The RNA tether model for human chromosomal translocation fragile zones.

One of the two chromosomal breakage events in recurring translocations in B cell neoplasms is often due to the recombination-activating gene complex (RAG complex) releasing DNA ends before end joining. The other break occurs in a fragile zone of 20-600 bp in a non-antigen receptor gene locus, with a more complex and intriguing set of mechanistic factors underlying such narrow fragile zones. These factors include activation-induced deaminase (AID), which acts only at regions of single-stranded DNA (ssDNA). Recent work leads to a model involving the tethering of AID to the nascent RNA as it emerges from the RNA polymerase. This mechanism may have relevance in class switch recombination (CSR) and somatic hypermutation (SHM), as well as broader relevance for other DNA enzymes.

Construction of high coverage whole-genome sequencing libraries from single colon crypts without DNA extraction or whole-genome amplification.

OBJECTIVE: Comprehensive and reliable genome-wide variant analysis of a small number of cells has been challenging due to genome coverage bias, PCR over-cycling, and the requirement of expensive technologies. To comprehensively identify genome alterations in single colon crypts that reflect genome heterogeneity of stem cells, we developed a method to construct whole-genome sequencing libraries from single colon crypts without DNA extraction, whole-genome amplification, or increased PCR enrichment cycles. RESULTS: We present post-alignment statistics of 81 single-crypts (each contains four- to eight-fold less DNA than the requirement of conventional methods) and 16 bulk-tissue libraries to demonstrate the consistent success in obtaining reliable coverage, both in depth (≥ 30X) and breadth (≥ 92% of the genome covered at ≥ 10X depth), of the human genome. These single-crypt libraries are of comparable quality as libraries generated with the conventional method using high quality and quantities of purified DNA. Conceivably, our method can be applied to small biopsy samples from many tissues and can be combined with single cell targeted sequencing to comprehensively profile cancer genomes and their evolution. The broad potential application of this method offers expanded possibilities in cost-effectively examining genome heterogeneity in small numbers of cells at high resolution.

Artemis inhibition as a therapeutic strategy for acute lymphoblastic leukemia.

As effective therapies for relapse and refractory B-cell acute lymphoblastic leukemia (B-ALL) remain problematic, novel therapeutic strategies are needed. Artemis is a key endonuclease in V(D)J recombination and nonhomologous end joining (NHEJ) of DNA double-strand break (DSB) repair. Inhibition of Artemis would cause chromosome breaks during maturation of RAG-expressing T- and B-cells. Though this would block generation of new B- and T-cells temporarily, it could be oncologically beneficial for reducing the proliferation of B-ALL and T-ALL cells by causing chromosome breaks in these RAG-expressing tumor cells. Currently, pharmacological inhibition is not available for Artemis. According to gene expression analyses from 207 children with high-risk pre-B acute lymphoblastic leukemias high Artemis expression is correlated with poor outcome. Therefore, we evaluated four compounds (827171, 827032, 826941, and 825226), previously generated from a large Artemis targeted drug screen. A biochemical assay using a purified Artemis:DNA-PKcs complex shows that the Artemis inhibitors 827171, 827032, 826941, 825226 have nanomolar IC50 values for Artemis inhibition. We compared these 4 compounds to a DNA-PK inhibitor (AZD7648) in three patient-derived B-ALL cell lines (LAX56, BLQ5 and LAX7R) and in two mature B-cell lines (3301015 and 5680001) as controls. We found that pharmacological Artemis inhibition substantially decreases proliferation of B-ALL cell lines while normal mature B-cell lines are not markedly affected. Inhibition of DNA-PKcs (which regulates Artemis) using the DNA-PK inhibitor AZD7648 had minor effects on these same primary patient-derived ALL lines, indicating that inhibition of V(D)J hairpin opening requires direct inhibition of Artemis, rather than indirect suppression of the kinase that regulates Artemis. Our data provides a basis for further evaluation of pharmacological Artemis inhibition of proliferation of B- and T-ALL.

The mRNA tether model for activation-induced deaminase and its relevance for Ig somatic hypermutation and class switch recombination.

