Q-Copy 2002 v1.0.342 serial key or number

• 2794 Accesses

• 4 Citations

• 8 Altmetric

• Metrics details

Background

Mitochondrial DNA copy number (mtDNA CN) exhibits interindividual and intercellular variation, but few genome-wide association studies (GWAS) of directly assayed mtDNA CN exist.

We undertook a GWAS of qPCR-assayed mtDNA CN in the Avon Longitudinal Study of Parents and Children (ALSPAC) and the UK Blood Service (UKBS) cohort. After validating and harmonising data, 5461 ALSPAC mothers (16–43 years at mtDNA CN assay) and 1338 UKBS females (17–69 years) were included in a meta-analysis. Sensitivity analyses restricted to females with white cell-extracted DNA and adjusted for estimated or assayed cell proportions. Associations were also explored in ALSPAC children and UKBS males.

Results

A neutrophil-associated locus approached genome-wide significance (rs709591 [MED24], β (change in SD units of mtDNA CN per allele) [SE] − 0.084 [0.016], p = 1.54e−07) in the main meta-analysis of adult females. This association was concordant in magnitude and direction in UKBS males and ALSPAC neonates. SNPs in and around ABHD8 were associated with mtDNA CN in ALSPAC neonates (rs10424198, β [SE] 0.262 [0.034], p = 1.40e−14), but not other study groups. In a meta-analysis of unrelated individuals (N = 11,253), we replicated a published association in TFAM (β [SE] 0.046 [0.017], p = 0.006), with an effect size much smaller than that observed in the replication analysis of a previous in silico GWAS.

Conclusions

In a hypothesis-generating GWAS, we confirm an association between TFAM and mtDNA CN and present putative loci requiring replication in much larger samples. We discuss the limitations of our work, in terms of measurement error and cellular heterogeneity, and highlight the need for larger studies to better understand nuclear genomic control of mtDNA copy number.

Mitochondria are the cellular organelles responsible for producing adenosine triphosphate (ATP), a ubiquitous substrate required for metabolism. ATP is the final product of the series of redox reactions that are facilitated by the complexes of the respiratory chain (RC), located on the cristae, the folded inner membrane of mitochondria.

Mitochondria possess their own genome (mtDNA), an extra-nuclear, double-stranded, circular DNA molecule of ~ 16.6 kb that is inherited maternally. Thirteen subunits contributing to complexes of the RC are encoded by mtDNA, and the entire mitochondrial genome is present at variable copy number in the cell. The relative copy number of mtDNA (mtDNA CN) may reflect differing energy requirements between cells: those from active tissues (e.g. liver, muscle, neuron) are observed to have higher mtDNA CNs compared to endothelial cells, which are comparatively quiescent [1,2,3].

Several nuclear genes are known to influence the regulation of mtDNA CN, and these are reviewed in detail elsewhere [1, 4,5,6]. These include POLG [4,5,6,7,8,9,10,11,12] and POLG2 [4, 5, 12], the catalytic and accessory subunits of DNA polymerase-gamma, the principal enzyme implicated in mtDNA replication. Other regulators include TFAM (mitochondrial transcription factor A) [4, 13,14,15,16,17], which initiates mtDNA replication, along with other factors TFB1M and TFB2M [4, 8, 18]. Regulators of these transcription factors include PGC-1α (peroxisome proliferators-activated receptor gamma coactivator 1 alpha) [4, 5, 8] and two nuclear respiratory factors (NRF-1, NRF-2) [4, 5, 8]. Moreover, maintenance of replication requires an adequate mitochondrial nucleotide supply [19]: nucleotides may be imported from the cytosol or salvaged by specific mitochondrial enzymes. Defective phosphorylation of deoxyribonucleosides by kinases encoded by DGUOK (deoxyguanosine kinase) and TK2 (thymidine kinase) leads to dysfunctional mitochondrial dNTP synthesis and key regulators of dNTP synthesis in the cytosol include the helicase C10orf2 (alias TWINK) [4,5,6,7,8,9,10, 12], along with thymidine phosphorylase (TYMP) [4, 5, 9] and the target of the p53-transcription factor, p53R2 (encoded by RRM2B) [4, 6, 7, 9, 12]. The role of succinyl CoA synthase deficiency as a cause of mtDNA depletion is less well understood, but mutations in the alpha and β subunits of succinyl CoA synthase genes (SUCLA2, SUCGL1) [6, 7, 9, 10] may be associated with mitochondrial nucleotide depletion [6].

