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.