Development and Validation of a Novel RNA Sequencing-Based Prognostic Score for Acute Myeloid Leukemia

Background

Recent progress in sequencing technologies allows us to explore comprehensive genomic and transcriptomic information to improve the current European LeukemiaNet (ELN) system of acute myeloid leukemia (AML).

Methods

We compared the prognostic value of traditional demographic and cytogenetic risk factors, genomic data in the form of somatic aberrations of 25 AML-relevant genes, and whole-transcriptome expression profiling (RNA sequencing) in 267 intensively treated AML patients (Clinseq-AML). Multivariable penalized Cox models (overall survival [OS]) were developed for each data modality (clinical, genomic, transcriptomic), together with an associated prognostic risk score.

Results

Of the three data modalities, transcriptomic data provided the best prognostic value, with an integrated area under the curve (iAUC) of a time-dependent receiver operating characteristic (ROC) curve of 0.73. We developed a prognostic risk score (Clinseq-G) from transcriptomic data, which was validated in the independent The Cancer Genome Atlas AML cohort (RNA sequencing, n = 142, iAUC = 0.73, comparing the high-risk group with the low-risk group, hazard ratio [HR]_OS = 2.42, 95% confidence interval [CI] = 1.51 to 3.88). Comparison between Clinseq-G and ELN score iAUC estimates indicated strong evidence in favor of the Clinseq-G model (Bayes factor = 26.78). The proposed model remained statistically significant in multivariable analysis including the ELN and other well-known risk factors (HR_os = 2.34, 95% CI = 1.30 to 4.22). We further validated the Clinseq-G model in a second independent data set (n = 458, iAUC = 0.66, adjusted HR_OS = 2.02, 95% CI = 1.33 to 3.08; adjusted HR_EFS = 2.10, 95% CI = 1.42 to 3.12).

Conclusions

Our results indicate that the Clinseq-G prediction model, based on transcriptomic data from RNA sequencing, outperforms traditional clinical parameters and previously reported models based on genomic biomarkers.

Publisher Correction: Evidence that breast cancer risk at the 2q35 locus is mediated through IGFBP5 regulation

This corrects the article DOI: 10.1038/ncomms5999.

Expression levels of long non-coding RNAs are prognostic for AML outcome

Background

Long non-coding RNA (lncRNA) expression has been implicated in a range of molecular mechanisms that are central in cancer. However, lncRNA expression has not yet been comprehensively characterized in acute myeloid leukemia (AML). Here, we assess to what extent lncRNA expression is prognostic of AML patient overall survival (OS) and determine if there are indications of lncRNA-based molecular subtypes of AML.

Methods

We performed RNA sequencing of 274 intensively treated AML patients in a Swedish cohort and quantified lncRNA expression. Univariate and multivariate time-to-event analysis was applied to determine association between individual lncRNAs with OS. Unsupervised statistical learning was applied to ascertain if lncRNA-based molecular subtypes exist and are prognostic.

Results

Thirty-three individual lncRNAs were found to be associated with OS (adjusted p value < 0.05). We established four distinct molecular subtypes based on lncRNA expression using a consensus clustering approach. LncRNA-based subtypes were found to stratify patients into groups with prognostic information (p value < 0.05). Subsequently, lncRNA expression-based subtypes were validated in an independent patient cohort (TCGA-AML). LncRNA subtypes could not be directly explained by any of the recurrent cytogenetic or mutational aberrations, although associations with some of the established genetic and clinical factors were found, including mutations in NPM1, TP53, and FLT3.

Conclusion

LncRNA expression-based four subtypes, discovered in this study, are reproducible and can effectively stratify AML patients. LncRNA expression profiling can provide valuable information for improved risk stratification of AML patients.

Electronic supplementary material

The online version of this article (10.1186/s13045-018-0596-2) contains supplementary material, which is available to authorized users.

Abstract P2-03-03: Molecular differences between screen-detected and interval breast cancers are largely explained by PAM50 subtypes

Purpose:Interval breast cancer is of clinical interest as it exhibits an aggressive phenotype and evades detection by screening mammography. A comprehensive picture of somatic changes that drive tumors to become symptomatic in the screening interval can improve understanding of the biology underlying these aggressive tumors.

Experimental design:Initiated in April 2013, Clinical Sequencing of Cancer in Sweden (Clinseq) is a scientific and clinical platform for the genomic profiling of cancer. The breast cancer pilot study consisted of women diagnosed with breast cancer between 2001-2012 in the Stockholm/Gotland regions. A subset of 318 breast tumors were sequenced, of which 113 were screen-detected and were 60 interval cancers.We applied targeted deep-sequencing of cancer-related genes, low-pass whole-genome sequencing and RNA-sequencing technology to characterize somatic differences in the genomic and transcriptomic architecture by interval cancer status. Mammographic density and PAM50 molecular subtypes were considered.

