linearly stable (or unstable) if a small superimposed disturbance or ‘perturbation’ decays (or grows) with time. We consider both static (steady state) and periodic (limit cycle) cell behaviors (e.g., a cell that oscillates between contracted and spread). Mathematically, we can calculate quantities that diagnose stability. These values are called eigenvalues (for steady states) and Floquet multipliers (for limit cycles).

As a control parameter is tuned past some ‘bifurcation value,’ the stability of a system may change. We distinguish between super- and subcritical bifurcations. In the former, arbitrarily small disturbances (like natural thermal fluctuations) would destabilize the system. In the latter case (i.e., when moving along the linearly stable solution toward the bifurcation point), only a sufficiently large disturbance (larger than some predictable threshold) would do so. For example, a small stimulus might fail to provoke a cell to round up or spread while a sufficiently large stimulus would do so, as shown in Box 3—figure 1A. Coexisting states (as in the case of A and B in Box 3—figure 1B) also generically come with a property called hysteresis, if two or more bifurcation branches are connected.

We exemplify this with the bifurcation diagram displayed in Box 3—figure 1B. In the region of stably coexisting states A and B (shaded blue), only a threshold stimulus causes state transition. Outside of this parameter regime (white), only state A or B exists. Tuning the control parameter up (horizontal green arrows) will eventually induce a transition from state B to state A (vertical red arrow). But it then takes considerable ‘detuning’ of the same parameter to get back to state B (red horizontal and green vertical arrows). Thus, the state transition depends on both state and history. We refer to the full excursion as a hysteresis loop. This is the mechanism that leads to coexisting oscillations and pulses when an excitable element is coupled to slow negative feedback from its ‘refractory’ partner variable (Kobayashi et al., 1994; Kobayashi et al., 1995). It can also explain periodic cell rounding-spreading if myosin activity (causing contraction) is coupled to a mechanosensitive cell element (sensing contraction) that suppresses myosin activity (e.g., see Zmurchok et al., 2018).

Box 3—figure 1

Download asset Open asset

The direction of the bifurcating solution branches, as shown in Box 3—figure 1A, is important. This distinction, termed super- and subcritical bifurcations, allows insights into how the evolution of the system depends on its initial conditions or on perturbations. In the super-critical case, a stable solution loses stability beyond the bifurcation point (‘at onset’ of the bifurcation), where even small noise can trigger a spontaneous transition to a different state of the system. The subcritical case is identified with threshold behavior since, before onset, the bifurcating solutions are unstable, as shown in Box 3—figure 1A to the left of the black dot. We refer to this as a nonlinear bifurcation. To switch to a new solution, perturbations have to be large enough (above threshold) to drive the system across the unstable branch (dashed curve shown in Box 3—figure 1). The subcritical case is thus associated with finite (as opposed to infinitesimal) perturbations. Also, it is often accompanied by hysteresis, where the system’s behavior is history-dependent, switching between states at different critical parameter values, depending on the direction in which the control parameter is varied (Box 3—figure 1). An intriguing biological example of experimentally observed hysteresis occurs in Rac-Rho-PAK signaling in breast-cancer cells as described in Byrne et al., 2016. In physics, hysteresis is associated with first-order phase transitions, for example, in super-cooling, where liquid water persists even below the freezing temperature.

While bifurcation diagrams are a highly informative summary, they can be challenging to construct and validate. Experimentally, one would have to run a controlled experiment many times, varying a single parameter (e.g., titrating the concentration of an inhibitor in small steps, both up and down in values) to observe whether transitions in behavior take place and where. (Recall that within a hysteresis region, gradual increase or decrease of a parameter results in transitions at different parameter values.) For a two-parameter bifurcation plot, even more experiments would be required, varying each of the two key parameters. This is prohibitively expensive in time and material, meaning that few such diagrams are found in purely experimental papers. Relying on a theoretical model, we can generate bifurcation diagrams ‘more easily’ since the dynamics can be simulated over and over again to produce even two-parameter bifurcation diagrams, as done, for example, in Arai et al., 2010; Shibata et al., 2012; Bolado-Carrancio et al., 2020; Staddon et al., 2022. With fast computers, this is a possible avenue that requires no sophisticated mathematical tools, yet, it is also challenging for several reasons:

• For realistic models, there are typically many control parameters, for example, ${\displaystyle N>10}$ in Michaud et al., 2022. Identifying a subset of two parameters to explore, and values for each of the remaining $N-2$, is a nontrivial task. Evidently, such parameter spaces are huge, and even well-informed searches in parameter space can fail to find the regimes of interest. Hence, brute-force computations can be frustrating and ineffective.

• Even after successfully arriving at the desired numerical bifurcation plot, the insights gained by such methods are limited, for example, it may be still difficult to determine the robustness of the model structure (i.e., the persistence of qualitative features in the dynamics if reaction terms are changed or upon addition of another variable). The benefit of leaning on rigorous methods instead is that they not only help to find the interesting transitions and relevant regions in parameter space, but also ‘peal away the layers of complexity’ close to those transitions.

Importantly, bifurcations may also act as organizing centers, affecting nearby dynamics, meaning that in the vicinity of such bifurcations, there exist additional behaviors that are also model-independent, and that are recognizable as generic ‘signatures’ of more complicated dynamics. This is particularly relevant in the case of so-called global bifurcations since the dynamic flow about these bifurcations is not limited to the proximity of a specific solution branch. In contrast to local bifurcations, however, analysis of global bifurcations requires deeper mathematical knowledge.

As an example, let us consider the characterization of patterns such as broad moving fronts versus localized (static or moving) pulses (see Figure 5). For example, an actin nucleation promoting factor (NPF) moving as a broad front could lead to wide, F-actin-rich protrusions, whereas in the latter case, pulses of NPF may engender localized F-actin sites such as neuronal dendritic spines. Such distinctions have clear significance to cell behavior. Now consider the key differences in these patterns: The front is a wave with values of NPF that vary over space from a low- to a high-density state across the wave (front to back), while the pulse corresponds to a variation from low to high and back to low NPF density. Mathematically, these waves are often depicted in so-called ‘phase-space’ diagrams as trajectories connecting states (see Figure 5). For example, heteroclinic bifurcations connect coexisting states while a homoclinic bifurcation is a closed loop connecting a state to itself. We can use mathematics to characterize the appearance or disappearance of these waves (and their speed) as global bifurcations, where trajectories of one kind (e.g., homoclinic, for a pulse) get broken and replaced by others (e.g., heteroclitic, for a front). As such, solutions emerging near global bifurcations are associated with dynamic trajectories (manifolds) that are generally far from any bifurcation onset and thus, their qualitative properties are generic (model-independent or structurally stable in a mathematical sense; Guckenheimer and Holmes, 2013). A prominent example is a propagating excitable pulse depicting an action potential in neural and cardiac systems (Murray, 2001; Murray, 2002; Keener and Sneyd, 2009). These solutions display a finite amplitude excursion, starting from a rest state and decaying back to the same state (i.e., a homoclinic loop). In contrast, the heteroclinic trajectories represent fronts that are associated, for example, with the invasion of one state into another, where the states may represent different species, concentrations, or cell types.

Next, we elaborate on the key role of global bifurcations in identifying robust features of wave propagation phenomena, relevant to our understanding of actin waves.

Figure 5

Download asset Open asset

Bifurcations and waves in spatially extended systems

After describing the basic ideas underlying the bifurcation theory of ODE systems, we now take into account space dependence that may arise due to transport and mechanical coupling. Mathematically, this is reflected in PDE-type model equations that inherently have infinite degrees of freedom due to spatial modes and dependence on boundary conditions. There are several rigorous methodologies for PDE analysis (e.g., multiple timescale reduction and singular perturbation theory). However, similarly to the phase-space method, also other advanced reduction methods can be efficiently applied in a few simple cases of two-variable toy models expressed via polynomial-type terms that mimic the realistic interactions, such as in the FitzHugh–Nagumo, Schnakenberg, and Gray–Scott reaction-diffusion systems. More involved and detailed models of actin dynamics are typically studied by direct numerical integration. Consequently, lacking theory, it is difficult to determine which of the solutions are sensitive to model details and which are robust (structurally stable).

Even though bifurcation theory for spatially extended systems is significantly less advanced than for ODEs (Cross and Hohenberg, 1993; Sandstede and Scheel, 2000; Pismen, 2006; Meron, 2015), in some cases, mathematics can still provide valuable insights. Such cases often rely on specific properties of solutions such as traveling waves or pulses, where PDEs can be reduced to ODEs that describe the (fixed) shape of a solution along a transformed coordinate (say $\xi$), instead of the spatiotemporal dependence (e.g., using a comoving frame transformation such as $\xi =x\pm vt$, where $v$ is the wave speed). One then seeks bifurcations that depict the onset and demise of such wave patterns. Following these ideas, most of the propagating actin patterns can be classified (using phase-space reduction) according to their oscillatory, excitable, or bistable properties (cf. Allard and Mogilner, 2013). Figure 5 summarizes these three classes; note that we exclude here the transient actin polymerization dynamics that converge to steady-state patterns, as discussed in Bhattacharya et al., 2020.

However, these classifications do not resolve the origin and the mechanisms of the emerging actin waves. For example, many of the observed actin patterns are of a spiral form. Yet, since spirals share universal properties (Sandstede and Scheel, 2020) and can form in oscillatory, excitable, and bistable systems, it is difficult to decipher to which dynamic class the system belongs. Moreover, there are also basic subtleties that are frequently overlooked when relying on ODE classifications, such as secondary (space-dependent) instabilities (Cross and Hohenberg, 1993). This directly impacts attempts to determine the lowest number of variables needed to describe a certain phenomenon in a qualitative (generic) manner. For example, as has been shown in Figure 4B, in ODEs a steady state can lose stability to oscillations via the so-called Hopf bifurcation while in PDEs, oscillations can arise via two different mechanisms, leading to three qualitatively different patterns. The first type is the natural extension to homogeneous oscillations (HO), a.k.a. Hopf bifurcation with a zero wavenumber (infinite wavelength since a wavenumber is proportional to the inverse of a wavelength). As for ODEs, also for PDEs the minimal setting is a two-variable system. The second type is a Hopf bifurcation with a finite wavenumber (or finite wavelength), where the resulting oscillations are counter-propagating traveling waves (TW) and standing waves (SW); for more details, see Knobloch, 1986. In Figure 5, we display these three oscillatory patterns (HO, TW, SW) in a spatially extended system. In contrast to the zero wavenumber Hopf bifurcation, the finite wavenumber Hopf bifurcation requires a three-variable system (Yochelis et al., 2008; Stich et al., 2009; Anma et al., 2012; Hata et al., 2014; Villar-Sepúlveda and Champneys, 2023). In Box 4, we provide additional heuristic details about the richness of the primary finite wavenumber Hopf bifurcation. More advanced related topics, mixed-mode solutions, secondary instabilities, and conserved quantities are detailed in Knobloch, 1986; Knobloch, 1992.

To summarize, while the bifurcation theory of PDEs is still a developing field, especially in nonlinear regimes, it has matured to study some of the actin wave phenomena, as will be briefly demonstrated by examples in the following. Yet, further extensions are needed, especially in the broad context of mechano-chemical feedback (e.g., see Ben Isaac et al., 2013), where nonlocal interactions, advective flows, and changing geometries are coupled to reaction-diffusion dynamics.

Box 4

Bifurcations of spatially extended oscillatory solutions

In Box 3, we showed how oscillatory behavior (limit cycles) can arise in time-dependent systems (ODE models); see also Figure 4B. Adding spatial coupling (PDE models) affects sensitivity to perturbations (e.g., to input stimuli), especially in systems with many interacting components (many model variables) characteristic of most biological systems. Such models can potentially exhibit a vast range of qualitative behaviors that are challenging to characterize and study.

Bifurcations in PDEs are currently only partly understood. Still, we know that oscillations can arise through bifurcations that produce one of several outcomes: Spatially uniform cycles or waves. The latter include traveling (TW) or standing (SW) waves (Figure 5). The birth of spatially uniform cycles is predicted by the same type of (Hopf) bifurcation in PDEs as in ODEs, but the appearance of TW/SW waves is simultaneous and has a bifurcation structure that has no analog in ODEs (a so-called finite-wavenumber Hopf bifurcation). Mathematically, we can predict that TW and SW can appear (emerge at a bifurcation point) in one of six possible ways (both super- and subcritical, see Box 4—figure 1), contrary to only two types of bifurcations for spatially uniform cycles (Knobloch, 1986). As shown in Box 3, subcritical bifurcations imply coexistence of unstable branches over a large range of parameter values. Such waves have a richer sensitivity not only to size but also to spatial forms of stimuli. Moreover, in many mass-conserved systems (e.g. conserved amount of total actin over the timescale of actin-wave formation; Beta et al., 2020), new behaviors can arise (e.g., TW that stall to form a polarized distribution) (Knobloch, 1992). Furthermore, domain size, shape, curvature, and boundary conditions are significant (Yochelis et al., 2022).