Activation-induced deaminase (AID) only deaminates cytosine within single-stranded DNA. Transcription is known to increase AID deamination on duplex DNA substrates during transcription. Using a purified T7 RNA polymerase transcription system, we recently found that AID deamination of a duplex DNA substrate is reduced if RNase A is added during transcription. This finding prompted us to consider that the mRNA tail may contribute to AID action at the nearby transcribed strand (TS) or non-transcribed strand (NTS) of DNA, which are transiently single-stranded in the wake of RNA polymerase movement. Here, we used a purified system to test whether a single-stranded oligonucleotide (oligo) consisting of RNA in the 5' portion and DNA in the 3' portion (i.e., 5'RNA-DNA3', also termed an RNA-DNA fusion substrate) could be deaminated equally efficiently as the same sequence when it is entirely DNA. We found that AID acts on the RNA-DNA fusion substrate and the DNA-only substrate with similar efficiency. Based on this finding and our recent observation on the importance of the mRNA tail, we propose a model in which the proximity and length of the mRNA tail provide a critical site for AID loading to permit a high local collision frequency with the NTS and TS in the transient wake of the RNA polymerase. When the mRNA tail is not present, we know that AID action drops to levels equivalent to when there is no transcription at all. This mRNA tether model explains several local and global features of Ig somatic hypermutation and Ig class switch recombination, while integrating structural and functional features of AID.

Preclinical Evaluation of a Novel Dual Targeting PI3Kδ/BRD4 Inhibitor, SF2535, in B-Cell Acute Lymphoblastic Leukemia.

The PI3K/Akt pathway-and in particular PI3Kδ-is known for its role in drug resistant B-cell acute lymphoblastic leukemia (B-ALL) and it is often upregulated in refractory or relapsed B-ALL. Myc proteins are transcription factors responsible for transcribing pro-proliferative genes and c-Myc is often overexpressed in cancers. The chromatin regulator BRD4 is required for expression of c-Myc in hematologic malignancies including B-ALL. Previously, combination of BRD4 and PI3K inhibition with SF2523 was shown to successfully decrease Myc expression. However, the underlying mechanism and effect of dual inhibition of PI3Kδ/BRD4 in B-ALL remains unknown. To study this, we utilized SF2535, a novel small molecule dual inhibitor which can specifically target the PI3Kδ isoform and BRD4. We treated primary B-ALL cells with various concentrations of SF2535 and studied its effect on specific pharmacological on-target mechanisms such as apoptosis, cell cycle, cell proliferation, and adhesion molecules expression usingin vitro and in vivo models. SF2535 significantly downregulates both c-Myc mRNA and protein expression through inhibition of BRD4 at the c-Myc promoter site and decreases p-AKT expression through inhibition of the PI3Kδ/AKT pathway. SF2535 induced apoptosis in B-ALL by downregulation of BCL-2 and increased cleavage of caspase-3, caspase-7, and PARP. Moreover, SF2535 induced cell cycle arrest and decreased cell counts in B-ALL. Interestingly, SF2535 decreased the mean fluorescence intensity (MFI) of integrin α4, α5, α6, and ß1 while increasing MFI of CXCR4, indicating that SF2535 may work through inside-out signaling of integrins. Taken together, our data provide a rationale for the clinical evaluation of targeting PI3Kδ/BRD4 in refractory or relapsed B-ALL using SF2535.

Mechanistic basis for chromosomal translocations at the E2A gene and its broader relevance to human B cell malignancies.

Analysis of translocation breakpoints in human B cell malignancies reveals that DNA double-strand breaks at oncogenes most frequently occur at CpG sites located within 20-600 bp fragile zones and depend on activation-induced deaminase (AID). AID requires single-stranded DNA (ssDNA) to act, but it has been unclear why or how this region transiently acquires a ssDNA state. Here, we demonstrate the ssDNA state in the 23 bp E2A fragile zone using several methods, including native bisulfite DNA structural analysis in live human pre-B cells. AID deamination within the E2A fragile zone does not require but is increased upon transcription. High C-string density, nascent RNA tails, and direct DNA sequence repeats prolong the ssDNA state of the E2A fragile zone and increase AID deamination at overlapping AID hotspots that contain the CpG sites at which breaks occur in patients. These features provide key insights into lymphoid fragile zones generally.

Two-stage Study of Familial Prostate Cancer by Whole-exome Sequencing and Custom Capture Identifies 10 Novel Genes Associated with the Risk of Prostate Cancer.

