Unexpected gene activation following CRISPR-Cas9-mediated genome editing.

The discovery of the Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) and its development as a genome editing tool has revolutionized the field of molecular biology. In the DNA damage field, CRISPR has brought an alternative to induce endogenous double-strand breaks (DSBs) at desired genomic locations and study the DNA damage response and its consequences. Many systems for sgRNA delivery have been reported in order to efficiently generate this DSB, including lentiviral vectors. However, some of the consequences of these systems are not yet well understood. Here, we report that lentiviral-based sgRNA vectors can integrate into the endogenous genomic target location, leading to undesired activation of the target gene. By generating a DSB in the regulatory region of the ABCB1 gene using a lentiviral sgRNA vector, we can induce the formation of Taxol-resistant colonies. We show that these colonies upregulate ABCB1 via integration of the EEF1A1 and the U6 promoters from the sgRNA vector. We believe that this is an unreported CRISPR/Cas9 on-target effect that researchers need to be aware of when using lentiviral vectors for genome editing.

Double-strand break toxicity is chromatin context independent.

Cells respond to double-strand breaks (DSBs) by activating DNA damage response pathways, including cell cycle arrest. We have previously shown that a single double-strand break generated via CRISPR/Cas9 is sufficient to delay cell cycle progression and compromise cell viability. However, we also found that the cellular response to DSBs can vary, independent of the number of lesions. This implies that not all DSBs are equally toxic, and raises the question if the location of a single double-strand break could influence its toxicity. To systematically investigate if DSB-location is a determinant of toxicity we performed a CRISPR/Cas9 screen targeting 6237 single sites in the human genome. Next, we developed a data-driven framework to design CRISPR/Cas9 sgRNA (crRNA) pools targeting specific chromatin features. The chromatin context was defined using ChromHMM states, Lamin-B1 DAM-iD, DNAseI hypersensitivity, and RNA-sequencing data. We computationally designed 6 distinct crRNA pools, each containing 10 crRNAs targeting the same chromatin state. We show that the toxicity of a DSB is highly similar across the different ChromHMM states. Rather, we find that the major determinants of toxicity of a sgRNA are cutting efficiency and off-target effects. Thus, chromatin features have little to no effect on the toxicity of a single CRISPR/Cas9-induced DSB.

Predicting patient response with models trained on cell lines and patient-derived xenografts by nonlinear transfer learning.

Preclinical models have been the workhorse of cancer research, producing massive amounts of drug response data. Unfortunately, translating response biomarkers derived from these datasets to human tumors has proven to be particularly challenging. To address this challenge, we developed TRANSACT, a computational framework that builds a consensus space to capture biological processes common to preclinical models and human tumors and exploits this space to construct drug response predictors that robustly transfer from preclinical models to human tumors. TRANSACT performs favorably compared to four competing approaches, including two deep learning approaches, on a set of 23 drug prediction challenges on The Cancer Genome Atlas and 226 metastatic tumors from the Hartwig Medical Foundation. We demonstrate that response predictions deliver a robust performance for a number of therapies of high clinical importance: platinum-based chemotherapies, gemcitabine, and paclitaxel. In contrast to other approaches, we demonstrate the interpretability of the TRANSACT predictors by correctly identifying known biomarkers of targeted therapies, and we propose potential mechanisms that mediate the resistance to two chemotherapeutic agents.

Widespread chromatin context-dependencies of DNA double-strand break repair proteins.

DNA double-strand breaks are repaired by multiple pathways, including non-homologous end-joining (NHEJ) and microhomology-mediated end-joining (MMEJ). The balance of these pathways is dependent on the local chromatin context, but the underlying mechanisms are poorly understood. By combining knockout screening with a dual MMEJ:NHEJ reporter inserted in 19 different chromatin environments, we identified dozens of DNA repair proteins that modulate pathway balance dependent on the local chromatin state. Proteins that favor NHEJ mostly synergize with euchromatin, while proteins that favor MMEJ generally synergize with distinct types of heterochromatin. Examples of the former are BRCA2 and POLL, and of the latter the FANC complex and ATM. Moreover, in a diversity of human cancer types, loss of several of these proteins alters the distribution of pathway-specific mutations between heterochromatin and euchromatin. Together, these results uncover a complex network of proteins that regulate MMEJ:NHEJ balance in a chromatin context-dependent manner.

