Kinetic Regulation of Coated Vesicle Secretion
(L. Foret & P. Sens)  Accepted PNAS (July 2008)
reprint

The secretion of vesicles for intracellular transport often rely on the aggregation of specialized membrane-bound proteins into a coat able to curve cell membranes. Mathematical modelling of this process is necessary if one wishes to fully to understand the Physical basis underlying the formation of coated vesicles and its Biological implication, e.g for the regulation of vesicle secretion and intracellular transport. The basic ingredients for such model are i) the existence of freely diffusing membrane-bound proteins that couple with the local curvature of the membrane, and ii) the tendency for these proteins to aggregate and form domains that concentrate this curvature. Models can then span the spectrum between a equilibrium theory (see our model for caveolae formation) in which protein aggregation is reversible and the membrane find its optimal energy state consisting of aggregates of a given size and curvature, and a model where aggregation is irreversible and eventually leads to the formation of extended protein patches, but in which the kinetics of protein aggregation might be complex. Neither model leaves much room for the fast and sensitive regulation of vesicle secretion which is often observed in cellular processes. We have previously shown that the phase behaviour of a mixed membrane can be strongly affected by the exchange of components between the membrane and the cytosol (e.g. via endo and exocytosis, see our Raft and Recycling model).  In three major classes of coated vesicles (COPI and COPII, trafficking between the Golgi and the Endoplasmic Reticulum, and Clathrin-coated vesicles and the plasma membrane) vesicle formation occurs at the expense of energy by way of GTP hydrolysis. More precisely, the assembly and disassembly of COP coat components follow the cycle of activation - inactivation of GTPase proteines, Sar1 for COPII and Arf1 for COPI. Once activated, the GTPases bind the membrane and recruit individual coatomer complexes, that later polymerize into coats. The inactivation of the GTPase, triggered by the hydrolysis of its bound GTP, leads to its unbinding from the membrane and to the coatomer disassembly. Strikingly, FRAP experiments suggest that the exchange kinetics of coat components is much faster than the rate of vesicle secretion. In other words, many futile coatomers are released to the cytosol during the expansion of a coat. So, while new membrane-bound proteins polymerize at the coat periphery, others within the coat disassemble and are expelled to the cytosol. Paradoxically, the consumption of energy via GTP hydrolysis seems to work against coat growth and to prevent vesicle formation. We have developed a quantitative model that shows that the turnover of coat components allows for a highly sensitive switching mechanism between a quiescent and a vesicle producing membrane, upon a slowing down of the exchange kinetics. We claim that the existence of this switching behaviour, also triggered by factors such as the presence of cargo and the membrane mechanical properties, allows for efficient regulation of vesicle secretion. We propose a model, supported by different experimental observation, in which vesiculation of secretory membranes is impaired by the energy consuming desorption of coat proteins, until the presence of cargo or other factors triggers a dynamical switch into a vesicle producing state. EM picture showing different stage of maturation  of a COP vesicle  Our model is based on the activation-inactivation cycle of coat proteins, which turn-over between the membrane and the cytosol at prescribed rates (Figure-a). When on the membrane, these proteins diffuse around and aggregate to form a growing coat, until their inactivation and desorption. The nucleation and growth of a protein coat is a kinetic process that competes with the energy-consuming turnover of coat components between the membrane and the cytosol. Our goal is to compare the flux of monomer leaving the membrane individually after inactivation (Joff) and the flux of secreted vesicles (Jv), for a given flux of incoming monomer (Jon) (Figure-b). Intuitively, one expects GTPase inactivation to decrease the rate of vesicle formation by reducing the lifetime of membrane-bound monomers. Our quantitative approach has revealed the existence of a discontinuous dynamical transition from a quiescent to a vesicle producing membrane. There exists a very narrow range of GTPase turnover within which the amount of secreted vesicles jumps from essentially zero to a finite value. In other words, the apparently counter-productive energy consumption that favors the unbinding of coat components provides secretory membranes with a highly sensitive switch to regulate vesicle release, triggered for instance by a variation of the cargo concentration or the mechanical tension of the membrane. We have shown that a secretory membrane can exist in either of two states. If the rate of GTP hydrolysis and coatomer inactivation (koff) is low, protein domains may grow and invaginate up to a critical size at which a mature vesicle is formed. One the other hand, if the coatomer turnover is fast, domains are arrested in their maturation by coatomer desorption, and almost no vesicle is produced. Instead, invaginated domains of intermediate size are stabilized by coatomer inactivation. Our crucial finding is that the transition between these two states is very sharp, highly non-linear. In physical terms, it is a first order dynamical transition, characterized by two critical thresholds for the coatomer recycling rate (Figure). Above the high threshold, recycling is too fast, and vesicles cannot form. Only stationary  coated pits can be seen on the membrane. Below the low threshold, the membrane is covered by coats of any size, including mature invaginated pits that eventually break-off as coated vesicles. In between, the dynamical switch between the two states involves an hysteretic cycles. Vesicule secretion is switched "on" at the low threshold and "off" at the high threshold. Switch-like behaviors are clearly advantageous for biological systems. The highly non-linear nature of a switch confers evident robustness with respect to the noisy environment, and allows for a precise regulation of the system's activity. In particular, the recycling rate is known to be influence by the presence of cargo at the membrane. Cargo stabilizes the coat proteins, and abundance of cargo is equivalent to a low recycling rate, thereby favoring the formation of vesicles. The nature of the dynamical transition allows vesicle production to be switched on only when a sufficient amount of cargo is present, thereby preventing the futile delivery of empty vesicles. Strikingly, analysis of COPI vesicles in  cells where arfGTPase are unable to hydrolyse GTP (when the vesicle production switch stays "on" regardless of the amount of available cargo) reveals a much lower cargo content than in normal cells (Pepperkok etal. J. Cell Science 113, 135-144 (2000)). An interesting consequence of the "Secretory Switch" that could be observed in-vivo  is the prediction that under constant cargo synthesis, a secretory membrane could exhibit periodic vesicle secretion. Considering that newly synthesized cargo is brought to the membrane at a steady rate and is removed by vesiculation, membrane-bound cargo accumulates in the no-secretion regime, thereby decreasing koff and moving the system toward the secretion regime. Vesicle production starts below the low turnover threshold, thereby reducing the amount of membrane-bound cargo and moving the system toward the quiescent state. Above the high threshold, vesicle secretion is switched off, letting the cargo accumulates until the high-density threshold is reached and vesicle production is resumed, starting a new cycle. At constant in-flux of cargo,  the system should thus periodically switch between quiescent phases and phases of vesicle secretion, following the hysteretic loop (Figure). The efficiency of the secretion machinery (the amount of cargo delivered per cycle) would then be thighly controlled by the width the the hysteretic loop.