RUV-III-NB: normalization of single cell RNA-seq data.

Normalization of single cell RNA-seq data remains a challenging task. The performance of different methods can vary greatly between datasets when unwanted factors and biology are associated. Most normalization methods also only remove the effects of unwanted variation for the cell embedding but not from gene-level data typically used for differential expression (DE) analysis to identify marker genes. We propose RUV-III-NB, a method that can be used to remove unwanted variation from both the cell embedding and gene-level counts. Using pseudo-replicates, RUV-III-NB explicitly takes into account potential association with biology when removing unwanted variation. The method can be used for both UMI or read counts and returns adjusted counts that can be used for downstream analyses such as clustering, DE and pseudotime analyses. Using published datasets with different technological platforms, kinds of biology and levels of association between biology and unwanted variation, we show that RUV-III-NB manages to remove library size and batch effects, strengthen biological signals, improve DE analyses, and lead to results exhibiting greater concordance with independent datasets of the same kind. The performance of RUV-III-NB is consistent and is not sensitive to the number of factors assumed to contribute to the unwanted variation.

A new normalization for Nanostring nCounter gene expression data.

The Nanostring nCounter gene expression assay uses molecular barcodes and single molecule imaging to detect and count hundreds of unique transcripts in a single reaction. These counts need to be normalized to adjust for the amount of sample, variations in assay efficiency and other factors. Most users adopt the normalization approach described in the nSolver analysis software, which involves background correction based on the observed values of negative control probes, a within-sample normalization using the observed values of positive control probes and normalization across samples using reference (housekeeping) genes. Here we present a new normalization method, Removing Unwanted Variation-III (RUV-III), which makes vital use of technical replicates and suitable control genes. We also propose an approach using pseudo-replicates when technical replicates are not available. The effectiveness of RUV-III is illustrated on four different datasets. We also offer suggestions on the design and analysis of studies involving this technology.

GRIDSS: sensitive and specific genomic rearrangement detection using positional de Bruijn graph assembly.

The identification of genomic rearrangements with high sensitivity and specificity using massively parallel sequencing remains a major challenge, particularly in precision medicine and cancer research. Here, we describe a new method for detecting rearrangements, GRIDSS (Genome Rearrangement IDentification Software Suite). GRIDSS is a multithreaded structural variant (SV) caller that performs efficient genome-wide break-end assembly prior to variant calling using a novel positional de Bruijn graph-based assembler. By combining assembly, split read, and read pair evidence using a probabilistic scoring, GRIDSS achieves high sensitivity and specificity on simulated, cell line, and patient tumor data, recently winning SV subchallenge #5 of the ICGC-TCGA DREAM8.5 Somatic Mutation Calling Challenge. On human cell line data, GRIDSS halves the false discovery rate compared to other recent methods while matching or exceeding their sensitivity. GRIDSS identifies nontemplate sequence insertions, microhomologies, and large imperfect homologies, estimates a quality score for each breakpoint, stratifies calls into high or low confidence, and supports multisample analysis.

Correction to: Serum microRNA is a biomarker for post-operative monitoring in glioma.

For the reference citation '[57]' in the second paragraph of the Results section of the original article there was no corresponding entry in the References section. It should have referred to the below mentioned article by Ebrahimkhani et al. (2018).

Serum microRNA is a biomarker for post-operative monitoring in glioma.

