Coactosin-like protein (CLP, or Cotl1), is an F-actin-binding protein, whose role in cancer is largely unknown. Here we show that CLP/Cotl1 is highly expressed in a rat epithelial breast cancer cell line (FE1.3) compared with its mesenchymal counterpart (FE1.2). Knockdown of CLP/Cotl1 in FE1.3 cells increased cell proliferation, whereas its overexpression in FE1.2 cells inhibited proliferation in culture and reduced tumor growth in xenograft assays in mice. Mechanistically, we identified two major pathways through which CLP/Cotl1 exerts its suppressive effects. First, CLP/Cotl1 re-expression in FE1.2 and in human MCF7 breast cancer cells induced expression of the growth-suppressor gene interleukin-24 (IL-24), which independently of p53 upregulates the tumor-suppressor genes p53 apoptosis effector related to PMP-22 (PERP) and p21. Second, overexpression of CLP/Cotl1 potentiated the growth-suppressive effect of transforming growth factor-β1 (TGFβ1), leading to downregulation of TGFβ-responsive genes vascular growth factor A/B (VEGFA/VEGFB), hypoxia inducing factor 1α (HIF-1α) and trombospondin 1 (TSP1), which mediate various hallmarks of cancer progression including angiogenesis, invasion and metastasis. CLP/Cotl1 inhibited TGFβ signaling via a non-canonical signaling involving IL-24-instigated inhibition of mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK) phosphorylation and subsequent post-transcriptional downregulation of SMAD2 and SMAD4. We also showed that CLP/COTL1 expression sensitizes breast cancer cells to chemotherapeutic drugs, and this was further enhanced by addition of exogenous TGFβ1. CLP/Cotl1 expression is lost in many human malignancies including prostate, uterine and breast cancers. Thus, our results uncover a novel tumor-suppressor role for CLP/Cotl1 and identify the downstream effectors interleukin 24 (IL-24)/PERP and IL-24/MAPK/ERK/TGFβ as potential targets for precision therapy.

Large-scale dimensional metrology usually requires a combination of multiple measurement systems, such as laser tracking, total station, laser scanning, coordinate measuring arm and video photogrammetry, etc. Often, the results from different measurement systems must be combined to provide useful results. The coordinate transformation is used to unify coordinate frames in combination; however, coordinate transformation uncertainties directly affect the accuracy of the final measurement results. In this paper, a novel method is proposed for improving the accuracy of coordinate transformation, combining the advantages of the best-fit least-square and radial basis function (RBF) neural networks. First of all, the configuration of coordinate transformation is introduced and a transformation matrix containing seven variables is obtained. Second, the 3D uncertainty of the transformation model and the residual error variable vector are established based on the best-fit least-square. Finally, in order to optimize the uncertainty of the developed seven-variable transformation model, we used the RBF neural network to identify the uncertainty of the dynamic, and unstructured, owing to its great ability to approximate any nonlinear function to the designed accuracy. Intensive experimental studies were conducted to check the validity of the theoretical results. The results show that the mean error of coordinate transformation decreased from 0.078 mm to 0.054 mm after using this method in contrast with the GUM method.