BACKGROUND: Family history of prostate cancer (PCa) is a well-known risk factor, and both common and rare genetic variants are associated with the disease. OBJECTIVE: To detect new genetic variants associated with PCa, capitalizing on the role of family history and more aggressive PCa. DESIGN, SETTING, AND PARTICIPANTS: A two-stage design was used. In stage one, whole-exome sequencing was used to identify potential risk alleles among affected men with a strong family history of disease or with more aggressive disease (491 cases and 429 controls). Aggressive disease was based on a sum of scores for Gleason score, node status, metastasis, tumor stage, prostate-specific antigen at diagnosis, systemic recurrence, and time to PCa death. Genes identified in stage one were screened in stage two using a custom-capture design in an independent set of 2917 cases and 1899 controls. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Frequencies of genetic variants (singly or jointly in a gene) were compared between cases and controls. RESULTS AND LIMITATIONS: Eleven genes previously reported to be associated with PCa were detected (ATM, BRCA2, HOXB13, FAM111A, EMSY, HNF1B, KLK3, MSMB, PCAT1, PRSS3, and TERT), as well as an additional 10 novel genes (PABPC1, QK1, FAM114A1, MUC6, MYCBP2, RAPGEF4, RNASEH2B, ULK4, XPO7, and THAP3). Of these 10 novel genes, all but PABPC1 and ULK4 were primarily associated with the risk of aggressive PCa. CONCLUSIONS: Our approach demonstrates the advantage of gene sequencing in the search for genetic variants associated with PCa and the benefits of sampling patients with a strong family history of disease or an aggressive form of disease. PATIENT SUMMARY: Multiple genes are associated with prostate cancer (PCa) among men with a strong family history of this disease or among men with an aggressive form of PCa.

Temporally uncoupled signal and coding joint formation in human V(D)J recombination.

In vertebrate antigen receptor gene rearrangement, V(D)J recombination events can occur by deletion or by inversion. For deletional events, the signal joint is deleted from the genome. Nearly half of the immunoglobulin light chain genes undergo V(D)J recombination in an inversional manner, and both signal and coding joint formation must occur to retain chromosomal integrity. But given the undetermined amount of pre-B and pre-T cell death that occurs during V(D)J recombination, the efficiency with which both joints are completed is not known, nor is the relative efficiency (balance) of signal versus coding joint formation. Signal joint formation only requires Ku and XRCC4:DNA ligase 4 of the nonhomologous DNA end joining repair pathway. Coding joint formation requires these proteins as well, but in addition requires Artemis and DNA-dependent protein kinase to open the hairpin DNA coding ends, which the RAG complex generated; and further processing is required because the hairpin opening generates incompatible 3' overhangs. Mutations in some of the end processing enzymes affect one, but only minimally the other joint. We have devised a precise cellular assay that does not have any cellular, enzymatic or biochemical selective bias to assess signal and coding joint formation independently, and it can detect intermediates for which one joint has formed but not the other. We find that intermediates with only one completed joint are more abundant than molecules with both joints completed. This indicates that either joint can form independent of the other and joint formation can be a relatively slow process.

DNA-PKcs chemical inhibition versus genetic mutation: Impact on the junctional repair steps of V(D)J recombination.

Spontaneous DNA-PKcs deficiencies in animals result in a severe combined immunodeficiency (SCID) phenotype because DNA-PKcs is required to activate Artemis for V(D)J recombination coding end hairpin opening. The impact on signal joint formation in these spontaneous mutant mammals is variable. Genetically engineered DNA-PKcs null mice and cells from them show a >1,000-fold reduction in coding joint formation and minimal reduction in signal joint formation during V(D)J recombination. Does chemical inhibition of DNA-PKcs mimic this phenotype? M3814 (also known as Nedisertib) is a potent DNA-PKcs inhibitor. We find here that M3814 causes a quantitative reduction in coding joint formation relative to signal joint formation. The sequences of signal and coding junctions were within normal limits, though rare coding joints showed novel features. The signal junctions generally did not show evidence of resection into the signal ends that is often seen in cells that have genetic defects in DNA-PKcs. Comparison of the chemical inhibition findings here with the known results for spontaneous and engineered DNA-PKcs mutant mammals is informative for considering pharmacologic small molecule inhibition of DNA-PKcs in various types of neoplasia.

Derivation of induced pluripotent stem cells from ferret somatic cells.

