The need for speed: migratory cells in tight spaces boost their molecular clock

Abstract

Cells migrating in spatial confinement exhibit higher intracellular calcium levels, which increases the oscillation frequency of a “molecular clock” that synchronizes guanine nucleotide exchange factor GEF-H1 and microtubule polymerization for more frequent bursts of speed.

Directed cell migration in confined spaces is crucial for tissue development, wound repair, and tumor dissemination (Yamada and Sixt, 2019). In response to asymmetric extracellular stimuli, cells polarize with a “front” and “back,” orchestrated spatially and temporally via a complex signaling network of Rho GTPases (e.g. Rac1, RhoA, etc.) (Ridley et al., 2003). Cells then migrate in a preferred direction by advancing their front via actin polymerization (regulated by Rac1), coordinated with retraction of the back via actomyosin contractility (regulated by RhoA). These RhoGTPases are further regulated by a host of guanine nucleotide exchange factors (GEFs), GTPase-activating proteins (GAPs), and guanine nucleotide dissociation inhibitors (GDIs). Nevertheless, a systems-level understanding of how cells make migratory decisions based on external cues remains nascent (Rappel and Edelstein-Keshet, 2017). For instance, cells maintain polarity and migrate more directionally within narrow channels, but frequently repolarize and migrate more randomly when unconstrained on flat surfaces. In this issue of Cell Systems, (Lee et al., 2022) show that cells in spatial confinement utilize a “molecular clock” to enhance their polarization and migration.

Lee et al. elegantly integrated experimental live cell imaging with mathematical modeling to uncover successive layers of this oscillatory signaling circuit (Figure 1A). First, they showed that spatial confinement increases intracellular Ca²⁺ concentration, which regulates oscillations of Rho associated guanine nucleotide exchange factor GEF-H1 as well as RhoA. Moreover, increased intracellular Ca²⁺ also increased microtubule depolymerization, which was also coupled back to GEF-H1. Finally, GEF-H1 was coupled to microtubules through the formin mDia1, which has also been implicated in maintaining cell polarity and actin polymerization (Li and Gundersen, 2008). It should be noted that there were distinct mechanisms associated with protein abundance and activity, which contributed to the rich complexity of these cyclic behaviors.

Figure 1.

(A) Confined cell migration controlled by a molecular clock dependent on intracellular calcium (Ca²⁺), GEF-H1, RhoA, mDia1, and microtubule polymerization. Continuous line denotes dependence on abundance, while dashed line denotes dependence on activity. Created with BioRender.com. (B) Genetically encoded Ca²⁺ biosensor (GCaMP) shows increased green fluorescence as mouse pancreatic endothelial cells (MS1) enter a narrow microchannel. Scale bar 20 μm. (C) Computational model of coupled GEF-H1 and RhoA oscillations, corroborating experimental measurements. Reproduced from Lee et al. Copyright Elsevier.

To elucidate the role of spatial confinement, cells incorporating fluorescent indicators were imaged within tightly confining channels with width and height comparable to the nucleus (< 10 μm). In particular, confined cells exhibited a striking increase in intracellular Ca²⁺ concentration relative to unconfined cells (Figure 1B). Further, RhoA activity was observed to be oscillatory for confined cells, correlating with bursts of migration speed. Pharmacological treatment of unconfined cells to increase intracellular Ca²⁺ concentration was sufficient to recapitulate this RhoA oscillation. Increased intracellular Ca²⁺ concentration was also observed to decrease microtubule stability, implicating the RhoA guanine exchange factor GEF-H1. Indeed, GEF-H1 activity exhibited oscillations in phase with RhoA (Figure 1C), indicative of a Ca²⁺ regulated timing mechanism that drives periodic migration dynamics.

