Spatial meta-transcriptomics reveal associations of intratumor bacteria burden with lung cancer cells showing a distinct oncogenic signature.

BACKGROUND: The lung intratumor microbiome influences lung cancer tumorigenesis and treatment responses, but detailed data on the extent, location, and effects of microbes within lung tumors are missing, information needed for improved prognosis and treatment. METHODS: To address this gap, we developed a novel spatial meta-transcriptomic method simultaneously detecting the expression level of 1,811 host genes and 3 microbe targets (bacteria, fungi, and cytomegalovirus). After rigorous validation, we analyzed the spatial meta-transcriptomic profiles of tumor cells, T cells, macrophages, other immune cells, and stroma in surgically resected tumor samples from 12 patients with early-stage lung cancer. RESULTS: Bacterial burden was significantly higher in tumor cells compared with T cells, macrophages, other immune cells, and stroma. This burden increased from tumor-adjacent normal lung and tertiary lymphoid structures to tumor cells to the airways, suggesting that lung intratumor bacteria derive from the latter route of entry. Expression of oncogenic ß-catenin was strongly correlated with bacterial burden, as were tumor histological subtypes and environmental factors. CONCLUSIONS: Intratumor bacteria were enriched with tumor cells and associated with multiple oncogenic pathways, supporting a rationale for reducing the local intratumor microbiome in lung cancer for patient benefit. TRIAL REGISTRATION NUMBER: NCT00242723, NCT02146170.

Artificial Intelligence-based Tumor Segmentation in Mouse Models of Lung Adenocarcinoma.

BACKGROUND: Mouse models are highly effective for studying the pathophysiology of lung adenocarcinoma and evaluating new treatment strategies. Treatment efficacy is primarily determined by the total tumor burden measured on excised tumor specimens. The measurement process is time-consuming and prone to human errors. To address this issue, we developed a novel deep learning model to segment lung tumor foci on digitally scanned hematoxylin and eosin (H&E) histology slides. METHODS: Digital slides of 239 mice from 9 experimental cohorts were split into training (n=137), validation (n=37), and testing cohorts (n=65). Image patches of 500×500 pixels were extracted from 5× and 10× magnifications, along with binary masks of expert annotations representing ground-truth tumor regions. Deep learning models utilizing DeepLabV3+ and UNet architectures were trained for binary segmentation of tumor foci under varying stain normalization conditions. The performance of algorithm segmentation was assessed by Dice Coefficient, and detection was evaluated by sensitivity and positive-predictive value (PPV). RESULTS: The best model on patch-based validation was DeepLabV3+ using a Resnet-50 backbone, which achieved Dice 0.890 and 0.873 on validation and testing cohort, respectively. This result corresponded to 91.3 Sensitivity and 51.0 PPV in the validation cohort and 93.7 Sensitivity and 51.4 PPV in the testing cohort. False positives could be reduced 10-fold with thresholding artificial intelligence (AI) predicted output by area, without negative impact on Dice Coefficient. Evaluation at various stain normalization strategies did not demonstrate improvement from the baseline model. CONCLUSIONS: A robust AI-based algorithm for detecting and segmenting lung tumor foci in the pre-clinical mouse models was developed. The output of this algorithm is compatible with open-source software that researchers commonly use.

Unexpected guests in the tumor microenvironment: microbiome in cancer.

Although intestinal microbiome have been established as an important biomarker and regulator of cancer development and therapeutic response, less is known about the role of microbiome at other body sites in cancer. Emerging evidence has revealed that the local microbiota make up an important part of the tumor microenvironment across many types of cancer, especially in cancers arising from mucosal sites, including the lung, skin and gastrointestinal tract. The populations of bacteria that reside specifically within tumors have been found to be tumor-type specific, and mechanistic studies have demonstrated that tumor-associated microbiota may directly regulate cancer initiation, progression and responses to chemo- or immuno-therapies. This review aims to provide a comprehensive review of the important literature on the microbiota in the cancerous tissue, and their function and mechanism of action in cancer development and treatment.

The distal cytoplasmic tail of the influenza A M2 protein dynamically extends from the membrane.

The influenza A M2 protein is a multifunctional membrane-associated homotetramer that orchestrates several essential events in the viral infection cycle. The monomeric subunits of the M2 homotetramer consist of an N-terminal ectodomain, a transmembrane domain, and a C-terminal cytoplasmic domain. The transmembrane domain forms a four-helix proton channel that promotes uncoating of virions upon host cell entry. The membrane-proximal region of the C-terminal domain forms a surface-associated amphipathic helix necessary for viral budding. The structure of the remaining ~34 residues of the distal cytoplasmic tail has yet to be fully characterized despite the functional significance of this region for influenza infectivity. Here, we extend structural and dynamic studies of the poorly characterized M2 cytoplasmic tail. We used SDSL-EPR to collect site-specific information on the mobility, solvent accessibility, and conformational properties of residues 61-70 of the full-length, cell-expressed M2 protein reconstituted into liposomes. Our analysis is consistent with the predominant population of the C-terminal tail dynamically extending away from the membranes surface into the aqueous medium. These findings provide insight into the hypothesis that the C-terminal domain serves as a sensor that regulates how M2 protein participates in critical events in the viral infection cycle.