Intracellular electric field and pH optimize protein localization and movement

Abstract

Mammalian cell function requires timely and accurate transmission of information from the cell membrane (CM) to the nucleus (N). These pathways have been intensively investigated and many critical components and interactions have been identified. However, the physical forces that control movement of these proteins have received scant attention. Thus, transduction pathways are typically presented schematically with little regard to spatial constraints that might affect the underlying dynamics necessary for protein-protein interactions and molecular movement from the CM to the N. We propose messenger protein localization and movements are highly regulated and governed by Coulomb interactions between: 1. A recently discovered, radially directed E-field from the NM into the CM and 2. Net protein charge determined by its isoelectric point, phosphorylation state, and the cytosolic pH. These interactions, which are widely applied in elecrophoresis, provide a previously unknown mechanism for localization of messenger proteins within the cytoplasm as well as rapid shuttling between the CM and N. Here we show these dynamics optimize the speed, accuracy and efficiency of transduction pathways even allowing measurement of the location and timing of ligand binding at the CM--previously unknown components of intracellular information flow that are, nevertheless, likely necessary for detecting spatial gradients and temporal fluctuations in ligand concentrations within the environment. The model has been applied to the RAF-MEK-ERK pathway and scaffolding protein KSR1 using computer simulations and in-vitro experiments. The computer simulations predicted distinct distributions of phosphorylated and unphosphorylated components of this transduction pathway which were experimentally confirmed in normal breast epithelial cells (HMEC).

Figure 1. The EGFR pathway and intracellular electric field.

(Left) Typical presentation of EGFR pathway. The proteins are not drawn to scale and, as a result, the limitation of random walk in allowing rapid and reliable transmission of information by random walk is underestimated. In fact, the distance from the cell membrane to the nucleus is about 1,000 protein diameters. (Right) Measurements of the intracellular electric field using nano-voltmeter from Tyner et al. . The 10 E-field values in the bar graph on the right are those within the respective 10 blue boxes on the left. Note the decline in the electric field with distance from the nuclear membrane as well as the local perturbation caused by the presence of a mitochondria on regions 5, 6, and 7.

Figure 2. Diffusion of messenger proteins resulting from ligand binding to a small focus of receptors on the cell membrane at time t = 0.

By random walk (top panels), there is broad dispersal of the messenger proteins through the cytoplasm. As a result, all information regarding the time and location of ligand binding is lost. By biased random walk (lower panels) due to Coulomb interactions of phophorylated, negatively charged messenger proteins with an intracytoplasmic electric field, the spatial location of ligand binding on the CM is projected onto the NM and the transition time is less than 0.1 second.

Figure 3. Spatial resolution of ligand binding on cell membrane.

Here we assume a cubicle cellular configuration to match the shape frequently seen on epithelial surfaces. We assume that the lower part of the cell is attached to a basement membrane and that ligand binding occurs uniformly and simultaneousl but only in the cell wall attached to the basement membrane (left images). The messenger proteins travel as a wave from the CM to the NM (middle images). The messenger proteins, due to directed motion, can project spatial information on the site of ligand binding onto the NM. This is evident in the recapitulation of the ligand binding pattern in the CM onto the NM in the right images.

Figure 4. Dynamic modeling of RAF, MEK and ERK after ligand binding.

In the initial state, RAF is clustered around the CM and MEK and ERK around the NM. When ligand binds the membrane receptor, RAF is phosphorylated. The negative chareges interact with the intracellular field resulting in rapid (<0.01 sec) movement of pRAF toward the NM. As it reaches the perinuclear region, pRAF encounters and phosphorylates several MEK proteins which, in turn, phosphorylate several ERK proteins. The rapid, direct movement of pRAF provides spatial and temporal information while the interactions with MEK and ERK amplify the signal at the NM. The pRAF is assumed to encounter a phosphorylase after about 30 seconds. The loss of negative charges causes RAF to return to it original isoelectric point with very rapid (<0.01 sec) return to the CM where it is again available for signal transduction.

Figure 5. Simulations of steady state distribution of MAPK proteins in culture conditions with continuous presence of EGF.

The model assumes the presence of scaffolding proteins KSR1 as outlined in the text. Top panel represents the physical characteristics of RAF, MEK and ERK both free and bound to KSR1 used in the computer simulations. Lower panels represent predicted steady state distribution of RAF, MEK, and ERK in normal cells assuming continuous presence of ligand at the cell membrane and assuming the presence of scaffolding protein KSR1. Middle panels are actual distribution observed in HMEK cells in culture using CYTOO chips so that every cell maintains roughly the same shape.

Figure 6

Simulations of steady state distribution of phosphorylated MAPK proteins in culture conditions with continuous presence of EGF. As in Figure 5, the model assumes the presence of scaffolding proteins KSR1 as outlined in the text. Top panel represents the physical characteristics of pRAF, pMEK and pERK both free and bound to KSR1 used in the computer simulations. Lower panels represent predicted steady state distribution of pRAF, pMEK, and pERK in normal cells assuming continuous presence of ligand at the cell membrane and assuming the presence of scaffolding protein KSR1. Middle panels are actual distribution observed in HMEK cells in culture using CYTOO chips so that every cell maintains the same shape. Arrows on the left middle panel denote the approximate location of the CM.

Figure 7. Predicted and measured steady state distribution of RAF and pRAF under typical culture condition with continuous ligand binding of EGFR on the cell membrane.

The left panels demonstrate typical pRAF and RAF distribution in cultured HMEK cells. A) An iso-rendered distribution of RAF (green) and pRAF (red) with B) the photomicrograph demonstrating RAF clustered around the CM and pRAF clustered around the NM, as predicted by the IEM simulations. In the right panels, computer simulations demonstrate that if protein movement is by diffusion alone, pRAF will exhibit a concentration gradient with highest levels near the CM where it is generated by interactions with RAS and lowest levels near the nucleus. RAF will be widely dispersed. In the IEM simulations, pRAF will, due to its negative charge, concentrate near the nucleus while RAF (with an IEP of 9.2) will cluster near the CM.