Juxtaposition of heterochromatic and euchromatic regions by chromosomal translocation mediates a heterochromatic long-range position effect associated with a severe neurological phenotype

Background The term "position effect" is used when the expression of a gene is deleteriously affected by an alteration in its chromosomal environment even though the integrity of the protein coding sequences is maintained. We describe a patient affected by epilepsy and severe neurodevelopment delay carrying a balanced translocation t(15;16)(p11.2;q12.1)dn that we assume caused a position effect as a result of the accidental juxtaposition of heterochromatin in the euchromatic region. Results FISH mapped the translocation breakpoints (bkps) to 15p11.2 within satellite III and the 16q12.1 euchromatic band within the ITFG1 gene. The expression of the genes located on both sides of the translocation were tested by means of real-time PCR and three, all located on der(16), were found to be variously perturbed: the euchromatic gene NETO2/BTCL2 was silenced, whereas VPS35 and SHCBP1, located within the major heterochromatic block of chromosome 16q11.2, were over-expressed. Pyrosequencing and chromatin immunoprecipitation of NETO2/BTCL2 and VPS35 confirmed the expression findings. Interphase FISH analysis showed that der(16) localised to regions occupied by the beta satellite heterochromatic blocks more frequently than der(15). Conclusions To the best of our knowledge, this is the first report of a heterochromatic position effect in humans caused by the juxtaposition of euchromatin/heterochromatin as a result of chromosomal rearrangement. The overall results are fully in keeping with the observations in Drosophila and suggest the occurrence of a human heterochromatin position effect associated with the nuclear repositioning of the der(16) and its causative role in the patient's syndromic phenotype.


Background
Patients with syndromic clinical phenotypes include an interesting subset of carriers of de novo balanced chromosomal rearrangements with no apparent loss or gain of genetic material. These abnormalities can be explained by the loss of function of a dose-sensitive gene disrupted by one rearrangement breakpoint [1][2][3]. The breakpoints of the balanced chromosomal rearrangements associated with a clinical phenotype are often molecularly mapped in an attempt to identify the disease-causing genes affected by the abnormality. Although the rearrangements may lead to the direct disruption of one or two genes, this is not always the case. It has also been shown that breakpoints occur outside the genes themselves and affect their regulation by causing a change in their position within the genome, a phenomenon known as the "position effect" (PE). Investigations of the PE have required intensive experimental effort because of the unique features of each case and the often scarcely defined length of the involved genomic region.
A number of potential mechanisms can be suggested to explain the position effects of chromosomal rearrangements in humans for a review see ref. [4]: 1) the separation of a gene from its enhancer, promoter or locus control region; 2) juxtaposition to an enhancer element from another gene; 3) the removal of the longrange insulator or boundary elements; 4) competition with another enhancer; 5) alterations in local chromatin structure; 6) alterations in nuclear organisation [5,6]; and 7) position effect variegation (PEV), when the genes are moved into close proximity to constitutive heterochromatin and their activity become unstable and leads to variegated patches of gene expression.
PEV was first observed in Drosophila and was so named because of the variegated pattern of expression of euchromatic genes transposed to locations near to pericentric heterochromatin as a result of natural or induced genetic rearrangements [7][8][9][10]. The chromosomal position effect can spread over distances of 1 Mb or more, and generally reflects a gradient of gene inactivation that is inversely correlated with distance [10], although it can also be affected by the local context of a gene [11].
Some genes are specifically adapted to be expressed exclusively in a heterochromatic context (heterochromatic genes) [9]. They are also influenced by PEV, but their behaviour is opposite to that of genes located in euchromatic chromosomal regions (euchromatic genes). The regulatory implication of genes residing in heterochromatic regions was first evidenced in a pioneering study that linked PEV to euchromatic/heterochromatic rearrangements: the variegation of the light gene, known as Dmel\lt, encoding a protein containing a zinc ion binding domain and residing at the heterochromatin side of the rearrangement breakpoint, was enhanced by an increased dose of Y chromosome constitutive heterochromatin, whereas the variegation of genes located on the euchromatic side of the breakpoint was suppressed [12]. Subsequently, it was found that a number of other heterochromatic genes show similar heterochromatic dependence [13][14][15], some of which are unique in terms of function and protein coding [16].
Sequencing projects have led to the discovery of hundreds of heterochromatic genes in Drosophila, plants and mammals, but no variegation effects have yet been reported in humans, although they have been observed in transgenic mice bearing incomplete functional gene domains that became inserted into heterochromatic regions [17].
Given the variability of chromosomal rearrangements, rare patients carrying unique chromosomal abnormalities often show single or multiple clinical signs that cannot be assigned to any recognisable syndrome. We here describe the molecular cytogenetic, genetic and epigenetic characterisation of a "private" balanced chromosomal translocation t(15;16)(p11.2;q12.1)dn carried by a patient affected by epilepsy and severe neurodevelopment delay. Fluorescence in situ hybridisation (FISH) breakpoint mapping and analyses of the quantitative expression and promoter epigenetic signatures of the genes at or near the breakpoints showed that a simple chromosomal rearrangement may involve multiple genes as a result of a long-range heterochromatic position effect.

