Search for: All records

Although synaptic evolution has been extensively studied, how axons first arose remains unexplored. Because evolution often occurs by coopting existing features, we review the evolutionary histories, biophysics, and cell biology of cytokinesis, cell crawling, and ciliogenesis to explore the origin of axons. Although we found that cilia and axons are outwardly similar, and growth cones strongly resemble the leading edge of crawling cells, the biophysical processes and the critical proteins that drive each seem weakly linked to axons as a structure. In contrast, the traction force machinery that pulls daughter cells apart during cytokinesis and the cytoskeletal organization of cytokinetic bridges appear to have a one-to-one correspondence to neuronal growth cones and axons. Based on these observations, we propose the hypothesis that axons evolved due to mutations that partially activated cytokinesis in an interphase cell. To rigorously test this hypothesis, we suggest conducting systematic phylogenetic analysis of the genes essential for each process, paired with molecular genetic studies in which critical genes are systematically disrupted. Doing so will provide a framework for understanding the relationship between diverse cellular processes, the early evolution of neurons, and insights that could potentially assist in treating cancer and promoting neuronal regeneration. 

Abstract Pre-training is a powerful paradigm in machine learning to pass information across models. For example, suppose one has a modest-sized dataset of images of cats and dogs and plans to fit a deep neural network to classify them. With pre-training, we start with a neural network trained on a large corpus of images of not just cats and dogs but hundreds of classes. We fix all network weights except the top layer(s) and fine tune on our dataset. This often results in dramatically better performance than training solely on our dataset. Here, we ask: ‘Can pre-training help the lasso?’. We propose a framework where the lasso is fit on a large dataset and then fine-tuned on a smaller dataset. The latter can be a subset of the original, or have a different but related outcome. This framework has a wide variety of applications, including stratified and multi-response models. In the stratified model setting, lasso pre-training first estimates coefficients common to all groups, then estimates group-specific coefficients during fine-tuning. Under appropriate assumptions, support recovery of the common coefficients is superior to the usual lasso trained on individual groups. This separate identification of common and individual coefficients also aids scientific understanding. 

We previously documented that individual microtubules in the axons of cultured juvenile rodent neurons consist of a labile domain and a stable domain and that experimental depletion of tau results in selective shortening and partial stabilization of the labile domain. After first confirming these findings in adult axons, we sought to understand the mechanism that accounts for the formation and maintenance of these microtubule domains. We found that fluorescent tau and MAP6 ectopically expressed in RFL-6 fibroblasts predominantly segregate on different microtubules or different domains on the same microtubule, with the tau-rich ones becoming more labile than in control cells and the MAP6-rich ones being more stable than in control cells. These and other experimental findings, which we studied further using computational modeling with tunable parameters, indicate that these two MAPs do not merely bind to pre-existing stable and labile domains but actually create stable and labile domains on microtubules. 

While the structural organization and molecular biology of neurons are well characterized, the physical process of axonal elongation remains elusive. The classic view posited elongation occurs through the deposition of cytoskeletal elements in the growth cone at the tip of a stationary array of microtubules. Yet, recent studies reveal axonal microtubules and docked organelles flow forward in bulk in the elongating axons ofAplysia, chick sensory, rat hippocampal, andDrosophilaneurons. Noting that the morphology, molecular components, and subcellular flow patterns of growth cones strongly resemble the leading edge of migrating cells and the polar regions of dividing cells, our working hypothesis is that axonal elongation utilizes the same physical mechanisms that drive cell crawling and cell division. As a test of that hypothesis, here we take experimental data sets of sub-cellular flow patterns in cells undergoing cytokinesis, mesenchymal migration, amoeboid migration, neuronal migration, and axonal elongation. We then apply active fluid theory to develop a biophysical model that describes the different sub-cellular flow profiles across these forms of motility and how this generates cell motility under low Reynolds numbers. The modeling suggests that mechanisms for generating motion are shared across these processes, and differences arise through modifications of sub-cellular adhesion patterns and the profiles of internal force generation. Collectively, this work suggests that ameboid and mesenchymal cell crawling may have arisen from processes that first developed to support cell division, that growth cone motility and cell crawling are closely related, and that neuronal migration and axonal elongation are fundamentally similar, differing primarily in the motion and strength of adhesion under the cell body. 

Clinical diagnosis typically incorporates physical examination, patient history, various laboratory tests, and imaging studies but makes limited use of the human immune system’s own record of antigen exposures encoded by receptors on B cells and T cells. We analyzed immune receptor datasets from 593 individuals to develop MAchine Learning for Immunological Diagnosis, an interpretive framework to screen for multiple illnesses simultaneously or precisely test for one condition. This approach detects specific infections, autoimmune disorders, vaccine responses, and disease severity differences. Human-interpretable features of the model recapitulate known immune responses to severe acute respiratory syndromecoronavirus2, influenza, and human immunodeficiency virus, highlight antigen-specific receptors, and reveal distinct characteristics of systemic lupus erythematosus and type-1 diabetes autoreactivity. This analysis framework has broad potential for scientific and clinical interpretation of immune responses.