Perturbations in 3D genome organization can promote acquired drug resistance.

Acquired drug resistance is a major problem in the treatment of cancer. hTERT-immortalized, untransformed RPE-1 cells can acquire resistance to Taxol by derepressing the ABCB1 gene, encoding for the multidrug transporter P-gP. Here, we investigate how the ABCB1 gene is derepressed. ABCB1 activation is associated with reduced H3K9 trimethylation, increased H3K27 acetylation, and ABCB1 displacement from the nuclear lamina. While altering DNA methylation and H3K27 methylation had no major impact on ABCB1 expression, nor did it promote resistance, disrupting the nuclear lamina component Lamin B Receptor did promote the acquisition of a Taxol-resistant phenotype in a subset of cells. CRISPRa-mediated gene activation supported the notion that lamina dissociation influences ABCB1 derepression. We propose a model in which nuclear lamina dissociation of a repressed gene allows for its activation, implying that deregulation of the 3D genome topology could play an important role in tumor evolution and the acquisition of drug resistance.

Glucocorticoid receptor triggers a reversible drug-tolerant dormancy state with acquired therapeutic vulnerabilities in lung cancer.

The glucocorticoid receptor (GR) regulates gene expression, governing aspects of homeostasis, but is also involved in cancer. Pharmacological GR activation is frequently used to alleviate therapy-related side-effects. While prior studies have shown GR activation might also have anti-proliferative action on tumours, the underpinnings of glucocorticoid action and its direct effectors in non-lymphoid solid cancers remain elusive. Here, we study the mechanisms of glucocorticoid response, focusing on lung cancer. We show that GR activation induces reversible cancer cell dormancy characterised by anticancer drug tolerance, and activation of growth factor survival signalling accompanied by vulnerability to inhibitors. GR-induced dormancy is dependent on a single GR-target gene, CDKN1C, regulated through chromatin looping of a GR-occupied upstream distal enhancer in a SWI/SNF-dependent fashion. These insights illustrate the importance of GR signalling in non-lymphoid solid cancer biology, particularly in lung cancer, and warrant caution for use of glucocorticoids in treatment of anticancer therapy related side-effects.

TUBB3 overexpression has a negligible effect on the sensitivity to taxol in cultured cell lines.

Microtubules are cellular targets for a variety of anticancer therapies because of their critical function in mitosis. Taxol belongs to a class of microtubule targeting agents that suppresses microtubule dynamics and interferes with the functioning of the mitotic spindle, thereby effectively blocking cell cycle progression of rapidly proliferating tumor cells. Despite its antitumor activity, drug resistance remains a common obstacle in improving its overall clinical efficacy. Previous studies have shown that the expression of a specific ß-tubulin isotype, ßIII-tubulin/TUBB3, is dysregulated in drug-refractory tumors. However, whether enhanced TUBB3 expression is directly involved in promoting taxol resistance remains a subject of debate. Here, we have used several approaches to assess the functional relation of TUBB3 overexpression and taxol resistance. First, we generated a number of taxol-resistant cell lines, to find that TUBB3 expression was elevated in a resistant cell line (RPE-20) derived from untransformed retinal pigment epithelial (RPE) cells, but the abundance of TUBB3 remained unchanged in four other cell lines after taxol treatment. However, although RPE-20 cells displayed enhanced TUBB3 levels, we find that simultaneous up-regulation of the P-glycoprotein (P-gP) drug-efflux pump is responsible for the resistance to taxol. Indeed, we could show that TUBB3 levels were dynamically regulated upon taxol exposure and withdrawal, unrelated to the resistance phenotype. Next, we generated cell lines in which we could induce robust overexpression of TUBB3 from its endogenous locus employing the CRISPRa system. We demonstrate that solely enhancing TUBB3 expression results in a very minor decrease in the sensitivity to taxol. This was further substantiated by selective depletion of TUBB3 in a series of breast cancer cell lines expressing high levels of TUBB3. We find that TUBB3 depletion had a minimal effect on the sensitivity to taxol in one of these cell lines, but had no effect in all of the others. Based on these findings we propose that TUBB3 overexpression can only marginally affect the sensitivity to taxol in cultured cell lines.