Results:In the crude analyses, TP53, PPP1R3A, and KMT2B were significantly more frequently mutated in interval cancers than in screen-detected cancers. Acquired somatic copy number aberrations with a frequency difference of at least 15% between the two groups included gains in 17q23-q25.3 and losses in 16q24.2. Gene expression analysis identified 447 significantly differentially expressed genes, of which 120 were replicated in an independent microarray dataset. After adjusting for PAM50, most differences were no longer significant.

Conclusions: Molecular differences by interval cancer status were observed, but they were largely explained by PAM50 subtypes. This work offer new insights into the biological differences between the two tumor groups.

Translational relevance: Although screen-detected cancers are biologically distinct from interval cancers in terms of somatic mutations, copy number aberrations and gene expression, most of the differences are no longer significant after adjusting for breast cancer intrinsic subtypes (PAM50). We also show that the molecular differences appear to form a spectrum from less aggressive (screen-detected) to more aggressive (interval) manifestations of the disease, which can be characterized by PAM50 subtypes, namely, luminal A, luminal B, HER2-enriched and basal-like, in that order. This work clarifies the picture on what type of breast cancer we are likely to identify through population-based screening, and what type of cancer we are likely to miss. Current knowledge of PAM50 subtype-specific risk factors need to be expanded as our findings might influence how we screen women with a higher risk of basal-like breast cancer for example, beyond known risk groups BRCA1 mutation carriers and women of African-American descent.

Citation Format: Czene K, Ivansson E, Klevebring D, Tobin NP, Lindström LS, Holm J, Prochazka G, Hilliges C, Palmgren J, Törnberg S, Humphreys K, Hartman J, Frisell J, Rantalainen M, Lindberg J, Hall P, Bergh J, Grönberg H, Li J. Molecular differences between screen-detected and interval breast cancers are largely explained by PAM50 subtypes [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P2-03-03.

Sequencing-based breast cancer diagnostics as an alternative to routine biomarkers

Sequencing-based breast cancer diagnostics have the potential to replace routine biomarkers and provide molecular characterization that enable personalized precision medicine. Here we investigate the concordance between sequencing-based and routine diagnostic biomarkers and to what extent tumor sequencing contributes clinically actionable information. We applied DNA- and RNA-sequencing to characterize tumors from 307 breast cancer patients with replication in up to 739 patients. We developed models to predict status of routine biomarkers (ER, HER2,Ki-67, histological grade) from sequencing data. Non-routine biomarkers, including mutations in BRCA1, BRCA2 and ERBB2(HER2), and additional clinically actionable somatic alterations were also investigated. Concordance with routine diagnostic biomarkers was high for ER status (AUC = 0.95;AUC(replication) = 0.97) and HER2 status (AUC = 0.97;AUC(replication) = 0.92). The transcriptomic grade model enabled classification of histological grade 1 and histological grade 3 tumors with high accuracy (AUC = 0.98;AUC(replication) = 0.94). Clinically actionable mutations in BRCA1, BRCA2 and ERBB2(HER2) were detected in 5.5% of patients, while 53% had genomic alterations matching ongoing or concluded breast cancer studies. Sequencing-based molecular profiling can be applied as an alternative to histopathology to determine ER and HER2 status, in addition to providing improved tumor grading and clinically actionable mutations and molecular subtypes. Our results suggest that sequencing-based breast cancer diagnostics in a near future can replace routine biomarkers.

An intergenic risk locus containing an enhancer deletion in 2q35 modulates breast cancer risk by deregulating IGFBP5 expression

Breast cancer is the most diagnosed malignancy and the second leading cause of cancer mortality in females. Previous association studies have identified variants on 2q35 associated with the risk of breast cancer. To identify functional susceptibility loci for breast cancer, we interrogated the 2q35 gene desert for chromatin architecture and functional variation correlated with gene expression. We report a novel intergenic breast cancer risk locus containing an enhancer copy number variation (enCNV; deletion) located approximately 400Kb upstream to IGFBP5, which overlaps an intergenic ERα-bound enhancer that loops to the IGFBP5 promoter. The enCNV is correlated with modified ERα binding and monoallelic-repression of IGFBP5 following oestrogen treatment. We investigated the association of enCNV genotype with breast cancer in 1,182 cases and 1,362 controls, and replicate our findings in an independent set of 62,533 cases and 60,966 controls from 41 case control studies and 11 GWAS. We report a dose-dependent inverse association of 2q35 enCNV genotype (percopy OR = 0.68 95%CI 0.55-0.83, P = 0.0002; replication OR = 0.77 95% CI 0.73-0.82, P = 2.1 × 10-19) and identify 13 additional linked variants (r2 > 0.8) in the 20Kb linkage block containing the enCNV (P = 3.2 × 10-15 - 5.6 × 10-17). These associations were independent of previously reported 2q35 variants, rs13387042/rs4442975 and rs16857609, and were stronger for ER-positive than ER-negative disease. Together, these results suggest that 2q35 breast cancer risk loci may be mediating their effect through IGFBP5.

