Plekhg5 controls the unconventional secretion of Sod1 by presynaptic secretory autophagy.

Increasing evidence suggests an essential function for autophagy in unconventional protein secretion (UPS). However, despite its relevance for the secretion of aggregate-prone proteins, the mechanisms of secretory autophagy in neurons have remained elusive. Here we show that the lower motoneuron disease-associated guanine exchange factor Plekhg5 drives the UPS of Sod1.

Fully closed and automated enrichment of primary blood dendritic cells for cancer immunotherapy.

Dendritic cell (DC) vaccination is a promising approach to induce tumor-specific immune responses in cancer patients. Until recently, most DC vaccines were based on in vitro-differentiated monocyte-derived DCs. However, through development of efficient isolation techniques, the use of primary blood dendritic cell subsets has come within reach.

Adjuvant dendritic cell therapy in stage IIIB/C melanoma: the MIND-DC randomized phase III trial.

Autologous natural dendritic cells (nDCs) treatment can induce tumor-specific immune responses and clinical responses in cancer patients. In this phase III clinical trial (NCT02993315), 148 patients with resected stage IIIB/C melanoma were randomized to adjuvant treatment with nDCs (n = 99) or placebo (n = 49). Active treatment consisted of intranodally injected autologous CD1c+ conventional and plasmacytoid DCs loaded with tumor antigens.

Blood DCs activated with R848 and poly(I:C) induce antigen-specific immune responses against viral and tumor-associated antigens.

Monocyte-derived Dendritic cells (DCs) have successfully been employed to induce immune responses against tumor-associated antigens in patients with various cancer entities. However, objective clinical responses have only been achieved in a minority of patients. Additionally, generation of GMP-compliant DCs requires time- and labor-intensive cell differentiation.

ProtTrans: Toward Understanding the Language of Life Through Self-Supervised Learning.

Computational biology and bioinformatics provide vast data gold-mines from protein sequences, ideal for Language Models (LMs) taken from Natural Language Processing (NLP). These LMs reach for new prediction frontiers at low inference costs. Here, we trained two auto-regressive models (Transformer-XL, XLNet) and four auto-encoder models (BERT, Albert, Electra, T5) on data from UniRef and BFD containing up to 393 billion amino acids.