Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 96 (4), 1275-92

Biochemistry on a Leash: The Roles of Tether Length and Geometry in Signal Integration Proteins

Affiliations

Biochemistry on a Leash: The Roles of Tether Length and Geometry in Signal Integration Proteins

David Van Valen et al. Biophys J.

Abstract

We use statistical mechanics and simple ideas from polymer physics to develop a quantitative model of proteins whose activity is controlled by flexibly tethered ligands and receptors. We predict how the properties of tethers influence the function of these proteins and demonstrate how their tether length dependence can be exploited to construct proteins whose integration of multiple signals can be tuned. One case study to which we apply these ideas is that of the Wiskott-Aldrich Syndrome Proteins as activators of actin polymerization. More generally, tethered ligands competing with those free in solution are common phenomena in biology, making this an important specific example of a widespread biological idea.

Figures

Figure 1
Figure 1
Biological examples of biochemistry on a leash. Receptors and ligands connected by a flexible tether are a common motif in biological systems. (a) Formin-mediated actin polymerization (3). In one model, formin proteins can use flexible tethers to grab actin monomers in solution and deliver them to the end of a growing actin filament. (b) Inactivation of Src tyrosine kinase (Courtesy of David Goodsell, Scripps Research Institute, La Jolla, CA). A flexible linker connects the SH2 domain and the tail portion of Src tyrosine kinase. It is thought that the binding of the phosphorylated tail to SH2 locks the protein in an inactive conformation. Theoretical studies suggest that the properties of the tether can influence the protein's function. (c) Activation of Arp2/3 by synthetic WASPs (59). WASPs link Arp2/3-mediated actin polymerization to chemical signals by controlling the delivery of the first actin monomer to the growing actin filament. Synthetic WASPs have an interesting structural motif where a receptor and ligand are connected by a flexible tether. The competition between the tethered ligand and ligands free in solution gives rise to varying levels of actin polymerization.
Figure 2
Figure 2
States and weights for the simple switch. The tethered ligand functions as an inhibitor and locks the protein into an inactive state when bound to the receptor. Ligands in solution can compete with cis-ligand to activate the switch. Statistical mechanics assigns statistical weights to each state and allows us to compute the probability of the protein to be in an active state, as shown in the ratio at the bottom.
Figure 3
Figure 3
Random walk model for flexible polypeptide chains. (a) In this model, the polypeptide chain is separated into statistically independent segments, called Kuhn segments, of length b. The orientation of two monomers is perfectly correlated if they are in the same segment and completely uncorrelated if they are in different segments. (b) The probability distribution for the end-to-end distance vector for a chain of length L = 20ξp computed with both the wormlike chain and random walk models. The computation for the wormlike chain was implemented using the method of Samuel and Sinha (60). A comparison between the two polymer models shows that the random walk model is an acceptable approximation of the wormlike chain model for long chains.
Figure 4
Figure 4
Physical model for the output domain of N-WASP. The VCA domain is thought to be a mixture of α-helices and unstructured amino acids. (a) Separation of structure into residues assumed to have secondary structure and residues assumed to be unstructured. Here, cylinders represent stable α-helices. (b) Model for binding of the VCA domain to Arp2/3 (59). In our model, binding of VCA to Arp2/3 fixes the location of the C and A domains. This allows us to treat the C and A domains as rigid cylinders and the rest of the VCA domain as flexible tethers.
Figure 5
Figure 5
Tether length dependence of the simple switch. (a) The probability of the switch being active as a function of ligand concentration and tether length. The parameters used for the case study on reprogrammed N-WASP are described in the main text. Zero additional amino acids corresponds to the experimental construct with a tether length of ∼24.4 nm. The additional amino acids used by Dueber et al. (17) consist of serine-glycine repeats. Assuming the relative activity is equivalent to pon, we can use experimental data from Dueber et al. to fit for unknown parameters in our model (17). (b) The tether length dependence of the effective concentration of cis-ligand. We find that for short tethers, an increase in tether length leads to higher effective concentrations and hence a lower probability of occupying an active state. The reverse effect is seen in the model for very long tethers.
Figure 6
Figure 6