Determining breast cancer histological grade from RNA-sequencing data

BACKGROUND

The histologic grade (HG) of breast cancer is an established prognostic factor. The grade is usually reported on a scale ranging from 1 to 3, where grade 3 tumours are the most aggressive. However, grade 2 is associated with an intermediate risk of recurrence, and carries limited information for clinical decision-making. Patients classified as grade 2 are at risk of both under- and over-treatment.

METHODS

RNA-sequencing analysis was conducted in a cohort of 275 women diagnosed with invasive breast cancer. Multivariate prediction models were developed to classify tumours into high and low transcriptomic grade (TG) based on gene- and isoform-level expression data from RNA-sequencing. HG2 tumours were reclassified according to the prediction model and a recurrence-free survival analysis was performed by the multivariate Cox proportional hazards regression model to assess to what extent the TG model could be used to stratify patients. The prediction model was validated in N=487 breast cancer cases from the The Cancer Genome Atlas (TCGA) data set. Differentially expressed genes and isoforms associated with HGs were analysed using linear models.

RESULTS

The classification of grade 1 and grade 3 tumours based on RNA-sequencing data achieved high accuracy (area under the receiver operating characteristic curve = 0.97). The association between recurrence-free survival rate and HGs was confirmed in the study population (hazard ratio of grade 3 versus 1 was 2.62 with 95 % confidence interval = 1.04-6.61). The TG model enabled us to reclassify grade 2 tumours as high TG and low TG gene or isoform grade. The risk of recurrence in the high TG group of grade 2 tumours was higher than in low TG group (hazard ratio = 2.43, 95 % confidence interval = 1.13-5.20). We found 8200 genes and 13,809 isoforms that were differentially expressed between HG1 and HG3 breast cancer tumours.

CONCLUSIONS

Gene- and isoform-level expression data from RNA-sequencing could be utilised to differentiate HG1 and HG3 tumours with high accuracy. We identified a large number of novel genes and isoforms associated with HG. Grade 2 tumours could be reclassified as high and low TG, which has the potential to reduce over- and under-treatment if implemented clinically.

Gene regulatory mechanisms underpinning prostate cancer susceptibility

Molecular characterization of genome-wide association study (GWAS) loci can uncover key genes and biological mechanisms underpinning complex traits and diseases. Here we present deep, high-throughput characterization of gene regulatory mechanisms underlying prostate cancer risk loci. Our methodology integrates data from 295 prostate cancer chromatin immunoprecipitation and sequencing experiments with genotype and gene expression data from 602 prostate tumor samples. The analysis identifies new gene regulatory mechanisms affected by risk locus SNPs, including widespread disruption of ternary androgen receptor (AR)-FOXA1 and AR-HOXB13 complexes and competitive binding mechanisms. We identify 57 expression quantitative trait loci at 35 risk loci, which we validate through analysis of allele-specific expression. We further validate predicted regulatory SNPs and target genes in prostate cancer cell line models. Finally, our integrated analysis can be accessed through an interactive visualization tool. This analysis elucidates how genome sequence variation affects disease predisposition via gene regulatory mechanisms and identifies relevant genes for downstream biomarker and drug development.

Study design requirements for RNA sequencing-based breast cancer diagnostics

Sequencing-based molecular characterization of tumors provides information required for individualized cancer treatment. There are well-defined molecular subtypes of breast cancer that provide improved prognostication compared to routine biomarkers. However, molecular subtyping is not yet implemented in routine breast cancer care. Clinical translation is dependent on subtype prediction models providing high sensitivity and specificity. In this study we evaluate sample size and RNA-sequencing read requirements for breast cancer subtyping to facilitate rational design of translational studies. We applied subsampling to ascertain the effect of training sample size and the number of RNA sequencing reads on classification accuracy of molecular subtype and routine biomarker prediction models (unsupervised and supervised). Subtype classification accuracy improved with increasing sample size up to N = 750 (accuracy = 0.93), although with a modest improvement beyond N = 350 (accuracy = 0.92). Prediction of routine biomarkers achieved accuracy of 0.94 (ER) and 0.92 (Her2) at N = 200. Subtype classification improved with RNA-sequencing library size up to 5 million reads. Development of molecular subtyping models for cancer diagnostics requires well-designed studies. Sample size and the number of RNA sequencing reads directly influence accuracy of molecular subtyping. Results in this study provide key information for rational design of translational studies aiming to bring sequencing-based diagnostics to the clinic.

