Wave of the future: involvement of actin polymerization in the regulation of tissue growth and shape

ABSTRACT

The WASP-family verprolin-homologous (Wave) complex enhances actin-related protein 2/3 (Arp2/3)-mediated actin polymerization. Recently, we have identified a novel Wave complex–actin–Wnt/β-catenin–Sox9 pathway that regulates epidermal morphogenesis and proliferation. These findings highlight a mechanism by which actin polymerization impacts crucial signaling events in vivo.

The actin cytoskeleton is a complex cellular structure that plays a role in many biological processes. A wide range of biochemical and mechanical signals controls the activity of dozens of actin-binding proteins that regulate actin polymerization, assembly into more than 20 distinct actin-rich structures, and disassembly. For more than 50 years, thousands of research papers highlight the roles of the actin cytoskeleton in structural and mechanical cellular processes (e.g., cell shape, cell adhesion, cell polarity, cell migration, cytokinesis, etc.). More recently, a growing body of work demonstrates that actin and its binding proteins also function in fundamental regulatory processes including signal transduction, gene transcription, and cell fate determination. Therefore, the actin cytoskeleton links between cell mechanical properties and structural organization, and cell fate decisions.^1,2

The murine skin epidermis is an ideal model system to study how the actin cytoskeleton orchestrates the functions mentioned above since its development involves changes in both cell shape (from cuboidal to squamous) and cell fate (from stem cells/progenitors to keratinized dead cells). The epidermis begins its development soon after gastrulation when cells of the surface ectoderm become committed to the epidermal lineage. Later on, the tissue stratifies as cells of the basal layer detach from the basement membrane, exit the cell cycle, move towards the surface of the body, differentiate and function as a barrier. In parallel, some basal layer cells will grow towards the underlying connective tissue (dermis) and give rise to hair follicles that cover most of the mammalian body.¹

Recently, we uncovered a novel function for the WASP-family verprolin-homologous (Wave) complex. The heteropentameric Wave complex is known to activate actin-related protein 2/3 (Arp 2/3) complex-mediated actin polymerization to execute cells spreading, migration, adhesion, and division. We have identified a novel regulatory function for the Wave complex in suppressing the Wnt/β-catenin – Sox9 pathway.³

To study the function of the Wave complex in the developing epidermis, we identified short hairpin RNAs (shRNAs) that deplete the function of two essential Wave complex proteins: Abi1 and Wasf2 (also known as Wave2; encoded by Abi1 and Wasf2 respectively). We used ultrasound-guided in utero injection of lentiviruses that harbors shAbi1 or shWasf2 to knockdown (KD) the function of the Wave complex at embryonic day (E)9 in which the epidermis consists of a single layer of stem cells/progenitors. One week later we collected the KD embryos that exhibited an open-eyes phenotype, which is a common defect in actin cytoskeleton mutants. As expected, we documented a decrease in Filamentous (F)-actin content in the epidermis of Wave complex loss-of-function embryos. Since the actin cytoskeleton plays a role in epidermal morphogenesis and differentiation,⁴ we asked whether Wave complex activity is essential in these processes. We found that epidermal architecture was abnormal without Abi1 or Wave2 and defects were detected in basement membrane organization and spindle orientation.⁵ In contrast, epidermal differentiation markers were readily detected in Wave complex loss-of-function epidermis. However, the basal layer, in which mitotic activity takes place, was thicker than normal. In line with this observation, quantification of cell proliferation confirmed that Wave complex loss-of-function epidermis exhibits hyper-proliferation. To complete the analysis of Wave complex mutants, we turned to the developing hair follicles. We found that while their number is normal, their growth was delayed without Wave complex activity. Together, these observations demonstrate that Wave complex activity regulates epidermal shape and growth.

To understand why hair follicle growth was delayed we analyzed the distribution of Sox9, a transcription factor that regulates hair follicle cell proliferation but cannot be detected in wild-type epidermis. Unexpectedly, in Wave complex loss-of-function Sox9+ cells were no longer restricted to the hair follicles. In Abi1 or Wasf2 KD epidermis, Sox9+ cells were readily detected in the epidermis. Ectopic Sox9 expression was previously reported in basal cell carcinoma, where its activity enhances cell stemness, disrupted basement membrane organization, and alter spindle orientation.⁶ Given the striking similarity between Sox9-dependent processes in basal cell carcinoma and Wave complex loss-of-function phenotype in the developing epidermis we asked whether the ectopic expression of Sox9 can explain Wave complex loss-of-function phenotype. Indeed, forced overexpression of Sox9 in the developing epidermis gave rise to defects analogous to Wave complex loss-of-function: hyperproliferation, defects in basement membrane assembly, and defects in spindle orientation. Together, these observations suggest that ectopic expression of Sox9 affects both tissue shape and growth.

Sox9+ cell specification requires canonical Wnt signaling activity as an upstream signal in both the developing epidermis⁷ and basal cell carcinoma.⁶ In wild-type epidermis, nuclear localization of β-catenin and Wnt reporter activity are restricted to the developing hair follicles. In contrast, without Wave complex activity ectopic β-catenin nuclear localization and Wnt reporter activity were detected in the epidermis. The major function of the Wave complex is to enhance actin polymerization. To determine if actin polymerization affects β-catenin localization we treated cultured keratinocytes with drugs that increase (jasplakinolide) or decrease (latrunculin) F-actin content. We found that latrunculin, but not jasplakinolide treatment increased β-catenin nuclear localization. To determine if a decrease in actin levels is enough to trigger Wnt/β-catenin signaling in vivo, we used shRNA to moderately deplete β-actin levels (encoded by Actb). In line with our tissue culture results, this treatment triggered ectopic Wnt signaling activity in the developing epidermis.

Together, our study uncovers a novel Wave complex–F-actin–Wnt–Sox9 pathway that regulates epidermal growth and shape. Without Wave complex activity F-actin content is downregulated. This is sufficient to increase β-catenin nuclear accumulation and activate Wnt signaling which induces Sox9 expression. Sox9 ectopic activity in the epidermis alters the epidermal architecture and induces hyperproliferation (Figure 1).

Figure 1.

Wave complex activity controls epidermal growth and shape.

Without WASP-family verprolin-homologous (Wave) complex activity filamentous (F)-actin content in the developing epidermis is down-regulated. This is enough to induce β-catenin nuclear localization and trigger Wnt signaling (red nuclei). Downstream to Wnt signaling ectopic expression of Sox9 takes place (yellow nuclei), and its activity alters the epidermal architecture and induces hyper-proliferation.

It is important to note that changes in actin expression and polymerization,⁸ Wnt signaling, and Sox9 expression are all common in many cancers.⁶ Recently, Jung et al. showed that a decrease in actin polymerization destabilizes adherens junctions and increases Wnt signaling that initiates colorectal cancer.⁹ Lian et al. demonstrated that the actin-binding proteins Flna (filaminA) and Fmn2 (formin2) regulate canonical Wnt signaling in neural progenitor cells by mediating the transportation and endocytosis of Wnt components.¹⁰ Given the profound complexity of both the actin cytoskeleton and Wnt/β-catenin biology, additional work is required to understand their crosstalk in health and disease fully.

Funding Statement

This work was supported by grants to C. L. from the Israel Science Foundation (1113/15) and the Israeli Centers for Research Excellence (I-CORE) Gene Regulation in Complex Human Disease (41/11).

Disclosure of Potential Conflicts of Interest

The authors declare no competing financial interests.