Past and Current Research
Cell Adhesion, the Cytoskeleton and Cancers.
As we discovered plakins that associate with actin and microtubule networks, we began to wonder why epidermal cells possess different integrins and cadherins that link to these other cytoskeletons rather than to keratin filaments. Thus, basal epidermal cells possess focal contacts that contain α3β1 integrins and adherens junctions that contain E-cadherin. Whereas the keratin-hemidesmosome-desmosome network is critical for mechanical integrity of the epidermal cells, the microtubule-actin-focal adhesion-adherens junction network appears to be essential for tissue architecture and function and for coordinating cellular movements during wound-healing.
In the late 1990's, we set up conditional gene targeting to knockout out genes in the stem cells and basal epidermal layer of mouse skin (Vasioukhin et al., 1999). One of the first genes we knocked out was α-catenin, an essential component for forming stable adherens junctions and linking them to the actin cytoskeleton (Vasioukhin et al., 2000). Surprisingly, the skin of the mice quickly resembled squamous cell carcinoma in situ, a precancerous lesion of the skin. When we got to the heart of the problem, we learned that α-catenin also regulates cell growth, by controlling interactions between cell-cell junctions and growth factor receptor pathways (Vasioukhin et al., 2001b). α-Catenin and p120-catenin also act to control NFkB activity, which links alterations in cell-cell adhesion with recruiting immune cells to the skin (Perez-Moreno et al., 2006; Kobielak et al., 2006; Perez-Moreno and Fuchs, 2006). This link could be important in wound-healing, where immune cells must be recruited to fight infections when cellular junctions and the epidermal barrier is severed. The link is also intriguing, as it may explain why mutations in catenins can lead to cancers that are frequently associated with a proinflammatory response.
Given the myriad of changes that occur upon loss of α-catenin, we realized that to delve deeper into mechanism, we needed a new method to be able to knockout α-catenin early in skin development and then monitor the temporal sequence of defects that arise. In the past several years, we've developed an early and efficient means of knocking down genes in mice by ultrasound-guided in utero infections of fluorescently traceable lentiviruses carrying Cre recombinase into the amniotic sac of mouse embryos (Beronja et al., 2010). Lentivirus infects only the first layer of cells it sees, which at E9.5 is the single-layered surface epithelium. The virus stably incorporates and propagates the desired genetic alterations, which allowed us to uncover new insights into α-catenin's role as a tumor suppressor. The relation between α-catenin mutations and human SCCs of the skin will be an interesting future avenue for study.
It has been over 20 years since researchers first noted a correlation between human cancers and alterations in the actin cytoskeleton. To begin to understand the role of the actin cytoskeleton in epithelial biology, we engineered mice expressing green fluorescent protein linked to α-actin. The beauty of skin as a model system is that we are then able to not only culture the normal epidermal keratinocytes to study actin dynamics during cell-cell adhesion, but we can also mate the mice on different genetic backgrounds, for instance the α-catenin null mice. Through imaging live cells by fluorescence microscopy, we've learned that like linking hands, adherens junctions integrate the actin cytoskeleton from cell to cell, across the entire tissue (Vaezi et al., 2002). In this way, a sheet of epidermal cells can move and function as if it were a single cell. Most recently, we've been exploring how actin dynamics impacts on tissue development, asymmetric cell division and planar cell polarity (PCP), a process controlling the angling of bristles in flies and hair follicles in mice (Luxenburg et al., 2011; Devenport et al., 2008; 2011).
Similar studies are ongoing with integrins. Some years ago, we ablated the β4 and β1 genes in mouse skin(Dowling et al., 1996; Raghavan et al., 2000). In the absence of β4 integrin, the epidermis of mice lose their anchorage to the underlying basement membrane, resulting in severe blistering, a condition known as junctional epidermolysis bullosa (JEB) in humans (Dowling et al., 1996). In the absence of β1, the epidermis can't assemble the basement membrane nor can it repair wounds properly (Raghavan et al., 2000). We've been focusing on how β1 works with the cytoskeleton to perform its tasks in epidermal keratinocytes. In the past year, we showed that the focal adhesion kinase (FAK), activated upon β1 signaling, promotes focal adhesion turnover and cell migration in the epidermis, and that TGFβ receptor signaling dampens the activity of FAK and places the brake on cell migration. Concomitantly, loss of FAK renders mice more resistant to SCC formation while loss of TGFβRII renders mice more prone to these cancers (Schober et al., 2007; Guasch et al., 2007).
Research over the prior decade has placed the laboratory squarely within the realm of focusing on how cytoskeletal and adhesive events go awry in tumorigenesis. To this end, we've exploited our mouse SCC models, which show an increased susceptibility (TGFβRII-null) and increased resistance (FAK-null) to tumor formation. We've shown that TGFβ and integrin/FAK signaling occur at the stroma-carcinoma interface, and play a critical role in regulating tumor growth (Guasch et al., 2007; Schober and Fuchs, 2011). Finally, we've now isolated and transcriptionally profiled the tumor-initiating, so-called cancer stem cells (CSCs) of SCCs. We've discovered that CSCs are high in integrins and their numbers and aggressiveness increase when cells are a) refractory to TGFβ signaling and b) able to upregulate FAK signaling. We've identified a unique signature for CSCs of SCCs, distinct from the normal stem cells of the skin.
This now provides a plethora of new information about the nature of these abundant cancers, and serves as a major avenue for future study. In the future, we will use our expertise in mouse genetics, knockdown and signaling/transcriptional profiling to elucidate the underlying mechanisms involved and to develop small molecule targets directed towards SCCs, the most common and life-threatening cancers world-wide. Understanding the complexities of tumor development and progression in the skin has become a major direction of my group.