Fine-mapping identifies two additional breast cancer susceptibility loci at 9q31.2

We recently identified a novel susceptibility variant, rs865686, for estrogen-receptor positive breast cancer at 9q31.2. Here, we report a fine-mapping analysis of the 9q31.2 susceptibility locus using 43 160 cases and 42 600 controls of European ancestry ascertained from 52 studies and a further 5795 cases and 6624 controls of Asian ancestry from nine studies. Single nucleotide polymorphism (SNP) rs676256 was most strongly associated with risk in Europeans (odds ratios [OR] = 0.90 [0.88-0.92]; P-value = 1.58 × 10(-25)). This SNP is one of a cluster of highly correlated variants, including rs865686, that spans ∼14.5 kb. We identified two additional independent association signals demarcated by SNPs rs10816625 (OR = 1.12 [1.08-1.17]; P-value = 7.89 × 10(-09)) and rs13294895 (OR = 1.09 [1.06-1.12]; P-value = 2.97 × 10(-11)). SNP rs10816625, but not rs13294895, was also associated with risk of breast cancer in Asian individuals (OR = 1.12 [1.06-1.18]; P-value = 2.77 × 10(-05)). Functional genomic annotation using data derived from breast cancer cell-line models indicates that these SNPs localise to putative enhancer elements that bind known drivers of hormone-dependent breast cancer, including ER-α, FOXA1 and GATA-3. In vitro analyses indicate that rs10816625 and rs13294895 have allele-specific effects on enhancer activity and suggest chromatin interactions with the KLF4 gene locus. These results demonstrate the power of dense genotyping in large studies to identify independent susceptibility variants. Analysis of associations using subjects with different ancestry, combined with bioinformatic and genomic characterisation, can provide strong evidence for the likely causative alleles and their functional basis.

Fine-scale mapping of the 5q11.2 breast cancer locus reveals at least three independent risk variants regulating MAP3K1

Genome-wide association studies (GWASs) have revealed SNP rs889312 on 5q11.2 to be associated with breast cancer risk in women of European ancestry. In an attempt to identify the biologically relevant variants, we analyzed 909 genetic variants across 5q11.2 in 103,991 breast cancer individuals and control individuals from 52 studies in the Breast Cancer Association Consortium. Multiple logistic regression analyses identified three independent risk signals: the strongest associations were with 15 correlated variants (iCHAV1), where the minor allele of the best candidate, rs62355902, associated with significantly increased risks of both estrogen-receptor-positive (ER(+): odds ratio [OR] = 1.24, 95% confidence interval [CI] = 1.21-1.27, ptrend = 5.7 × 10(-44)) and estrogen-receptor-negative (ER(-): OR = 1.10, 95% CI = 1.05-1.15, ptrend = 3.0 × 10(-4)) tumors. After adjustment for rs62355902, we found evidence of association of a further 173 variants (iCHAV2) containing three subsets with a range of effects (the strongest was rs113317823 [pcond = 1.61 × 10(-5)]) and five variants composing iCHAV3 (lead rs11949391; ER(+): OR = 0.90, 95% CI = 0.87-0.93, pcond = 1.4 × 10(-4)). Twenty-six percent of the prioritized candidate variants coincided with four putative regulatory elements that interact with the MAP3K1 promoter through chromatin looping and affect MAP3K1 promoter activity. Functional analysis indicated that the cancer risk alleles of four candidates (rs74345699 and rs62355900 [iCHAV1], rs16886397 [iCHAV2a], and rs17432750 [iCHAV3]) increased MAP3K1 transcriptional activity. Chromatin immunoprecipitation analysis revealed diminished GATA3 binding to the minor (cancer-protective) allele of rs17432750, indicating a mechanism for its action. We propose that the cancer risk alleles act to increase MAP3K1 expression in vivo and might promote breast cancer cell survival.

Evidence that breast cancer risk at the 2q35 locus is mediated through IGFBP5 regulation

GWAS have identified a breast cancer susceptibility locus on 2q35. Here we report the fine mapping of this locus using data from 101,943 subjects from 50 case-control studies. We genotype 276 SNPs using the 'iCOGS' genotyping array and impute genotypes for a further 1,284 using 1000 Genomes Project data. All but two, strongly correlated SNPs (rs4442975 G/T and rs6721996 G/A) are excluded as candidate causal variants at odds against >100:1. The best functional candidate, rs4442975, is associated with oestrogen receptor positive (ER+) disease with an odds ratio (OR) in Europeans of 0.85 (95% confidence interval=0.84-0.87; P=1.7 × 10(-43)) per t-allele. This SNP flanks a transcriptional enhancer that physically interacts with the promoter of IGFBP5 (encoding insulin-like growth factor-binding protein 5) and displays allele-specific gene expression, FOXA1 binding and chromatin looping. Evidence suggests that the g-allele confers increased breast cancer susceptibility through relative downregulation of IGFBP5, a gene with known roles in breast cell biology.

