Heavy Ion Minibeam Therapy: Side Effects in Normal Brain

The purpose of this work was to investigate whether minibeam therapy with heavy ions might offer improvements of the therapeutic ratio for the treatment of human brain cancers. To assess neurotoxicity, we irradiated normal juvenile rats using 120 MeV lithium-7 ions at an absorbed integral dose of 20 Gy. Beams were configured either as a solid parallel circular beam or as an array of planar parallel minibeams having 300-micron width and 1-mm center-to-center spacing within a circular array. We followed animals for 6 months after treatment and utilized behavioral testing and immunohistochemical studies to investigate the resulting cognitive impairment and chronic pathologic changes. We found both solid-beam therapy and minibeam therapy to result in cognitive impairment compared with sham controls, with no apparent reduction in neurotoxicity using heavy ion minibeams instead of solid beams under the conditions of this study.

Physical therapy exerts sub-additive and suppressive effects on intracerebral neural stem cell implantation in a rat model of stroke

Intracerebral cell therapy (CT) is emerging as a new therapeutic paradigm for stroke. However, the impact of physical therapy (PT) on implanted cells and their ability to promote recovery remains poorly understood. To address this translational issue, a clinical-grade neural stem cell (NSC) line was implanted into peri-infarct tissue using MRI-defined injection sites, two weeks after stroke. PT in the form of aerobic exercise (AE) was administered 5 × per week post-implantation using a paradigm commonly applied in patients with stroke. A combined AE and CT exerted sub-additive therapeutic effects on sensory neglect, whereas AE suppressed CT effects on motor integration and grip strength. Behavioral testing emerged as a potentially major component for task integration. It is expected that this study will guide and inform the incorporation of PT in the design of clinical trials evaluating intraparenchymal NSCs implantation for stroke.

The Small Molecule Alpha-Synuclein Aggregator, FN075, Enhances Alpha-Synuclein Pathology in Subclinical AAV Rat Models

Animal models of Parkinson’s disease, in which the human α-synuclein transgene is overexpressed in the nigrostriatal pathway using viral vectors, are widely considered to be the most relevant models of the human condition. However, although highly valid, these models have major limitations related to reliability and variability, with many animals exhibiting pronounced α-synuclein expression failing to demonstrate nigrostriatal neurodegeneration or motor dysfunction. Therefore, the aim of this study was to determine if sequential intra-nigral administration of AAV-α-synuclein followed by the small α-synuclein aggregating molecule, FN075, would enhance or precipitate the associated α-synucleinopathy, nigrostriatal pathology and motor dysfunction in subclinical models. Rats were given unilateral intra-nigral injections of AAV-α-synuclein (either wild-type or A53T mutant) followed four weeks later by a unilateral intra-nigral injection of FN075, after which they underwent behavioral testing for lateralized motor functionality until they were sacrificed for immunohistological assessment at 20 weeks after AAV administration. In line with expectations, both of the AAV vectors induced widespread overexpression of human α-synuclein in the substantia nigra and striatum. Sequential administration of FN075 significantly enhanced the α-synuclein pathology with increased density and accumulation of the pathological form of the protein phosphorylated at serine 129 (pS129-α-synuclein). However, despite this enhanced α-synuclein pathology, FN075 did not precipitate nigrostriatal degeneration or motor dysfunction in these subclinical AAV models. In conclusion, FN075 holds significant promise as an approach to enhancing the α-synuclein pathology in viral overexpression models, but further studies are required to determine if alternative administration regimes for this molecule could improve the reliability and variability in these models.

Co-Ultra PEALut Enhances Endogenous Repair Response Following Moderate Traumatic Brain Injury

This study aimed to assess the neuro-regenerative properties of co-ultramicronized PEALut (Glialia®), composed of palmitoylethanolamide (PEA) and the flavonoid luteolin (Lut), in an in vivo model of traumatic brain injury (TBI) and patients affected by moderate TBI. An increase in neurogenesis was seen in the mice at 72 h and 7 days after TBI. The co-ultra PEALut treatment helped the neuronal reconstitution process to restore the basal level of both novel and mature neurons; moreover, it induced a significant upregulation of the neurotrophic factors, which ultimately led to progress in terms of memory recall during behavioral testing. Moreover, our preliminary findings in a clinical trial suggested that Glialia® treatment facilitated neural recovery on working memory. Thus, co-ultra PEALut (Glialia®) could represent a valuable therapeutic agent for intensifying the endogenous repair response in order to better treat TBI.

Beyond Accuracy: Behavioral Testing of NLP Models with Checklist (Extended Abstract)

Although measuring held-out accuracy has been the primary approach to evaluate generalization, it often overestimates the performance of NLP models, while alternative approaches for evaluating models either focus on individual tasks or on specific behaviors. Inspired by principles of behavioral testing in software engineering, we introduce CheckList, a task-agnostic methodology for testing NLP models. CheckList includes a matrix of general linguistic capabilities and test types that facilitate comprehensive test ideation, as well as a software tool to generate a large and diverse number of test cases quickly. We illustrate the utility of CheckList with tests for three tasks, identifying critical failures in both commercial and state-of-art models. In a user study, a team responsible for a commercial sentiment analysis model found new and actionable bugs in an extensively tested model. In another user study, NLP practitioners with CheckList created twice as many tests, and found almost three times as many bugs as users without it.