Next, Lee et al. investigated microtubule dynamics by fluorescence imaging of the plus end protein EB1 after pharmacological manipulation of intracellular Ca²⁺ concentration. For instance, increased Ca²⁺ concentration decreased the length of persistently growing microtubules, although average growth rate was unaffected. Thus, the authors concluded that increasing Ca²⁺ concentration increased microtubule depolymerization. Nevertheless, this mechanism was not sufficient to explain the oscillatory dynamics of GEF-H1 abundance. Lee et al. considered whether GEF-H1 complexed with microtubules would confer some protection against degradation, using further pharmacological treatment. They showed that microtubule depolymerization would dissociate these complexes, making GEF-H1 more vulnerable to proteasomal degradation and decreasing abundance. Subsequently, GEF-H1 abundance increased in the second phase of oscillation, which was shown to occur due to increased GEF-H1 synthesis in response to a sustained increase in Ca²⁺ concentration.

Oscillatory dynamics in biological systems are typically regulated by negative feedback loops. Thus, in addition to GEF-H1 activity and abundance being regulated by microtubule stability, Lee et al. investigated whether microtubule polymerization was regulated, in turn, by GEF-H1. Knockdown of GEF-H1 also resulted in an increase in the length of persistently growing microtubules, without affecting average growth rate. This multifaceted mechanism was shown to be mediated by the formin mDia1, implicated in Rho-dependent microtubule stabilization as well as actin polymerization. Altogether, GEF-H1 abundance was positively controlled by microtubule stabilization via negative regulation of mDia1 abundance. Second, GEF-H1 activity was negatively controlled by microtubule stabilization via positive regulation of mDia1 activity. This molecular clock was activated and regulated by intracellular Ca²⁺ concentration, which impacted microtubule stability as well as GEF-H1 expression.

A system of coupled differential equations accounting for Ca²⁺, microtubules, GEF-H1, RhoA, and mDia1 concentration was sufficient to quantitatively capture these cyclic behaviors. For instance, mDia1 activity was expected to be oscillatory with a phase shift relative to RhoA activity, which was experimentally consistent with the observed bursts of migration speed and a delay time before maximum F-actin polymerization. Moreover, the frequency and amplitude of GEF-H1 oscillations in the mathematical model were highly sensitive to microtubule stability, which was verified experimentally by pharmacological perturbation of microtubule acetylation, targeting acetyltransferarse p300.

A computational motor-clutch model was then used to explicitly link cytoskeletal dynamics to cell migration (Bangasser et al., 2017), showing that increasing GEF-H1 oscillation frequency would result in increased migration speed with more repeated bursts. Experimentally, inhibition of p300 increased RhoA oscillation frequency, resulting in enhancement of oscillatory changes in speed as well as overall distance traveled. In comparison, inhibition of mDia1 decreased oscillation frequency with fewer oscillatory changes and decreased distance traveled. Thus, Lee et al. established a systems-level understanding of both biochemical kinetics and biophysical dynamics that govern confined cell migration.

Overall, this work represents an experimental and computational tour-de-force that reveals why cells sustain their polarity and migration in confined spaces. Lee et al. take advantage of microfabricated channels with tightly controlled geometries to ensure that cells experienced nearly identical confinement conditions. Nevertheless, cells still exhibited appreciable heterogeneity in their intracellular signaling dynamics and migratory behavior. The interplay of stochastic “noise” relative to deterministic inputs to control cell migration decisions remains an intriguing area for further study, particularly when coupled with more complex microenvironmental cues such as spatial or temporal gradients (SenGupta et al., 2021). In addition to individual cell migration, Lee et al. showed that groups of cells migrating collectively in confinement also exhibit transient bursts of intracellular calcium. Previous work from these authors has suggested that calcium signaling coordinates leader and follower cells during tumor organoid invasion (Ellison et al., 2016), which might act through a similar mechanism to synchronize RhoA and microtubule dynamics across many cell lengths. Finally, these results implicate GEF-H1 as a key regulator of confined cell migration, that can be controlled via pharmacological treatment. Indeed, recent work elsewhere has shown that T-cells more effectively infiltrate fibrotic extracellular matrix after loss of GEF-H1 or destabilization of microtubules (Tabdanov et al., 2021). Thus, therapeutic manipulation of GEF-H1 could potentially occur in two ways – either to augment immune cell migration into a disease site, or to impede tumor cell dissemination.