Ferrets are an attractive mammalian model for several diseases, especially those affecting the lungs, liver, brain, and kidneys. Many chronic human diseases have been difficult to model in rodents due to differences in size and cellular anatomy. This is particularly the case for the lung, where ferrets provide an attractive mammalian model of both acute and chronic lung diseases, such as influenza, cystic fibrosis, A1A emphysema, and obliterative bronchiolitis, closely recapitulating disease pathogenesis, as it occurs in humans. As such, ferrets have the potential to be a valuable preclinical model for the evaluation of cell-based therapies for lung regeneration and, likely, for other tissues. Induced pluripotent stem cells (iPSCs) provide a great option for provision of enough autologous cells to make patient-specific cell therapies a reality. Unfortunately, they have not been successfully created from ferrets. In this study, we demonstrate the generation of ferret iPSCs that reflect the primed pluripotent state of human iPSCs. Ferret fetal fibroblasts were reprogrammed and acquired core features of pluripotency, having the capacity for self-renewal, multilineage differentiation, and a high-level expression of the core pluripotency genes and pathways at both the transcriptional and protein level. In conclusion, we have generated ferret pluripotent stem cells that provide an opportunity for advancing our capacity to evaluate autologous cell engraftment in ferrets.

Data-adaptive multi-locus association testing in subjects with arbitrary genealogical relationships.

Genome-wide sequencing enables evaluation of associations between traits and combinations of variants in genes and pathways. But such evaluation requires multi-locus association tests with good power, regardless of the variant and trait characteristics. And since analyzing families may yield more power than analyzing unrelated individuals, we need multi-locus tests applicable to both related and unrelated individuals. Here we describe such tests, and we introduce SKAT-X, a new test statistic that uses genome-wide data obtained from related or unrelated subjects to optimize power for the specific data at hand. Simulations show that: a) SKAT-X performs well regardless of variant and trait characteristics; and b) for binary traits, analyzing affected relatives brings more power than analyzing unrelated individuals, consistent with previous findings for single-locus tests. We illustrate the methods by application to rare unclassified missense variants in the tumor suppressor gene BRCA2, as applied to combined data from prostate cancer families and unrelated prostate cancer cases and controls in the Multi-ethnic Cohort (MEC). The methods can be implemented using open-source code for public use as the R-package GATARS (Genetic Association Tests for Arbitrarily Related Subjects) .

Novel Lines of Evidence for the Asymmetric Strand Displacement Model of Mitochondrial DNA Replication.

The mitochondrial genome, which consists of 16,569 bp of DNA with a cytosine-rich light (L) strand and a heavy (H) strand, exists as a multicopy closed circular genome within the mitochondrial matrix. The machinery for replication of the mammalian mitochondrial genome is distinct from that for replication of the nuclear genome. Three models have been proposed for mitochondrial DNA (mtDNA) replication, and one of the key differences among them is whether extensive single-stranded regions exist on the H strand. Here, three different methods that can detect single-stranded DNA (ssDNA) are utilized to identify the presence, location, and abundance of ssDNA on mtDNA. Importantly, none of these newly described methods involve the complication of prior mtDNA fractionation. The H strand was found to have extensive single-stranded regions with a profile consistent with the strand displacement model of mtDNA replication, whereas single strandedness was predominantly absent on the L strand. These findings are consistent with the in vivo occupancy of mitochondrial single-stranded DNA binding protein reported previously and provide strong new qualitative and quantitative evidence for the asymmetric strand displacement model of mtDNA replication.

gsSKAT: Rapid gene set analysis and multiple testing correction for rare-variant association studies using weighted linear kernels.

Next-generation sequencing technologies have afforded unprecedented characterization of low-frequency and rare genetic variation. Due to low power for single-variant testing, aggregative methods are commonly used to combine observed rare variation within a single gene. Causal variation may also aggregate across multiple genes within relevant biomolecular pathways. Kernel-machine regression and adaptive testing methods for aggregative rare-variant association testing have been demonstrated to be powerful approaches for pathway-level analysis, although these methods tend to be computationally intensive at high-variant dimensionality and require access to complete data. An additional analytical issue in scans of large pathway definition sets is multiple testing correction. Gene set definitions may exhibit substantial genic overlap, and the impact of the resultant correlation in test statistics on Type I error rate control for large agnostic gene set scans has not been fully explored. Herein, we first outline a statistical strategy for aggregative rare-variant analysis using component gene-level linear kernel score test summary statistics as well as derive simple estimators of the effective number of tests for family-wise error rate control. We then conduct extensive simulation studies to characterize the behavior of our approach relative to direct application of kernel and adaptive methods under a variety of conditions. We also apply our method to two case-control studies, respectively, evaluating rare variation in hereditary prostate cancer and schizophrenia. Finally, we provide open-source R code for public use to facilitate easy application of our methods to existing rare-variant analysis results.

REVEL: An Ensemble Method for Predicting the Pathogenicity of Rare Missense Variants.

