The identification of stem-cell-like cancer cells through conventional methods that depend on stem-cell markers is often unreliable. We developed a mechanical method of selecting tumourigenic cells by culturing single cancer cells in fibrin matrices of ~100 Pa in stiffness. When cultured within these gels, primary human cancer cells or single cancer cells from mouse or human cancer cell lines grew within a few days into individual round colonies that resembled embryonic stem-cell colonies. Subcutaneous or intravenous injection of 10 or 100 fibrin-cultured cells in syngeneic or severe-combined-immunodeficiency mice led to the formation of solid tumours at the site of injection or at the distant lung organ much more efficiently than control cancer cells selected using conventional surface marker methods or cultured on conventional rigid dishes or on soft gels. Remarkably, as few as 10 such cells were able to survive and form tumours in the lungs of wild-type non-syngeneic mice.

Evolutionary algorithms (EAs) have been applied with success to many numerical and combinatorial optimization problems in recent years. However, they often lose their effectiveness and advantages when applied to large and complex problems, e.g., those with high dimensions. Although cooperative coevolution has been proposed as a promising framework for tackling high-dimensional optimization problems, only limited studies were reported by decomposing a high-dimensional problem into single variables (dimensions). Such methods of decomposition often failed to solve nonseparable problems, for which tight interactions exist among different decision variables. In this paper, we propose a new cooperative coevolution framework that is capable of optimizing large scale nonseparable problems. A random grouping scheme and adaptive weighting are introduced in problem decomposition and coevolution. Instead of conventional evolutionary algorithms, a novel differential evolution algorithm is adopted. Theoretical analysis is presented in this paper to show why and how the new framework can be effective for optimizing large nonseparable problems. Extensive computational studies are also carried out to evaluate the performance of newly proposed algorithm on a large number of benchmark functions with up to 1000 dimensions. The results show clearly that our framework and algorithm are effective as well as efficient for large scale evolutionary optimisation problems. We are unaware of any other evolutionary algorithms that can optimize 1000-dimension nonseparable problems as effectively and efficiently as we have done.

Cellular microparticles are vesicular plasma membrane fragments with a diameter of 100-1,000 nanometres that are shed by cells in response to various physiological and artificial stimuli. Here we demonstrate that tumour cell-derived microparticles can be used as vectors to deliver chemotherapeutic drugs. We show that tumour cells incubated with chemotherapeutic drugs package these drugs into microparticles, which can be collected and used to effectively kill tumour cells in murine tumour models without typical side effects. We describe several mechanisms involved in this process, including uptake of drug-containing microparticles by tumour cells, synthesis of additional drug-packaging microparticles by these cells that contribute to the cytotoxic effect and the inhibition of drug efflux from tumour cells. This study highlights a novel drug delivery strategy with potential clinical application.

Multiobjective evolutionary algorithms (MOEAs) have been widely used in real-world applications. However, most MOEAs based on Pareto-dominance handle many-objective problems (MaOPs) poorly due to a high proportion of incomparable and thus mutually nondominated solutions. Recently, a number of many-objective evolutionary algorithms (MaOEAs) have been proposed to deal with this scalability issue. In this article, a survey of MaOEAs is reported. According to the key ideas used, MaOEAs are categorized into seven classes: relaxed dominance based, diversity-based, aggregation-based, indicator-based, reference set based, preference-based, and dimensionality reduction approaches. Several future research directions in this field are also discussed.

Cytokine release syndrome (CRS) counteracts the effectiveness of chimeric antigen receptor (CAR) T cell therapy in cancer patients, but the mechanism underlying CRS remains unclear. Here, we show that tumor cell pyroptosis triggers CRS during CAR T cell therapy. We find that CAR T cells rapidly activate caspase 3 in target cells through release of granzyme B. The latter cleaves gasdermin E (GSDME), a pore-forming protein highly expressed in B leukemic and other target cells, which results in extensive pyroptosis. Consequently, pyroptosis-released factors activate caspase 1 for GSDMD cleavage in macrophages, which results in the release of cytokines and subsequent CRS. Knocking out GSDME, depleting macrophages, or inhibiting caspase 1 eliminates CRS occurrence in mouse models. In patients, GSDME and lactate dehydrogenase levels are correlated with the severity of CRS. Notably, we find that the quantity of perforin/granzyme B used by CAR T cells rather than existing CD8+ T cells is critical for CAR T cells to induce target cell pyroptosis.