Current mathematical methods are not yet sufficiently developed to decipher bifurcation structures that explain the appearance of spatially oscillatory dynamics just from experimental cell images or movies. This is especially true for 2D and 3D deforming cells (motivating 1D ‘cell in a channel’ experiments that simplify the geometry, e.g., Hadjitheodorou et al., 2021). So far, we can use mathematics to exclude some mechanisms. For example, we know that a finite wavenumber Hopf bifurcation to TW and SW cannot arise in two variable RD models (Villar-Sepúlveda and Champneys, 2023). A third component is required to account for trains of pulses emanating from a single localized stimulus (e.g., see Yochelis et al., 2008; Yochelis et al., 2015). (Localized optogenetic activation of NPF results in a sequence of filopodia moving bidirectionally away from the stimulated site.) More exotic dynamics include jumping spatially localized oscillations (oscillons) (Yang et al., 2006; Knobloch et al., 2021), observed, for example, in the pattern-forming Belousov–Zhabotinsky chemical reaction (Cherkashin et al., 2008). Hence, oversimplifying a system may result in loss of fundamental dynamic behavior.

Box 4—figure 1

Download asset Open asset

Insights obtained using bifurcation theory

Linking cell responses to model mechanism

In recent years, new tools have been developed for prodding, stimulating, and manipulating cell behavior, such as optogenetics, where light signals activate one or another regulator in a highly localized manner inside a cell (see O’Neill and Gautam, 2014; O’Neill et al., 2016; Meshik et al., 2018; Hadjitheodorou et al., 2021). This has resulted in rich data sets connecting the strength, size, and timing of the stimulus to resultant cell responses.

The question is how to understand cell responses from such data sets, and what such responses tell us about the regulatory circuits that govern those responses. For example, in O’Neill et al., 2016, an optogenetic signal was used to repolarize macrophages by photoactivating a Rho-GEF (activating the GTPase Rho at the front of the cell). In some cases, this stimulus will repolarize a cell, whose initial Rho activity was concentrated in the rear. This reversal of polarity was quantified in two models using PDE bifurcation analysis (Buttenschön and Edelstein-Keshet, 2022), mapping out how the signal intensity, coupled to the inherent cell parameters, elicits a response (Figure 6A). In some regimes, the cell does not repolarize (i), in others it loses polarization entirely (iii), while in a wedge-shaped region (ii), it repolarizes in a direction opposite to its initial polarity. The analysis revealed distinct response ‘signatures’ depending on the assumed underlying polarization model. Such synergy between models, analysis, and experimental data can be used to distinguish competing models for the underlying regulatory circuits. The mathematical analysis demonstrated that cells with mutually antagonistic (Rac-Rho) polarity GTPases (as in Holmes and Edelstein-Keshet, 2016) have much simpler routes to repolarization than cells with a single dominant polarity GTPase (as in the wave-pinning model by Mori et al., 2008). It also gave important insights into how cell repolarization takes place. Similar coordination between experiments that provoke dynamic actin waves and mathematical PDE bifurcation analysis of the proposed models could, in future, help to refine our understanding of those dynamic phenomena.

Figure 6

Download asset Open asset

To date, a full PDE bifurcation analysis has not yet been applied to most biochemically regulated actin wave models. The shortcut of LPA has been used to explore how parameters such as basal GTPase activation rate (GEF activity) and F-actin negative feedback (GAP activity) affect the formation of various dynamic patterns (Holmes et al., 2012; Mata et al., 2013). One example of a two-parameter LPA bifurcation diagram from Liu et al., 2021 is shown in Figure 6B. In principle, these predictions can be tested against data that inhibits/overexpresses an F-actin-associated GAP (like RGA3/4) or similarly manipulates the corresponding GEF.

Identification of circular dorsal ruffles as front phenomena

As has been shown throughout this review, actin waves are abundant in cells and arise in a wide variety of forms. However, since in experiments these waves are fully developed, that is, the instability at the onset of wave formation is typically not observed, there is an inherent difficulty to classify them and determine whether they arise via the same generic mechanism or whether there are fundamentally different coexisting classes of wave patterns that are generated via distinct biochemical circuits (cf. Michaud et al., 2022).

CDRs are ring-shaped actin waves on the dorsal side of cells that are related to endocytosis (engulfment of external material). Typical phase-contrast microscopy images of CDRs are shown in Figure 7. As failures in CDR development are believed to be associated with cancerous phenotypes, a mechanistic understanding of their dynamics is of prime importance (Bloomfield and Kay, 2016). Previously, it was assumed that CDRs belong to the class of excitable waves, as annihilation upon collision frequently occurs (Zeng et al., 2011; Bernitt et al., 2015). However, this hypothesis was challenged by the observation of reflections of CDRs at the cell boundary, which are not expected in an excitable system.

This inconsistency has stimulated new research, combining live-cell imaging, mathematical modeling, and bifurcation analysis (Bernitt et al., 2017). Using confocal fluorescence microscopy recordings to quantify the actin density along the dorsal cell cortex, a significant difference in the actin density between the interior and the exterior of CDRs was observed, as shown in Figure 7B. The two regions of different actin density clearly indicate a bistable situations, where both the low and the high actin density regimes are stable solutions of the actin system. Moreover, the actin profile across the CDR border connects actin-rich and -poor regions and, therefore, corresponds to a generic ‘front’ connecting these two states (with an additional localized overshoot at the border). This evidence for bistability motivated a simple mass-conserved reaction-diffusion model in which actin waves naturally arise from the interplay of F-actin, G-actin, cortical actin, and an F-actin inhibitor. An analysis of this model confirmed the reflections of CDRs at boundaries and also accounted for the different propagation speeds of expanding and contracting rings observed in the experiments. They are related to two distinct coexisting front states, a property that naturally arises in bistable systems but is not expected in excitable media. Moreover, bifurcation analysis of front stability that was followed by experiments suggested conditions under which CDR failure can occur, giving rise to either disordered wave dynamics or front pinning at the cell boundary, thus, providing fresh insights into CDR dynamics and regulation. Beyond the mechanistic understanding of CDRs, these results may, in future, allow us to explain the link between the dynamics of CDRs and cancer progression by providing a framework to understand the roles of different cytoskeletal components and signaling factors (Orth and McNiven, 2006; Hoon et al., 2012; Itoh and Hasegawa, 2013).

Figure 7

Download asset Open asset

Actin waves appear in a wide variety of cell types and conditions. In some but not all such systems, these waves are correlated with certain cell functions, for example, edge protrusion (in motile cells) or cytokinesis. In other systems, the correlation with cell function is less clear, but we can use the existence, shapes, frequencies, and other wave characteristics to understand underlying processes at a more basic level of organization (i.e., the chemical and mechanical interactions that lead to those waves).

As surveyed here, different modeling approaches have been proposed to describe actin waves, some tied directly to experimental findings, and some primarily phenomenological. A challenge has been to strike the right balance between the level of biological detail and mathematical tractability. It is evident that the key driving molecular components may differ from one cell type to another (or between experimental conditions) since some cells or cell states lack effectors that are central in others. However, we often see related dynamic circuits (even in distinct molecular signaling networks), as, for example, in the case of a bistable switch driven around a cycle by negative feedback.

We believe that mathematical methods (including bifurcation analysis) are useful to truly characterize and synthesize the common features and differences between classes of models and, hence, between the distinct mechanisms leading to actin waves. This task, however, is nontrivial given the fact that much of the current model complexity is beyond the reach of existing mathematical methods, emphasizing the need (i) to further develop those mathematical tools and (ii) foster close consultation between biologists and mathematicians to identify the backbone components.

Theoretical investigations of actin waves are still driven primarily by simulations. However, even simulations that demonstrate biologically realistic behavior do not necessarily point at the correct underlying mechanism and mostly fall short of mapping out full regimes of behavior. To substantially advance the field, a grand synthesis of existing models is required, describing the common aspects, and identifying their unique vs. shared features. This type of synthesis would contribute to our understanding of universal (versus specific) cellular mechanisms, conserved mechanisms, and those that arose via convergent evolution in different cell types. Such collaborative work of biologists and mathematicians would also identify a minimal set of models, biologically relevant, and yet sufficiently compact to be analyzed mathematically in full. Needless to say, the analysis and the formulation of novel mathematical methods would also contribute broadly to the theory of PDEs, the physics of active matter, nonlinear chemical reactions/kinetics, and the biology of eukaryotic cells.

We conclude by listing some challenges for future research in this field. First, for the modeling community, we advocate for the development of a comprehensive ‘atlas’ of (small) regulatory modules that can (or cannot) account for observed waves. Here, we refer to generic circuits, whose dynamics have been well-documented, with modifications that remove or add specific behavior. Examples of such work include Kholodenko, 2006, where a combinatorial set of variants of feedback connectivity is mapped out to determine which ones result in a relaxation oscillator. The multiple (small) circuits discussed by Tyson et al., 2003 contribute to a similar flavor, helping to identify specific regulatory circuit topologies with known behavior. We also believe that new models that are proposed should be accompanied by comparisons to those that already exist.

Among the mathematical challenges, we can identify several areas that, in our opinion, merit highlighting:

• Apparently, PDE bifurcation analysis is still a specialty that is less familiar in biological sciences. This merits further development of pedagogical tools to emerge as a ready-to-use method (in contrast to ODE bifurcation analysis, for which many tools for nonexperts exist; e.g., see Strogatz, 2018).

• In communities that work at the border of computational biology, bifurcation theory is frequently employed, however, mostly at the linear analysis level. Yet, since most of the observed phenomena are far from the linear limit, nonlinear bifurcation theory relying on global bifurcations should be advanced to provide interpretations and mechanistic understandings.

• In classical models of biological wave patterns, for example, in the FitzHugh–Nagumo, Gray–Scott, Gierer–Meinhardt, or Schnakenberg models, there are no ‘conserved quantities.’ However, as demonstrated in this review, in intracellular actin waves, some quantities are conserved on the timescale of wave formation (switch-like signaling proteins, actin in various forms, etc.) The effect of mass conservation has been discussed, for example, in Jilkine and Edelstein-Keshet, 2011; Halatek and Frey, 2018; Brauns et al., 2020; Champneys et al., 2021, but its role in wave formation, nonlinear oscillatory instabilities, and pattern selection is still limited and should be considerably advanced (Beta et al., 2020).

• Analysis of pattern-forming mechanisms in reaction-diffusion systems traditionally relies on large, non-deforming, and isotropic domains, typically employing periodic or no-flux boundary conditions. But, for motile cells with continuously changing shapes, the boundaries bilaterally interact with the reaction-diffusion dynamics in the interior in known and unknown ways. While progress has been made in computations of such problems (see section ‘Membrane confinement, deformations, and mechano-chemical feedbacks’), more mathematical theory is needed to fully understand the coupling between domain shape and internal reaction-diffusion dynamics, even more so in the presence of anisotropy (directional forces, such as externally applied electric fields [Yang et al., 2022], or coupling to microtubules [Dogterom and Koenderink, 2019]).

We hope that the survey provided here will stimulate future cross-disciplinary work to address these challenges and advance the field of intracellular pattern formation with its many biological and medical applications.

1. 1. Abercrombie M
   2. Joan E
   3. Heaysman M
   4. Pegrum SM

   (1970) The locomotion of fibroblasts in culture: II."Ruffling"

   Experimental Cell Research 60:437–444.

   https://doi.org/10.1016/0014-4827(70)90537-9

   • Google Scholar
2. 1. Allard J
   2. Mogilner A

   (2013) Traveling waves in actin dynamics and cell motility

   Current Opinion in Cell Biology 25:107–115.

   https://doi.org/10.1016/j.ceb.2012.08.012

   • PubMed
   • Google Scholar
3. Book

   1. Alon U

   (2006) An Introduction to Systems Biology: Design Principles of Biological Circuits.