Evaluation of exome sequencing to estimate tumor burden in plasma

Accurate estimation of systemic tumor load from the blood of cancer patients has enormous potential. One avenue is to measure the presence of cell-free circulating tumor DNA in plasma. Various approaches have been investigated, predominantly covering hotspot mutations or customized, patient-specific assays. Therefore, we investigated the utility of using exome sequencing to monitor circulating tumor DNA levels through the detection of single nucleotide variants in plasma. Two technologies, claiming to offer efficient library preparation from nanogram levels of DNA, were evaluated. This allowed us to estimate the proportion of starting molecules measurable by sequence capture (<5%). As cell-free DNA is highly fragmented, we designed and provide software for efficient identification of PCR duplicates in single-end libraries with a varying size distribution. On average, this improved sequence coverage by 38% in comparison to standard tools. By exploiting the redundant information in PCR-duplicates the background noise was reduced to ∼1/35000. By applying our optimized analysis pipeline to a simulation analysis, we determined the current sensitivity limit to ∼1/2400, starting with 30 ng of cell-free DNA. Subsequently, circulating tumor DNA levels were assessed in seven breast- and one prostate cancer patient. One patient carried detectable levels of circulating tumor DNA, as verified by break-point specific PCR. These results demonstrate exome sequencing on cell-free DNA to be a powerful tool for disease monitoring of metastatic cancers. To enable a broad implementation in the diagnostic settings, the efficiency limitations of sequence capture and the inherent noise levels of the Illumina sequencing technology must be further improved.

Sequencing of breast cancer stem cell populations indicates a dynamic conversion between differentiation states in vivo

Introduction

The cancer stem cell model implies a hierarchical organization within breast tumors maintained by cancer stem-like cells (CSCs). Accordingly, CSCs are a subpopulation of cancer cells with capacity for self-renewal, differentiation and tumor initiation. These cells can be isolated through the phenotypic markers CD44+/CD24-, expression of ALDH1 and an ability to form nonadherent, multicellular spheres in vitro. However, controversies to describe the stem cell model exist; it is unclear whether the tumorigenicity of CSCs in vivo is solely a proxy for a certain genotype. Moreover, in vivo evidence is lacking to fully define the reversibility of CSC differentiation.

Methods

In order to answer these questions, we undertook exome sequencing of CSCs from 12 breast cancer patients, along with paired primary tumor samples. As suggested by stem classical cell biology, we assumed that the number of mutations in the CSC subpopulation should be lower and distinct compared to the differentiated tumor cells with higher proliferation.

Results

Our analysis revealed that the majority of somatic mutations are shared between CSCs and bulk primary tumor, with similar frequencies in the two.

Conclusions

The data presented here exclude the possibility that CSCs are only a phenotypic consequence of certain somatic mutations, that is a distinct and non-reversible population of cells. In addition, our results imply that CSCs must be a population of cells that can dynamically switch from differentiated tumor cells, and vice versa. This finding increases our understanding of CSC function in tumor heterogeneity and the importance of identifying drugs to counter de-differentiation rather than targeting CSCs.

CDK-mediated activation of the SCF(FBXO) (28) ubiquitin ligase promotes MYC-driven transcription and tumourigenesis and predicts poor survival in breast cancer

SCF (Skp1/Cul1/F-box) ubiquitin ligases act as master regulators of cellular homeostasis by targeting key proteins for ubiquitylation. Here, we identified a hitherto uncharacterized F-box protein, FBXO28 that controls MYC-dependent transcription by non-proteolytic ubiquitylation. SCF^FBXO28 activity and stability are regulated during the cell cycle by CDK1/2-mediated phosphorylation of FBXO28, which is required for its efficient ubiquitylation of MYC and downsteam enhancement of the MYC pathway. Depletion of FBXO28 or overexpression of an F-box mutant unable to support MYC ubiquitylation results in an impairment of MYC-driven transcription, transformation and tumourigenesis. Finally, in human breast cancer, high FBXO28 expression and phosphorylation are strong and independent predictors of poor outcome. In conclusion, our data suggest that SCF^FBXO28 plays an important role in transmitting CDK activity to MYC function during the cell cycle, emphasizing the CDK-FBXO28-MYC axis as a potential molecular drug target in MYC-driven cancers, including breast cancer.