Clinical presentation
The proband is the first and only daughter of healthy and unrelated parents, and was born in the 39 th g.w. after an uncomplicated pregnancy with a birthweight of 3480 g (50 th centile), a length of 50 cm (50 th centile), and a head circumference of 33.5 cm (10 th centile). At birth, she was affected by neonatal respiratory distress and hypotonia.
During her first day of life, she experienced brief partial seizures with secondary generalisation that were treated with phenobarbital (PB) but recurred several times a week until she was two months old, when they started appearing as clusters of infantile axial spasms. The spasms disappeared after treatment with valproic acid (VPA), and the baby started to smile and gained control of her head. At the age of six months, the spasms returned and became daily until she was 30 months old, despite the administration of adrenocorticotropin hormone (ACTH) and gamma-vinyl GABA (GVG). Interictal electroencephalography (EEG) revealed the presence of frequent paroxysms involving the temporo-parietal regions of both hemispheres, with left predominance in the absence of a hypsarrhythmic pattern.
Subsequently, the spasms became progressively less frequent with VPA treatment, occurring only a few times a month during sleep or upon awakening. They finally disappeared when the patient was five years old, and interictal EEG revealed normal findings during wakefulness and rare focal spikes on the parietal regions only during sleep.
From the age of four years, the patient's clinical picture progressively improved. Now aged 12 years, she can sit alone, has hypotonic-dystonic cerebral palsy, is averbal, and suffers from very severe cognitive impairment, but her relationships are relatively good.
Cerebral MRI at the age of three years revealed a left temporo-parietal focal abnormality suggesting a prenatal insult with micropolygyria.
A physical examination at the age of 10 years showed a length of 120 cm (< 3 rd centile), a weight of 20 kg (< 3 rd centile), and a head circumference of 49 cm (< 3 rd centile), together with brachyplagiocephaly, wide palpebral fissures with mild lateral ectropion and long eyelashes, thin fingers, and severe thoracic scoliosis.
The patient was originally referred to our laboratory because of suspected Angelman syndrome, which was ruled out as SNRPN methylation analysis showed a biparental pattern and no UBE3A gene mutation was found. Rett syndrome was ruled out by means of mutation screening of the MeCp2 and CDKL5 genes.
FISH characterisation narrowed the 15p breakpoint region to within 15p11.2 satellite III (D15Z1) on the basis of the fact that probe D15Z1 gave a signal on both derivative chromosomes 15 and 16 ( Figures 1A and 1D), whereas RPC1-98 C19 and RPC1-21I10 (containing beta satellite sequences) only gave a signal on der(16) (Figures 1A and 1D), as did the rDNA probe RPC5-1174A5 ( Figures 1B and 1D). As expected, the α-satellite probes D15Z3 and D15Z4 gave a signal on the centromere of der (15) (Figures 1C and 1D).
The 16q translocation bkp was mapped to 16q12.1 using a BAC contig and was localised within the ITFG1 gene, which is disrupted: the contiguous BACs showed different hybridisation patterns, with CTD-3056I17 only giving a signal on der(16) ( Figure 1E), CTD-2027D24 giving a signal on both derivative chromosomes ( Figure  1F), and RP11-474B12 giving a signal only on der(15) (data not shown). In order to refine the localisation of the 16q breakpoint, a 7 kb long-range PCR genomic fragment, covering the distal part of IVS12, was hybridised with patient metaphases. This probe only gave a signal on der(16), the intensity of which was comparable with that of the normal chromosome 16, thus allowing the bkp region to be approximately narrowed to a 12 kb interval (chr16:47268711-47280772, hg19) ( Figure 1G and 1H). All of the probes and the FISH results are shown in Additional file 1: Table S1.
In order to investigate whether the translocation involving the 15p heterochromatic blocks favoured a relocalisation of the 16q euchromatic bkp regions in a heterochromatic environment, we performed tripled i-FISH  experiments using patient lymphoblast cells derived from three different cultures and hybridised with probes D15Z1, GABRB3 and RPC1-98 C19. For each experiment, at least 100 nuclei were analysed by evaluating the co-localisation of the 16q bkps with regions recognised by the probe RPC1-98 C19, which contains betasatellite DNA.
We observed der(16) co-localisation with heterochromatic regions more frequently than der(15) co-localisation in all of the experiments ( Figure 2).