   Chapman and Hall/CRC.

   https://doi.org/10.1201/9781420011432

   • Google Scholar
4. 1. Alonso S
   2. Bär M
   3. Echebarria B

   (2016) Nonlinear physics of electrical wave propagation in the heart: A review

   Reports on Progress in Physics 79:096601.

   https://doi.org/10.1088/0034-4885/79/9/096601

   • PubMed
   • Google Scholar
5. 1. Alonso S
   2. Stange M
   3. Beta C

   (2018) Modeling random crawling, membrane deformation and intracellular polarity of Motile Amoeboid cells

   PLOS ONE 13:e0201977.

   https://doi.org/10.1371/journal.pone.0201977

   • PubMed
   • Google Scholar
6. 1. Alt W
   2. Brosteanu O
   3. Hinz B
   4. Kaiser HW

   (1995) Patterns of spontaneous motility in videomicrographs of human epidermal keratinocytes (HEK)

   Biochemistry and Cell Biology = Biochimie et Biologie Cellulaire 73:441–459.

   https://doi.org/10.1139/o95-051

   • PubMed
   • Google Scholar
7. 1. Anma A
   2. Sakamoto K
   3. Yoneda T

   (2012) Unstable subsystems cause Turing instability

   Kodai Mathematical Journal 35:215–247.

   https://doi.org/10.2996/kmj/1341401049

   • Google Scholar
8. 1. Arai Y
   2. Shibata T
   3. Matsuoka S
   4. Sato MJ
   5. Yanagida T
   6. Ueda M

   (2010) Self-organization of the phosphatidylinositol lipids signaling system for random cell migration

   PNAS 107:12399–12404.

   https://doi.org/10.1073/pnas.0908278107

   • Google Scholar
9. 1. Asano Y
   2. Nagasaki A
   3. Uyeda TQP

   (2008) Correlated waves of actin filaments and PIP₃ in Dictyostelium cells

   Cell Motility and the Cytoskeleton 65:923–934.

   https://doi.org/10.1002/cm.20314

   • PubMed
   • Google Scholar
10. 1. Avila Ponce de León MA
    2. Félix B
    3. Othmer HG

    (2021) A phosphoinositide-based model of actin waves in frustrated phagocytosis

    Journal of Theoretical Biology 527:110764.

    https://doi.org/10.1016/j.jtbi.2021.110764

    • PubMed
    • Google Scholar
11. 1. Bailey J
    2. Potter KJ
    3. Verchere CB
    4. Edelstein-Keshet L
    5. Coombs D

    (2011) Reverse engineering an amyloid aggregation pathway with dimensional analysis and scaling

    Physical Biology 8:066009.

    https://doi.org/10.1088/1478-3975/8/6/066009

    • PubMed
    • Google Scholar
12. 1. Bailles A
    2. Gehrels EW
    3. Lecuit T

    (2022) Mechanochemical principles of spatial and temporal patterns in cells and tissues

    Annual Review of Cell and Developmental Biology 38:321–347.

    https://doi.org/10.1146/annurev-cellbio-120420-095337

    • PubMed
    • Google Scholar
13. 1. Baker RE
    2. Gaffney EA
    3. Maini PK

    (2008) Partial differential equations for self-organization in cellular and developmental biology

    Nonlinearity 21:R251–R290.

    https://doi.org/10.1088/0951-7715/21/11/R05

    • Google Scholar
14. 1. Banerjee T
    2. Biswas D
    3. Pal DS
    4. Miao Y
    5. Iglesias PA
    6. Devreotes PN

    (2022) Spatiotemporal dynamics of membrane surface charge regulates cell polarity and migration

    Nature Cell Biology 24:1499–1515.

    https://doi.org/10.1038/s41556-022-00997-7

    • PubMed
    • Google Scholar
15. 1. Barnhart EL
    2. Lee KC
    3. Keren K
    4. Mogilner A
    5. Theriot JA

    (2011) An adhesion-dependent switch between mechanisms that determine motile cell shape

    PLOS Biology 9:e1001059.

    https://doi.org/10.1371/journal.pbio.1001059

    • PubMed
    • Google Scholar
16. 1. Barnhart EL
    2. Allard J
    3. Lou SS
    4. Theriot JA
    5. Mogilner A

    (2017) Adhesion-dependent wave generation in crawling cells

    Current Biology 27:27–38.

    https://doi.org/10.1016/j.cub.2016.11.011

    • PubMed
    • Google Scholar
17. 1. Bates PW
    2. Jin J

    (2014) Global dynamics of boundary droplets

    Discrete Contin. Dyn. Syst 34:1–17.

    https://doi.org/10.3934/dcds.2014.34.1

    • Google Scholar
18. 1. Bement WM
    2. Leda M
    3. Moe AM
    4. Kita AM
    5. Larson ME
    6. Golding AE
    7. Pfeuti C
    8. Su K-C
    9. Miller AL
    10. Goryachev AB
    11. von Dassow G

    (2015) Activator–inhibitor coupling between Rho signalling and actin assembly makes the cell cortex an excitable medium

    Nature Cell Biology 17:1471–1483.

    https://doi.org/10.1038/ncb3251

    • PubMed
    • Google Scholar
19. 1. Ben Isaac E
    2. Manor U
    3. Kachar B
    4. Yochelis A
    5. Gov NS

    (2013) Linking actin networks and cell membrane via a reaction-diffusion-elastic description of nonlinear filopodia initiation

    Physical Review E 88:022718.

    https://doi.org/10.1103/PhysRevE.88.022718

    • PubMed
    • Google Scholar
20. 1. Bernitt E
    2. Koh CG
    3. Gov N
    4. Döbereiner H-G

    (2015) Dynamics of actin waves on patterned substrates: A quantitative analysis of circular dorsal ruffles

    PLOS ONE 10:e0115857.

    https://doi.org/10.1371/journal.pone.0115857

    • PubMed
    • Google Scholar
21. 1. Bernitt E
    2. Döbereiner H-G
    3. Gov NS
    4. Yochelis A

    (2017) Fronts and waves of actin polymerization in a bistability-based mechanism of circular dorsal ruffles

    Nature Communications 8:15863.

    https://doi.org/10.1038/ncomms15863

    • PubMed
    • Google Scholar
22. 1. Beta C
    2. Kruse K

    (2017) Intracellular oscillations and waves

    Annual Review of Condensed Matter Physics 8:239–264.

    https://doi.org/10.1146/annurev-conmatphys-031016-025210

    • Google Scholar
23. 1. Beta C
    2. Gov NS
    3. Yochelis A

    (2020) Why a large-scale mode can be essential for understanding intracellular actin waves

    Cells 9:1533.

    https://doi.org/10.3390/cells9061533

    • PubMed
    • Google Scholar
24. 1. Bhattacharya S
    2. Banerjee T
    3. Miao Y
    4. Zhan H
    5. Devreotes PN
    6. Iglesias PA

    (2020) Traveling and standing waves mediate pattern formation in cellular protrusions

    Science Advances 6:eaay7682.

    https://doi.org/10.1126/sciadv.aay7682

    • PubMed
    • Google Scholar
25. 1. Blanchoin L
    2. Amann KJ
    3. Higgs HN
    4. Marchand JB
    5. Kaiser DA
    6. Pollard TD

    (2000) Direct observation of dendritic actin filament networks nucleated by Arp2/3 complex and WASP/Scar proteins

    Nature 404:1007–1011.

    https://doi.org/10.1038/35010008

    • PubMed
    • Google Scholar
26. 1. Bloomfield G
    2. Kay RR

    (2016) Uses and abuses of macropinocytosis

    Journal of Cell Science 129:2697–2705.

    https://doi.org/10.1242/jcs.176149

    • PubMed
    • Google Scholar
27. 1. Bolado-Carrancio A
    2. Rukhlenko OS
    3. Nikonova E
    4. Tsyganov MA
    5. Wheeler A
    6. Garcia-Munoz A
    7. Kolch W
    8. von Kriegsheim A
    9. Kholodenko BN

    (2020) Periodic propagating waves coordinate RhoGTPase network dynamics at the leading and trailing edges during cell migration

    eLife 9:e58165.

    https://doi.org/10.7554/eLife.58165

    • PubMed
    • Google Scholar
28. 1. Brauns F
    2. Halatek J
    3. Frey E

    (2020) Phase-space geometry of mass-conserving reaction-diffusion dynamics

    Physical Review X 10:1036.

    https://doi.org/10.1103/PhysRevX.10.041036

    • Google Scholar
29. 1. Breña–Medina V
    2. Champneys AR
    3. Grierson C
    4. Ward MJ

    (2014) Mathematical modeling of plant root hair initiation: Dynamics of localized patches

    SIAM Journal on Applied Dynamical Systems 13:210–248.

    https://doi.org/10.1137/120902264

    • Google Scholar
30. 1. Bretschneider T
    2. Diez S
    3. Anderson K
    4. Heuser J
    5. Clarke M
    6. Müller-Taubenberger A
    7. Köhler J
    8. Gerisch G

    (2004) Dynamic actin patterns and Arp2/3 assembly at the substrate-attached surface of motile cells

    Current Biology 14:1–10.

    https://doi.org/10.1016/j.cub.2003.12.005

    • PubMed
    • Google Scholar
31. 1. Bretschneider T
    2. Anderson K
    3. Ecke M
    4. Müller-Taubenberger A
    5. Schroth-Diez B
    6. Ishikawa-Ankerhold HC
    7. Gerisch G

    (2009) The three-dimensional dynamics of actin waves, a model of cytoskeletal self-organization

    Biophysical Journal 96:2888–2900.

    https://doi.org/10.1016/j.bpj.2008.12.3942

    • PubMed
    • Google Scholar
32. 1. Brzeska H
    2. Pridham K
    3. Chery G
    4. Titus MA
    5. Korn ED

    (2014) The association of myosin IB with actin waves in Dictyostelium requires both the plasma membrane-binding site and actin-binding region in the myosin tail

    PLOS ONE 9:e94306.

    https://doi.org/10.1371/journal.pone.0094306

    • PubMed
    • Google Scholar
33. 1. Brzeska H
    2. Bagnoli M
    3. Korn ED
    4. Titus MA

    (2020) Dictyostelium myosin 1F and myosin 1E inhibit actin waves in a lipid-binding-dependent and motor-independent manner

    Cytoskeleton 77:295–302.

    https://doi.org/10.1002/cm.21627

    • PubMed
    • Google Scholar
34. 1. Buccione R
    2. Orth JD
    3. McNiven MA

    (2004) Foot and mouth: Podosomes, invadopodia and circular dorsal ruffles

    Nature Reviews. Molecular Cell Biology 5:647–657.

    https://doi.org/10.1038/nrm1436

    • PubMed
    • Google Scholar
35. 1. Buttenschön A
    2. Edelstein-Keshet L

    (2022) Cell repolarization: A bifurcation study of spatio-temporal perturbations of polar cells

    Bulletin of Mathematical Biology 84:114.

    https://doi.org/10.1007/s11538-022-01053-z

    • PubMed
    • Google Scholar
36. 1. Byrne KM
    2. Monsefi N
    3. Dawson JC
    4. Degasperi A
    5. Bukowski-Wills J-C
    6. Volinsky N
    7. Dobrzyński M
    8. Birtwistle MR
    9. Tsyganov MA
    10. Kiyatkin A
    11. Kida K
    12. Finch AJ
    13. Carragher NO
    14. Kolch W
    15. Nguyen LK
    16. von Kriegsheim A
    17. Kholodenko BN

    (2016) Bistability in the Rac1, PAK, and RhoA signaling network drives actin cytoskeleton dynamics and cell motility switches

    Cell Systems 2:38–48.

    https://doi.org/10.1016/j.cels.2016.01.003

    • PubMed
    • Google Scholar
37. 1. Camley BA
    2. Zhao Y
    3. Li B
    4. Levine H
    5. Rappel WJ

    (2013) Periodic migration in a physical model of cells on micropatterns

    Physical Review Letters 111:158102.

    https://doi.org/10.1103/PhysRevLett.111.158102

    • PubMed
    • Google Scholar
38. 1. Camley BA
    2. Zhao Y
    3. Li B
    4. Levine H
    5. Rappel WJ

    (2017) Crawling and turning in a minimal reaction-diffusion cell motility model: Coupling cell shape and biochemistry

    Physical Review E 95:012401.

    https://doi.org/10.1103/PhysRevE.95.012401

    • PubMed
    • Google Scholar
39. 1. Campbell EJ
    2. Bagchi P

    (2018) A computational model of amoeboid cell motility in the presence of obstacles

    Soft Matter 14:5741–5763.

    https://doi.org/10.1039/c8sm00457a

    • PubMed
    • Google Scholar
40. 1. Cao Y
    2. Ghabache E
    3. Miao Y
    4. Niman C
    5. Hakozaki H
    6. Reck-Peterson SL
    7. Devreotes PN
    8. Rappel WJ

    (2019a) A minimal computational model for three-dimensional cell migration

    Journal of the Royal Society, Interface 16:20190619.

    https://doi.org/10.1098/rsif.2019.0619

    • PubMed
    • Google Scholar
41. 1. Cao Y
    2. Ghabache E
    3. Rappel WJ

    (2019b) Plasticity of cell migration resulting from mechanochemical coupling

    eLife 8:e48478.

    https://doi.org/10.7554/eLife.48478

    • PubMed
    • Google Scholar
42. 1. Carlsson AE

    (2010) Actin dynamics: From nanoscale to microscale

    Annual Review of Biophysics 39:91–110.

    https://doi.org/10.1146/annurev.biophys.093008.131207

    • PubMed
    • Google Scholar
43. 1. Case LB
    2. Waterman CM

    (2011) Adhesive F-actin waves: A novel integrin-mediated adhesion complex coupled to ventral actin polymerization

    PLOS ONE 6:e26631.

    https://doi.org/10.1371/journal.pone.0026631

    • PubMed
    • Google Scholar
44. 1. Champneys AR
    2. Al Saadi F
    3. Breña–Medina VF
    4. Grieneisen VA
    5. Marée AFM
    6. Verschueren N
    7. Wuyts B

    (2021) Bistability, wave pinning and localisation in natural reaction–diffusion systems