PURPOSE: A circulating biomarker has potential to provide more accurate information for glioma progression post treatment, however no such biomarker is currently available. We aimed to discover a microRNA serum biomarker for longitudinal monitoring of glioma patients. METHODS: A prospectively collected cohort of 91 glioma patients and 17 healthy controls underwent pre and post-operative serum miRNA profiling using Nanostring®. Differentially expressed miRNAs were discovered using a machine learning random forest analysis. Candidate miRNAs were then assessed by droplet digital PCR in 11 patients with multiple follow up samples and compared to tumor volume based on magnetic resonance imaging. RESULTS: A 9-gene miRNA signature was identified that could distinguish between glioma and healthy controls with 99.8% accuracy. Two miRNAs miR-223 and miR-320e, best demonstrated dynamic changes that correlated closely with tumor volume in LGG and GBM respectively. Importantly, miRNA levels did not increase in two cases of pseudo-progression, indicating the potential utility of this test in guiding treatment decisions. CONCLUSIONS: We identified a highly accurate 9-miRNA signature associated with glioma serum. Additionally, we observed dynamic changes in specific miRNAs correlating with tumor volume over long-term follow up. These results support a large prospective validation study of serum miRNA biomarkers in glioma.

Reducing Artifactual EGFR T790M Mutations in DNA from Formalin-Fixed Paraffin-Embedded Tissue by Use of Thymine-DNA Glycosylase.

BACKGROUND: False-positive EGFR T790M mutations have been reported in formalin-fixed lung tumors, but the cause of the false positives has not been identified. The T790M mutation results from a C>T change at the cytosine of a CpG dinucleotide. The presence or absence of methylation at this cytosine has different consequences following deamination, resulting in a thymine or uracil, respectively, both of which however result in an artifactual change. Uracil-DNA glycosylase (UDG) can be used to eliminate DNA templates with uracil residues but is not active against artifactual thymines. We therefore investigated the use of thymine-DNA glycosylase (TDG) to reduce artifactual T790M mutations. METHODS: Formalin-fixed normal lung tissues and lung squamous cell carcinomas were tested to measure the frequency of false-positive EGFR mutations by use of droplet digital PCR before and after treatment with either UDG or TDG. Methylation at the cytosine at EGFR T790 was assessed by pyrosequencing and by analysis of public databases. RESULTS: Artifactual EGFR T790M mutations were detected in all of the archival formalin-fixed normal lung and lung squamous cell carcinomas at mutant allele frequencies of 1% or lower. The cytosine at EGFR T790 showed high levels of methylation in all lung cancer samples and normal tissues. Pretreatment of the formalin-fixed DNA with either UDG or TDG reduced the false EGFR T790M mutations, but a greater reduction was seen with the TDG treatment. CONCLUSIONS: Both U:G and T:G lesions in formalin-fixed tissue are sources of false-positive EGFR T790M mutations. This is the first report of the use of TDG to reduce sequence artifacts in formalin-fixed DNA and is applicable to the accurate detection of mutations arising at methylated cytosines.

Transketolase-like 1 ectopic expression is associated with DNA hypomethylation and induces the Warburg effect in melanoma cells.

BACKGROUND: The metabolism of cancer cells is often reprogrammed by dysregulation of metabolic enzymes. Transketolase-like 1 (TKTL1) is a homodimeric transketolase linking the pentose-phosphate pathway with the glycolytic pathway. It is generally silenced at a transcriptional level in somatic tissues. However, in human cancers its expression is associated with the acquisition of a glycolytic phenotype (the Warburg effect) by cancer cells that contributes to the progression of malignant tumors. In melanoma, defective promoter methylation results in the expression of genes and their products that can affect the tumor cell's phenotype including the modification of immune and functional characteristics. The present study evaluates the role of TKTL1 as a mediator of disease progression in melanoma associated with a defective methylation phenotype. METHODS: The expression of TKTL1 in metastatic melanoma tumors and cell lines was analysed by qRT-PCR and immunohistochemistry. The promoter methylation status of TKTL1 in melanoma cells was evaluated by quantitative methylation specific PCR. Using qRT-PCR, the effect of a DNA demethylating agent 5-aza-2'-deoxycytidine (5aza) on the expression of TKTL1 was examined. Biochemical and molecular analyses such as glucose consumption, lactate production, invasion, proliferation and cell cycle progression together with ectopic expression and siRNA mediated knockdown were used to investigate the role of TKTL1 in melanoma cells. RESULTS: Expression of TKTL1 was highly restricted in normal adult tissues and was overexpressed in a subset of metastatic melanoma tumors and derived cell lines. The TKTL1 promoter was activated by hypomethylation and treatment with 5aza induced TKTL1 expression in melanoma cells. Augmented expression of TKTL1 in melanoma cells was associated with a glycolytic phenotype. Loss and gain of function studies revealed that TKTL1 contributed to enhanced invasion of melanoma cells. CONCLUSIONS: Our data provide evidence for an important role of TKTL1 in aerobic glycolysis and tumor promotion in melanoma that may result from defective promoter methylation. This epigenetic change may enable the natural selection of tumor cells with a metabolic phenotype and thereby provide a potential therapeutic target for a subset of melanoma tumors with elevated TKTL1 expression.