To our knowledge, few genome-wide scans of mtDNA CN have been published, and those that exist are of relatively small sample size [16, 20, 21] or use in silico proxies for mtDNA CN without actual biological measurements [17]. We had access to directly assayed mtDNA CN in a diverse set of study groups and so performed hypothesis-generating genome-wide association studies (GWAS) in ~ 14,000 individual participants from the Avon Longitudinal Study of Parents and Children (ALSPAC) and the UK Blood Service (UKBS) cohort. For our main analyses, the two most comparable study groups of adult females were combined in a joint analysis (N = 6799, approximately 10 times larger than previous GWAS of directly assayed mtDNA CN) [20, 21], with results from the other groups presented as opportunistic, secondary analyses. It is known that cellular heterogeneity contributes to mtDNA CN: granulocytes have relatively few mitochondria, whereas lymphocytes are rich in mitochondria, and therefore in mtDNA [22]. Since we also had access to data on white cell proportions, estimated from methylation data in ALSPAC, and assayed directly in UKBS, we performed sensitivity analyses that considered DNA source (whole blood/white cells), and controlled for white cell proportions. Finally, we extracted two SNPs that were robustly related to mtDNA CN in a recent GWAS of mtDNA CN measured in silico [17], and compared our results to those published associations.

Cohort details

ALSPAC is a prospective cohort of mothers and their children. Between 1991 and 1992, 14,541 women living in the former county of Avon, UK, were recruited during pregnancy, of whom 13,761 were enrolled into the study (women were aged between 16 and 43 years at recruitment when samples for mtDNA CN analyses were obtained). Further details are available in the cohort profile papers [23, 24], and the study website contains details of all data that are available through a fully searchable data dictionary: http://www.bristol.ac.uk/alspac/researchers/our-data/. Ethical approval for the study was obtained from the ALSPAC Ethics and Law Committee and the Local Research Ethics Committees.

The UK Blood Service control group is part of the Wellcome Trust Case Control Consortium 2. The UK National Blood Service (UKBS) consists of 3091 unrelated, healthy individuals (aged 17–69 years when samples for mtDNA CN assay were obtained), recruited between September 2005 and February 2006. Informed consent was obtained from all participants in accordance with protocols approved by the Peterborough & Fenland Local Research Ethics Committee in September 2005.

DNA samples

Blood samples used for mtDNA CN assay were collected from ALSPAC mothers during routine antenatal care. Children included in this study had DNA sampled at either birth (from cord blood, these individuals are hereafter referred to as ALSPAC ‘neonates’) or at a follow-up research clinic assessment at mean age 7 (range 6–9 years, hereafter ALSPAC ‘6–9-year-olds’). Antenatal DNA from mothers was extracted using a phenol-chloroform method [25]. DNA from ALSPAC children was extracted using a phenol-chloroform (ALSPAC neonates) or salting-out method (ALSPAC 6–9-year-olds) [25]. DNA sources used for the mtDNA CN assay varied by age group: DNA from whole blood was used for 6–9-year-olds, ALSPAC mothers’ DNA was extracted from whole blood or white cells, and ALSPAC neonates had DNA extracted from white cells, as described previously [25].

UKBS blood samples were separated by density centrifugation, and white blood cells were retained to perform DNA extractions, as previously described, using a guanidine-chloroform-based method [26, 27]. Thus, in UKBS, the DNA source was white blood cells for all participants. Blood composition information for UKBS samples was provided by Willem Ouwehand at the University of Cambridge as part of an on-going collaboration with Patrick Chinnery. These details are also summarised in Table 1.

Genotype data

ALSPAC

ALSPAC mothers were genotyped on the Illumina Human660W-Quad array (Illumina, San Diego, CA, USA) at the Centre Nationale du Génotypage (CNG). ALSPAC children were genotyped with the Illumina HumanHap550-Quad array, by the Wellcome Trust Sanger Institute, Cambridge, UK, and the Laboratory Corporation of America, Burlington, NC, USA, using support from 23andMe. Genotypes were called using Illumina GenomeStudio®. Quality control (QC) was performed using PLINK v1.07 [28], phasing using ShapeIT (v2.r644) [29], and imputation was to the Haplotype Reference Consortium (v1.0), performed using IMPUTE (v3) (http://mathgen.stats.ox.ac.uk/impute/impute.html). The genome build used was GRCh37. Further details of genotype QC are given in Additional file 1.