    Physica D 416:132735.

    https://doi.org/10.1016/j.physd.2020.132735

    • Google Scholar
45. 1. Cherkashin AA
    2. Vanag VK
    3. Epstein IR

    (2008) Discontinuously propagating waves in the bathoferroin-catalyzed Belousov–Zhabotinsky reaction incorporated into a microemulsion

    The Journal of Chemical Physics 128:204508.

    https://doi.org/10.1063/1.2924119

    • PubMed
    • Google Scholar
46. 1. Chhabra ES
    2. Higgs HN

    (2007) The many faces of actin: Matching assembly factors with cellular structures

    Nature Cell Biology 9:1110–1121.

    https://doi.org/10.1038/ncb1007-1110

    • PubMed
    • Google Scholar
47. 1. Crawford JD

    (1991) Introduction to bifurcation theory

    Reviews of Modern Physics 63:991–1037.

    https://doi.org/10.1103/RevModPhys.63.991

    • Google Scholar
48. 1. Cross MC
    2. Hohenberg PC

    (1993) Pattern formation outside of equilibrium

    Reviews of Modern Physics 65:851–1112.

    https://doi.org/10.1103/RevModPhys.65.851

    • Google Scholar
49. 1. Desroches M
    2. Rinzel J
    3. Rodrigues S

    (2022) Classification of bursting patterns: A tale of two ducks

    PLOS Computational Biology 18:e1009752.

    https://doi.org/10.1371/journal.pcbi.1009752

    • PubMed
    • Google Scholar
50. 1. Devreotes PN
    2. Bhattacharya S
    3. Edwards M
    4. Iglesias PA
    5. Lampert T
    6. Miao Y

    (2017) Excitable signal transduction networks in directed cell migration

    Annual Review of Cell and Developmental Biology 33:103–125.

    https://doi.org/10.1146/annurev-cellbio-100616-060739

    • PubMed
    • Google Scholar
51. 1. Diz-Muñoz A
    2. Fletcher DA
    3. Weiner OD

    (2013) Use the force: Membrane tension as an organizer of cell shape and motility

    Trends in Cell Biology 23:47–53.

    https://doi.org/10.1016/j.tcb.2012.09.006

    • PubMed
    • Google Scholar
52. 1. Diz-Muñoz A
    2. Thurley K
    3. Chintamen S
    4. Altschuler SJ
    5. Wu LF
    6. Fletcher DA
    7. Weiner OD

    (2016) Membrane tension acts through PLD2 and mTORC2 to limit actin network assembly during neutrophil migration

    PLOS Biology 14:e1002474.

    https://doi.org/10.1371/journal.pbio.1002474

    • PubMed
    • Google Scholar
53. 1. Döbereiner HG
    2. Dubin-Thaler BJ
    3. Hofman JM
    4. Xenias HS
    5. Sims TN
    6. Giannone G
    7. Dustin ML
    8. Wiggins CH
    9. Sheetz MP

    (2006) Lateral membrane waves constitute a universal dynamic pattern of motile cells

    Physical Review Letters 97:038102.

    https://doi.org/10.1103/PhysRevLett.97.038102

    • PubMed
    • Google Scholar
54. 1. Dogterom M
    2. Koenderink GH

    (2019) Actin–microtubule crosstalk in cell biology

    Nature Reviews. Molecular Cell Biology 20:38–54.

    https://doi.org/10.1038/s41580-018-0067-1

    • PubMed
    • Google Scholar
55. 1. Dominguez R
    2. Holmes KC

    (2011) Actin structure and function

    Annual Review of Biophysics 40:169–186.

    https://doi.org/10.1146/annurev-biophys-042910-155359

    • PubMed
    • Google Scholar
56. 1. Doubrovinski K
    2. Kruse K

    (2011) Cell motility resulting from spontaneous polymerization waves

    Physical Review Letters 107:258103.

    https://doi.org/10.1103/PhysRevLett.107.258103

    • PubMed
    • Google Scholar
57. 1. Dreher A
    2. Aranson IS
    3. Kruse K

    (2014) Spiral actin-polymerization waves can generate amoeboidal cell crawling

    New Journal of Physics 16:055007.

    https://doi.org/10.1088/1367-2630/16/5/055007

    • Google Scholar
58. 1. Driscoll MK
    2. McCann C
    3. Kopace R
    4. Homan T
    5. Fourkas JT
    6. Parent C
    7. Losert W

    (2012) Cell shape dynamics: From waves to migration

    PLOS Computational Biology 8:e1002392.

    https://doi.org/10.1371/journal.pcbi.1002392

    • PubMed
    • Google Scholar
59. 1. Driscoll MK
    2. Sun X
    3. Guven C
    4. Fourkas JT
    5. Losert W

    (2014) Cellular contact guidance through dynamic sensing of nanotopography

    ACS Nano 8:3546–3555.

    https://doi.org/10.1021/nn406637c

    • PubMed
    • Google Scholar
60. 1. Driscoll MK
    2. Losert W
    3. Jacobson K
    4. Kapustina M

    (2015) Spatiotemporal relationships between the cell shape and the actomyosin cortex of periodically protruding cells

    Cytoskeleton 72:268–281.

    https://doi.org/10.1002/cm.21229

    • PubMed
    • Google Scholar
61. 1. Ecke M
    2. Prassler J
    3. Tanribil P
    4. Müller-Taubenberger A
    5. Körber S
    6. Faix J
    7. Gerisch G

    (2020) Formins specify membrane patterns generated by propagating actin waves

    Molecular Biology of the Cell 31:373–385.

    https://doi.org/10.1091/mbc.E19-08-0460

    • PubMed
    • Google Scholar
62. Book

    1. Edelstein-Keshet L

    (2005) Mathematical Models in Biology

    SIAM.

    https://doi.org/10.1137/1.9780898719147

    • Google Scholar
63. 1. Enculescu M
    2. Sabouri-Ghomi M
    3. Danuser G
    4. Falcke M

    (2010) Modeling of protrusion phenotypes driven by the actin-membrane interaction

    Biophysical Journal 98:1571–1581.

    https://doi.org/10.1016/j.bpj.2009.12.4311

    • PubMed
    • Google Scholar
64. Book

    1. Ermentrout GB
    2. Terman DH

    (2010) Mathematical Foundations of Neuroscience

    New York, NY: springer.

    https://doi.org/10.1007/978-0-387-87708-2

    • Google Scholar
65. 1. Etienne-Manneville S
    2. Hall A

    (2002) Rho GTPases in cell biology

    Nature 420:629–635.

    https://doi.org/10.1038/nature01148

    • PubMed
    • Google Scholar
66. 1. Fang X
    2. Kruse K
    3. Lu T
    4. Wang J

    (2019) Nonequilibrium physics in biology

    Reviews of Modern Physics 91:5004.

    https://doi.org/10.1103/RevModPhys.91.045004

    • Google Scholar
67. 1. Fitzhugh R

    (1961) Impulses and physiological states in theoretical models of nerve membrane

    Biophysical Journal 1:445–466.

    https://doi.org/10.1016/s0006-3495(61)86902-6

    • PubMed
    • Google Scholar
68. 1. Flemming S
    2. Font F
    3. Alonso S
    4. Beta C

    (2020) How cortical waves drive fission of motile cells

    PNAS 117:6330–6338.

    https://doi.org/10.1073/pnas.1912428117

    • Google Scholar
69. 1. Flyvbjerg H
    2. Jobs E
    3. Leibler S

    (1996) Kinetics of self-assembling microtubules: An "inverse problem" in biochemistry

    PNAS 93:5975–5979.

    https://doi.org/10.1073/pnas.93.12.5975

    • Google Scholar
70. 1. Fukushima S
    2. Matsuoka S
    3. Ueda M

    (2019) Excitable dynamics of Ras triggers spontaneous symmetry breaking of PIP₃ signaling in motile cells

    Journal of Cell Science 132:jcs224121.

    https://doi.org/10.1242/jcs.224121

    • PubMed
    • Google Scholar
71. 1. Ganusov VV

    (2016) Strong inference in mathematical modeling: A method for robust science in the twenty-first century

    Frontiers in Microbiology 7:1131.

    https://doi.org/10.3389/fmicb.2016.01131

    • PubMed
    • Google Scholar
72. 1. Gebert J
    2. Radde N
    3. Weber GW

    (2007) Modeling gene regulatory networks with piecewise linear differential equations

    European Journal of Operational Research 181:1148–1165.

    https://doi.org/10.1016/j.ejor.2005.11.044

    • Google Scholar
73. 1. Gerhardt M
    2. Ecke M
    3. Walz M
    4. Stengl A
    5. Beta C
    6. Gerisch G

    (2014a) Actin and PIP₃ waves in giant cells reveal the inherent length scale of an excited state

    Journal of Cell Science 127:4507–4517.

    https://doi.org/10.1242/jcs.156000

    • PubMed
    • Google Scholar
74. 1. Gerhardt M
    2. Walz M
    3. Beta C

    (2014b) Signaling in chemotactic amoebae remains spatially confined to stimulated membrane regions

    Journal of Cell Science 127:5115–5125.

    https://doi.org/10.1242/jcs.161133

    • PubMed
    • Google Scholar
75. 1. Gerisch G
    2. Bretschneider T
    3. Müller-Taubenberger A
    4. Simmeth E
    5. Ecke M
    6. Diez S
    7. Anderson K

    (2004) Mobile actin clusters and traveling waves in cells recovering from actin depolymerization

    Biophysical Journal 87:3493–3503.

    https://doi.org/10.1529/biophysj.104.047589

    • PubMed
    • Google Scholar
76. 1. Gerisch G
    2. Ecke M
    3. Schroth-Diez B
    4. Gerwig S
    5. Engel U
    6. Maddera L
    7. Clarke M

    (2009) Self-organizing actin waves as planar phagocytic cup structures

    Cell Adhesion & Migration 3:373–382.

    https://doi.org/10.4161/cam.3.4.9708

    • Google Scholar
77. 1. Gerisch G

    (2010) Self-organizing actin waves that simulate phagocytic cup structures

    PMC Biophysics 3:7.

    https://doi.org/10.1186/1757-5036-3-7

    • PubMed
    • Google Scholar
78. 1. Gerisch G
    2. Ecke M
    3. Wischnewski D
    4. Schroth-Diez B

    (2011) Different modes of state transitions determine pattern in the phosphatidylinositide-actin system

    BMC Cell Biology 12:42.

    https://doi.org/10.1186/1471-2121-12-42

    • PubMed
    • Google Scholar
79. 1. Gerisch G
    2. Schroth-Diez B
    3. Müller-Taubenberger A
    4. Ecke M

    (2012) PIP₃ waves and PTEN dynamics in the emergence of cell polarity

    Biophysical Journal 103:1170–1178.

    https://doi.org/10.1016/j.bpj.2012.08.004

    • PubMed
    • Google Scholar
80. 1. Ghabache E
    2. Cao Y
    3. Miao Y
    4. Groisman A
    5. Devreotes PN
    6. Rappel W-J

    (2021) Coupling traction force patterns and actomyosin wave dynamics reveals mechanics of cell motion

    Molecular Systems Biology 17:e10505.

    https://doi.org/10.15252/msb.202110505

    • PubMed
    • Google Scholar
81. 1. Gholami A
    2. Enculescu M
    3. Falcke M

    (2012) Membrane waves driven by forces from actin filaments

    New Journal of Physics 14:115002.

    https://doi.org/10.1088/1367-2630/14/11/115002

    • Google Scholar
82. 1. Giannone G
    2. Dubin-Thaler BJ
    3. Döbereiner HG
    4. Kieffer N
    5. Bresnick AR
    6. Sheetz MP

    (2004) Periodic lamellipodial contractions correlate with rearward actin waves

    Cell 116:431–443.

    https://doi.org/10.1016/s0092-8674(04)00058-3

    • PubMed
    • Google Scholar
83. 1. Giannone G
    2. Dubin-Thaler BJ
    3. Rossier O
    4. Cai Y
    5. Chaga O
    6. Jiang G
    7. Beaver W
    8. Döbereiner H-G
    9. Freund Y
    10. Borisy G
    11. Sheetz MP

    (2007) Lamellipodial actin mechanically links myosin activity with adhesion-site formation

    Cell 128:561–575.

    https://doi.org/10.1016/j.cell.2006.12.039

    • Google Scholar
84. 1. Glogauer M
    2. Hartwig J
    3. Stossel T

    (2000) Two pathways through Cdc42 couple the N-formyl receptor to actin nucleation in permeabilized human neutrophils

    The Journal of Cell Biology 150:785–796.

    https://doi.org/10.1083/jcb.150.4.785

    • PubMed
    • Google Scholar
85. 1. Goryachev AB
    2. Leda M
    3. Miller AL
    4. von Dassow G
    5. Bement WM

    (2016) How to make a static cytokinetic furrow out of traveling excitable waves

    Small GTPases 7:65–70.