The vast majority of coding variants are rare, and assessment of the contribution of rare variants to complex traits is hampered by low statistical power and limited functional data. Improved methods for predicting the pathogenicity of rare coding variants are needed to facilitate the discovery of disease variants from exome sequencing studies. We developed REVEL (rare exome variant ensemble learner), an ensemble method for predicting the pathogenicity of missense variants on the basis of individual tools: MutPred, FATHMM, VEST, PolyPhen, SIFT, PROVEAN, MutationAssessor, MutationTaster, LRT, GERP, SiPhy, phyloP, and phastCons. REVEL was trained with recently discovered pathogenic and rare neutral missense variants, excluding those previously used to train its constituent tools. When applied to two independent test sets, REVEL had the best overall performance (p < 10-12) as compared to any individual tool and seven ensemble methods: MetaSVM, MetaLR, KGGSeq, Condel, CADD, DANN, and Eigen. Importantly, REVEL also had the best performance for distinguishing pathogenic from rare neutral variants with allele frequencies <0.5%. The area under the receiver operating characteristic curve (AUC) for REVEL was 0.046-0.182 higher in an independent test set of 935 recent SwissVar disease variants and 123,935 putatively neutral exome sequencing variants and 0.027-0.143 higher in an independent test set of 1,953 pathogenic and 2,406 benign variants recently reported in ClinVar than the AUCs for other ensemble methods. We provide pre-computed REVEL scores for all possible human missense variants to facilitate the identification of pathogenic variants in the sea of rare variants discovered as sequencing studies expand in scale.

Post hoc Analysis for Detecting Individual Rare Variant Risk Associations Using Probit Regression Bayesian Variable Selection Methods in Case-Control Sequencing Studies.

Rare variants (RVs) have been shown to be significant contributors to complex disease risk. By definition, these variants have very low minor allele frequencies and traditional single-marker methods for statistical analysis are underpowered for typical sequencing study sample sizes. Multimarker burden-type approaches attempt to identify aggregation of RVs across case-control status by analyzing relatively small partitions of the genome, such as genes. However, it is generally the case that the aggregative measure would be a mixture of causal and neutral variants, and these omnibus tests do not directly provide any indication of which RVs may be driving a given association. Recently, Bayesian variable selection approaches have been proposed to identify RV associations from a large set of RVs under consideration. Although these approaches have been shown to be powerful at detecting associations at the RV level, there are often computational limitations on the total quantity of RVs under consideration and compromises are necessary for large-scale application. Here, we propose a computationally efficient alternative formulation of this method using a probit regression approach specifically capable of simultaneously analyzing hundreds to thousands of RVs. We evaluate our approach to detect causal variation on simulated data and examine sensitivity and specificity in instances of high RV dimensionality as well as apply it to pathway-level RV analysis results from a prostate cancer (PC) risk case-control sequencing study. Finally, we discuss potential extensions and future directions of this work.

Sensitivity and specificity of immunoprecipitation of DNA containing 5-Methylcytosine.

BACKGROUND: Attempts to enrich or identify DNA with cytosine methylation have been commonly carried out using anti-5-methylcytosine or anti-MBD2 (methyl-CpG binding domain protein 2) antibody in immunoprecipitation (IP) assays. However, a careful and systematic control experiment to examine the sensitivity and specificity of this approach has not been reported. It is of critical importance to understand the potential pitfalls of this approach and to avoid potential misinterpretation of findings. FINDINGS: We found that increased concentration of antibody used in the assay increased the amount of overall DNA captured as expected. The increased number of methylated cytosines in/on the DNA fragment also increased the amount of DNA captured by the antibody. Importantly, the antibody can bind to some fully unmethylated DNA fragments, even when fully methylated DNA is present in the same experiment. CONCLUSION: The sensitivity of anti-5-methylcytosine antibody and anti-MBD2 antibody/MBD2 binding varies with the number of methylated cytosines on the DNA target. The specificity of these antibodies can also vary for different DNA target sequences. DNA fragments with fewer CpG sites may not bind to these antibodies even when all are methylated while DNA fragments with more CpG sites may bind to the antibodies when only some of these sites are methylated. More importantly, binding of DNA to these antibodies does not always indicate the presence of DNA methylation. It is clear that false positive and false negative findings can be easily reached even though it does not nullify these convenient and simple methods completely. Great caution should be taken for the interpretation of IP results using these antibodies and rigorous confirmation by sodium bisulfite sequencing is essential.

Complexities due to single-stranded RNA during antibody detection of genomic rna:dna hybrids.