Abstract 2296: Sequencing of breast cancer stem cells and corresponding primary bulk tumor reveals dominantly overlapping mutational spectra

Breast cancer cells with phenotypical similarities to adult mammary stem cells have been shown to initiate and maintain tumor growth and therefore named cancer stem cells (CSC). Phenotypic markers, such as CD44+/CD24-, ALDH1 High and PKH26High are all characterizing breast cancer cells with stem-like traits and ability to initiate tumor growth. There is substantial overlap of these markers with the biological capacity to form non-adherent cell spheres in vitro, named mammospheres. Based on the CSC concept, two alternative and opposing explanations have been proposed to describe the tumorigenic process governed by CSC; (1) the CSCs are an early-tumor initiating clone with capacity of asymmetric cell-divisions and subsequent differentiation to yield intratumoral heterogeneity; (2) the CSC phenotype is dynamic and reversible and can be adopted by any cancer epithelial cell through epigenetic de-differentiation.

To understand which of these models is the predominant, governing the development of breast cancer, we collected paired bulk tumor samples, in vitro purified mammospheres and normal leukocyte DNA from ten patients. DNA from mammospheres were amplified to increase the input DNA amount and in duplicates to control for the introduction of novel mutations. The sequencing was run with Illumina Hiseq 2000 using an exome capture kit. The sequence raw data was aligned to the human genome (hg19). Average target coverage was 53x. Mutations found in only one of the sequenced mammosphere amplicons and neither in the bulk tumor were considered artifacts by the DNA amplification and excluded from further analysis.

Across all patients, 80% of the mutations are on average shared between the tumor and CSC-fraction (range 45%-91%). We hypothesize that if the two cell populations, differentiated cells and CSC, were strictly separated, the number of mutations in the CSC would be very low due to low replication rate of that cell type compared to the continuously proliferating epithelial cells. In such a scenario, very few mutations would also be shared between CSC and bulk tumor cells. In contrast, we see a large degree of shared mutations. A cell-state model where inter-conversion between cell types occurs, would explain the large degree of shared mutations. As mutations occur in the rapidly dividing epithelial cells, they are propagated to daughter cells that are occasionally converted into CSCs. We also investigated the allele frequencies of the mutations discovered. There is no trend to larger frequencies in bulk tumor or in CSC, instead the frequencies are mostly equal in the two cellular components, pointing towards an inter-conversion of bulk tumor cells and CSC. Our data points towards the support of a model where the epithelial cells of the bulk tumor in vivo are able to phenotypically revert into CSC with stem cell properties and vise versa (model 2).

Citation Format: Gustaf Rosin, Daniel Klevebring, Ran Ma, Johan Lindberg, Lisa Viberg, Juha Kere, Kamila Czene, Claes Lindh, Irma Fredriksson, Jonas Bergh, Johan Hartman. Sequencing of breast cancer stem cells and corresponding primary bulk tumor reveals dominantly overlapping mutational spectra. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 2296. doi:10.1158/1538-7445.AM2013-2296

Functional variants at the 11q13 risk locus for breast cancer regulate cyclin D1 expression through long-range enhancers

Analysis of 4,405 variants in 89,050 European subjects from 41 case-control studies identified three independent association signals for estrogen-receptor-positive tumors at 11q13. The strongest signal maps to a transcriptional enhancer element in which the G allele of the best candidate causative variant rs554219 increases risk of breast cancer, reduces both binding of ELK4 transcription factor and luciferase activity in reporter assays, and may be associated with low cyclin D1 protein levels in tumors. Another candidate variant, rs78540526, lies in the same enhancer element. Risk association signal 2, rs75915166, creates a GATA3 binding site within a silencer element. Chromatin conformation studies demonstrate that these enhancer and silencer elements interact with each other and with their likely target gene, CCND1.

Contribution of antibody-based protein profiling to the human Chromosome-centric Proteome Project (C-HPP)

A gene-centric Human Proteome Project has been proposed to characterize the human protein-coding genes in a chromosome-centered manner to understand human biology and disease. Here, we report on the protein evidence for all genes predicted from the genome sequence based on manual annotation from literature (UniProt), antibody-based profiling in cells, tissues and organs and analysis of the transcript profiles using next generation sequencing in human cell lines of different origins. We estimate that there is good evidence for protein existence for 69% (n = 13985) of the human protein-coding genes, while 23% have only evidence on the RNA level and 7% still lack experimental evidence. Analysis of the expression patterns shows few tissue-specific proteins and approximately half of the genes expressed in all the analyzed cells. The status for each gene with regards to protein evidence is visualized in a chromosome-centric manner as part of a new version of the Human Protein Atlas ( www.proteinatlas.org ).