    https://doi.org/10.1080/21541248.2016.1168505

    • PubMed
    • Google Scholar
86. 1. Gov NS

    (2018) Guided by curvature: Shaping cells by coupling curved membrane proteins and cytoskeletal forces

    Philosophical Transactions of the Royal Society B 373:20170115.

    https://doi.org/10.1098/rstb.2017.0115

    • Google Scholar
87. 1. Graessl M
    2. Koch J
    3. Calderon A
    4. Kamps D
    5. Banerjee S
    6. Mazel T
    7. Schulze N
    8. Jungkurth JK
    9. Patwardhan R
    10. Solouk D
    11. Hampe N
    12. Hoffmann B
    13. Dehmelt L
    14. Nalbant P

    (2017) An excitable Rho GTPase signaling network generates dynamic subcellular contraction patterns

    Journal of Cell Biology 216:4271–4285.

    https://doi.org/10.1083/jcb.201706052

    • Google Scholar
88. Book

    1. Guckenheimer J
    2. Holmes P

    (2013) Nonlinear Oscillations, Dynamical Systems, and Bifurcations of Vector Fields

    Springer Science & Business Media.

    https://doi.org/10.1007/978-1-4612-1140-2

    • Google Scholar
89. 1. Guetta-Terrier C
    2. Monzo P
    3. Zhu J
    4. Long H
    5. Venkatraman L
    6. Zhou Y
    7. Wang P
    8. Chew SY
    9. Mogilner A
    10. Ladoux B
    11. Gauthier NC

    (2015) Protrusive waves guide 3D cell migration along nanofibers

    The Journal of Cell Biology 211:683–701.

    https://doi.org/10.1083/jcb.201501106

    • PubMed
    • Google Scholar
90. 1. Hadjitheodorou A
    2. Bell GRR
    3. Ellett F
    4. Shastry S
    5. Irimia D
    6. Collins SR
    7. Theriot JA

    (2021) Directional reorientation of migrating neutrophils is limited by suppression of receptor input signaling at the cell rear through myosin II activity

    Nature Communications 12:6619.

    https://doi.org/10.1038/s41467-021-26622-z

    • PubMed
    • Google Scholar
91. 1. Halatek J
    2. Brauns F
    3. Frey E

    (2018) Self-organization principles of intracellular pattern formation

    Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 373:1747.

    https://doi.org/10.1098/rstb.2017.0107

    • PubMed
    • Google Scholar
92. 1. Halatek J
    2. Frey E

    (2018) Rethinking pattern formation in reaction–diffusion systems

    Nature Physics 14:507–514.

    https://doi.org/10.1038/s41567-017-0040-5

    • Google Scholar
93. 1. Hata S
    2. Nakao H
    3. Mikhailov AS

    (2014) Sufficient conditions for wave instability in three-component reaction–diffusion systems

    Progress of Theoretical and Experimental Physics 2014:13A01.

    https://doi.org/10.1093/ptep/ptt102

    • Google Scholar
94. 1. Helenius J
    2. Ecke M
    3. Müller DJ
    4. Gerisch G

    (2018) Oscillatory switches of dorso-ventral polarity in cells confined between two surfaces

    Biophysical Journal 115:150–162.

    https://doi.org/10.1016/j.bpj.2018.05.025

    • PubMed
    • Google Scholar
95. 1. Hilgemann DW

    (2007) Local PIP₂ signals: When, where, and how

    Pflugers Archiv 455:55–67.

    https://doi.org/10.1007/s00424-007-0280-9

    • PubMed
    • Google Scholar
96. 1. Hodge RG
    2. Ridley AJ

    (2016) Regulating Rho GTPases and their regulators

    Nature Reviews. Molecular Cell Biology 17:496–510.

    https://doi.org/10.1038/nrm.2016.67

    • PubMed
    • Google Scholar
97. 1. Hodgkin AL
    2. Huxley AF

    (1952) A quantitative description of membrane current and its application to conduction and excitation in nerve

    The Journal of Physiology 117:500–544.

    https://doi.org/10.1113/jphysiol.1952.sp004764

    • Google Scholar
98. 1. Holmes WR
    2. Carlsson AE
    3. Edelstein-Keshet L

    (2012) Regimes of wave type patterning driven by refractory actin feedback: Transition from static polarization to dynamic wave behaviour

    Physical Biology 9:046005.

    https://doi.org/10.1088/1478-3975/9/4/046005

    • PubMed
    • Google Scholar
99. 1. Holmes WR

    (2014) An efficient, nonlinear stability analysis for detecting pattern formation in reaction diffusion systems

    Bulletin of Mathematical Biology 76:157–183.

    https://doi.org/10.1007/s11538-013-9914-6

    • PubMed
    • Google Scholar
100. 1. Holmes WR
     2. Mata MA
     3. Edelstein-Keshet L

     (2015) Local perturbation analysis: A computational tool for biophysical reaction-diffusion models

     Biophysical Journal 108:230–236.

     https://doi.org/10.1016/j.bpj.2014.11.3457

     • PubMed
     • Google Scholar
101. 1. Holmes WR
     2. Edelstein-Keshet L

     (2016) Analysis of a minimal Rho-GTPase circuit regulating cell shape

     Physical Biology 13:046001.

     https://doi.org/10.1088/1478-3975/13/4/046001

     • PubMed
     • Google Scholar
102. 1. Holmes WR
     2. Golding AE
     3. Bement WM
     4. Edelstein-Keshet L

     (2016) A mathematical model of GTPase pattern formation during single-cell wound repair

     Interface Focus 6:20160032.

     https://doi.org/10.1098/rsfs.2016.0032

     • PubMed
     • Google Scholar
103. 1. Honda G
     2. Saito N
     3. Fujimori T
     4. Hashimura H
     5. Nakamura MJ
     6. Nakajima A
     7. Sawai S

     (2021) Microtopographical guidance of macropinocytic signaling patches

     PNAS 118:50.

     https://doi.org/10.1073/pnas.2110281118

     • Google Scholar
104. 1. Hoon JL
     2. Wong WK
     3. Koh CG

     (2012) Functions and regulation of circular dorsal ruffles

     Molecular and Cellular Biology 32:4246–4257.

     https://doi.org/10.1128/MCB.00551-12

     • PubMed
     • Google Scholar
105. 1. Hörning M
     2. Shibata T

     (2019) Three-dimensional cell geometry controls excitable membrane signaling in Dictyostelium cells

     Biophysical Journal 116:372–382.

     https://doi.org/10.1016/j.bpj.2018.12.012

     • PubMed
     • Google Scholar
106. 1. Houk AR
     2. Jilkine A
     3. Mejean CO
     4. Boltyanskiy R
     5. Dufresne ER
     6. Angenent SB
     7. Altschuler SJ
     8. Wu LF
     9. Weiner OD

     (2012) Membrane tension maintains cell polarity by confining signals to the leading edge during neutrophil migration

     Cell 148:175–188.

     https://doi.org/10.1016/j.cell.2011.10.050

     • PubMed
     • Google Scholar
107. 1. Huang CH
     2. Tang M
     3. Shi C
     4. Iglesias PA
     5. Devreotes PN

     (2013) An excitable signal Integrator couples to an idling cytoskeletal oscillator to drive cell migration

     Nature Cell Biology 15:1307–1316.

     https://doi.org/10.1038/ncb2859

     • PubMed
     • Google Scholar
108. 1. Imoto D
     2. Saito N
     3. Nakajima A
     4. Honda G
     5. Ishida M
     6. Sugita T
     7. Ishihara S
     8. Katagiri K
     9. Okimura C
     10. Iwadate Y
     11. Sawai S

     (2021) Comparative mapping of crawling-cell morphodynamics in deep learning-based feature space

     PLOS Computational Biology 17:e1009237.

     https://doi.org/10.1371/journal.pcbi.1009237

     • PubMed
     • Google Scholar
109. 1. Inagaki N
     2. Katsuno H

     (2017) Actin waves: Origin of cell polarization and migration

     Trends in Cell Biology 27:515–526.

     https://doi.org/10.1016/j.tcb.2017.02.003

     • PubMed
     • Google Scholar
110. 1. Iron D
     2. Ward MJ

     (2000) The dynamics of boundary spikes for a nonlocal reaction-diffusion model

     European Journal of Applied Mathematics 11:491–514.

     https://doi.org/10.1017/S0956792500004253

     • Google Scholar
111. 1. Itoh T
     2. Hasegawa J

     (2013) Mechanistic insights into the regulation of circular dorsal ruffle formation

     Journal of Biochemistry 153:21–29.

     https://doi.org/10.1093/jb/mvs138

     • PubMed
     • Google Scholar
112. 1. Jasnin M
     2. Beck F
     3. Ecke M
     4. Fukuda Y
     5. Martinez-Sanchez A
     6. Baumeister W
     7. Gerisch G

     (2019) The architecture of traveling actin waves revealed by cryo-electron tomography

     Structure 27:1211–1223.

     https://doi.org/10.1016/j.str.2019.05.009

     • PubMed
     • Google Scholar
113. 1. Jiang BH
     2. Liu LZ

     (2008) PI3K/PTEN signaling in tumorigenesis and angiogenesis

     Biochimica et Biophysica Acta - Proteins and Proteomics 1784:150–158.

     https://doi.org/10.1016/j.bbapap.2007.09.008

     • Google Scholar
114. 1. Jilkine A
     2. Edelstein-Keshet L

     (2011) A comparison of mathematical models for polarization of single eukaryotic cells in response to guided cues

     PLOS Computational Biology 7:e1001121.

     https://doi.org/10.1371/journal.pcbi.1001121

     • PubMed
     • Google Scholar
115. 1. Kabsch W
     2. Vandekerckhove J

     (1992) Structure and function of actin

     Annual Review of Biophysics and Biomolecular Structure 21:49–76.

     https://doi.org/10.1146/annurev.bb.21.060192.000405

     • PubMed
     • Google Scholar
116. 1. Karma A

     (2013) Physics of cardiac arrhythmogenesis

     Annual Review of Condensed Matter Physics 4:313–337.

     https://doi.org/10.1146/annurev-conmatphys-020911-125112

     • Google Scholar
117. Book

     1. Keener J
     2. Sneyd J

     (2009) Mathematical Physiology: I: Cellular Physiology

     New York, NY: Springer.

     https://doi.org/10.1007/978-0-387-75847-3

     • Google Scholar
118. 1. Khamviwath V
     2. Hu J
     3. Othmer HG

     (2013) A continuum model of actin waves in Dictyostelium discoideum

     PLOS ONE 8:e64272.

     https://doi.org/10.1371/journal.pone.0064272

     • PubMed
     • Google Scholar
119. 1. Kholodenko BN

     (2006) Cell-signalling dynamics in time and space

     Nature Reviews. Molecular Cell Biology 7:165–176.

     https://doi.org/10.1038/nrm1838

     • PubMed
     • Google Scholar
120. 1. Killich T
     2. Plath PJ
     3. Wei X
     4. Bultmann H
     5. Rensing L
     6. Vicker MG

     (1993) The locomotion, shape and pseudopodial dynamics of unstimulated Dictyostelium cells are not random

     Journal of Cell Science 106 (Pt 4):1005–1013.

     https://doi.org/10.1242/jcs.106.4.1005

     • PubMed
     • Google Scholar
121. 1. Knobloch E

     (1986) Oscillatory convection in binary mixtures

     Physical Review A 34:1538–1549.

     https://doi.org/10.1103/PhysRevA.34.1538

     • Google Scholar
122. Book

     1. Knobloch E

     (1992) Pattern Formation in Complex Dissipative Systems

     World scientific.

     https://doi.org/10.1142/9789814538039

     • Google Scholar
123. 1. Knobloch E
     2. Uecker H
     3. Yochelis A

     (2021) Origin of jumping oscillons in an excitable reaction-diffusion system

     Physical Review E 104:L062201.

     https://doi.org/10.1103/PhysRevE.104.L062201

     • PubMed
     • Google Scholar
124. 1. Knoch F
     2. Tarantola M
     3. Bodenschatz E
     4. Rappel WJ

     (2014) Modeling self-organized spatio-temporal patterns of PIP₃ and PTEN during spontaneous cell polarization

     Physical Biology 11:046002.

     https://doi.org/10.1088/1478-3975/11/4/046002

     • PubMed
     • Google Scholar
125. 1. Kobayashi R
     2. Ohta T
     3. Hayase Y

     (1994) Self-organized pulse generator in a reaction-diffusion system

     Physical Review E 50:R3291–R3294.

     https://doi.org/10.1103/PhysRevE.50.R3291

     • Google Scholar
126. 1. Kobayashi R
     2. Ohta T
     3. Hayase Y

     (1995) Self-organized pulse generator

     Physica D 84:162–170.

     https://doi.org/10.1016/0167-2789(95)00027-2

     • Google Scholar
127. 1. Kölsch V
     2. Charest PG
     3. Firtel RA

     (2008) The regulation of cell motility and chemotaxis by phospholipid signaling

     Journal of Cell Science 121:551–559.

