When COS7 fibroblasts and other cells were exposed to UVC irradiation and cold shock at 4°C for 5 min, rapid upregulation and nuclear accumulation of NOS2, p53, WWOX, and TRAF2 occurred in 10–30 min. By time-lapse microscopy, an enlarging gas bubble containing nitric oxide (NO) was formed in the nucleus in each cell that finally popped out to cause “bubbling death”. Bubbling occurred effectively at 4 and 22°C, whereas DNA fragmentation was markedly blocked at 4°C. When temperature was increased to 37°C, bubbling was retarded and DNA fragmentation occurred in 1 hr, suggesting that bubbling death is switched to apoptosis with increasing temperatures. Bubbling occurred prior to nuclear uptake of propidium iodide and DAPI stains. Arginine analog Nω-LAME inhibited NO synthase NOS2 and significantly suppressed the bubbling death. Unlike apoptosis, there were no caspase activation and flip-over of membrane phosphatidylserine (PS) during bubbling death. Bubbling death was significantly retarded in Wwox knockout MEF cells, as well as in cells overexpressing TRAF2 and dominant-negative p53. Together, UV/cold shock induces bubbling death at 4°C and the event is switched to apoptosis at 37°C. Presumably, proapoptotic WWOX and p53 block the protective TRAF2 to execute the bubbling death.

The family of WW domain-containing proteins contains over 2000 members. The small WW domain module is responsible, in part, for protein/protein binding interactions and signaling. Many of these proteins are located at the membrane/cytoskeleton area, where they act as adaptors to receive signals from the cell surface. In this review, we provide molecular insights regarding recent novel findings on signaling from the cell surface toward WW domain-containing oxidoreductase, known as WWOX, FOR or WOX1. More specifically, transforming growth factor beta 1 utilizes cell surface hyaluronidase Hyal-2 (hyaluronoglucosaminidase 2) as a cognate receptor for signaling with WWOX and Smad4 to control gene transcription, growth and death. Complement C1q alone, bypassing the activation of classical pathway, signals a novel event of apoptosis by inducing microvillus formation and WWOX activation. Deficiency in these signaling events appears to favorably support cancer growth.

Not all leukemia T cells are susceptible to high levels of phorbol myristate acetate (PMA)-mediated apoptosis. At micromolar levels, PMA induces apoptosis of Jurkat T cells by causing mitochondrial polarization/de-polarization, release of cytosolic granules, and DNA fragmentation. Chemical inhibitors U0126 and PD98059 block mitogen-activated protein kinase kinase 1 (MEK1)-mediated phosphorylation of extracellular signal-regulated kinase (ERK) and prevent apoptosis. Mechanistically, proapoptotic tumor suppressor WOX1 (also named WWOX or FOR) physically interacts with MEK1, in part, in the lysosomes in Jurkat cells. PMA induces the dissociation, which leads to relocation of MEK1 to lipid rafts and WOX1 to the mitochondria for causing apoptosis. U0126 inhibits PMA-induced dissociation of WOX1/MEK1 complex and supports survival of Jurkat cells. In contrast, less differentiated Molt-4 T cells are resistant to PMA-induced dissociation of the WOX1/MEK1 complex and thereby are refractory to apoptosis. U0126 overturns the resistance for enhancing apoptosis in Molt-4 cells. Together, the in vivo MEK1/WOX1 complex is a master on/off switch for apoptosis in leukemia T cells.

This study proposed a hybrid neural network model that combines a self-organizing map (SOM) and back-propagation neural networks (BPNNs) to model the rainfall-runoff process in a physically interpretable manner and to accurately forecast typhoon floods. The SOM and a two-stage clustering scheme were applied to group hydrologic data into four clusters, each of which represented a meaningful hydrologic component of the rainfall-runoff process. BPNNs were constructed for each cluster to achieve high forecasting capability. The physical hybrid neural network model was used to forecast typhoon flood discharges in Wu River in Taiwan by using two types of rainfall data. The clustering results demonstrated that the rainfall-runoff process was favorably described by the sequence of derived clusters. The flood forecasting results indicated that the proposed hybrid neural network model has good forecasting capability, and the performance of the models using the two types of rainfall data is similar. In addition, the derived lagged inputs are hydrologically meaningful, and the number and activation function of the hidden nodes can be rationally interpreted. This study also developed a traditional, single BPNN model trained using the whole calibration data for comparison with the hybrid neural network model. The proposed physical hybrid neural network model outperformed the traditional neural network model in forecasting the peak discharges and low flows.