Digital PCR of Genomic Rearrangements for Monitoring Circulating Tumour DNA.

Identifying circulating tumour DNA (ctDNA) for monitoring of cancer therapy is dependent on the development of readily designed, sensitive cancer-specific DNA markers. Genomic rearrangements that are present in the vast majority of cancers provide such markers.Tumour DNA isolated from two fresh-frozen lung tumours underwent whole genome sequencing. Genomic rearrangements were detected using a new computational algorithm, GRIDSS. Four genomic rearrangements from each tumour were chosen for further study using rearrangement-specific primers. Six of the eight rearrangements tested were identified as tumour-specific, the remaining two were present in the germline. ctDNA was quantified using digital PCR for the tumour genomic rearrangements in patient blood. Interestingly, one of the patients had no detectable ctDNA either prior to or post surgery although the rearrangements were readily detectable in the tumour DNA.This study demonstrates the feasibility of using digital PCR based on genomic rearrangements for the monitoring of minimal residual disease. In addition, whole genome sequencing provided further information enabling therapeutic choices including the identification of a cryptic EGFR exon 19 deletion in one patient and the identification of a high somatic mutation load in the other patient. This approach can be used as a model for all cancers with rearranged genomes.

Removing unwanted variation from large-scale RNA sequencing data with PRPS.

Accurate identification and effective removal of unwanted variation is essential to derive meaningful biological results from RNA sequencing (RNA-seq) data, especially when the data come from large and complex studies. Using RNA-seq data from The Cancer Genome Atlas (TCGA), we examined several sources of unwanted variation and demonstrate here how these can significantly compromise various downstream analyses, including cancer subtype identification, association between gene expression and survival outcomes and gene co-expression analysis. We propose a strategy, called pseudo-replicates of pseudo-samples (PRPS), for deploying our recently developed normalization method, called removing unwanted variation III (RUV-III), to remove the variation caused by library size, tumor purity and batch effects in TCGA RNA-seq data. We illustrate the value of our approach by comparing it to the standard TCGA normalizations on several TCGA RNA-seq datasets. RUV-III with PRPS can be used to integrate and normalize other large transcriptomic datasets coming from multiple laboratories or platforms.

ahctf1 and kras mutations combine to amplify oncogenic stress and restrict liver overgrowth in a zebrafish model of hepatocellular carcinoma.

The nucleoporin (NUP) ELYS, encoded by AHCTF1, is a large multifunctional protein with essential roles in nuclear pore assembly and mitosis. Using both larval and adult zebrafish models of hepatocellular carcinoma (HCC), in which the expression of an inducible mutant kras transgene (krasG12V) drives hepatocyte-specific hyperplasia and liver enlargement, we show that reducing ahctf1 gene dosage by 50% markedly decreases liver volume, while non-hyperplastic tissues are unaffected. We demonstrate that in the context of cancer, ahctf1 heterozygosity impairs nuclear pore formation, mitotic spindle assembly, and chromosome segregation, leading to DNA damage and activation of a Tp53-dependent transcriptional programme that induces cell death and cell cycle arrest. Heterozygous expression of both ahctf1 and ranbp2 (encoding a second nucleoporin), or treatment of heterozygous ahctf1 larvae with the nucleocytoplasmic transport inhibitor, Selinexor, completely blocks krasG12V-driven hepatocyte hyperplasia. Gene expression analysis of patient samples in the liver hepatocellular carcinoma (LIHC) dataset in The Cancer Genome Atlas shows that high expression of one or more of the transcripts encoding the 10 components of the NUP107-160 subcomplex, which includes AHCTF1, is positively correlated with worse overall survival. These results provide a strong and feasible rationale for the development of novel cancer therapeutics that target ELYS function and suggest potential avenues for effective combinatorial treatments.