Array CGH analysis
The analysis excluded the gain or loss of any genomic region encompassing the der(15)/der(16) breakpoints and flanking regions, including the loci monitored for gene expression. It also revealed a loss on chromosome 9 and a gain on the X chromosome, neither of which were listed in the reference Database of Genomic Variants http://projects.tcag.ca/variation, a catalogue of structural variations identified in healthy subjects. In detail, high-resolution a-CGH analysis identified a paternal deletion of at least 28 kb at 9p24.1 (chr9:6,215,819-6,244,723, hg19) and a maternal duplication of at least 500 kb at Xp22.2 (chrX:11,686,237-12,187,337, hg19) (Additional file 2: Figure S1). Two genes, MSL3 and RFMPD4/PDZD10 map to Xp22.2, the latter covering a duplication bkp and hence being hypothetically interrupted. The 9p deletion disrupts the IL33 gene.

XCI pattern
In order to verify whether the maternally derived CNV localised on chromosome X could be involved in the patient's phenotype, X chromosome inactivation assay was performed analysing AR and DXS6673E loci. While the AR locus was uninformative, the DXS6673e locus displayed a random X inactivation in the mother (59: 41) whereas the patient showed a preferential X inactivation of the maternal allele (82:18; Additional file 3: Figure  S2). These results indicate no role of this CNV in the patient phenotype.

Gene expression
Real-time PCR using primers positioned on the opposite sides of the bkp within ITFG1 IVS12 asses that the amount of the 5' portion of the transcript was comparable with that of the control and half that of its 3' portion, thus confirming the FISH results ( Figure 3).
Given the complex syndromic phenotype of the patient, we tested whether the breakage and juxtaposition of 15p constitutive heterochromatin to the 16q12.1 euchromatic region disturbed the expression of the The large and small arrows respectively indicate der(16) and der(15). Chromosome 15 is not shown but it occupies the region in which all three hybridization signals are located, with the green signal larger than those identified on the two derivatives. At the bottom left a drawing showing the expected hybridization pattern allowing to recognize the three different chromosomes (der (16), der (15) and chromosome 15) on nucleus. The picture above shows only the expected pattern of hybridization on the three chromosomes in the drawing below is also shown acrocentric regions recognized by the probe containing beta-satellite DNA. At the bottom right graphs A, B, and C show the rate of co-localisation of der(16) (dark grey) and der(15) (light grey) to the regions targeted by the beta satellite. The results average of three experiments is shown on the right.  intact genes flanking the 16q bkp. We first analysed four genes whose expression profiles and functions suggested a possible causative role: VPS35 and NETO2/BTCL2, which respectively map to 16q11.2 heterochromatic and 16q12.1 euchromatic bands, and are localised 0.5 Mb and 62 kb from the bkp on der (16), and SIAH1 and CBLN1, which both map to 16q12.1 and are localised 1.1 Mb and 2 Mb from the bkp on der(15). Real-time PCR revealed halved expression of NETO2/BTCL2 and the over-expression of VPS35, both located on der (16), whereas the expression of SIAH1 was comparable with that of the control (Figure 3). The expression of the brain-specific CBLN1 gene was not quantifiable because it was low in all of lymphoblast cell lines.
In order to establish whether other more distant genes were transcriptionally perturbed, we also analysed the expression of the der (16)

