In Science this week, an international team of scientists describes a metagenomic-based technique that generated models of 614 previously known protein family structures. Despite technological advances, there are about 5,200 protein families with unknown structures outside the range of comparative modeling. To address this knowledge gap, the investigators combined metagenomic data with calculations that predict protein structure based on residue interaction. They used the method to confirm the structure of 27 already known protein families, improving accuracy in 14 cases, while reducing accuracy in only one. They then used it to generate models of 614 protein families with unknown structures, of which about 140 represent new protein folds.
And in Science Translational Medicine, a multi-institute research team reports a new approach for monitoring the efficacy of cancer immunotherapies. Cancer treatments that use engineered CD8+ cytotoxic T lymphocytes (CTLs) to harness a patient's immune system against the disease show great promise, but once they are administered it is difficult to know whether they have localized to a tumor or not. In their study, the investigators modified CTLs to take up a fluorescent marker, allowing them to monitor their location when given to a handful of patients being treated for high-grade gliomas. The researchers were able to track the CTLs cells as they migrated to malignant lesions, as well as other non-target locations. "Further optimization of this imaging approach for monitoring in vivo cell trafficking should greatly benefit various cell-based therapies for cancer," the researchers write.