     https://doi.org/10.1242/jcs.023333

     • PubMed
     • Google Scholar
128. 1. Kortholt A
     2. van Haastert PJM

     (2008) Highlighting the role of Ras and Rap during Dictyostelium chemotaxis

     Cellular Signalling 20:1415–1422.

     https://doi.org/10.1016/j.cellsig.2008.02.006

     • PubMed
     • Google Scholar
129. 1. Kulawiak DA
     2. Löber J
     3. Bär M
     4. Engel H

     (2019) Active poroelastic two-phase model for the motion of Physarum microplasmodia

     PLOS ONE 14:e0217447.

     https://doi.org/10.1371/journal.pone.0217447

     • PubMed
     • Google Scholar
130. Book

     1. Kuznetsov YA

     (2004) Elements of Applied Bifurcation Theory

     New York, NY: Springer.

     https://doi.org/10.1007/978-1-4757-3978-7

     • Google Scholar
131. 1. Lam Hui K
     2. Kwak SI
     3. Upadhyaya A

     (2014) Adhesion-dependent modulation of actin dynamics in Jurkat T cells

     Cytoskeleton 71:119–135.

     https://doi.org/10.1002/cm.21156

     • PubMed
     • Google Scholar
132. 1. Landino J
     2. Leda M
     3. Michaud A
     4. Swider ZT
     5. Prom M
     6. Field CM
     7. Bement WM
     8. Vecchiarelli AG
     9. Goryachev AB
     10. Miller AL

     (2021) Rho and F-actin self-organize within an artificial cell cortex

     Current Biology 31:5613–5621.

     https://doi.org/10.1016/j.cub.2021.10.021

     • PubMed
     • Google Scholar
133. 1. Lange M
     2. Prassler J
     3. Ecke M
     4. Müller-Taubenberger A
     5. Gerisch G

     (2016) Local Ras activation, PTEN pattern, and global actin flow in the chemotactic responses of oversized cells

     Journal of Cell Science 129:3462–3472.

     https://doi.org/10.1242/jcs.191148

     • Google Scholar
134. 1. Le Goff T
     2. Liebchen B
     3. Marenduzzo D

     (2016) Pattern formation in polymerizing actin flocks: Spirals, spots, and waves without nonlinear chemistry

     Physical Review Letters 117:238002.

     https://doi.org/10.1103/PhysRevLett.117.238002

     • PubMed
     • Google Scholar
135. Book

     1. Lin CC
     2. Segel LA

     (1988) Mathematics Applied to Deterministic Problems in the Natural Sciences

     SIAM.

     https://doi.org/10.1137/1.9781611971347

     • Google Scholar
136. 1. Liu Y
     2. Rens EG
     3. Edelstein-Keshet L

     (2021) Spots, stripes, and spiral waves in models for static and motile cells

     Journal of Mathematical Biology 82:1–38.

     https://doi.org/10.1007/s00285-021-01550-0

     • Google Scholar
137. 1. Lou SS
     2. Diz-Muñoz A
     3. Weiner OD
     4. Fletcher DA
     5. Theriot JA

     (2015) Myosin light chain kinase regulates cell polarization independently of membrane tension or Rho kinase

     Journal of Cell Biology 209:275–288.

     https://doi.org/10.1083/jcb.201409001

     • Google Scholar
138. Preprint

     1. Lutton JE
     2. Coker HLE
     3. Paschke P
     4. Munn CJ
     5. King JS
     6. Bretschneider T
     7. Kay RR

     (2022) The Formation and Closure of Macropinocytic Cups in a Model System

     bioRxiv.

     https://doi.org/10.1101/2022.10.07.511330

     • Google Scholar
139. 1. Machacek M
     2. Danuser G

     (2006) Morphodynamic profiling of protrusion phenotypes

     Biophysical Journal 90:1439–1452.

     https://doi.org/10.1529/biophysj.105.070383

     • PubMed
     • Google Scholar
140. 1. Machesky LM
     2. Insall RH

     (1999) Signaling to actin dynamics

     The Journal of Cell Biology 146:267–272.

     https://doi.org/10.1083/jcb.146.2.267

     • PubMed
     • Google Scholar
141. 1. Marchesin V
     2. Castro-Castro A
     3. Lodillinsky C
     4. Castagnino A
     5. Cyrta J
     6. Bonsang-Kitzis H
     7. Fuhrmann L
     8. Irondelle M
     9. Infante E
     10. Montagnac G
     11. Reyal F
     12. Vincent-Salomon A
     13. Chavrier P

     (2015) ARF6–JIP3/4 regulate endosomal tubules for MT1-MMP exocytosis in cancer invasion

     The Journal of Cell Biology 211:339–358.

     https://doi.org/10.1083/jcb.201506002

     • PubMed
     • Google Scholar
142. 1. Marée AFM
     2. Jilkine A
     3. Dawes A
     4. Grieneisen VA
     5. Edelstein-Keshet L

     (2006) Polarization and movement of keratocytes: A multiscale modelling approach

     Bulletin of Mathematical Biology 68:1169–1211.

     https://doi.org/10.1007/s11538-006-9131-7

     • PubMed
     • Google Scholar
143. 1. Marée AFM
     2. Grieneisen VA
     3. Edelstein-Keshet L

     (2012) How cells integrate complex stimuli: The effect of feedback from phosphoinositides and cell shape on cell polarization and motility

     PLOS Computational Biology 8:e1002402.

     https://doi.org/10.1371/journal.pcbi.1002402

     • PubMed
     • Google Scholar
144. 1. Masters TA
     2. Sheetz MP
     3. Gauthier NC

     (2016) F-actin waves, actin cortex disassembly and focal exocytosis driven by actin-Phosphoinositide positive feedback

     Cytoskeleton 73:180–196.

     https://doi.org/10.1002/cm.21287

     • PubMed
     • Google Scholar
145. 1. Mata MA
     2. Dutot M
     3. Edelstein-Keshet L
     4. Holmes WR

     (2013) A model for intracellular actin waves explored by nonlinear local perturbation analysis

     Journal of Theoretical Biology 334:149–161.

     https://doi.org/10.1016/j.jtbi.2013.06.020

     • Google Scholar
146. 1. Meron E

     (1992) Pattern formation in excitable media

     Physics Reports 218:1–66.

     https://doi.org/10.1016/0370-1573(92)90098-K

     • Google Scholar
147. Book

     1. Meron E

     (2015) Nonlinear Physics of Ecosystems

     CRC Press.

     https://doi.org/10.1201/b18360

     • Google Scholar
148. 1. Meshik X
     2. O’Neill PR
     3. Gautam N

     (2018) Optogenetic control of cell migration

     Cell Migration: Methods and Protocols 1749:313–324.

     https://doi.org/10.1007/978-1-4939-7701-7

     • Google Scholar
149. 1. Miao Y
     2. Bhattacharya S
     3. Edwards M
     4. Cai H
     5. Inoue T
     6. Iglesias PA
     7. Devreotes PN

     (2017) Altering the threshold of an excitable signal transduction network changes cell migratory modes

     Nature Cell Biology 19:329–340.

     https://doi.org/10.1038/ncb3495

     • Google Scholar
150. 1. Miao Y
     2. Bhattacharya S
     3. Banerjee T
     4. Abubaker-Sharif B
     5. Long Y
     6. Inoue T
     7. Iglesias PA
     8. Devreotes PN

     (2019) Wave patterns organize cellular protrusions and control cortical dynamics

     Molecular Systems Biology 15:e8585.

     https://doi.org/10.15252/msb.20188585

     • PubMed
     • Google Scholar
151. 1. Michaud A
     2. Swider ZT
     3. Landino J
     4. Leda M
     5. Miller AL
     6. von Dassow G
     7. Goryachev AB
     8. Bement WM

     (2021) Cortical excitability and cell division

     Current Biology 31:R553–R559.

     https://doi.org/10.1016/j.cub.2021.02.053

     • Google Scholar
152. 1. Michaud A
     2. Leda M
     3. Swider ZT
     4. Kim S
     5. He J
     6. Landino J
     7. Valley JR
     8. Huisken J
     9. Goryachev AB
     10. von Dassow G
     11. Bement WM

     (2022) A versatile cortical pattern-forming circuit based on Rho, F-actin, Ect2 and RGA-3/4

     The Journal of Cell Biology 221:e202203017.

     https://doi.org/10.1083/jcb.202203017

     • PubMed
     • Google Scholar
153. 1. Michaux JB
     2. Robin FB
     3. McFadden WM
     4. Munro EM

     (2018) Excitable RhoA dynamics drive pulsed contractions in the early C. elegans embryo

     The Journal of Cell Biology 217:4230–4252.

     https://doi.org/10.1083/jcb.201806161

     • PubMed
     • Google Scholar
154. 1. Millius A
     2. Dandekar SN
     3. Houk AR
     4. Weiner OD

     (2009) Neutrophils establish rapid and robust WAVE complex polarity in an actin-dependent fashion

     Current Biology 19:253–259.

     https://doi.org/10.1016/j.cub.2008.12.044

     • PubMed
     • Google Scholar
155. 1. Mogilner A
     2. Oster G

     (2003) Force generation by actin polymerization II: The elastic ratchet and tethered filaments

     Biophysical Journal 84:1591–1605.

     https://doi.org/10.1016/S0006-3495(03)74969-8

     • Google Scholar
156. 1. Moldenhawer T
     2. Moreno E
     3. Schindler D
     4. Flemming S
     5. Holschneider M
     6. Huisinga W
     7. Alonso S
     8. Beta C

     (2022) Spontaneous transitions between amoeboid and keratocyte-like modes of migration

     Frontiers in Cell and Developmental Biology 10:898351.

     https://doi.org/10.3389/fcell.2022.898351

     • PubMed
     • Google Scholar
157. 1. Moreno E
     2. Flemming S
     3. Font F
     4. Holschneider M
     5. Beta C
     6. Alonso S

     (2020) Modeling cell crawling strategies with a bistable model: From amoeboid to fan-shaped cell motion

     Physica D 412:132591.

     https://doi.org/10.1016/j.physd.2020.132591

     • Google Scholar
158. 1. Mori Y
     2. Jilkine A
     3. Edelstein-Keshet L

     (2008) Wave-pinning and cell polarity from a bistable reaction-diffusion system

     Biophysical Journal 94:3684–3697.

     https://doi.org/10.1529/biophysj.107.120824

     • PubMed
     • Google Scholar
159. Book

     1. Murray JD

     (2001)

     Mathematical Biology II: Spatial Models and Biomedical Applications

     New York: Springer.

     • Google Scholar
160. Book

     1. Murray JD

     (2002) Mathematical Biology: I. An Introduction

     New York, NY: Springer.

     https://doi.org/10.1007/b98868

     • Google Scholar
161. 1. Nagumo J
     2. Arimoto S
     3. Yoshizawa S

     (1962) An active pulse transmission line simulating nerve axon

     Proceedings of the IRE 50:2061–2070.

     https://doi.org/10.1109/JRPROC.1962.288235

     • Google Scholar
162. 1. Nakamura N
     2. Shibata T

     (2015) Bifurcation analysis of a self-organizing signaling system for eukaryotic chemotaxis

     Japan Journal of Industrial and Applied Mathematics 32:807–828.

     https://doi.org/10.1007/s13160-015-0185-5

     • Google Scholar
163. 1. Naoz M
     2. Gov NS

     (2020) Cell-substrate patterns driven by curvature-sensitive actin polymerization: Waves and podosomes

     Cells 9:782.

     https://doi.org/10.3390/cells9030782

     • PubMed
     • Google Scholar
164. 1. Neilson MP
     2. Veltman DM
     3. van Haastert PJM
     4. Webb SD
     5. Mackenzie JA
     6. Insall RH

     (2011) Chemotaxis: A feedback-based computational model robustly predicts multiple aspects of real cell behaviour

     PLOS Biology 9:e1000618.

     https://doi.org/10.1371/journal.pbio.1000618

     • PubMed
     • Google Scholar
165. 1. Newby JM
     2. Bressloff PC

     (2010) Quasi-steady state reduction of molecular motor-based models of directed intermittent search

     Bulletin of Mathematical Biology 72:1840–1866.

     https://doi.org/10.1007/s11538-010-9513-8

     • PubMed
     • Google Scholar
166. 1. Nishikawa M
     2. Hörning M
     3. Ueda M
     4. Shibata T

     (2014) Excitable signal transduction induces both spontaneous and directional cell asymmetries in the phosphatidylinositol lipid signaling system for eukaryotic chemotaxis

     Biophysical Journal 106:723–734.

     https://doi.org/10.1016/j.bpj.2013.12.023

     • PubMed
     • Google Scholar
167. 1. Nobes CD
     2. Hall A

     (1995) Rho, RAC, and Cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia

     Cell 81:53–62.

     https://doi.org/10.1016/0092-8674(95)90370-4

     • PubMed
     • Google Scholar
168. 1. Nonaka S
     2. Naoki H
     3. Ishii S

     (2011) A multiphysical model of cell migration integrating reaction–diffusion, membrane and cytoskeleton

     Neural Networks 24:979–989.

     https://doi.org/10.1016/j.neunet.2011.06.009

     • PubMed
     • Google Scholar
169. 1. O’Neill PR
     2. Gautam N

     (2014) Subcellular optogenetic inhibition of G proteins generates signaling gradients and cell migration

     Molecular Biology of the Cell 25:2305–2314.

     https://doi.org/10.1091/mbc.E14-04-0870

     • PubMed
     • Google Scholar
170. 1. O’Neill PR
     2. Kalyanaraman V
     3. Gautam N

     (2016) Subcellular optogenetic activation of Cdc42 controls local and distal signaling to drive immune cell migration

     Molecular Biology of the Cell 27:1442–1450.

     https://doi.org/10.1091/mbc.E15-12-0832

     • PubMed
     • Google Scholar
171. 1. Orth JD
     2. McNiven MA

     (2006) Get off my back! Rapid receptor Internalization through circular dorsal ruffles

     Cancer Research 66:11094–11096.

     https://doi.org/10.1158/0008-5472.CAN-06-3397

     • Google Scholar
172. 1. Pal DS
     2. Li X
     3. Banerjee T
     4. Miao Y
     5. Devreotes PN

     (2019) The excitable signal transduction networks: Movers and shapers of eukaryotic cell migration

     The International Journal of Developmental Biology 63:407–416.

     https://doi.org/10.1387/ijdb.190265pd

     • PubMed
     • Google Scholar
173. 1. Peleg B
     2. Disanza A
     3. Scita G
     4. Gov N

     (2011) Propagating cell-membrane waves driven by curved activators of actin polymerization

     PLOS ONE 6:e18635.

     https://doi.org/10.1371/journal.pone.0018635

     • PubMed
     • Google Scholar
174. Book

     1. Pismen L

     (2006)

     Patterns and Interfaces in Dissipative Dynamics

     Springer.