Epigenetic characterisation of the promoter regions
It is well known that H3mK9, H3mK27 and CpG site methylation occurs in silenced gene promoters, whereas the acetylation of histones H3 and H4 and H3 lysine 4 methylation correlate with transcriptionally active genes. In order to assess the epigenetic status of the intact genes flanking both sides of the 16q breakpoint, lymphoblastoid cell lines from the patient and controls were investigated for their DNA methylation status and chromatin modifications in the regulatory regions of the genes whose expression was disturbed or slightly perturbed. In particular, we tested the promoter regions of the NETO2/BTCL2, SIAH1, CYLD and RBL2 genes.
The promoter methylation analysis was made using a pyrosequencing assay that quantitatively measures the methylation of all CpG sites. In the normal samples, NETO2/BTCL2, SIAH1, CYLD and RBL2 were demethylated at all of the analysed CpG sites; in the patient, and in line with the gene expression results, we found a NETO2 promoter methylation pattern indicating the epigenetic silencing of one allele (mean 50% methylation) (Figure 4), whereas all of the other gene promoters located on der(15) (SIAH1, CYLD and RBL2) showed a methylation pattern comparable with that of the normal samples ( Figure 4).
We then tested seven different modifications on the histone tails: H3, H4 acetylation, and H3 di-and trimethylation on lysine 4 (associated with an active chromatin state), and H3 di-and tri-methylation on lysine 9 and H3 di-methylation on lysine 27, which are specific for transcriptionally silenced chromatin. The ChIP assays showed H3 K9 di-and tri-methylation and H3 K27 di-methylation in the NETO2/BTCL2 promoter (Additional file 4: Table S2), whereas VPS35 and SIAH1 showed the same qualitative histone code as the control. As VPS35 was up-regulated, we used Real-time PCR on chromatin immunoprecipitates to assess the quantitative changes in the epigenetic modifications related to the trascriptionally competent chromatin, and found increased H4 acetylation and H3 tri-methylation on lysine 4 ( Figure 5).

Discussion
We believe that this may be the first description of a human disease involving a heterochromatic PE, by means of which intact euchromatic and heterochromatic genes are respectively silenced or enhanced as a result of euchromatin/heterochromatin juxtaposition to constitutive heterochromatin. Although our molecular analysis could not assay variegation in the classical sense of PEV, at least three genes showed perturbed expression and behaviour equivalent to that of Drosophila genes in the same chromosomal context [8]. As in Drosophila, we found the over-expression of heterochromatic genes (VPS35 and SHCBP1) located at 16q11.2, and the under-expression of the euchromatic gene (NETO2/ BTCL2) located at 16q12.1. In keeping with the expression results and the findings observed in Drosophila [18], the silenced NETO2/BTCL2 "euchromatic" gene on der(16) showed promoter DNA methylation and H3 K9 di-and tri-methylation/H3 K27 di-methylation histone modifications, whereas the over-expressed VPS35 "heterochromatic" gene showed an increase in H4 acetylation and H3 tri-methylation on lysine 4 in comparison with the control. These findings confirm that these chromatin modifications act as universal markers of active/inactive genes in both the heterochromatic and euchromatic bands [18]. We did not observe any epigenetic silencing of the assayed der(15) genes SIAH1, CYLD and RBL2. The expression profiles and epigenetic findings consistently revealed the presence of an alteration only on one side of the translocation breakpoint, on the der(16) chromosome. Interestingly but reasonably expected, we also observed that der (16), owning chromosome 15 satellites, is more frequently associated with acrocentric chromosomes beta satellite heterochromatic blocks than der (15). In line with the results of recent studies leading  to the conclusion that the reorganisation of chromosomal territories may be an important step in the process of gene silencing [6], our overall findings suggest that the different behaviour of the two derivative chromosomes may be caused by the different nuclear re-localisation of der (16).
Treatment with VPA, acting also as HDAC inhibitor, could modify the epigenetic profile of the patient [19,20], nevertheless the epigenetic signature observed in the derivative chromosomes is coherent with a PE mechanism and not altogether justifiable by the drug treatment.
In terms of genotype-phenotype correlations, our assumption is that most of the observed clinical signs of our patient can be attributed to the de novo balanced translocation. The clinical phenotype of the patient can be interpreted as an effect of the disruption of the ITFG1 gene combined with the heterochromatic PE effect (assuming that the genes in question are dose sensitive).
The 16q translocation bkp disrupts the ITFG1 gene (an inhibitory regulator of PP2A that plays a role in modulating kinase levels in the ATM/ATR signalling pathway), leading to a truncated transcript that is not degraded by nonsense-mediated RNA decay [21]. Moreover, some of the perturbed genes can be plausibly considered causative candidates on the basis of their expression profiles and functions. These include NETO2/BTCL2, which is highly expressed in fetal brain and may play a role in the development and/or maintenance of neuronal circuitry [22,23] as a similar gene in rats encodes a protein that modulates glutamate signalling in the brain by regulating kainate receptor function [24,25]. The over-expressed VPS35 is also highly expressed in the brain, and encodes an essential component of the retromer complex [26,27]. The investigation of other plausible candidates (e.g. CBLN1) was precluded by their lack of expression in lymphoblastoid cell lines.
The imbalance in the X chromosome inherited from the patient mother could contribute to the patient's phenotype as it contains MSL3 gene and RFMPD4/PDZD10 gene, acting respectively as chromatin remodelling/transcription regulating factor [28,29] and as positive regulator of dendritic spine morphogenesis, with a role in the maintenance of excitatory synaptic transmission [30] Nonetheless, the X inactivation analysis showed a random X inactivation in the mother and a preferential maternal X inactivation in the daughter, excluding a role for this CNV in the patient's phenotype.