Topologies of complex switches. This figure outlines the topologies of the complex switches discussed in this work and highlights the relevant tether lengths. The statistical mechanical model used to examine the simple switch can be extended to examine constructs with multiple tethered receptor-ligand pairs. Note that there are two distinct mechanisms for signal integration. For class I switches, the interaction between the black receptor-ligand pair serves to control the length of the white receptor's tether. The tether length dependence is then crucial for this construct's signal integration behavior. For class II switches, the interaction between the black receptor-ligand pair serves to bring the white receptor-ligand pair into closer proximity. This likely increases the effective concentration of white cis-ligand and leads to cooperativity for the binding of the two inputs.
Figure 7
Figure 7
States and weights for class I and class II complex switches. Statistical mechanics can be used to compute the thermodynamic weight of each state and the probability the switch occupies an active state. A complete loop must be formed to inactivate the switch. (a) Class I switches have been observed to display antagonistic gating in experiments. (b) Class II switches display cooperative integration in experiments.
Figure 8
Figure 8
Nuclear magnetic resonance structure of the SH3 domain. We can estimate the effective concentration of cis-ligand seen by an SH3 domain in class I switches by appealing to a structure of the bound complex, identification number 1PRM in the Protein Data Bank (61). This structure can also be used to compute the effective concentration of cis-ligand seen by a PDZ domain in class II switches.
Figure 9
Figure 9
Activation profile for an antagonistic class I switch. For this profile, KB = 660 μM and KW = 8 μM. The color represents pon, with low values being blue and high values being black. When [white ligand] is increased the switch is activated, reflected by a color shift of blue to yellow or orange. When [black ligand] is increased, the switch is deactivated. This effect is best seen at high values of [white ligand]. Because the role of the black receptor/cis-ligand pair is to determine the length of white receptor's tether, the simple scaling properties of the simple switch provide a physical mechanism for antagonistic integration. The removal of SH3's interaction with its cis-ligand lengthens the white receptor's tether. The longer tether length for L2 increases the effective concentration of cis-ligand seen by the white receptor and represses the switch.
Figure 10
Figure 10
Activation profile for class II switches. The color represents pon, with low values being blue and high values being black. (a) This construct exhibits the same behavior as an AND gate; both inputs are necessary to achieve maximal activation. Dissociation constants used were KB = 1000 μM and KW = 8 μM. (b) This construct has behavior consistent with an OR gate; either input is sufficient to achieve maximum activation. Dissociation constants used were KB = 1000 μM and KW = 100 μM.
Figure 11
Figure 11
Kinetic model of the simple switch. This toy model assumes simple chemical kinetics for all reactions. Ligands free in solution bind and unbind to the receptor with rate constants k+ and k while the tethered ligand binds and unbinds with rate constants kc and k. Switches in an active state are consumed at a rate R to generate product.
Figure 12
Figure 12
Potential landscape for the cis-ligand. The probability distribution of the distance between the receptor and its tethered ligand can be used to define an entropic potential. The cis-ligand is then assumed to undergo one-dimensional diffusion inside this potential well. We then use the Fokker-Planck equation to find the average time needed for the cis-ligand to encounter the absorbing boundary assuming it starts at the bottom of the potential well. The inverse of this mean first passage time is an estimate for the collision rate.
Figure 13
Figure 13
Simulation of the full kinetic model. (a) The probabilities and concentrations for S, SL, and ST as a function of time for R = .034 s−1. In the full kinetic model, the switches achieve a rapid preequilibria between the three states S, SL, and ST. After the initial transient, the probabilities for each of the three states is constant and equal to the probability predicted by equilibrium statistical mechanics. (b) The probabilities and concentrations for S, SL, and ST as a function of time for R = 1000 s−1. Because the timescale of the productive reaction is the same as that for ligand unbinding, the system never reaches an effective equilibrium between the states of the simple switch. For this choice of parameters, the considerations of equilibrium statistical mechanics provide a poor description for the behavior of this system.

Similar articles

See all similar articles

Cited by 17 PubMed Central articles

See all "Cited by" articles

Publication types

LinkOut - more resources

Feedback