Transcriptomic profiling reveals three molecular phenotypes of adenocarcinoma at the gastroesophageal junction.

Cancers occurring at the gastroesophageal junction (GEJ) are classified as predominantly esophageal or gastric, which is often difficult to decipher. We hypothesized that the transcriptomic profile might reveal molecular subgroups which could help to define the tumor origin and behavior beyond anatomical location. The gene expression profiles of 107 treatment-naïve, intestinal type, gastroesophageal adenocarcinomas were assessed by the Illumina-HTv4.0 beadchip. Differential gene expression (limma), unsupervised subgroup assignment (mclust) and pathway analysis (gage) were undertaken in R statistical computing and results were related to demographic and clinical parameters. Unsupervised assignment of the gene expression profiles revealed three distinct molecular subgroups, which were not associated with anatomical location, tumor stage or grade (p > 0.05). Group 1 was enriched for pathways involved in cell turnover, Group 2 was enriched for metabolic processes and Group 3 for immune-response pathways. Patients in group 1 showed the worst overall survival (p = 0.019). Key genes for the three subtypes were confirmed by immunohistochemistry. The newly defined intrinsic subtypes were analyzed in four independent datasets of gastric and esophageal adenocarcinomas with transcriptomic data available (RNAseq data: OCCAMS cohort, n = 158; gene expression arrays: Belfast, n = 63; Singapore, n = 191; Asian Cancer Research Group, n = 300). The subgroups were represented in the independent cohorts and pooled analysis confirmed the prognostic effect of the new subtypes. In conclusion, adenocarcinomas at the GEJ comprise three distinct molecular phenotypes which do not reflect anatomical location but rather inform our understanding of the key pathways expressed.

ContrastivePose: A contrastive learning approach for self-supervised feature engineering for pose estimation and behavorial classification of interacting animals.

In recent years, supervised machine learning models trained on videos of animals with pose estimation data and behavior labels have been used for automated behavior classification. Applications include, for example, automated detection of neurological diseases in animal models. However, we identify two potential problems of such supervised learning approach. First, such models require a large amount of labeled data but the labeling of behaviors frame by frame is a laborious manual process that is not easily scalable. Second, such methods rely on handcrafted features obtained from pose estimation data that are usually designed empirically. In this paper, we propose to overcome these two problems using contrastive learning for self-supervised feature engineering on pose estimation data. Our approach allows the use of unlabeled videos to learn feature representations and reduce the need for handcrafting of higher-level features from pose positions. We show that this approach to feature representation can achieve better classification performance compared to handcrafted features alone, and that the performance improvement is due to contrastive learning on unlabeled data rather than the neural network architecture. The method has the potential to reduce the bottleneck of scarce labeled videos for training and improve performance of supervised behavioral classification models for the study of interaction behaviors in animals.

Mutational signature dynamics shaping the evolution of oesophageal adenocarcinoma.

A variety of mutational processes drive cancer development, but their dynamics across the entire disease spectrum from pre-cancerous to advanced neoplasia are poorly understood. We explore the mutagenic processes shaping oesophageal adenocarcinoma tumorigenesis in 997 instances comprising distinct stages of this malignancy, from Barrett Oesophagus to primary tumours and advanced metastatic disease. The mutational landscape is dominated by the C[T > C/G]T substitution enriched signatures SBS17a/b, which are linked with TP53 mutations, increased proliferation, genomic instability and disease progression. The APOBEC mutagenesis signature is a weak but persistent signal amplified in primary tumours. We also identify prevalent alterations in DNA damage repair pathways, with homologous recombination, base and nucleotide excision repair and translesion synthesis mutated in up to 50% of the cohort, and surprisingly uncoupled from transcriptional activity. Among these, the presence of base excision repair deficiencies show remarkably poor prognosis in the cohort. In this work, we provide insights on the mutational aetiology and changes enabling the transition from pre-neoplastic to advanced oesophageal adenocarcinoma.

Interfacial energetics approach for analysis of endothelial cell and segmental polyurethane interactions.

Understanding the physicochemical interactions between endothelial cells and biomaterials is vital for regenerative medicine applications. Particularly, physical interactions between the substratum interface and spontaneously deposited biomacromolecules as well as between the induced biomolecular interface and the cell in terms of surface energetics are important factors to regulate cellular functions. In this study, we examined the physical interactions between endothelial cells and segmental polyurethanes (PUs) using l-tyrosine based PUs to examine the structure-property relations in terms of PU surface energies and endothelial cell organization. Since, contact angle analysis used to probe surface energetics provides incomplete interpretation and understanding of the physical interactions, we sought a combinatorial surface energetics approach utilizing water contact angle, Zisman's critical surface tension (CST), Kaelble's numerical method, and van Oss-Good-Chaudhury theory (vOGCT), and applied to both substrata and serum adsorbed matrix to correlate human umbilical vein endothelial cell (HUVEC) behavior with surface energetics of l-tyrosine based PU surfaces. We determined that, while water contact angle of substratum or adsorbed matrix did not correlate well with HUVEC behavior, overall higher polarity according to the numerical method as well as Lewis base character of the substratum explained increased HUVEC interaction and monolayer formation as opposed to organization into networks. Cell interaction was also interpreted in terms of the combined effects of substratum and adsorbed matrix polarity and Lewis acid-base character to determine the effect of PU segments.

Hybrid cross-linking characteristics of hydrogel control stem cell fate.

Controlling hydrogel structures by combination of physical and chemical cross-links provides a novel system to regulate (stem) cell fate. In this study, we designed a polyethylene glycol (PEG)-based hydrogel where the polymer chains contain both physical and chemical cross-linking units in the same chain with self-assembling L-tyrosine-based dipeptides and photopolymerizable polyacrylate groups, respectively. It is shown that hydrogel architectures derived from this polymer are correlated to the cross-linking mechanisms. Combination of these cross-links controls three-dimensional gel architecture to regulate stem cell behavior in these hydrogels. Particularly, interaction of mesenchymal stem cells with the hydrogel enabled cellular aggregation to enhance chondrogenic differentiation as observed from the deposition of chondrogenic matrix. Increased chondrogenesis was due to enhanced cell-cell adhesion, which was mediated by gel morphology. This study shows the interplay of physical and chemical cross-links in hydrogels to regulate stem cell function and provides a novel molecular engineering tool for controlling hydrogel properties.

Hydrophilic polyurethane matrix promotes chondrogenesis of mesenchymal stem cells.

Segmental polyurethanes exhibit biphasic morphology and can control cell fate by providing distinct matrix guided signals to increase the chondrogenic potential of mesenchymal stem cells (MSCs). Polyethylene glycol (PEG) based hydrophilic polyurethanes can deliver differential signals to MSCs through their matrix phases where hard segments are cell-interactive domains and PEG based soft segments are minimally interactive with cells. These coordinated communications can modulate cell-matrix interactions to control cell shape and size for chondrogenesis. Biphasic character and hydrophilicity of polyurethanes with gel like architecture provide a synthetic matrix conducive for chondrogenesis of MSCs, as evidenced by deposition of cartilage-associated extracellular matrix. Compared to monophasic hydrogels, presence of cell interactive domains in hydrophilic polyurethanes gels can balance cell-cell and cell-matrix interactions. These results demonstrate the correlation between lineage commitment and the changes in cell shape, cell-matrix interaction, and cell-cell adhesion during chondrogenic differentiation which is regulated by polyurethane phase morphology, and thus, represent hydrophilic polyurethanes as promising synthetic matrices for cartilage regeneration.