Conclusion
Our results, which were obtained from genetic and epigenetic studies, suggest the occurrence in human of the PE due to an euchromatin/constitutive heterochromatin juxtaposition. To our knowledge this is the first report of a human disease involving this kind of PE mechanism.
Despite our extensive characterisation, the multiple genes that may be involved prevent a full understanding of the pathogenetic mechanism. However, we can reasonably assume that the heterochromatic PE in our case leads to a situation that is comparable with that observed in contiguous gene syndromes. Other carriers of balanced translocations with a breakpoint in heterochromatic sequences need to be similarly studied in order to be able to generalise the contribution of heterochromatic PE to human pathology.

Materials and methods
Cytogenetic and FISH analyses QFQ-banded metaphases prepared from peripheral blood lymphocytes using standard procedures were analysed, and their karyotype was described in accordance with ISCN (2009) [31]. The probes used for the FISH characterisation of the 15p bkp were the RPC1-98 C19 and RPC1-21I10 clones, provided by Invitrogene (cat. N. RPCI1.C), both containing beta-satellite DNA and targeting 15p11 and 15p13 heterochromatic bands [32]; the RP5-1174A5 clone, which contains the DNA/rRNA coding sequences (NOR regions) http://www.biologia.uniba. it/rmc/5-alfoidi/dJ1174A5.html has been selected from RPCI-5 PACs library provided by the YAC Screening Centre (Dibit, HSR, Milan, Italy); the D15Z1 satellite III probe mixed with the GABRB3 probe (Abbott, cat N. 05 J22-015); and the pMC15 (D15Z3) and pTRA25 (D15Z4) clones containing specific chromosome 15 alpha satellite sequences [33] were kindly provided by Prof. Mariano Rocchi, University of Bari. The contiguous BACs, belonging to the RPCI-11 and the CTD libraries and used for the FISH characterisation of the 16q bkp, were selected using the UCSC Genome Browser (http://genome.ucsc.edu/, hg19) and provided by Prof. Mariano Rocchi (RPCI-11 clones) and by Invitrogene (CITB Human D BAC clones, cat. N. 96012D) The 16p bkp was also characterised by means of a 7 kb probe covering part of ITFG1 intron 12, which was constructed by means of long-range PCR using the Expand 20KB PLUS PCR System (Roche, cat N. 11811002001) as previously described [34]. The primer pair (fw 5-GGCACCATCTTGGCACACTGCAACCTC3-and rev 5-ATTAGGAACCAGGCCGCACGACAGGAG 3-) was designed from the genomic contig NT_010498, and the amplicon sequences (chr16:47,280,772-47,287,772, hg19) were checked by BLAT against the human sequence in order to ensure amplification specifity. Ten nanograms of clone CTD-2027D24 were used as a template. The 7 kb PCR product was run in 1% low-melting agarose and subsequently purified. All of the BAC/PAC probes were nick-translation labelled with biotin (Roche, cat N. 11093070910), digoxigenin (Roche, cat N. 11093088910) or CY3-dUTP (Amersham, cat N.PAS3022), whereas the PCR product probe was labelled by means of oligopriming (Prime-It Fluor Fluorescence labelling kit, Stratagene, cat. N. 300380). All of the FISH probes were used following the protocols of Lichter and Cremer with minor modifications [35]. Each of the BAC clones was previously tested on chromosomal spreads from normal controls, and only those giving a signal on the chromosomal band indicated by the Genome Browser (UCSC) were used. The specificity of the probe obtained by means of long-range PCR was validated by a FISH experiment that gave a single signal on chromosome 16q11.2. Digital images were obtained using a Leica DMR epifluorescence microscope (Leica Imaging Systems Ltd) equipped with a CCD camera (Cohu Inc). DEAC, FITC, Cy3, and DAPI fluorescence signals were detected using specific filters. The images were recorded, pseudocolored, and merged using QFISH software (Leica Imaging Systems Ltd), and finally edited using Adobe Photoshop CS4 (Adobe Systems).