     • Google Scholar
175. 1. Plazen L
     2. Khadra A

     (2023) Excitable dynamics in a molecularly-explicit model of cell motility: Mixed-mode oscillations and beyond

     Journal of Theoretical Biology 564:111450.

     https://doi.org/10.1016/j.jtbi.2023.111450

     • PubMed
     • Google Scholar
176. 1. Plazen L
     2. Rahbani JA
     3. Brown CM
     4. Khadra A

     (2023) Polarity and mixed-mode oscillations may underlie different patterns of cellular migration

     Scientific Reports 13:4223.

     https://doi.org/10.1038/s41598-023-31042-8

     • PubMed
     • Google Scholar
177. 1. Poincaré H

     (1885) Sur l’équilibre d’une masse fluide animée d’un mouvement de rotation

     Bulletin Astronomique 2:109–118.

     https://doi.org/10.3406/bastr.1885.2592

     • Google Scholar
178. 1. Povarova OI
     2. Antifeeva IA
     3. Fonin AV
     4. Turoverov KK
     5. Kuznetsova IM

     (2023) The role of liquid–liquid phase separation in actin polymerization

     International Journal of Molecular Sciences 24:3281.

     https://doi.org/10.3390/ijms24043281

     • PubMed
     • Google Scholar
179. 1. Radszuweit M
     2. Alonso S
     3. Engel H
     4. Bär M

     (2013) Intracellular mechanochemical waves in an active poroelastic model

     Physical Review Letters 110:138102.

     https://doi.org/10.1103/PhysRevLett.110.138102

     • PubMed
     • Google Scholar
180. 1. Ruthel G
     2. Banker G

     (1998) Actin-dependent anterograde movement of growth-cone-like structures along growing hippocampal axons: A novel form of axonal transport

     Cell Motility and the Cytoskeleton 40:160–173.

     https://doi.org/10.1002/(SICI)1097-0169(1998)40:2<160::AID-CM5>3.0.CO;2-J

     • PubMed
     • Google Scholar
181. 1. Ryan GL
     2. Petroccia HM
     3. Watanabe N
     4. Vavylonis D

     (2012a) Excitable actin dynamics in lamellipodial protrusion and retraction

     Biophysical Journal 102:1493–1502.

     https://doi.org/10.1016/j.bpj.2012.03.005

     • PubMed
     • Google Scholar
182. 1. Ryan GL
     2. Watanabe N
     3. Vavylonis D

     (2012b) A review of models of fluctuating protrusion and retraction patterns at the leading edge of motile cells

     Cytoskeleton 69:195–206.

     https://doi.org/10.1002/cm.21017

     • PubMed
     • Google Scholar
183. 1. Saito N
     2. Sawai S

     (2021) Three-dimensional morphodynamic simulations of macropinocytic cups

     IScience 24:103087.

     https://doi.org/10.1016/j.isci.2021.103087

     • PubMed
     • Google Scholar
184. 1. Sandstede B
     2. Scheel A

     (2000) Absolute and convective instabilities of waves on unbounded and large bounded domains

     Physica D 145:233–277.

     https://doi.org/10.1016/S0167-2789(00)00114-7

     • Google Scholar
185. Preprint

     1. Sandstede B
     2. Scheel A

     (2020) Spiral Waves: Linear and Nonlinear Theory

     arXiv.

     https://arxiv.org/abs/2002.10352

     • Google Scholar
186. 1. Schroth-Diez B
     2. Gerwig S
     3. Ecke M
     4. Hegerl R
     5. Diez S
     6. Gerisch G

     (2009) Propagating waves separate two states of actin organization in living cells

     HFSP Journal 3:412–427.

     https://doi.org/10.2976/1.3239407

     • PubMed
     • Google Scholar
187. Book

     1. Segel LA
     2. Edelstein-Keshet L

     (2013) A Primer on Mathematical Models in Biology

     Philadelphia, PA: SIAM.

     https://doi.org/10.1137/1.9781611972504

     • Google Scholar
188. 1. Sept D
     2. Carlsson AE

     (2014) Modeling large-scale dynamic processes in the cell: Polarization, waves, and division

     Quarterly Reviews of Biophysics 47:221–248.

     https://doi.org/10.1017/S0033583514000079

     • PubMed
     • Google Scholar
189. 1. Shao D
     2. Rappel WJ
     3. Levine H

     (2010) Computational model for cell morphodynamics

     Physical Review Letters 105:108104.

     https://doi.org/10.1103/PhysRevLett.105.108104

     • PubMed
     • Google Scholar
190. 1. Shi C
     2. Huang CH
     3. Devreotes PN
     4. Iglesias PA

     (2013) Interaction of motility, directional sensing, and polarity modules recreates the behaviors of chemotaxing cells

     PLOS Computational Biology 9:e1003122.

     https://doi.org/10.1371/journal.pcbi.1003122

     • PubMed
     • Google Scholar
191. 1. Shibata T
     2. Nishikawa M
     3. Matsuoka S
     4. Ueda M

     (2012) Modeling the self-organized phosphatidylinositol lipid signaling system in chemotactic cells using quantitative image analysis

     Journal of Cell Science 125:5138–5150.

     https://doi.org/10.1242/jcs.108373

     • PubMed
     • Google Scholar
192. 1. Shibata T
     2. Nishikawa M
     3. Matsuoka S
     4. Ueda M

     (2013) Intracellular encoding of spatiotemporal guidance cues in a self-organizing signaling system for chemotaxis in Dictyostelium cells

     Biophysical Journal 105:2199–2209.

     https://doi.org/10.1016/j.bpj.2013.09.024

     • PubMed
     • Google Scholar
193. 1. Shlomovitz R
     2. Gov NS

     (2007) Membrane waves driven by actin and myosin

     Physical Review Letters 98:168103.

     https://doi.org/10.1103/PhysRevLett.98.168103

     • PubMed
     • Google Scholar
194. 1. Spiering D
     2. Hodgson L

     (2011) Dynamics of the Rho-family small GTPases in actin regulation and motility

     Cell Adhesion & Migration 5:170–180.

     https://doi.org/10.4161/cam.5.2.14403

     • Google Scholar
195. 1. Staddon MF
     2. Munro EM
     3. Banerjee S

     (2022) Pulsatile contractions and pattern formation in excitable actomyosin cortex

     PLOS Computational Biology 18:e1009981.

     https://doi.org/10.1371/journal.pcbi.1009981

     • PubMed
     • Google Scholar
196. 1. Stankevicins L
     2. Ecker N
     3. Terriac E
     4. Maiuri P
     5. Schoppmeyer R
     6. Vargas P
     7. Lennon-Duménil AM
     8. Piel M
     9. Qu B
     10. Hoth M
     11. Kruse K
     12. Lautenschläger F

     (2020) Deterministic actin waves as generators of cell polarization cues

     PNAS 117:826–835.

     https://doi.org/10.1073/pnas.1907845117

     • PubMed
     • Google Scholar
197. 1. Stich M
     2. Mikhailov AS
     3. Kuramoto Y

     (2009) Self-organized pacemakers and bistability of pulses in an excitable medium

     Physical Review E 79:026110.

     https://doi.org/10.1103/PhysRevE.79.026110

     • PubMed
     • Google Scholar
198. Book

     1. Strogatz SH

     (2018) Nonlinear Dynamics and Chaos with Student Solutions Manual

     Florida, United States: CRC press.

     https://doi.org/10.1201/9780429399640

     • Google Scholar
199. 1. Sun X
     2. Driscoll MK
     3. Guven C
     4. Das S
     5. Parent CA
     6. Fourkas JT
     7. Losert W

     (2015) Asymmetric nanotopography biases cytoskeletal dynamics and promotes unidirectional cell guidance

     PNAS 112:12557–12562.

     https://doi.org/10.1073/pnas.1502970112

     • Google Scholar
200. 1. Svitkina TM
     2. Borisy GG

     (1999) Arp2/3 complex and actin depolymerizing factor/cofilin in dendritic organization and treadmilling of actin filament array in lamellipodia

     Journal of Cell Biology 145:1009–1026.

     https://doi.org/10.1083/jcb.145.5.1009

     • Google Scholar
201. 1. Swanson JA

     (2008) Shaping cups into phagosomes and macropinosomes

     Nature Reviews Molecular Cell Biology 9:639–649.

     https://doi.org/10.1038/nrm2447

     • Google Scholar
202. 1. Swider ZT
     2. Michaud A
     3. Leda M
     4. Landino J
     5. Goryachev AB
     6. Bement WM
     7. Wallingford J

     (2022) Cell cycle and developmental control of cortical excitability in Xenopus laevis

     Molecular Biology of the Cell 33:ar73.

     https://doi.org/10.1091/mbc.E22-01-0025

     • PubMed
     • Google Scholar
203. 1. Tamemoto N
     2. Noguchi H

     (2020) Pattern formation in reaction–diffusion system on membrane with mechanochemical feedback

     Scientific Reports 10:19582.

     https://doi.org/10.1038/s41598-020-76695-x

     • PubMed
     • Google Scholar
204. 1. Tamemoto N
     2. Noguchi H

     (2021) Reaction-diffusion waves coupled with membrane curvature

     Soft Matter 17:6589–6596.

     https://doi.org/10.1039/d1sm00540e

     • PubMed
     • Google Scholar
205. 1. Tan TH
     2. Liu J
     3. Miller PW
     4. Tekant M
     5. Dunkel J
     6. Fakhri N

     (2020) Topological turbulence in the membrane of a living cell

     Nature Physics 16:657–662.

     https://doi.org/10.1038/s41567-020-0841-9

     • Google Scholar
206. 1. Taniguchi D
     2. Ishihara S
     3. Oonuki T
     4. Honda-Kitahara M
     5. Kaneko K
     6. Sawai S

     (2013) Phase geometries of two-dimensional excitable waves govern self-organized morphodynamics of amoeboid cells

     PNAS 110:5016–5021.

     https://doi.org/10.1073/pnas.1218025110

     • Google Scholar
207. 1. Tarama M
     2. Mori K
     3. Yamamoto R

     (2022) Mechanochemical subcellular-element model of crawling cells

     Frontiers in Cell and Developmental Biology 10:1046053.

     https://doi.org/10.3389/fcell.2022.1046053

     • PubMed
     • Google Scholar
208. 1. Tomba C
     2. Braïni C
     3. Bugnicourt G
     4. Cohen F
     5. Friedrich BM
     6. Gov NS
     7. Villard C

     (2017) Geometrical determinants of neuronal actin waves

     Frontiers in Cellular Neuroscience 11:86.

     https://doi.org/10.3389/fncel.2017.00086

     • PubMed
     • Google Scholar
209. 1. Toriyama M
     2. Shimada T
     3. Kim KB
     4. Mitsuba M
     5. Nomura E
     6. Katsuta K
     7. Sakumura Y
     8. Roepstorff P
     9. Inagaki N

     (2006) Shootin1: A protein involved in the organization of an asymmetric signal for neuronal polarization

     The Journal of Cell Biology 175:147–157.

     https://doi.org/10.1083/jcb.200604160

     • PubMed
     • Google Scholar
210. 1. Tran DX
     2. Sato D
     3. Yochelis A
     4. Weiss JN
     5. Garfinkel A
     6. Qu Z