Synthetic Epigenetic Reprogramming of Mesenchymal to Epithelial States Using the CRISPR/dCas9 Platform in Triple Negative Breast Cancer.

Epithelial-mesenchymal transition (EMT) is a reversible transcriptional program invoked by cancer cells to drive cancer progression. Transcription factor ZEB1 is a master regulator of EMT, driving disease recurrence in poor-outcome triple negative breast cancers (TNBCs). Here, this work silences ZEB1 in TNBC models by CRISPR/dCas9-mediated epigenetic editing, resulting in highly-specific and nearly complete suppression of ZEB1 in vivo, accompanied by long-lasting tumor inhibition. Integrated "omic" changes promoted by dCas9 linked to the KRAB domain (dCas9-KRAB) enabled the discovery of a ZEB1-dependent-signature of 26 genes differentially-expressed and -methylated, including the reactivation and enhanced chromatin accessibility in cell adhesion loci, outlining epigenetic reprogramming toward a more epithelial state. In the ZEB1 locus transcriptional silencing is associated with induction of locally-spread heterochromatin, significant changes in DNA methylation at specific CpGs, gain of H3K9me3, and a near complete erasure of H3K4me3 in the ZEB1 promoter. Epigenetic shifts induced by ZEB1-silencing are enriched in a subset of human breast tumors, illuminating a clinically-relevant hybrid-like state. Thus, the synthetic epi-silencing of ZEB1 induces stable "lock-in" epigenetic reprogramming of mesenchymal tumors associated with a distinct and stable epigenetic landscape. This work outlines epigenome-engineering approaches for reversing EMT and customizable precision molecular oncology approaches for targeting poor outcome breast cancers.

tidybulk: an R tidy framework for modular transcriptomic data analysis.

Recently, efforts have been made toward the harmonization of transcriptomic data structures and workflows using the concept of data tidiness, to facilitate modularisation. We present tidybulk, a modular framework for bulk transcriptional analyses that introduces a tidy transcriptomic data structure paradigm and analysis grammar. Tidybulk covers a wide variety of analysis procedures and integrates a large ecosystem of publicly available analysis algorithms under a common framework. Tidybulk decreases coding burden, facilitates reproducibility, increases efficiency for expert users, lowers the learning curve for inexperienced users, and bridges transcriptional data analysis with the tidyverse. Tidybulk is available at R/Bioconductor bioconductor.org/packages/tidybulk .

The Ratio of Exhausted to Resident Infiltrating Lymphocytes Is Prognostic for Colorectal Cancer Patient Outcome.

