The gene encoding the rat glycosylation enzyme beta1-4-N-acetylglucosaminyltransferase III (GnTIII) was cloned and coexpressed in a recombinant production Chinese hamster ovary (CHO) cell line expressing a chimeric mouse/human anti-CD20 IgG1 antibody. The new cell lines expressed high levels of antibody and have growth kinetics similar to that of the parent. Relative QPCR showed the cell lines to express varying levels of mRNA. High-performance liquid chromatography (HPLC) analysis showed the enzyme to have added bisecting N-acetylglucosamine (GlcNAc) residues in most (48% to 71%) of the N-linked oligosaccharides isolated from antibody preparations purified from the cell lines. In an ADCC assay the new antibody preparations promoted killing of CD20-positive target cells at approximately 10- to 20-fold lower concentrations than the parent. This activity was blocked using an anti-Fc gamma RIII antibody, supporting the role of Fc gamma RIII binding in this increase. In addition, cell binding assays showed the modified antibody bound better to Fc gamma RIII-expressing cells. The increase in ADCC activity is therefore likely due to an increased affinity of the modified antibody for the Fc gamma RIII receptor.

Cancer metastases arise from a multi-step process that requires metastasizing tumor cells to adapt to signaling input from varying tissue environments [1]. As an early metastatic event, cancer cell dissemination occurs through different migration programs, including multicellular, collective, and single-cell mesenchymal or amoeboid migration [2-4]. Migration modes can interconvert based on changes in cell adhesion, cytoskeletal mechanotransduction [5], and/or proteolysis [6], most likely under the control of transcriptional programs such as the epithelial-to-mesenchymal transition (EMT) [7, 8]. However, how plasticity of tumor cell migration and EMT is spatiotemporally controlled and connected upon challenge by the tumor microenvironment remains unclear. Using 3D cultures of collectively invading breast and head and neck cancer spheroids, here we identify hypoxia, a hallmark of solid tumors [9], as an inducer of the collective-to-amoeboid transition (CAT), promoting the dissemination of amoeboid-moving single cells from collective invasion strands. Hypoxia-induced amoeboid detachment was driven by hypoxia-inducible factor 1 (HIF-1), followed the downregulation of E-cadherin, and produced heterogeneous cell subsets whose phenotype and migration were dependent (∼30%) or independent (∼70%) of Twist-mediated EMT. EMT-like and EMT-independent amoeboid cell subsets showed stable amoeboid movement over hours as well as leukocyte-like traits, including rounded morphology, matrix metalloproteinase (MMP)-independent migration, and nuclear deformation. Cancer cells undergoing pharmacological stabilization of HIFs retained their constitutive ability for early metastatic seeding in an experimental model of lung metastasis, indicating that hypoxia-induced CAT enhances cell release rather than early organ colonization. Induced by metabolic challenge, amoeboid movement may thus constitute a common endpoint of both EMT-dependent and EMT-independent cancer dissemination programs.

The gene encoding the rat glycosylation enzyme beta1-4-N-acetylglucosaminyltransferase III (GnTIII) was cloned and coexpressed in a recombinant production Chinese hamster ovary (CHO) cell line expressing a chimeric mouse/human anti-CD20 IgG1 antibody. The new cell lines expressed high levels of antibody and have growth kinetics similar to that of the parent. Relative QPCR showed the cell lines to express varying levels of mRNA. High-performance liquid chromatography (HPLC) analysis showed the enzyme to have added bisecting N-acetylglucosamine (GlcNAc) residues in most (48% to 71%) of the N-linked oligosaccharides isolated from antibody preparations purified from the cell lines. In an ADCC assay the new antibody preparations promoted killing of CD20-positive target cells at approximately 10- to 20-fold lower concentrations than the parent. This activity was blocked using an anti-Fc gamma RIII antibody, supporting the role of Fc gamma RIII binding in this increase. In addition, cell binding assays showed the modified antibody bound better to Fc gamma RIII-expressing cells. The increase in ADCC activity is therefore likely due to an increased affinity of the modified antibody for the Fc gamma RIII receptor.

Hypoxia, through hypoxia inducible factor (HIF), drives cancer cell invasion and metastatic progression in various cancer types, leading to poor prognosis. In epithelial cancer, hypoxia further induces the transition to amoeboid cancer cell dissemination, yet the molecular mechanisms, relevance for metastasis, and effective interventions to combat hypoxia-induced amoeboid reprogramming remain unclear. Here, we identify calpain-2 as key regulator and anti-metastasis target of hypoxia-induced transition from collective to amoeboid dissemination of breast and head and neck (HN) carcinoma cells. Hypoxia-induced amoeboid dissemination occurred through low ECM-adhesive, bleb-based amoeboid movement, which effectively invaded into 3D collagen with low-oxidative and -glycolytic energy metabolism, revealing an microenvironmentally-induced, energy-conserving dissemination route in epithelial cancers. Hypoxia-induced calpain-2 mediated amoeboid conversion by deactivating beta1 integrins, through enzymatic cleavage of the focal adhesion adaptor protein talin-1.Consequently, targeted downregulation of calpain-2 or pharmacological intervention restored talin-1 integrity, beta1 integrin engagement and reverted blebbing-amoeboid to elongated phenotypes under hypoxia. Calpain-2 activity was required for hypoxia-induced blebbing-amoeboid conversion in the orthotopic mouse dermis, and upregulated in invasive HN tumor xenografts in vivo, and attenuation of calpain activity prevented hypoxia-induced metastasis to the lungs. This identifies the calpain-2/talin-1/beta1 integrin axis as mechanosignaling program and promising intervention target of plasticity of cancer cell invasion and metastasis formation in epithelial cancers under hypoxia. IntroductionCancer cell invasion initiates a multistep cascade to metastasis, which converts local neoplasia into a life-threatening systemic disease. 1-3 Invading cancer cells migrate away from the primary tumor and penetrate blood and lymph vessels, followed by systemic spreading and metastatic colonization of distant organs. 1,3 For tissue invasion, cancer cells deploy a range of collective and individual cell 3 migration strategies. Collective invasion of multicellular groups occurs when cells are held together by cell-cell adhesion, whereas single-cell migration lacks cell-cell cohesion and connectivity. 4 Mesenchymal single-cell migration depends on effective integrin-mediated adhesion to the extracellular matrix (ECM), which supports spindle-like cell elongation and directs matrix metalloproteases (MMPs) for ECM remodeling and path generation. 5 Amoeboid movement of roundish or ellipsoid cells engages only weak adhesion to the ECM, lacks ECM remodeling and, instead deploys kinetic deformation of the cell body for passage through 3D tissue. 6 Besides the cell shape and the strength of ECM interactions, protrusion types differ between amoeboid migration modes. Amoeboid-moving leukocytes and cancer cells develop actin-rich pseudopodia and filopodia at the leading edge, which generate protrusive...

Highlights d Hypoxia/HIF signaling converts collective invasion to blebbing-amoeboid migration d Calpain-2 cleaves talin-1, inactivates b1 integrins, and induces amoeboid conversion d Amoeboid movement occurs with reduced oxidative metabolism and glycolysis d Calpain-2 inhibition prevents amoeboid transition and HIFinduced metastasis