The Dynamical Systems Properties of the HOG Signaling Cascade

Abstract

The High Osmolarity Glycerol (HOG) MAP kinase pathway in the budding yeast Saccharomyces cerevisiae is one of the best characterized model signaling pathways. The pathway processes external signals of increased osmolarity into appropriate physiological responses within the yeast cell. Recent advances in microfluidic technology coupled with quantitative modeling, and techniques from reverse systems engineering have allowed yet further insight into this already well-understood pathway. These new techniques are essential for understanding the dynamical processes at play when cells process external stimuli into biological responses. They are widely applicable to other signaling pathways of interest. Here, we review the recent advances brought by these approaches in the context of understanding the dynamics of the HOG pathway signaling.

Figure 1

The canonical structure of a MAPK cascade. We used the Systems Biology Graphical Notation (SBGN) [4] to represent the interactions between the MAP Kinases. Activations of MAPK occur through enzymatic phosphorylation and ATP consumption. Interactions with other components and in particular with phosphatases are not shown. In the case of the HOG pathway in yeast, dual phosphorylation of the final MAPK (Hog1p) occurs within a few minutes after an hyper-osmotic stimulus.

Figure 2

The HOG pathway. View of the main molecular actors involved in the hyperosmotic glycerol pathway (see text for more details). Two branches led by Sho1p and Sln1p are sensitive to high osmolarity and lead to the activation of Pbs2p and Hog1p after a hyperosmotic shock. Hog1p has both a cytoplasmic and a nuclear role, with different timescales, that correspond to a fast non transcriptional response and a longer response involving transcription when dealing with strong hyper osmotic shock. The yeast pictures at the bottom show nuclear localization of Hog1p tagged by GFP after a moderate hyper-osmotic shock (Sorbitol, 1 M). Colocalization with the nucleus is seen on the overlay pictures between the GFP channel (Hog1p) and the RFP channel (Htb2p). Note that localization is transient and reversible if the cell is put back into isotonic conditions.

Figure 3

Sequential sketch of yeast adaptation to a hyperosmotic shock. The evolution with time of the size, phosphorylation of Hog1p, and internal concentration of glycerol are schematically represented in the center of the picture. (1) After an increase of the external osmolarity (green), a first mechanical response corresponds to a rapid loss of water (blue arrow). It leads to a decrease of the cell size and a loss of turgor pressure. (2) HOG osmosensors (blue) activate the pathway and eventually lead to the phosphorylation of Hog1p. (3) Hog1PP induces several processes: (a) Inactivation of the glycerol channel Fps1p preventing glycerol leakage; (b) direct or indirect activation of cytoplasmic actors, for example, 6-phosphofructo-2-kinase (Pfk2p) involved in glycerol synthesis; (c) translocation in the nucleus. Note that there are other targets of Hog1p such as Sic1p, Hsl1p, Nha1p, and Tok1p. (4) Nuclear Hog1PP induces a large transcriptional response. In particular, the gene GPD1 leading to glycerol synthesis is upregulated. Negative feedbacks (glycerol production, phosphorylation of Sho1p, etc.) allow inhibition of the pathway activity. (5) Increase of the internal glycerol leads to water influx and progressive cell size recovery while Hog1p is exported from the nucleus. (6) Pathway is off, and turgor pressure and cell size are restored. The cell is adapted to its new environment.

Figure 4

Different microfluidics techniques to control the chemical environment of single yeast cells while imaging them through microscopy. (a) Microfluidic system as described in Hersen et al. [11]. Yeast cells are fixed in the channel by the lectin protein Concanavalin A. One inlet is filled with an iso-osmotic media (blue) and the other with a hyperosmotic media (orange). By tightly controlling the pressure in each inlet, it is possible to create a periodic shock on the cells. (b) Optical tweezers system (red) as described by Eriksson et al. permits to control the cells position in the channel with two fluids flowing side by side [12]. (c) The system developed by Charvin et al. uses a dialysis membrane (green) to trap cells on top of a soft PDMS slice [13]. (d) Multilayer microfluidic device [14]. The top layer (green) is used to capture cells. By controlling the pressure inside this channel, cells can be optimally trapped while subjected to periodic shocks. The bottom layer is used to culture cells.

Figure 5

Schematic representation of the Hog pathway models of Mettetal et al. [10] and Zi et al. [15]. Pictures are redrawn from original figures of these papers. Top: (a) Diagrammatic representation of Mettetal's model. Au(t) represents the osmolarity applied at time t and the variables x and y can be identified with the intracellular glycerol concentration and the enrichment of Hog1 in the nucleus. The model contains a feedback depending on Hog1p (with strength β) and one, which is independent of Hog1p (strength α). The equations for this model read $\dot{y}=(A_{0}u-x)-\gamma y$ and $\dot{x}=\alpha (A_{0}u-x)+\beta y$. (b) The same model, interpreted in biological terms. The export of osmolytes is regulated by a mechanism, which does not depend on the MAPK pathway (e.g., closure of Fps1p) and by a mechanism depending on Hog1p activation. (c) Diagram of the model structure proposed by Zi et al. The model includes a simplified version of the MAPK pathway as well as two different feedbacks induced by activated Hog1p (a slow transcriptional and a fast nontranscriptional). Both of these feedbacks act by increasing the production of glycerol.