Cytoplasmic gradients of cytoskeleton regulators and sub-cellular morphogenesis

Eric Karsenti, Rafael Carazo-Salas, Philip Niethamer,
Maiwen Caudron and Philippe Bastiaens
Cell Biology and Biophysics Program
EMBL
D-69117, Heidelberg, Germany
Karsenti@embl-heidelberg.de

Introduction
Cell shape, motility and division are determined by the organization in three dimensions of essentially two types of polymers: microtubules and microfilaments. These polymers have in common three important properties: they are dynamic, asymmetric and dissipate energy in the form of ATP or GTP hydrolysis during their assembly. Moreover, they interact with a large number of so called associated molecules. Some regulate their dynamic and physical properties others are motor molecules able to move in a unidirectional manner along the surface lattice of the polymer.

Microtubules are tubes with a diameter of 25nm undergoing dynamic instability. They can grow then shrink and grow again. Two kinds of motors are associated with them. Some move towards the so-called minus end of microtubules (mostly Dynein and a few kinesins) others move towards the plus end (most kinesins). In interphase, microtubules organize the cytoplasmic space by providing oriented tracks on which kinesin and dynein motors move vesicles, organize the endoplasmic reticulum and position the nucleus and Golgi apparatus. During cell migration, microtubules play an important role in stabilizing directional cell movements that is achieved by the microfilament system. During cell differentiation, microtubules participate in the establishment of various cell shapes together with the microfilaments. During mitosis, microtubules reorganize into the bipolar spindle that distributes the chromosomes to the two daughter cells.

One important question concerns therefore the mechanism by which those cytoskeleton systems self-organize into various dynamic patterns. Here we report some recent results indicating that gradients of regulators of microtubule dynamics and nucleation exist both in interphase and mitosis and act as positional cues to determine where microtubules should grow and cellular areas in which they have a preferential stability. 

Results
During interphase, fibroblastic cells move in response to chemotactic factors that activate intracellular signaling cascades. Rac1 and Pak1, become activated at leading edges of cells made motile by a wound in the monolayer or by the exposure to growth-factors. More recently, it has been shown that the Rac1-Pak1 pathway mediates in growth and increased stability of “pioneer” microtubules into lamellipodial membrane protrusions. (1) Stathmin/op18 belongs to a class of proteins that negatively regulate microtubule dynamic. It is a 17-kD cytoplasmic protein with a complex phosphorylation pattern. In response to extracellular stimuli or during mitosis, stathmin becomes phosphorylated on up to four residues. Stathmin binds two tubulin heterodimers per molecule forming a trimeric complex (T₂S-complex). According to this property, stathmin‘s inhibiting effect on microtubule growth is believed to derive from its ability to sequester tubulin, thereby decreasing the concentration of free heterodimers available for polymerization. A sequestration independent mechanism has also been proposed. Phosphorylation inactivates the inhibitory effect of stathmin on microtubule growth, probably by inhibiting the interaction between tubulin and stathmin.

During mitosis, stathmin becomes highly phosphorylated due to one or several factors present on mitotic chromatin. One factor has been identified as Polokinase1. Depletion of Polokinase1 inhibits chromatin-induced stathmin hyperphosphorylation and spindle assembly in mitotic Xenopus egg extracts. It has also been shown that during interphase, stathmin is phosphorylated via the Rac1-Pak1-pathway. We have therefore developed a method to examine the interaction between Stathmin and tubulin in vivo by developing a FRET sensor. Stahmin is a globular unstructured protein. However upon binding to tubulin, it elongates and becomes stiff. In order to observe the spatial regulation of stathmin in cells, a fluorescent stathmin fusion protein COPY (CFP-OP18-YFP) was constructed), with ECFP and citrine fused to the N- and C-terminus respectively. Similarly, the fusion proteins for the phosphosite-deficient Ser-to-Ala (COPY-aaa) and the pseudo-phosphorylated Ser-to-Glu (COPY-eee) mutants of COPY were generated. This FRET-based sensor system can report tubulin binding to stathmin since the free stathmin molecule in solution is flexible whereas the stathmin-tubulin complex is stiff and longitudinally extended. As a consequence, the two fluorophores should move apart from each other, leading to a decrease in FRET detected by a decrease in the YFP/CFP-emission ratio. Using this sensor we have shown that there is a strong gradient of interaction between statmin and tubulin in the leading edges of migrating cells transfected with the sensor. The proportion of stathmin bound to tubulin is high in the cell body and low in the lamellipodia, indicating that microtubules should be more stable in the lamellipodia, an observation that fits with previous results (1). Moreover, this gradient is abolished in cells transfected with the same sensor built around the triple phosphorylation mutant of stathmin, indicating that the low interaction of Stathmin with tubulin in lamellipodia is due to stathmin phosphorylation in this area. 

We then looked into mitotic cells transfected with the wild type stathmin sensor. A gradient of low stathmin-tubulin interaction was found around mitotic chromosomes. This gradient has a radius of about 5-10 µm and is not visible in cells transfected with the triple phosphorylation mutant sensor of stathmin. Moreover, this mutant blocks spindle assembly. This shows that local phosphorylation of stathmin around chromosomes is necessary for the preferential assembly of microtubules around the chromosomes and formation of a functional bipolar spindle. 

Therefore both in mitosis and in interphase, phosphorylation gradients of stathmin regulate locally the growth of microtubules. Although the kinases and phosphatases involved may be different (polo kinase and PP2A in mitosis, Pak1 and unknown phosphatases in interphase) the principle is likely to be the same. Phosphorylation gradients likely reflect a steady state phosphorylation ratio of the substrate that is constantly being phosphorylated and dephosphorylated. The average dephosphorylation rate of proteins is fairly high, with a halftime of about 6-10 min in interphase and 1-2 min in mitosis. The activity of at least one of the two enzymes, the kinase or the phosphatase has to be localized. The counteracting activity can be non-localized and maintains a steady-state phosphorylation gradient of the diffusing substrate.

A similar principle seems to maintain the previously described steady-state gradient of Ran-GTP around mitotic chromosomes (2). Here, by contrast, the chromatin-localized activity of the GEF RCC1 is opposed by the global counteracting activity of Ran-GAP, which replenishes the cytoplasmic pool of inactive Ran-GDP. The established concentration gradient of Ran-GTP around chromosomes allows the spatial regulation of the mitotic spindle. We have recently shown that microtubule nucleation occurs through the local activation of a protein (TPX2) through its release from importins under the action of the local generation of RANGTP around chromatin in Xenopus mitotic egg extracts. Moreover, a long-range gradient of regulators seems to affect microtubule plus end dynamics, resulting in the preferential movement of centrosomal asters towards mitotic chromatin in Xenopus egg extracts (3). It seems therefore that gradients of regulators are widely used in cells to generate morphogenetic fields that direct the organization of microtubules in specific patterns. Such gradients are probably widely used to determine local cytoplasmic states and where different sub-cellular assemblies of molecules should take place. They are probably also essential in all directed cell motility processes.

References

1. T. Wittman, G. M. Bokoch, C. Waterman-Storer, J. Cell Biol. 161, 845 (2003).
2. E. Karsenti, I. Vernos, Science 294, 543-7. (2001).
3. R. E. Carazo-Salas, E. Karsenti, Curr Biol 13, 1728-1733 (Sep 30, 2003).