Lysosomal sequestration of anti-cancer compounds reduces drug availability at intracellular target sites, thereby limiting drug-sensitivity and inducing chemoresistance. For hepatocellular carcinoma (HCC), sorafenib (SF) is the first line systemic treatment, as well as a simultaneous activator of autophagy-induced drug resistance. The purpose of this study is to elucidate how combination therapy with the FDA-approved photosensitizer verteporfin (VP) can potentiate the antitumor effect of SF, overcoming its acquired resistance mechanisms. HCC cell lines and patient-derived in vitro and in vivo preclinical models were used to identify the molecular mechanism of action of VP alone and in combination with SF. We demonstrate that SF is lysosomotropic and increases the total number of lysosomes in HCC cells and patient-derived xenograft model. Contrary to the effect on lysosomal stability by SF, VP is not only sequestered in lysosomes, but induces lysosomal pH alkalinization, lysosomal membrane permeabilization (LMP) and tumor-selective proteotoxicity. In combination, VP-induced LMP potentiates the antitumor effect of SF, further decreasing tumor proliferation and progression in HCC cell lines and patient-derived samples in vitro and in vivo. Our data suggest that combination of lysosome-targeting compounds, such as VP, in combination with already approved chemotherapeutic agents could open a new avenue to overcome chemo-insensitivity caused by passive lysosomal sequestration of anti-cancer drugs in the context of HCC.

Colorectal cancer, along with its high potential for recurrence and metastasis, is a major health burden. Uncovering proteins and pathways required for tumor cell growth is necessary for the development of novel targeted therapies. Ajuba is a member of the LIM domain family of proteins whose expression is positively associated with numerous cancers. Our data shows that Ajuba is highly expressed in human colon cancer tissue and cell lines. Publicly available data from The Cancer Genome Atlas shows a negative correlation between survival and Ajuba expression in patients with colon cancer. To investigate its function, we transduced SW480 human colon cancer cells, with lentiviral constructs to knockdown or overexpress Ajuba protein. The transcriptome of the modified cell lines was analyzed by RNA sequencing. Among the pathways enriched in the differentially expressed genes, were cell proliferation, migration and differentiation. We confirmed our sequencing data with biological assays; cells depleted of Ajuba were less proliferative, more sensitive to irradiation, migrated less and were less efficient in colony formation. In addition, loss of Ajuba expression decreased the tumor burden in a murine model of colorectal metastasis to the liver. Taken together, our data supports that Ajuba promotes colon cancer growth, migration and metastasis and therefore is a potential candidate for targeted therapy.

Assessing similarity is highly important for bioinformatics algorithms to determine correlations between biological information. A common problem is that similarity can appear by chance, particularly for low expressed entities. This is especially relevant in single-cell RNA-seq (scRNA-seq) data because read counts are much lower compared to bulk RNA-seq. Recently, a Bayesian correlation scheme that assigns low similarity to genes that have low confidence expression estimates has been proposed to assess similarity for bulk RNA-seq. Our goal is to extend the properties of the Bayesian correlation in scRNA-seq data by considering three ways to compute similarity. First, we compute the similarity of pairs of genes over all cells. Second, we identify specific cell populations and compute the correlation in those populations. Third, we compute the similarity of pairs of genes over all clusters, by considering the total mRNA expression. We demonstrate that Bayesian correlations are more reproducible than Pearson correlations. Compared to Pearson correlations, Bayesian correlations have a smaller dependence on the number of input cells. We show that the Bayesian correlation algorithm assigns high similarity values to genes with a biological relevance in a specific population. We conclude that Bayesian correlation is a robust similarity measure in scRNA-seq data.