GWAS were run separately in 5461 ALSPAC mothers, 3647 6–9-year-olds, and 2102 neonates (see Additional file 1 for details of selection into the study). Relatedness within each group of participants (mothers, neonates, and 6–9-year-olds) was assessed by identical-by-descent (IBD) proportions from a genetic relatedness matrix, calculated using the GCTA standard algorithm [30], based on 1.1 million HapMap3 best-guess tag SNPs, present at a combined allele frequency of > 0.01 and imputation quality > 0.8 in 17,842 individuals. Within each group (mothers, 6–9-year-olds, and neonates), participants were unrelated (IBD > 0.125; i.e. first-cousin level). A subset of children was related to the 5461 ALSPAC mothers: there were 1611 mother/6–9-year-old pairs related at IBD > 0.125 (1570 pairs IBD > 0.45) and 869 mother-neonate pairs related at IBD > 0.125 (839 IBD > 0.45). For some sensitivity analyses, a GWAS of a subset of 2833 mothers, who are unrelated to any 6–9-year-olds or neonates at IBD > 0.125, is used. SNPs were filtered by MAF < 0.01 and imputation score < 0.8 in all study groups [31], leaving 7,360,988; 7,410,776; and 7,361,275 SNPs in ALSPAC mothers, 6–9-year-olds, and neonates, respectively.

UKBS

Chromosome 17q gains are almost invariably present in high-risk neuroblastoma cases. Here, we perform an integrative epigenomics search for dosage-sensitive transcription factors on 17q marked by H3K27ac defined super-enhancers and identify TBX2 as top candidate gene. We show that TBX2 is a constituent of the recently established core regulatory circuitry in neuroblastoma with features of a cell identity transcription factor, driving proliferation through activation of p21-DREAM repressed FOXM1 target genes. Combined MYCN/TBX2 knockdown enforces cell growth arrest suggesting that TBX2 enhances MYCN sustained activation of FOXM1 targets. Targeting transcriptional addiction by combined CDK7 and BET bromodomain inhibition shows synergistic effects on cell viability with strong repressive effects on CRC gene expression and p53 pathway response as well as several genes implicated in transcriptional regulation. In conclusion, we provide insight into the role of the TBX2 CRC gene in transcriptional dependency of neuroblastoma cells warranting clinical trials using BET and CDK7 inhibitors.

Neuroblastoma (NB) is a cancer of the developing sympatho-adrenergic nervous system and is the most common malignancy diagnosed in children during their first years of life¹. Sequencing revealed a relatively silent mutational landscape with only ALK activating mutations being identified in up to 10% of primary cases as well as de novo secondary or emerging subclonal ALK mutations in relapsed cases^2,3. Further, in relapsed cases additional RAS-MAPK pathway driving mutations are enriched^4,5. In contrast to mutations, DNA copy number alterations are remarkably recurrent in NB, including focal amplification of the MYCN oncogene in approximately half of the high-stage patients⁶ and large 17q segmental gains occurring in the majority of both MYCN amplified and non-amplified high stage tumors^7,8,9. The finding of recurrent gains of the syntenic human 17q region in MYCN driven NB mouse tumors further supports the putative functional importance of this genomic aberration in NB¹⁰. Investigating dosage-sensitive genes affected by recurrent copy number alterations can offer new insights into tumor biology as was illustrated in ependymoma where multiple dosage-affected genes, located within large chromosomal regions of recurrent gains and losses, were shown to act as oncogenes or tumor suppressors through installing a so-called cellular state driven through one or more altered cellular functions¹¹.

Given the recently proposed role of a core regulatory circuitry (CRC)¹² consisting of several super-enhancer (SE) marked¹³ transcription factor constituents in NB^14,15,16, we decided to search for dosage-sensitive SE marked transcription factors encoding genes residing on chromosome 17q. The ‘T-box 2 transcription factor’ (TBX2), hitherto not reported to be implicated in NB, was prioritized as transcription factor with top-ranked SE score in NB cell lines and with expression levels highly correlated with survival outcome in NB tumors. TBX2 is a member of the T-box family of transcription factors with an important role during embryogenesis and morphogenesis^17,18 and is overexpressed in several cancer entities including melanoma, breast, and pancreatic cancer^19,20,21. The oncogenic effect of TBX2 overexpression has been attributed to its role in proliferation as well as inducing epithelial-to-mesenchymal transition (EMT) and senescence bypass²². Based on integrated analysis of TBX2 occupancy as determined by ChIP-sequencing and transcriptome analysis upon knockdown (KD), we propose TBX2 as a novel bona fide constituent of the recently reported CRC in NB^14,15,16.

To investigate the role of TBX2 in this CRC, functional analyses were performed showing the implication of TBX2 in cell cycle, proliferation, and downstream E2F-FOXM1 signaling. Finally, we demonstrate that combined pharmacological targeting of transcriptional addiction using a BET and CDK7 inhibitor, yields synergistic effects on TBX2 downregulation leading to massive apoptosis.