Immunotherapy success in colorectal cancer is mainly limited to patients whose tumors exhibit high microsatellite instability (MSI). However, there is variability in treatment outcomes within this group, which is in part driven by the frequency and characteristics of tumor-infiltrating immune cells. Indeed, the presence of specific infiltrating immune-cell subsets has been shown to correlate with immunotherapy response and is in many cases prognostic of treatment outcome. Tumor-infiltrating lymphocytes (TIL) can undergo distinct differentiation programs, acquiring features of tissue-residency or exhaustion, a process during which T cells upregulate inhibitory receptors, such as PD-1, and lose functionality. Although residency and exhaustion programs of CD8+ T cells are relatively well studied, these programs have only recently been appreciated in CD4+ T cells and remain largely unknown in tumor-infiltrating natural killer (NK) cells. In this study, we used single-cell RNA sequencing (RNA-seq) data to identify signatures of residency and exhaustion in colorectal cancer-infiltrating lymphocytes, including CD8+, CD4+, and NK cells. We then tested these signatures in independent single-cell data from tumor and normal tissue-infiltrating immune cells. Furthermore, we used versions of these signatures designed for bulk RNA-seq data to explore tumor-intrinsic mutations associated with residency and exhaustion from TCGA data. Finally, using two independent transcriptomic datasets from patients with colon adenocarcinoma, we showed that combinations of these signatures, in particular combinations of NK-cell activity signatures, together with tumor-associated signatures, such as TGFß signaling, were associated with distinct survival outcomes in patients with colon adenocarcinoma.

Melanoma brain metastases that progress on BRAF-MEK inhibitors demonstrate resistance to ipilimumab-nivolumab that is associated with the Innate PD-1 Resistance Signature (IPRES).

BACKGROUND: Melanoma brain metastases (MBMs) are a challenging clinical problem with high morbidity and mortality. Although first-line dabrafenib-trametinib and ipilimumab-nivolumab have similar intracranial response rates (50%-55%), central nervous system (CNS) resistance to BRAF-MEK inhibitors (BRAF-MEKi) usually occurs around 6 months, and durable responses are only seen with combination immunotherapy. We sought to investigate the utility of ipilimumab-nivolumab after MBM progression on BRAF-MEKi and identify mechanisms of resistance. METHODS: Patients who received first-line ipilimumab-nivolumab for MBMs or second/third line ipilimumab-nivolumab for intracranial metastases with BRAFV600 mutations with prior progression on BRAF-MEKi and MRI brain staging from March 1, 2015 to June 30, 2018 were included. Modified intracranial RECIST was used to assess response. Formalin-fixed paraffin-embedded samples of BRAFV600 mutant MBMs that were naïve to systemic treatment (n=18) or excised after progression on BRAF-MEKi (n=14) underwent whole transcriptome sequencing. Comparative analyses of MBMs naïve to systemic treatment versus BRAF-MEKi progression were performed. RESULTS: Twenty-five and 30 patients who received first and second/third line ipilimumab-nivolumab, were included respectively. Median sum of MBM diameters was 13 and 20.5 mm for the first and second/third line ipilimumab-nivolumab groups, respectively. Intracranial response rate was 75.0% (12/16), and median progression-free survival (PFS) was 41.6 months for first-line ipilimumab-nivolumab. Efficacy of second/third line ipilimumab-nivolumab after BRAF-MEKi progression was poor with an intracranial response rate of 4.8% (1/21) and median PFS of 1.3 months. Given the poor activity of ipilimumab-nivolumab after BRAF-MEKi MBM progression, we performed whole transcriptome sequencing to identify mechanisms of drug resistance. We identified a set of 178 differentially expressed genes (DEGs) between naïve and MBMs with progression on BRAF-MEKi treatment (p value <0.05, false discovery rate (FDR) <0.1). No distinct pathways were identified from gene set enrichment analyses using Kyoto Encyclopedia of Genes and Genomes, Gene Ontogeny or Hallmark libraries; however, enrichment of DEG from the Innate Anti-PD1 Resistance Signature (IPRES) was identified (p value=0.007, FDR=0.03). CONCLUSIONS: Second-line ipilimumab-nivolumab for MBMs after BRAF-MEKi progression has poor activity. MBMs that are resistant to BRAF-MEKi that also conferred resistance to second-line ipilimumab-nivolumab showed enrichment of the IPRES gene signature.

The Hippo pathway oncoprotein YAP promotes melanoma cell invasion and spontaneous metastasis.