SummaryThe liver is exemplar to study tissue regeneration due to its inherent ability of repair and regrowth. It replaces its lost or injured tissue by the proliferation, interaction and temporal coordination of multiple residential cell types. Until now we lacked a detailed description of the specific contributions of each cell type to the regenerative process, and therefore analyzed mouse livers 0, 3, 6, and 24 hours following two-thirds partial hepatectomy (PHx) by single cell RNA-sequencing (scRNA-seq) and mass cytometry. Our resulting genome wide temporal atlas contains the time dependent transcriptional changes in hepatocytes, endothelial cells, bone marrow-derived macrophages (BMDM) and Kupffer cells. In addition, it describes the cell specific contribution of mitogenic growth factors from biliary epithelial, endothelial and stellate cells as well as chemokines and cytokines from BMDM and granulocytes. And interestingly, Kupffer cells as opposed to hepatocytes emerged as the first cell to proliferate presenting a new dynamic in the liver following PHx. Here, we provide a robust data set at cellular resolution to uncover new elements and revisit current dogmas on the mechanisms underlying liver regeneration. To facilitate access to the data, we have launched the portal www.phxatlas.ch in which the scRNA-seq data can be visualized.

AbstractAssessing similarity is highly important for bioinformatics algorithms to determine correlations between biological information. A common problem is that similarity can appear by chance, particularly for low expressed entities. This is especially relevant in single cell RNA-seq (scRNA-seq) data because read counts are much lower compared to bulk RNA-seq.Recently, a Bayesian correlation scheme, that assigns low similarity to genes that have low confidence expression estimates, has been proposed to assess similarity for bulk RNA-seq. Our goal is to extend the properties of the Bayesian correlation in scRNA-seq data by considering 3 ways to compute similarity. First, we compute the similarity of pairs of genes over all cells. Second, we identify specific cell populations and compute the correlation in those populations. Third, we compute the similarity of pairs of genes over all clusters, by considering the total mRNA expression.We demonstrate that Bayesian correlations are more reproducible than Pearson correlations. Compared to Pearson correlations, Bayesian correlations have a smaller dependence on the number of input cells. We show that the Bayesian correlation algorithm assigns high similarity values to genes with a biological relevance in a specific population.We conclude that Bayesian correlation is a robust similarity measure in scRNA-seq data.

No abstract

Background The maintenance of genome stability is a key process to slow aging. One of the mechanisms ensuring this stability is the correct coordination of origins of replication (ORI), resulting in the successful transmission of DNA. Previously, we mapped and compared ORI firing between young and aged mice in vivo, using a regenerating liver mouse model. We confirmed a decreased hepatocyte ORI efficiency in aged mice, known to have impaired liver regeneration. ORI firing proved to be fully rescued when blocking the aged mice's ATR serine/threonine-protein kinase, suggesting that the DNA replication checkpoint actively mediates ORI firing impairment upon DNA damage detection. Aims To explore DNA damage differences between young and aged mice regenerating livers. Methods To induce proliferation, mice were subjected to partial hepatectomy (PH) and liver sections were harvested at different timepoints. Immunohistochemistry staining (IHC) were used to address proliferation and DNA damage, using Ki67 and serine 139 phosphorylated histone H2AX (g-H2AX) as markers, respectively. Results We confirmed a lack of Ki67 signal in young and aged mice prior to PH. The signal takes off at 24-28h post-PH and reaches its peak at 36-48h, which is significantly lower in aged mice. After 48h post-PH, young hepatocytes’ Ki67 reaches its basal level 120h post-PH. However, aged hepatocytes’ Ki67 is maintained at low levels overtime. Next, we compared DNA damage kinetics between young and aged mice livers. Both mouse groups present an increase of g-H2AX upon PH, higher in young mice. The g-H2AX signal decreases in young hepatocytes after 48h post-PH until disappearing 120h post PH. Aged mice hepatocytes maintain the g-H2AX rates overtime after 48h post-PH. Conclusions Our data suggests that hepatocytes develop DNA damage upon proliferation, which is able to be resolved in young mice hepatocytes but remains present in aged livers, ultimately leading to impaired liver regeneration.

No abstract