MeaningBERT: assessing meaning preservation between sentences.

In the field of automatic text simplification, assessing whether or not the meaning of the original text has been preserved during simplification is of paramount importance. Metrics relying on n-gram overlap assessment may struggle to deal with simplifications which replace complex phrases with their simpler paraphrases. Current evaluation metrics for meaning preservation based on large language models (LLMs), such as BertScore in machine translation or QuestEval in summarization, have been proposed. However, none has a strong correlation with human judgment of meaning preservation. Moreover, such metrics have not been assessed in the context of text simplification research. In this study, we present a meta-evaluation of several metrics we apply to measure content similarity in text simplification. We also show that the metrics are unable to pass two trivial, inexpensive content preservation tests. Another contribution of this study is MeaningBERT (https://github.com/GRAAL-Research/MeaningBERT), a new trainable metric designed to assess meaning preservation between two sentences in text simplification, showing how it correlates with human judgment. To demonstrate its quality and versatility, we will also present a compilation of datasets used to assess meaning preservation and benchmark our study against a large selection of popular metrics.

Impact of Concanavalin-A-Mediated Cytoskeleton Disruption on Low-Density Lipoprotein Receptor-Related Protein-1 Internalization and Cell Surface Expression in Glioblastomas.

The low-density lipoprotein receptor-related protein 1 (LRP-1) is a multiligand endocytic receptor, which plays a pivotal role in controlling cytoskeleton dynamics during cancer cell migration. Its rapid endocytosis further allows efficient clearance of extracellular ligands. Concanavalin-A (ConA) is a lectin used to trigger in vitro physiological cellular processes, including cytokines secretion, nitric oxide production, and T-lymphocytes activation. Given that ConA exerts part of its effects through cytoskeleton remodeling, we questioned whether it affected LRP-1 expression, intracellular trafficking, and cell surface function in grade IV U87 glioblastoma cells. Using flow cytometry and confocal microscopy, we found that loss of the cell surface 600-kDa mature form of LRP-1 occurs upon ConA treatment. Consequently, internalization of the physiological α2-macroglobulin and the synthetic angiopep-2 ligands of LRP-1 was also decreased. Silencing of known mediators of ConA, such as the membrane type-1 matrix metalloproteinase, and the Toll-like receptors (TLR)-2 and TLR-6 was unable to rescue ConA-mediated LRP-1 expression decrease, implying that the loss of LRP-1 was independent of cell surface relayed signaling. The ConA-mediated reduction in LRP-1 expression was emulated by the actin cytoskeleton-disrupting agent cytochalasin-D, but not by the microtubule inhibitor nocodazole, and required both lysosomal- and ubiquitin-proteasome system-mediated degradation. Our study implies that actin cytoskeleton integrity is required for proper LRP-1 cell surface functions and that impaired trafficking leads to specialized compartmentation and degradation. Our data also strengthen the biomarker role of cell surface LRP-1 functions in the vectorized transport of therapeutic angiopep bioconjugates into brain cancer cells.

PAX2 is activated by estradiol in breast cancer cells of the luminal subgroup selectively, to confer a low invasive phenotype.

BACKGROUND: Metastasis is the leading cause of death among breast cancer patients. Identifying key cellular factors controlling invasion and metastasis of breast cancer cells should pave the way to new therapeutic strategies efficiently interfering with the metastatic process. PAX2 (paired box 2) transcription factor is expressed by breast cancer cells in vivo and recently, it was shown to negatively regulate the expression of ERBB2 (erythroblastic leukemia viral oncogene homolog 2, HER-2/neu), a well-documented pro-invasive and pro-metastastic gene, in luminal/ERalpha-positive (ERα+) breast cancer cells. The objective of the present study was to investigate a putative role for PAX2 in the control of luminal breast cancer cells invasion, and to begin to characterize its regulation. RESULTS: PAX2 activity was higher in cell lines from luminal compared to non-luminal subtype, and activation of PAX2 by estradiol was selectively achieved in breast cancer cell lines of the luminal subtype. This process was blocked by ICI 182780 and could be antagonized by IGF-1. Knockdown of PAX2 in luminal MCF-7 cells completely abrogated estradiol-induced downregulation of ERBB2 and decrease of cell invasion, whereas overexpression of PAX2 in these cells enhanced estradiol effects on ERBB2 levels and cell invasion. CONCLUSIONS: The study demonstrates that PAX2 activation by estradiol is selectively achieved in breast cancer cells of the luminal subtype, via ERα, and identifies IGF-1 as a negative regulator of PAX2 activity in these cells. Further, it reveals a new role for PAX2 in the maintenance of a low invasive behavior in luminal breast cancer cells upon exposure to estradiol, and shows that overexpression and activation of PAX2 in these cells is sufficient to reduce their invasive ability.

Asymmetric lower extremity muscle activity on a gallium scan.

A 17-year-old girl with a history of Hodgkin disease showed increased gallium accumulation in the region of the right hip and reduced uptake in the muscles of the right lower extremity. Increased gallium accumulation in the right hip was attributed to avascular necrosis and the reduced gallium activity was attributed to disuse atrophy of the right lower extremity secondary to avascular necrosis (confirmed on the CT scan).