TBX2 is a super-enhancer marked transcription factor on 17q

CRCs consisting of SE marked master transcription factors were recently shown to be dysregulated in NB through MYCN-dependent transcriptional amplification^14,16 causing transcriptional addiction²³. Given the highly recurrent chromosome 17q gain in high-risk human NBs and MYCN-driven mouse NBs, we hypothesized that one or more dosage-sensitive CRC transcription factors map to 17q thus rendering a selective advantage to tumors cells exhibiting 17q gain. To identify such transcription factors, we determined SE scores using the LILY algorithm¹⁵ based on the intensity of H3K27ac marks in 26 NB cell lines with 17q gain, two non-malignant neural crest cell lines and the breast cancer cell line MCF-7 as non-embryonal control (gene prioritization strategy is depicted in Fig. 1a, b and Supplementary Fig. 1a, b). We identified a total of 176 SE clusters on 17q of which six were present in at least 20 NB cell lines (Supplementary Fig. 1c). These six SE clusters are located in the vicinity of 86 candidate genes of interest, including 11 transcription factors²⁴, of which 5 are actively transcribed in NB cells, i.e. TBX2, RARA, SP2, NFE2L1, and VEZF1.

Next, we assessed the expression levels of these transcription factors in relation to patient survival in two independent NB tumor cohorts (GSE85047 n = 276, GSE62564 n = 498) and observed the strongest association with overall and progression-free survival for TBX2²⁵ (Fig. 1c and Supplementary Fig. 1d, Kaplan–Meier analysis). Moreover, TBX2 is marked by a SE in all investigated NB cell lines, but not in the human neural crest line (hNCC) and the MCF-7 breast cancer cell line (Fig. 1b and Supplementary Fig. 1b).

Of further interest, the highest expression levels for TBX2 were observed in NB cell lines and primary tumors compared to other tumor entities, based on the online pan-cancer analysis in the CCLE database (cancer cell lines) and R2 platform (primary tumors and normal tissues) (Fig. 1d). TBX2 expression levels were also high in normal embryonic tissues in keeping with the established role of TBX2 in early development. Taken together, our data suggest a possible important role for TBX2 as hitherto unrecognized transcriptional regulator in NB tumor development.

4C-seq defines TBX2 promotor—super-enhancer interactions

To provide further evidence for a functional role of the assigned SE for TBX2 gene regulation, three different viewpoints residing in the SE region (20 kb and 260 kb up TSS) or the promoter site of TBX2 (4.5 kb up TSS) were selected for 4C-sequencing in NB cell lines SK-N-AS and CLB-GA (Fig. 2a and Supplementary Fig. 2a). Reciprocal interactions were observed between the two viewpoints in the SE region and the promoter region, as well as interaction with a region more upstream of TBX2 (400 kb up TSS). These results are in line with the proposed TBX2 regulation according to the associated SE region as determined by H3K27ac mapping.

Of further interest, TBX2 maps to the border of a topologically associated domain (TAD)²⁶ (Supplementary Fig. 2b) and has been associated with a bi-directionally transcribed topological anchor point (tap)RNA TBX2-AS1²⁷. These positional conserved tapRNAs are located at chromatin loop anchor points and borders of TADs and show strong coordinated expression with their associated nearby protein-coding gene²⁸. Indeed, we found a strong correlation between the expression levels of the tapRNA TBX2-AS1 and TBX2 in a large cohort of NB tumors (n = 79, Supplementary Fig. 2c). In addition to the potential regulatory connection of TBX2-AS1 and TBX2, we also found strong correlation with expression levels of the PPM1D gene which maps within a 1.5 Mb distance from the TBX2 locus (Supplementary Fig. 2d). In summary, the above findings support a physical interaction between the TBX2 locus and its nearby SE and suggest that the proposed chromatin looping drives TBX2 expression.

TBX2 is a copy-number affected dosage-sensitive gene

Next, we investigated in more detail the genomic aberrations that account for the high TBX2 expression in NB (Fig. 1d). We first analyzed the CCLE database and found that NB was the tumor entity exhibiting the most frequent gains for the TBX2 locus, the highest expression and lowest methylation levels (Figs. 1d and 2b, c). Next, we assessed the effect of DNA copy number alterations on TBX2 expression levels using an ANOVA analysis in the NRC tumor dataset (n = 218, GSE85047). Only in high-stage disease (stage 3 and 4), we observed significantly increased TBX2 expression levels due to increased TBX2 copy number (p = 6.004e−5, log2 ratio > 0.3) (Fig. 2d). We specifically looked for rare TBX2 encompassing amplicons in a series of 556 high-risk NB cases²⁹ and detected a single MYCN-amplified case with an additional 1.076 Mb focal 17q23.2 amplification (Fig. 2e and Supplementary Fig. 2e) encompassing only six protein-coding genes including the transcription factors TBX2 and TBX4. Of further note, a previously reported focal high level 1.8 Mb gain of a chromosome 17q23 segment in the NB cell line MP-N-TS also encompasses the TBX2 locus^30,31. Taken together, our data indicate that TBX2 is a dosage-sensitive transcription factor affected by the common segmental 17q gains and rare amplification events in NB.