BACKGROUND: Long genomic R-loops in eukaryotes were first described at the immunoglobulin heavy chain locus switch regions using bisulfite sequencing and functional studies. A mouse monoclonal antibody called S9.6 has been used for immunoprecipitation (IP) to identify R-loops, based on the assumption that it is specific for RNA:DNA over other nucleic acid duplexes. However, recent work has demonstrated that a variable domain of S9.6 binds AU-rich RNA:RNA duplexes with a KD that is only 5.6-fold weaker than for RNA:DNA duplexes. Most IP protocols do not pre-clear the genomic nucleic acid with RNase A to remove free RNA. Fold back of ssRNA can readily generate RNA:RNA duplexes that may bind the S9.6 antibody, and adventitious binding of RNA may also create short RNA:DNA regions. Here we investigate whether RNase A is needed to obtain reliable IP with S9.6. FINDINGS: As our test locus, we chose the most well-documented site for kilobase-long mammalian genomic R-loops, the immunoglobulin heavy chain locus (IgH) class switch regions. The R-loops at this locus can be induced by using cytokines to stimulate transcription from germline transcript promoters. We tested IP using S9.6 with and without various RNase treatments. The RNase treatments included RNase H to destroy the RNA in an RNA:DNA duplex and RNase A to destroy single-stranded (ss) RNA to prevent it from binding S9.6 directly (as duplex RNA) and to prevent the ssRNA from annealing to the genome, resulting in adventitious RNA:DNA hybrids. We find that optimal detection of RNA:DNA duplexes requires removal of ssRNA using RNase A. Without RNase A treatment, known regions of R-loop formation containing RNA:DNA duplexes can not be reliably detected. With RNase A treatment, a signal can be detected over background, but only within a limited 2 or 3-fold range, even with a stable kilobase-long genomic R-loop. CONCLUSION: Any use of the S9.6 antibody must be preceded by RNase A treatment to remove free ssRNA that may compete for the S9.6 binding by forming RNA:RNA regions or short, transient RNA:DNA duplexes. Caution should be used when interpreting S9.6 data, and confirmation by independent structural and functional methods is essential.

Effect of CpG dinucleotides within IgH switch region repeats on immunoglobulin class switch recombination.

Immunoglobulin (Ig) heavy chains undergo class switch recombination (CSR) to change the heavy chain isotype from IgM to IgG, A or E. The switch regions are several kilobases long, repetitive, and G-rich on the nontemplate strand. They are also relatively depleted of CpG (also called CG) sites for unknown reasons. Here we use synthetic switch regions at the IgH switch alpha (Sα) locus to test the effect of CpG sites and to try to understand why the IgH switch sequences evolved to be relatively depleted of CpG. We find that even just two CpG sites within an 80 bp synthetic switch repeat iterated 15 times (total switch region length of 1200 bp containing 30 CpG sites) are sufficient to dramatically reduce both Ig CSR and transcription through the switch region from the upstream Iα sterile transcript promoter, which is the promoter that directs transcripts through the Sα region. De novo DNA methylation occurs at the four CpG sites in and around the Iα promoter when each 80 bp Iα switch repeat contains the two CpG sites. Thus, a relatively low density of CpG sites within the switch repeats can induce upstream CpG methylation at the IgH alpha locus, and cause a substantial decrease in transcription from the sterile transcript promoter. This effect is likely the reason that switch regions evolved to contain very few CpG sites. We discuss these findings as they relate to DNA methylation and to Ig CSR.

Human lymphoid translocation fragile zones are hypomethylated and have accessible chromatin.

Chromosomal translocations are a hallmark of hematopoietic malignancies. CG motifs within translocation fragile zones (typically 20 to 600 bp in size) are prone to chromosomal translocation in lymphomas. Here we demonstrate that the CG motifs in human translocation fragile zones are hypomethylated relative to the adjacent DNA. Using a methyltransferase footprinting assay on isolated nuclei (in vitro), we find that the chromatin at these fragile zones is accessible. We also examined in vivo accessibility using cellular expression of a prokaryotic methylase. Based on this assay, which measures accessibility over a much longer time interval than is possible with in vitro methods, these fragile zones were found to be more accessible than the adjacent DNA. Because DNA within the fragile zones can be methylated by both cellular and exogenous methyltransferases, the fragile zones are predominantly in a duplex DNA conformation. These observations permit more-refined models for why these zones are 100- to 1,000-fold more prone to undergo chromosomal translocation than the adjacent regions.