     (2009) Bifurcation and chaos in a model of cardiac early afterdepolarizations

     Physical Review Letters 102:258103.

     https://doi.org/10.1103/PhysRevLett.102.258103

     • PubMed
     • Google Scholar
211. 1. Tyson JJ
     2. Keener JP

     (1988) Singular perturbation theory of traveling waves in excitable media (a review)

     Physica D 32:327–361.

     https://doi.org/10.1016/0167-2789(88)90062-0

     • Google Scholar
212. 1. Tyson JJ
     2. Chen KC
     3. Novak B

     (2003) Sniffers, Buzzers, Toggles and blinkers: Dynamics of regulatory and signaling pathways in the cell

     Current Opinion in Cell Biology 15:221–231.

     https://doi.org/10.1016/s0955-0674(03)00017-6

     • PubMed
     • Google Scholar
213. 1. Vanderlei B
     2. Feng JJ
     3. Edelstein-Keshet L

     (2011) A computational model of cell polarization and motility coupling mechanics and biochemistry

     Multiscale Modeling & Simulation 9:1420–1443.

     https://doi.org/10.1137/100815335

     • Google Scholar
214. 1. Van Keymeulen A
     2. Wong K
     3. Knight ZA
     4. Govaerts C
     5. Hahn KM
     6. Shokat KM
     7. Bourne HR

     (2006) To stabilize neutrophil polarity, PIP3 and Cdc42 augment RhoA activity at the back as well as signals at the front

     The Journal of Cell Biology 174:437–445.

     https://doi.org/10.1083/jcb.200604113

     • PubMed
     • Google Scholar
215. 1. Veksler A
     2. Gov NS

     (2009) Calcium-actin waves and oscillations of cellular membranes

     Biophysical Journal 97:1558–1568.

     https://doi.org/10.1016/j.bpj.2009.07.008

     • Google Scholar
216. 1. Veltman DM
     2. Williams TD
     3. Bloomfield G
     4. Chen BC
     5. Betzig E
     6. Insall RH
     7. Kay RR

     (2016) A plasma membrane template for macropinocytic cups

     eLife 5:e20085.

     https://doi.org/10.7554/eLife.20085

     • PubMed
     • Google Scholar
217. 1. Verkhovsky AB

     (2015) The mechanisms of spatial and temporal patterning of cell-edge Dynamics

     Current Opinion in Cell Biology 36:113–121.

     https://doi.org/10.1016/j.ceb.2015.09.001

     • Google Scholar
218. 1. Verschueren N
     2. Champneys A

     (2017) A model for cell polarization without mass conservation

     SIAM Journal on Applied Dynamical Systems 16:1797–1830.

     https://doi.org/10.1137/16M1093847

     • Google Scholar
219. 1. Vicker MG
     2. Xiang W
     3. Plath PJ
     4. Wosniok W

     (1997) Pseudopodium extension and amoeboid locomotion in Dictyostelium discoideum: Possible autowave behaviour of F-actin

     Physica D 101:317–332.

     https://doi.org/10.1016/S0167-2789(96)00224-2

     • Google Scholar
220. 1. Vicker MG

     (2000) Reaction–diffusion waves of actin filament polymerization/depolymerization in Dictyostelium pseudopodium extension and cell locomotion

     Biophysical Chemistry 84:87–98.

     https://doi.org/10.1016/s0301-4622(99)00146-5

     • PubMed
     • Google Scholar
221. 1. Vicker MG

     (2002a) F-actin assembly in Dictyostelium cell locomotion and shape oscillations propagates as a self-organized reaction–diffusion wave

     FEBS Letters 510:5–9.

     https://doi.org/10.1016/S0014-5793(01)03207-0

     • Google Scholar
222. 1. Vicker MG

     (2002b) Eukaryotic cell locomotion depends on the propagation of self-organized reaction–diffusion waves and oscillations of actin filament assembly

     Experimental Cell Research 275:54–66.

     https://doi.org/10.1006/excr.2001.5466

     • PubMed
     • Google Scholar
223. 1. Villar-Sepúlveda E
     2. Champneys AR

     (2023) Conditions for Turing and wave instabilities in reaction–diffusion systems

     Journal of Mathematical Biology 86:57.

     https://doi.org/10.1007/s00285-023-01884-x

     • PubMed
     • Google Scholar
224. 1. Wasnik V
     2. Mukhopadhyay R

     (2014) Modeling the dynamics of dendritic actin waves in living cells

     Physical Review E 90:052707.

     https://doi.org/10.1103/PhysRevE.90.052707

     • PubMed
     • Google Scholar
225. 1. Watanabe N
     2. Madaule P
     3. Reid T
     4. Ishizaki T
     5. Watanabe G
     6. Kakizuka A
     7. Saito Y
     8. Nakao K
     9. Jockusch BM
     10. Narumiya S

     (1997) p140mDia, a mammalian homolog of Drosophila diaphanous, is a target protein for Rho small GTPase and is a ligand for profilin

     The EMBO Journal 16:3044–3056.

     https://doi.org/10.1093/emboj/16.11.3044

     • Google Scholar
226. 1. Wei J

     (1999) Existence, stability and metastability of point condensation patterns generated by the Gray-Scott system

     Nonlinearity 12:593–616.

     https://doi.org/10.1088/0951-7715/12/3/011

     • Google Scholar
227. 1. Weiner OD
     2. Marganski WA
     3. Wu LF
     4. Altschuler SJ
     5. Kirschner MW

     (2007) An actin-based wave generator organizes cell motility

     PLOS Biology 5:e221.

     https://doi.org/10.1371/journal.pbio.0050221

     • PubMed
     • Google Scholar
228. 1. Welch MD
     2. Mullins RD

     (2002) Cellular control of actin nucleation

     Annual Review of Cell and Developmental Biology 18:247–288.

     https://doi.org/10.1146/annurev.cellbio.18.040202.112133

     • PubMed
     • Google Scholar
229. 1. Whitelam S
     2. Bretschneider T
     3. Burroughs NJ

     (2009) Transformation from spots to waves in a model of actin pattern formation

     Physical Review Letters 102:198103.

     https://doi.org/10.1103/PhysRevLett.102.198103

     • PubMed
     • Google Scholar
230. 1. Winans AM
     2. Collins SR
     3. Meyer T

     (2016) Waves of actin and microtubule polymerization drive microtubule-based transport and neurite growth before single axon formation

     eLife 5:e12387.

     https://doi.org/10.7554/eLife.12387

     • PubMed
     • Google Scholar
231. 1. Winkler B
     2. Aranson IS
     3. Ziebert F

     (2019) Confinement and substrate topography control cell migration in a 3D computational model

     Communications Physics 2:1–11.

     https://doi.org/10.1038/s42005-019-0185-x

     • Google Scholar
232. 1. Wu Min
     2. Wu X
     3. De Camilli P

     (2013) Calcium oscillations-coupled conversion of actin travelling waves to standing oscillations

     PNAS 110:1339–1344.

     https://doi.org/10.1073/pnas.1221538110

     • PubMed
     • Google Scholar
233. 1. Wu Z
     2. Su M
     3. Tong C
     4. Wu M
     5. Liu J

     (2018) Membrane shape-mediated wave propagation of cortical protein dynamics

     Nature Communications 9:136.

     https://doi.org/10.1038/s41467-017-02469-1

     • PubMed
     • Google Scholar
234. 1. Wu M.
     2. Liu J

     (2021) Mechanobiology in cortical waves and oscillations

     Current Opinion in Cell Biology 68:45–54.

     https://doi.org/10.1016/j.ceb.2020.08.017

     • PubMed
     • Google Scholar
235. 1. Xiong Y
     2. Huang CH
     3. Iglesias PA
     4. Devreotes PN

     (2010) Cells navigate with a local-excitation, global-inhibition-biased excitable network

     PNAS 107:17079–17086.

     https://doi.org/10.1073/pnas.1011271107

     • Google Scholar
236. 1. Xiong D
     2. Xiao S
     3. Guo S
     4. Lin Q
     5. Nakatsu F
     6. Wu M

     (2016) Frequency and amplitude control of cortical oscillations by phosphoinositide waves

     Nature Chemical Biology 12:159–166.

     https://doi.org/10.1038/nchembio0416-304b

     • PubMed
     • Google Scholar
237. 1. Yang L
     2. Zhabotinsky AM
     3. Epstein IR

     (2006) Jumping solitary waves in an autonomous reaction–diffusion system with subcritical wave instability

     Phys. Chem. Chem. Phys 8:4647–4651.

     https://doi.org/10.1039/B609214D

     • Google Scholar
238. 1. Yang Q
     2. Miao Y
     3. Campanello LJ
     4. Hourwitz MJ
     5. Abubaker-Sharif B
     6. Bull AL
     7. Devreotes PN
     8. Fourkas JT
     9. Losert W

     (2022) Cortical waves mediate the cellular response to electric fields

     eLife 11:e73198.

     https://doi.org/10.7554/eLife.73198

     • PubMed
     • Google Scholar
239. 1. Yao B
     2. Donoughe S
     3. Michaux J
     4. Munro E

     (2022) Modulating RhoA effectors induces transitions to oscillatory and more wavelike RhoA dynamics in Caenorhabditis elegans zygotes

     Molecular Biology of the Cell 33:ar58.

     https://doi.org/10.1091/mbc.E21-11-0542

     • PubMed
     • Google Scholar
240. 1. Yochelis A
     2. Knobloch E
     3. Xie Y
     4. Qu Z
     5. Garfinkel A

     (2008) Generation of finite wave trains in excitable media

     Europhysics Letters 83:64005p1–64005p6.

     https://doi.org/10.1209/0295-5075/83/64005

     • PubMed
     • Google Scholar
241. 1. Yochelis A
     2. Knobloch E
     3. Köpf MH

     (2015) Origin of finite pulse trains: Homoclinic snaking in excitable media

     Physical Review E 91:032924.

     https://doi.org/10.1103/PhysRevE.91.032924

     • PubMed
     • Google Scholar
242. 1. Yochelis A
     2. Beta C
     3. Gov NS

     (2020) Excitable Solitons: Annihilation, crossover, and nucleation of pulses in mass-conserving activator-inhibitor media

     Physical Review E 101:022213.

     https://doi.org/10.1103/PhysRevE.101.022213

     • PubMed
     • Google Scholar
243. 1. Yochelis A
     2. Flemming S
     3. Beta C

     (2022) Versatile patterns in the actin cortex of motile cells: Self-organized pulses can coexist with macropinocytic ring-shaped waves

     Physical Review Letters 129:088101.

     https://doi.org/10.1103/PhysRevLett.129.088101

     • PubMed
     • Google Scholar
244. 1. Zeng Y
     2. Lai T
     3. Koh CG
     4. LeDuc PR
     5. Chiam KH

     (2011) Investigating circular dorsal ruffles through varying substrate stiffness and mathematical modeling

     Biophysical Journal 101:2122–2130.

     https://doi.org/10.1016/j.bpj.2011.09.047

     • PubMed
     • Google Scholar
245. 1. Zhan H
     2. Bhattacharya S
     3. Cai H
     4. Iglesias PA
     5. Huang CH
     6. Devreotes PN

     (2020) An excitable Ras/PI3K/ERK signaling network controls migration and oncogenic transformation in epithelial cells

     Developmental Cell 54:608–623.

     https://doi.org/10.1016/j.devcel.2020.08.001

     • PubMed
     • Google Scholar
246. 1. Zmurchok C
     2. Bhaskar D
     3. Edelstein-Keshet L

     (2018) Coupling mechanical tension and GTPase signaling to generate cell and tissue dynamics

     Physical Biology 15:046004.

     https://doi.org/10.1088/1478-3975/aab1c0

     • PubMed
     • Google Scholar

Author details

1. Carsten Beta

   Institute of Physics and Astronomy, University of Potsdam, Potsdam, Germany

   Contribution

   Conceptualization, Visualization, Writing – original draft, Writing – review and editing

   For correspondence

   beta@uni-potsdam.de

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-0100-1043

2. Leah Edelstein-Keshet

   Department of Mathematics, University of British Columbia, Vancouver, BC, Canada

   Contribution

   Conceptualization, Visualization, Writing – original draft, Writing – review and editing

   For correspondence

   keshet@math.ubc.ca

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-2233-690X

3. Nir Gov

   Department of Chemical and Biological Physics, Weizmann Institute of Science, Rehovot, Israel

   Contribution

   Writing – review and editing

   For correspondence

   nir.gov@weizmann.ac.il

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-7774-1139

4. Arik Yochelis

   1. Swiss Institute for Dryland Environmental and Energy Research, Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev, Sede Boqer Campus, Midreshet Ben-Gurion, Israel
   2. Department of Physics, Ben-Gurion University of the Negev, Be’er Sheva, Israel

   Contribution

   Conceptualization, Visualization, Writing – original draft, Writing – review and editing

   For correspondence

   yochelis@bgu.ac.il

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-1516-0766

• 1,300

  views

• 257

  downloads

• 11

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.