Library preparation and multiplex capture for massive parallel sequencing applications made efficient and easy

During the recent years, rapid development of sequencing technologies and a competitive market has enabled researchers to perform massive sequencing projects at a reasonable cost. As the price for the actual sequencing reactions drops, enabling more samples to be sequenced, the relative price for preparing libraries gets larger and the practical laboratory work becomes complex and tedious. We present a cost-effective strategy for simplified library preparation compatible with both whole genome- and targeted sequencing experiments. An optimized enzyme composition and reaction buffer reduces the number of required clean-up steps and allows for usage of bulk enzymes which makes the whole process cheap, efficient and simple. We also present a two-tagging strategy, which allows for multiplex sequencing of targeted regions. To prove our concept, we have prepared libraries for low-pass sequencing from 100 ng DNA, performed 2-, 4- and 8-plex exome capture and a 96-plex capture of a 500 kb region. In all samples we see a high concordance (>99.4%) of SNP calls when comparing to commercially available SNP-chip platforms.

Prognostic implications of estrogen receptor pattern of both tumors in contralateral breast cancer

Estrogen receptor (ER) status is important for breast cancer survival, it is however unclear how prognosis of contralateral breast cancer (CBC) is affected by ER-status of the two tumors. We conducted a large, population-based study of ER-status of both tumors in CBC patients and its influence on prognosis. The cohort consisted of all women diagnosed with CBC in Stockholm, Sweden during 1976-2005, with information on ER-status from medical records (N = 933). Prognosis was modeled as incidence rates of distant metastasis via Poisson regression. The proportion of CBCs with both cancers of the same ER-status was significantly larger than expected by chance. For synchronous (simultaneous) cancers the prognosis was significantly affected by the combined ER-status of both tumors (p = 0.01). Compared to unilateral breast cancer patients the incidence rate ratio (IRR) for patients with double ER-positive tumors was 1.25 (95 % CI: 0.88-1.76), for ER-discordant tumors 2.19 (95 % CI: 1.18-4.08) and for double ER-negative tumors 3.95 (95 % CI: 1.77-8.81). For metachronous (non-simultaneous) cancers, women with double ER-positive tumors had similarly bad prognosis (IRR = 2.95; 95 % CI: 2.39-3.64) as women with double ER-negative tumors (IRR = 2.88; 95 % CI: 1.83-4.52). Both shorter time span between first and second cancer and endocrine therapy for the first cancer further worsened prognosis of women with double ER-positive metachronous CBC. For synchronous CBC patients, ER-pattern of both tumors is an important prognosticator, while among metachronous CBC patients, double ER-positive tumors confer equally bad prognosis as double ER-negative cancers. Our results indicate that this might be due to endocrine therapy resistance.

Antibody-based protein profiling of the human chromosome 21

The Human Proteome Project has been proposed to create a knowledge-based resource based on a systematical mapping of all human proteins, chromosome by chromosome, in a gene-centric manner. With this background, we here describe the systematic analysis of chromosome 21 using an antibody-based approach for protein profiling using both confocal microscopy and immunohistochemistry, complemented with transcript profiling using next generation sequencing data. We also describe a new approach for protein isoform analysis using a combination of antibody-based probing and isoelectric focusing. The analysis has identified several genes on chromosome 21 with no previous evidence on the protein level, and the isoform analysis indicates that a large fraction of human proteins have multiple isoforms. A chromosome-wide matrix is presented with status for all chromosome 21 genes regarding subcellular localization, tissue distribution, and molecular characterization of the corresponding proteins. The path to generate a chromosome-specific resource, including integrated data from complementary assay platforms, such as mass spectrometry and gene tagging analysis, is discussed.

Defining the transcriptome and proteome in three functionally different human cell lines

An essential question in human biology is how cells and tissues differ in gene and protein expression and how these differences delineate specific biological function. Here, we have performed a global analysis of both mRNA and protein levels based on sequence-based transcriptome analysis (RNA-seq), SILAC-based mass spectrometry analysis and antibody-based confocal microscopy. The study was performed in three functionally different human cell lines and based on the global analysis, we estimated the fractions of mRNA and protein that are cell specific or expressed at similar/different levels in the cell lines. A highly ubiquitous RNA expression was found with >60% of the gene products detected in all cells. The changes of mRNA and protein levels in the cell lines using SILAC and RNA ratios show high correlations, even though the genome-wide dynamic range is substantially higher for the proteins as compared with the transcripts. Large general differences in abundance for proteins from various functional classes are observed and, in general, the cell-type specific proteins are low abundant and highly enriched for cell-surface proteins. Thus, this study shows a path to characterize the transcriptome and proteome in human cells from different origins.