Array comparative genome hybridisation (CGH) analysis
For array CGH analysis genomic DNA was extracted from proband's and parent's whole blood using the Dneasy Blood & Tissue Kit (Sigma-Aldrich, cat. N. NA2020) according to the manufacturer's instructions. Pooled DNA from the peripheral blood of 10 healthy donors (Promega, cat. N. G1521), sex-matched to the samples, was used as a reference. Genome scan was performed by the Human Genome CGH Microarray Kit 244 K (Agilent Technologies, cat. N. G4411B). Briefly, 3 μg of DNA from the test and the normal reference were processed according to the manufacturer's protocol. Images were extracted using Agilent Feature Extraction software 9.1 and analysed by DNA Analytics 4.0 software. A log ratio plot between test and reference genomic DNA was assigned so that aberrations in test DNA copy number at a particular locus are observed as the deviation of the ratio plots from a modal value of 0. Aberration calls were identified by the ADM-2 algorithm.

X inactivation analysis
The X-inactivation pattern was assessed using the Humara Androgen Receptor locus and the DXS6673E locus methylation assay as previously described [36,37]. PCR products were run by capillary electrophoresis and XCI values determined in heterozygous cases using the formula previously reported [38]. This formula calculates the XCI pattern as (d1/u1)/(d1/u1 + d2/u2), where d1 and d2 represent the two peak areas from the digested sample and u1 and u2 are the corresponding areas of the alleles obtained from undigested DNA.
The investigated genes were SHCBP1, VPS35, ORC6L, GPT2, DNA JA2, NETO2/BTCL2, ITFG1, ABCC12, SIAH1, N4BP, CBLN1, ADCY7, BRD7, CYLD, SALL1, RBL2, HERPUD, and CKLF. The primer pairs designed for each gene, which spanned coding sequences, are listed in Additional file 5: Table S3. Quantitative RT-PCR was performed using the comparative threshold cycle method of the Syber-Green (SYBR ® Green PCR Master Mix -Applied Biosystem, cat N. 4309155) protocol, as described by Livak and Aarskog and Vedeler [39,40]. The RT-PCR and data analyses were carried out using an ABI PRISM 7900HT Sequence Detection System (Applied Biosystem). The estimated expression of all of the genes was computed from three independent retrotranscriptions obtained from three different RNA extractions; the threshold cycle number (Ct) values were obtained from three replicas, and normalised to the endogenous HGPRT control gene. RNA from three healthy human controls (one female and two males) were used as the calibrator sample.
Ct values were determined for all of the PCR reactions, and a comparative Ct method was used to calculate the copy numbers of the expressed transcripts. The relative quantity of expressed transcript was calculated as RQ = 2-(ΔΔCt ± standard deviation [SD]) [39,40].

Methylation analysis
The methylation status of the NETO2/BTCL2, SIAH1, RBL2 and CYLD promoter regions was analysed by means of pyrosequencing technology. The bisulfite conversion of genomic DNA (1 μg) was obtained using the EZ DNA methylation kit (Zymo Research, cat N. D5002). After bisulfite treatment, PCR was carried out in a final volume of 50 μl with 2.5 units of Promega Go-Taq polymerase (Promega, cat N. M8301). The primers for modified sequences are summarised in Additional file 4: Table S3. The PCR conditions were 45 cycles of 95°C for 30 sec, 52°C for 30 sec, and 72°C for 30 sec, followed by 72°C for 5 min; 30 μl of the PCR product were used for the pyrosequencing assay carried out using the sequencing primers ( Figure 3).
The pyrosequencing reactions were performed in a PSQ HS 96 System (Biotage, Uppsala, Sweden) by using Pyro Gold Reagent kits (Qiagene, cat. N. 972804). Methylation was quantified using Pyro Q-CpG Software (Biotage), which calculates the ratio of converted C's (T's) to unconverted C's at each CpG site and expresses this as a percentage of methylation. To assess the methylation pattern under normal conditions, we analysed a control lymphoblastoid cell line and different normal DNA from peripheral blood lymphocytes.
Quantitative RT-PCR was performed on immunoprecipitated chromatin as previously described in order to verify the quantitative changes in histone modifications on the VPS35 promoter. The primer sequences specific for the VPS35, NETO2/BTCL2, SIAH1, PGK1 and HPRT genes (the last two used as controls) are shown in Additional file 6: Table S4.