Melanoma is a deadly form of skin cancer that accounts for a disproportionally large proportion of cancer-related deaths in younger people. Compared with most other skin cancers, a feature of melanoma is its high metastatic capacity, although the mechanisms that confer this are not well understood. The Hippo pathway is a key regulator of organ growth and cell fate that is deregulated in many cancers. To analyse the Hippo pathway in cutaneous melanoma, we generated a transcriptional signature of melanoma cells that overexpressed YAP, the key downstream Hippo pathway oncoprotein. YAP-mediated transcriptional activity varied in melanoma cell lines but did not cluster with known genetic drivers of melanomagenesis such as BRAF and NRAS mutations. Instead, it correlated strongly with published gene expression profiles linked to melanoma cell invasiveness and varied throughout the metastatic cascade in melanoma patient tumours. Consistent with this, YAP was both necessary and sufficient for melanoma cell invasion in vitro. In vivo, YAP promoted spontaneous melanoma metastasis, whilst the growth of YAP-expressing primary tumours was impeded. Finally, we identified the YAP target genes AXL, THBS1 and CYR61 as key mediators of YAP-induced melanoma cell invasion. These data suggest that YAP is a critical regulator of melanoma metastasis.

LRH-1 expression patterns in breast cancer tissues are associated with tumour aggressiveness.

The significance and regulation of liver receptor homologue 1 (LRH-1, NR5A2), a tumour-promoting transcription factor in breast cancer cell lines, is unknown in clinical breast cancers. This study aims to determine LRH-1/NR5A2 expression in breast cancers and relationship with DNA methylation and tumour characteristics. In The Cancer Genome Atlas breast cancer cohort NR5A2 expression was positively associated with intragenic CpG island methylation (1.4-fold expression for fully methylated versus not fully methylated, p=0.01) and inversely associated with promoter CpG island methylation (0.6-fold expression for fully methylated versus not fully methylated, p=0.036). LRH-1 immunohistochemistry of 329 invasive carcinomas and ductal carcinoma in situ (DCIS) was performed. Densely punctate/coarsely granular nuclear reactivity was significantly associated with high tumour grade (p<0.005, p=0.033 in invasive carcinomas and DCIS respectively), negative estrogen receptor status (p=0.008, p=0.038 in overall cohort and invasive carcinomas, respectively), negative progesterone receptor status (p=0.003, p=0.013 in overall cohort and invasive carcinomas, respectively), HER2 amplification (overall cohort p=0.034) and non-luminal intrinsic subtype (p=0.018, p=0.038 in overall cohort and invasive carcinomas, respectively). These significant associations of LRH-1 protein expression with tumour phenotype suggest that LRH-1 is an important indicator of tumour biology in breast cancers and may be useful in risk stratification.

A Panel of Cancer Testis Antigens and Clinical Risk Factors to Predict Metastasis in Colorectal Cancer.

Colorectal cancer (CRC) is the third common carcinoma with a high rate of mortality worldwide and several studies have investigated some molecular and clinicopathological markers for diagnosis and prognosis of its malignant phenotypes. The aim of this study is to evaluate expression frequency of PAGE4, SCP-1, and SPANXA/D cancer testis antigen (CTA) genes as well as some clinical risk markers to predict liver metastasis of colorectal cancer patients. The expression frequency of PAGE4, SCP-1, and SPANXA/D cancer/testis antigen (CTA) genes was obtained using reverse transcription polymerase chain reaction (RT-PCR) assay in 90 colorectal tumor samples including both negative and positive liver metastasis tumors. Statistical analysis was performed to assess the association of three studied genes and clinical risk factors with CRC liver metastasis. The frequency of PAGE4 and SCP-1 genes expression was significantly higher in the primary tumours with liver metastasis when statistically compared with primary tumors with no liver metastasis (P < 0.05). Among all clinical risk factors studied, the lymph node metastasis and the depth of invasion were statistically correlated with liver metastasis of CRC patients. In addition, using multiple logistic regression, we constructed a model based on PAGE4 and lymph node metastasis to predict liver metastasis of CRC.