The regulation of MacMARCKS expression by integrin β3

Reciprocal integrin/integrin antagonism through kindlin-2 and Rho GTPases regulates cell cohesion and collective migration

Collective cell behaviour during embryogenesis and tissue repair requires the coordination of intercellular junctions, cytoskeleton-dependent shape changes controlled by Rho GTPases, and integrin-dependent cell-matrix adhesion. Many different integrins are simultaneously expressed during wound healing, embryonic development, and sprouting angiogenesis, suggesting that there is extensive integrin/integrin cross-talk to regulate cell behaviour. Here, we show that fibronectin-binding β1 and β3 integrins do not act synergistically, but rather antagonize each other during collective cell processes in neuro-epithelial cells, placental trophoblasts, and endothelial cells. Reciprocal β1/β3 antagonism controls RhoA activity in a kindlin-2-dependent manner, balancing cell spreading, contractility, and intercellular adhesion. In this way, reciprocal β1/β3 antagonism controls cell cohesion and cellular plasticity to switch between extreme and opposing states, including epithelial versus mesenchymal-like phenotypes and collective versus individual cell migration. We propose that integrin/integrin antagonism is a universal mechanism to effectuate social cellular interactions, important for tissue morphogenesis, endothelial barrier function, trophoblast invasion, and sprouting angiogenesis.

Photoreceptor phagocytosis is mediated by phosphoinositide signaling

Circadian oscillations in peripheral tissues, such as the retinal compartment of the eye, are critical to anticipating changing metabolic demands. Circadian shedding of retinal photoreceptor cell discs with subsequent phagocytosis by the neighboring retinal pigmented epithelium (RPE) is essential for removal of toxic metabolites and lifelong survival of these postmitotic neurons. Defects in photoreceptor phagocytosis can lead to severe retinal pathology, but the biochemical mechanisms remain poorly defined. By first documenting a 2.8-fold burst of photoreceptor phagocytosis events in the mouse eye in the morning compared with the afternoon by serial block face imaging, we established time points to assess transcriptional readouts by RNA sequencing (RNA-Seq). We identified 365 oscillating protein-coding transcripts that implicated the phosphoinositide lipid signaling network mediating the discrete steps of photoreceptor phagocytosis. Moreover, examination of overlapping cistromic sites by core clock transcription factors and promoter elements of these effector genes provided a functional basis for the circadian cycling of these transcripts. RNA-Seq also revealed oscillating expression of 16 long intergenic noncoding RNAs and key histone modifying enzymes critical for circadian gene expression. Our phenotypic and genotypic characterization reveals a complex global landscape of overlapping and temporally controlled networks driving the essential circadian process in the eye.

The prognostic value of MARCKS-like 1 in lymph node-negative breast cancer

There is a need for new biomarkers to more correctly identify node-negative breast cancer patients with a good or bad prognosis. Myristoylated alanine-rich C kinase substrate like-1 (MARCKSL1) is a membrane-bound protein that is associated with cell spreading, integrin activation and exocytosis. Three hundred and five operable T(1,2)N(0)M(0) lymph node-negative breast cancer patients (median follow-up time 121 months, range 10-178 months) were evaluated for MARCKSL1 expression by immunohistochemistry and quantitative real-time PCR. The results were compared with classical prognosticators (age, tumor diameter, grade, estrogen receptor, and proliferation), using single (Kaplan-Meier) and multivariate survival analysis (Cox model). Forty-seven patients (15 %) developed distant metastases. With single and multivariate analysis of all features, MARCKSL1 protein expression was the strongest prognosticator (P < 0.001, HR = 5.1, 95 % CI = 2.7-9.8). Patients with high MARCKSL1 expression (n = 23) showed a 44 % survival versus 88 % in patients with low expression at 15-year follow-up. mRNA expression of MARCKSL1 in formalin fixed paraffin-embedded tissue was also prognostic (P = 0.002, HR = 3.6, 95 % CI = 1.5-8.3). However, the prognostic effect of high and low was opposite from the protein expression, i.e., low expression (relative expression ≤ 0.0264, n = 76) showed a 79 % survival versus 92 % in those with high expression of MARCKSL1 mRNA. Multivariate analysis of all features with distant metastases free survival as the end-point showed that the combination of MARCKSL1 protein and phosphohistone H3 (PPH3) has the strongest independent prognostic value. Patients with high expression (≥13) of PPH3 and high MARCKSL1 protein had 45 % survival versus 78 % survival for patients with low MARCKSL1 protein expression and high expression (≥13) of PPH3. In conclusion, MARCKSL1 has strong prognostic value in lymph node-negative breast cancer patients, especially in those with high proliferation.

Investigation into the mechanism regulating MRP localization

The major PKC substrates MARCKS and MacMARCKS (MRP) are membrane-binding proteins implicated in cell spreading, integrin activation and exocytosis. According to the myristoyl-electrostatic switch model the co-operation between the myristoyl moiety and the positively charged effector domain (ED) is an essential mechanism by which proteins bind to membranes. Loss of the electrostatic interaction between the ED and phospholipids, such as Ptdins(4,5)P2, results in the translocation of such proteins to the cytoplasm. While this model has been extensively tested for the binding of MARCKS far less is known about the mechanisms regulating MRP localization. We demonstrate that after phosphorylation, MRP is relocated to the intracellular membranes of late endosomes and lysosomes. MRP binds to all membranes via its myristoyl moiety, but for its localization at the plasma membrane the ED is also required. Although the ED of MRP can bind to Ptdins(4,5)P2 in vitro, this binding is not essential for its retention at or targeting to the plasma membrane. We conclude that the co-operation between the myristoyl moiety and the ED is not required for the binding to membranes in general but that it is essential for the targeting of MRP to the plasma membrane in a Ptdins(4,5)P2-independent manner.