Analysis of transcript and protein overlap in a human osteosarcoma cell line

Background

An interesting field of research in genomics and proteomics is to compare the overlap between the transcriptome and the proteome. Recently, the tools to analyse gene and protein expression on a whole-genome scale have been improved, including the availability of the new generation sequencing instruments and high-throughput antibody-based methods to analyze the presence and localization of proteins. In this study, we used massive transcriptome sequencing (RNA-seq) to investigate the transcriptome of a human osteosarcoma cell line and compared the expression levels with in situ protein data obtained in-situ from antibody-based immunohistochemistry (IHC) and immunofluorescence microscopy (IF).

Results

A large-scale analysis based on 2749 genes was performed, corresponding to approximately 13% of the protein coding genes in the human genome. We found the presence of both RNA and proteins to a large fraction of the analyzed genes with 60% of the analyzed human genes detected by all three methods. Only 34 genes (1.2%) were not detected on the transcriptional or protein level with any method. Our data suggest that the majority of the human genes are expressed at detectable transcript or protein levels in this cell line. Since the reliability of antibodies depends on possible cross-reactivity, we compared the RNA and protein data using antibodies with different reliability scores based on various criteria, including Western blot analysis. Gene products detected in all three platforms generally have good antibody validation scores, while those detected only by antibodies, but not by RNA sequencing, generally consist of more low-scoring antibodies.

Conclusion

This suggests that some antibodies are staining the cells in an unspecific manner, and that assessment of transcript presence by RNA-seq can provide guidance for validation of the corresponding antibodies.

In-depth transcriptome analysis reveals novel TARs and prevalent antisense transcription in human cell lines

Several recent studies have indicated that transcription is pervasive in regions outside of protein coding genes and that short antisense transcripts can originate from the promoter and terminator regions of genes. Here we investigate transcription of fragments longer than 200 nucleotides, focusing on antisense transcription for known protein coding genes and intergenic transcription. We find that roughly 12% to 16% of all reads that originate from promoter and terminator regions, respectively, map antisense to the gene in question. Furthermore, we detect a high number of novel transcriptionally active regions (TARs) that are generally expressed at a lower level than protein coding genes. We find that the correlation between RNA-seq data and microarray data is dependent on the gene length, with longer genes showing a better correlation. We detect high antisense transcriptional activity from promoter, terminator and intron regions of protein-coding genes and identify a vast number of previously unidentified TARs, including putative novel EGFR transcripts. This shows that in-depth analysis of the transcriptome using RNA-seq is a valuable tool for understanding complex transcriptional events. Furthermore, the development of new algorithms for estimation of gene expression from RNA-seq data is necessary to minimize length bias.

Genome-wide profiling of populus small RNAs

Background

Short RNAs, and in particular microRNAs, are important regulators of gene expression both within defined regulatory pathways and at the epigenetic scale. We investigated the short RNA (sRNA) population (18-24 nt) of the transcriptome of green leaves from the sequenced Populus trichocarpa using a concatenation strategy in combination with 454 sequencing.

Results

The most abundant size class of sRNAs were 24 nt. Long Terminal Repeats were particularly associated with 24 nt sRNAs. Additionally, some repetitive elements were associated with 22 nt sRNAs. We identified an sRNA hot-spot on chromosome 19, overlapping a region containing both the proposed sex-determining locus and a major cluster of NBS-LRR genes. A number of phased siRNA loci were identified, a subset of which are predicted to target PPR and NBS-LRR disease resistance genes, classes of genes that have been significantly expanded in Populus. Additional loci enriched for sRNA production were identified and characterised. We identified 15 novel predicted microRNAs (miRNAs), including miRNA*sequences, and identified a novel locus that may encode a dual miRNA or a miRNA and short interfering RNAs (siRNAs).

Conclusions

The short RNA population of P. trichocarpa is at least as complex as that of Arabidopsis thaliana. We provide a first genome-wide view of short RNA production for P. trichocarpa and identify new, non-conserved miRNAs.

Validation of serum protein profiles by a dual antibody array approach

In recent years, affinity-based technologies have become important tools for serum profiling to uncover protein expression patterns linked to disease state or therapeutic effects. In this study, we describe a path towards the production of an antibody microarray to allow protein profiling of biotinylated human serum samples with reproducible sensitivity in the picomolar range. With the availability of growing numbers of affinity reagents, protein profiles are to be validated in efficient manners and we describe a cross-platform strategy based on data concordance with a suspension bead array to interrogate the identical set of antibodies with the same cohort of serum samples. Comparative analysis enabled to screen for high-performing antibodies, which were displaying consistent results across the two platforms and targeting known serum components. Moreover, data processing methods such as sample referencing and normalization were evaluated for their effects on inter-platform agreement. Our work suggests that mutual validation of protein expression profiles using alternative microarray platforms holds great potential in becoming an important and valuable component in affinity-based high-throughput proteomic screenings as it allows to narrow down the number of discovered targets prior to orthogonal, uniplexed validation approaches.