Ras Pathway Proteins
Ras proteins play critical roles in normal physiology and in several disease states. Our laboratory studies these molecules in learning and memory and in cancer.
Figure 1. The diagram presents an overview of signaling via Ras pathway proteins.
Ras-GRF Proteins in Synaptic Plasticity, Learning and Memory
Ras-GRF1 and Ras-GRF2 are two closely related exchange factors expressed highly in the brain that have the capacity to activate both Ras and Rac in response to calcium influx into cells. Despite their similar overall structure Ras-GRF1 and Ras-GRF2 mediate opposing forms of synaptic plasticity in the highly investigated CA1 region of the hippocampus that are known to contribute to learning and memory.
Figure 2. Ras-GRF1 and Ras-GRF2 mediate opposing forms of synaptic plasticity.
GRF1 mediates NMDA receptor-induced LTD, while GRF2 mediates NMDA receptor induced LTP. At least part of this difference is due to the fact that although both proteins can activate both Ras and Rac effectors, GRF1 predominantly activates p38 MAP kinase, while GRF2 predominantly mediates Erk MAP kinase. A hint at the mechanism underlying this difference is our observation that the two exchange factors mediate the action of different sub-classes of NMDA receptors. GRF1 and GRF2 also make distinct contributions to hippocampus-mediated learning and memory.
Figure 3. Signaling that mediates LTP and LTD utilize different GRFs.
Transgenerational epigenetic effects of an "enriched" environment during youth on synaptic plasticity, learning and memory.
We discovered that the signaling network responsible for promoting LTP in the hippocampus of mice is modulated by the degree of stimulation in their environment. In particular, exposure of adolescent but not adult mice to 2-weeks of an enriched environment unlocks an otherwise latent NMDA receptor/p38 Map kinase signaling cascade (see Figure 7 below) that contributes to LTP induction. Moreover, this environmentally induced signaling cascade can compensate for the loss of Erk signaling that occurs when GRF2 is lacking in mice. Interestingly, a similar phenomenon does not occur in adult mice.
Figure 4. Signaling cascades that contribute to LTP induction depend upon the environment to which an animal is exposed (see Li et al, 2006 
).
This phenomenon lasts for at least 3 months in enriched juvenile mice, whereupon it wanes and is absent by 6 months of age. Remarkably, this effect is passed on to the offspring of these enriched mice (see Arai et al, 2009 ). To define how this phenomenon is passed on to offspring of enriched adolescent mice, high-throughput genome-wide screening for EE-induced epigenetic marks on genes in offspring of enriched mice is being performed in collaboration with the Alex Meissner’s lab at Harvard, funded by a recently obtained Challenge grant from the NIH. You can read more about this work in commentaries in Nature News and Scientific American .
Ral GTPases in Ras-mediated Tumorigenesis
Constitutively activated Ras proteins are frequently found in many forms of human cancer, including colon, pancreatic and epidermal squamous cell carcinoma. How they promote oncogenesis through stimulation of their multiple effector proteins (Raf, PI3-K, Ral-GEFS, and others) in each of these cancers remains poorly understood. Evidence from various model systems of cancer supports the idea that Ral-GEFs and their downstream targets, RalA and RalB, can play significant positive roles in Ras-mediated oncogenesis. However, there are hints that RalA and/or RalB may sometimes suppress specific phenomena associated with transformation.
In order to understand how Ral GTPases functions in squamous carcinoma of the skin, where activated Ras is often an important component, we used a bioengineered human tissue model human skin. The overall goal of our work is to reveal how the Ral signaling cascade contributes to cancer in both tumor cells and adjacent stromal fibroblasts.
Figure 5. Engineering human skin tissue.
This system allows us to manipulate the Ral signaling cascade in either epithelial cells (HaCaT keratinocytes) of the epidermis or fibroblasts of the dermis. This work is done in collaboration with the laboratory of Dr. Jonathan Garlick who runs the Tufts Center for Integrated Tissue Engineering, http://dental.tufts.edu/CITE/. In these tissues, expression on oncogenic 12V-Ras (II-4 cells) is not sufficient to induce invasive behavior. Invasiveness is observed if E-cadherin function is also inhibited by expression of a dominant negative E-cadherin (II-4-dnE-cad).
Figure 6. Loss of E-cadherin function conferred by expression of a dominant negative E-cadherin correlates with invasiveness.
Ral in Keratinocytes
Ral proteins, RalA and RalB, are known to regulate the exocyst, a protein complex involved in vesicle sorting. As such, Ral proteins may influence polarity of epithelial cells to influence Ras-mediated tumorigenesis. Our recently published study (Sowalsky et al 2010) shows that in keratinocytes RalA actually SUPPRESSES, rather than promotes, Ras-induced tumorigenesis. This is most likely due to the ability of RalA, but not RalB, to stimulate exocyst function and enhance delivery of the tumor suppressor E-cadherin to the basolateral surface of epithelial cells.
Figure 7. Engineered tissues containing RalA knockdown, Ras-expressing keratinocytes look similar to those generated by knocking down E-cadherin levels and generate highly aggressive tumors when transplanted to mice.
Ral in Dermal Fibroblasts
Ral proteins may have distinct exocyst-mediated functions in stromal fibroblasts by influencing secretion of factors that influence tumor progression of neighboring epithelial cells. Studies on the consequence of blocking Ral function in stromal fibroblasts in these engineered tissues are under way.