TBX2 is a core regulatory circuitry constituent in NB

To gain further insight into the TBX2-controlled regulatory network, we assessed TBX2 DNA occupancy by ChIP-sequencing and ATAC-sequencing in the NB cell line IMR-32. A total of 557 significant (adj.P.val < 0.05) TBX2 binding sites were identified and motif analysis confirmed enrichment for a TBX motif (AGGTGTGA, p = 1e−41), supporting the validity of our ChIP-seq data (Supplementary Data 1). In total, 81, 28, and 94% of TBX2 binding sites in IMR-32 respectively overlap H3K27ac, H3K4me3, and ATAC-sequencing peaks (Fisher test p < 2.2e−16, Fig. 3a, b), which confirms the binding of TBX2 to active promotor and enhancer regions. Moreover, respectively 41 and 30% of the TBX2 ChIP-seq peaks are found intergenic or are annotated to lncRNAs (Supplementary Fig. 3a, b), and 19% overlap with the SEs annotated in the cell line IMR-32 (Fisher test p < 2.2e−16, Supplementary Fig. 3c).

The recent reports on distinct CRCs in NB^14,15 prompted us to investigate the possible involvement of TBX2. In line with the recent finding of invasion of MYCN into non-canonical E-boxes at enhancers, motif analysis of TBX2-bound regions showed that the non-canonical MYC(N) E-box motif CANNTG was found to be highly enriched (Binomial test p = 1e−63) as well as motifs for GATA(1/2/3/4), PHOX2(A/B), HAND(1/2) and neuronal lineage-specific marker genes such as ASCL1, ISL1 and MEIS(1/2)^32,33 (Supplementary Data 1). We integrated the ChIP-seq tracks for TBX2 with those reported for GATA3^15,34, HAND2, PHOX2B¹⁵, and MYCN (this study) in NB cell lines and observed overlap of TBX2 peak summits with binding sites of these CRC transcription factors (Fig. 3b), thus supporting the notion that TBX2 is indeed actively taking part in this CRC. Overlap of PHOX2B, HAND2, and GATA3 binding with the TBX2 peaks was predominantly observed in enhancer regions (Fig. 3b and Supplementary Fig. 3d). The integration of TBX2 into this CRC is further confirmed by the observation of auto-regulation by binding of the TBX2 transcription factor to its own SE constituent and binding of at least three CRC members including GATA3, HAND2 and PHOX2B within this SE constituent (Fig. 3c). In addition, TBX2 is binding the SE constituents of the other CRC members PHOX2B, GATA3, and HAND2, amongst others, as shown in Supplementary Fig. 3e. Finally, TBX2 expression is positively correlated with GATA3, HAND2, and PHOX2B expression levels as well as with those of other potential CRC genes important in development in a NB tumor cohort (n = 283, Supplementary Fig. 3f). Taken together, our data suggest that TBX2 is part of the recently described CRC together with HAND2, GATA3, and PHOX2B.

TBX2 controls E2F-FOXM1 driven cell cycle and proliferation

To unravel the role of TBX2 within the CRC in NB cells, we performed TBX2 KD with two shRNAs and a non-targeting control and subsequent gene expression profiling in the NB cell line IMR-5/75 (Supplementary Fig. 4a). A total of 1055 and 1326 genes were differentially down and upregulated, respectively (adj.p.val < 0.05, Supplementary Data 2), including the upregulated gene CDKN1A, which is a known target gene repressed by TBX2³⁵. Gene set enrichment analysis (GSEA) on the downregulated genes upon TBX2 KD in IMR-5/75 cells revealed enrichment (FDR < 0.01) for the hallmark and gene ontology gene sets involved in cell cycle including G2/M checkpoint, E2F, MYC(N) targets, mitosis, and DNA replication (Fig. 4a, Supplementary Data 3) and enrichment was shown for TP53 pathway among the upregulated genes. Using iRegulon, designed to detect transcription factors, targets and motifs/tracks from a set of genes