4.1 Introduction

The importance of the integrity of chromosome structure and its association with human disease is one of the oldest and most studied topics in clinical genetics. As early as 1959, cytogenetic studies in humans linked specific genetic or genomic disorders and intellectual disability syndromes to changes in chromosomal ploidy, translocations, and DNA duplications and deletions (LeJeune et al. 1959; Ford et al. 1959; Jacobs and Strong 1959; Stankiewicz and Lupski 2002; Iafrate et al. 2004). The discovery of copy-number variants (CNVs) by microarray and sequencing technologies expanded the catalogue of genetic variation between individuals to test such associations at higher resolution (Iafrate et al. 2004; Sebat et al. 2004; Hinds et al. 2006; Conrad et al. 2006, 2010; Korbel et al. 2007; Stankiewicz and Lupski 2010; International HapMap 3 Consortium et al. 2010; Carvalho and Lupski 2016).

Over the years, analysis of disease-related structural rearrangements has illuminated genes that are mutated in various human developmental disorders (Zhang et al. 2009; Theisen and Shaffer 2010; Nambiar and Raghavan 2011; Higgins et al. 2008). Such chromosome aberrations can directly disrupt gene sequences, affect gene dosage, generate gene fusions, unmask recessive alleles, reveal imprinted genes, or result in alterations of gene expression through additional mechanisms such as position effects (Zhang et al. 2009). The latter is particularly important for the study of apparently balanced chromosome abnormalities (BCAs), such as translocations and inversions, often found outside of the hypothesized disease-causing genes (reviewed in (Kleinjan and Van Heyningen 2005)).

Position effects were first identified in Drosophila melanogaster, where chromosomal inversions placing white+ near centric heterochromatin caused mosaic red/white eye patterns.(Weiler and Wakimoto 1995) In humans, BCAs can induce position effects through disruption of a gene’s long-range transcriptional control (i.e., enhancer-promoter interactions, insulator influence, etc.), or its placement in regions with different local chromatin environments as observed in the classical Drosophila position effect variegation (reviewed in (Kleinjan and Van Heyningen 2005; Zhang and Wolynes 2015; Spielmann and Mundlos 2016)). Examples of position effect genes include paired box gene 6 (PAX6 [MIM: 607108]), for which downstream chromosome translocations affect its cis-regulatory control and produce aniridia (AN [MIM: 106210]);(Fantes et al. 1995; Kleinjan et al. 2001) twist family bHLH transcription factor 1 (TWIST1 [MIM: 601622]), where downstream translocations and inversions are associated with Saethre-Chotzen syndrome (SCS [MIM: 101400]);(Cai et al. 2003) paired like homeodomain 2 (PITX2 [MIM: 601542]) for which translocations are associated with Axenfeld-Rieger syndrome type 1 (RIEG1 [MIM: 180500]);(Flomen et al. 1998; Trembath et al. 2004) SRY-box 9 (SOX9 [MIM: 608160]), where translocation breakpoints located up to 900 Kilobases (Kb) upstream and 1.3 Megabases (Mb) downstream are associated with campomelic dysplasia (CMPD [MIM: 114290]),(Velagaleti et al. 2005) in addition to several others.(Kleinjan and Van Heyningen 2005; Kleinjan and Heyningen 1998; Lupski and Stankiewicz 2005)

The availability of genome sequencing in the clinical setting has generated a need for rapid prediction and interpretation of structural variants, especially those pertaining to de novo non-coding rearrangements in individual subjects. With the development and subsequent branching of the chromosome conformation capture (3C) technique ((Dekker et al. 2002), reviewed in (Wit and Laat 2012)), regulatory issues such as alteration of long-range transcriptional control and position effects can now be predicted in terms of chromosome organization. The high resolution view of chromosome architecture in diverse human cell lines and tissues(Lieberman-Aiden et al. 2009; Fullwood et al. 2009; Dixon et al. 2012, 2015; Sanyal et al. 2012; Phillips-Cremins et al. 2013; Rao et al. 2014; Mifsud et al. 2015) has allowed molecular assessment of the disruption of regulatory chromatin contacts by pathogenic structural variants and single nucleotide changes; examples include the study of limb malformations,(Lupiáñez et al. 2015) leukemia,(Gröschel et al. 2014) and obesity,(Claussnitzer et al. 2015) among others.(Visser et al. 2012; Roussos et al. 2014; Giorgio et al. 2015; Oldridge et al. 2015; Ibn-Salem et al. 2014; Ordulu et al. 2016) These examples underscore the importance of chromatin interactions in quantitative and temporal control of gene expression, which can greatly enhance our power to predict pathologic consequences.

To test the feasibility of prediction and clinical interpretation of position effects of non-coding chromosome rearrangements, we analyzed 17 subjects from the Developmental Gene Anatomy Project (DGAP)(Higgins et al. 2008; Ligon et al. 2005; Kim and Marcotte 2008; Lu et al. 2007; Redin et al. 2017) with de novo non-coding BCAs classified as variants of uncertain significance (VUS). Using publicly available chromatin contact information, annotated and predicted regulatory elements, and correlation between phenotypes observed in DGAP subjects and those associated with neighboring genes, we reliably predicted candidate genes exhibiting mis-regulated expression in DGAP-derived lymphoblastoid cell lines (LCLs). These results suggest that many VUS are likely to be further interpretable via long-range effects, and warrant their routine assessment and integration in clinical diagnosis.